- general attitude/philosophy (prelude)
- gag buffs/nerfs
- accuracy
- facilities
- cogs and laff
- toontasks
- gag tracks
- trees
- odds and ends
- appendix (a.k.a. random garbage)
TTO/TTR (henceforth TTR, because TTO no longer exists) is an excellent game, but suffers (as so many games do) from some design decisions that unfortunately were not quite prescient enough to see into the far future. There are issues of balance: between gag tracks, between cogs and toons, &c. There are issues of overly restrictive gameplay, and gameplay not adapted to the ways that players have shaped it. It has historically been generally wise to adapt games to the ways that players have intentionally adjusted them for the purpose of fun. This document catalogues some very concrete proposed changes that typically err on the side of only that which is needed to improve the game, rather than change the game unnecessarily or add to it. This document seeks not to imagine a new game, but rather to scratch the surface of an already-existing game (viz. Toontown) and reveal the superior game that was within it all along.[1]
Some of the changes proposed here are meant to be combined with other such changes, but some may be designed without other changes in mind, for the sake of independence and flexibility. This means — among other things — that this document is not a blueprint for a particular, single, version of Toontown with some choice combination of all of the proposed changes discussed within this document. Rather, when concretely forming an actual version of Toontown, careful judgement and deliberation should be applied to the concerns, reasoning, and proposals contained within this document.
While the changes made here are obviously, first and foremost, changes made to balance the game and make it more fun, there are also some aesthetic considerations to fill in the gaps. Mostly, this comes in the form of what numbers are chosen in cases where numeric values are being balanced. The tendency is to favor numbers that are more “round” (read: regular, or its synonym, 5-smooth), and to favor sequences and mathematical functions that are more “beautiful”. In this document, heeding these aesthetic concerns is referred to as “rationalization” (both in the mathematical sense, from “ratio”, and the usual sense of “to give a rationale”). Again, these concerns are always strictly secondary to the more important concerns of game balance & fun.
Other servers besides TTR exist, but are often dedicated to “adding to” the base game and/or “modernizing” it. While these are certainly noble goals, they often amount to addition for the sake of addition (e.g. entirely new cog types, entirely new gag tracks, &c.) and revisionism for the same of revisionism (a kind of pathological uniqueness). This document instead focuses on TTR-as-essentially-TTO, that is, the “base game” in more or less the same state gameplay-wise as TTO in 2013. This acts as a starting point, or even tabula rasa (a tabula rasa that, like a newborn human mind, is not really a “blank” slate at all…).
Suggestions/corrections/additions/typo-fixes/&c. are encouraged. You can file an issue and/or submit a pull request and/or contact me directly.
By contributing, you agree to license your contributions under the same license as this work (see the “legal” section).
This work is licensed to you (or to anyone else) under the terms of the Creative Commons Attribution-ShareAlike 4.0 International license (or any later version of the same license, at your option).
footnotes for “general attitude/philosophy (prelude)”
- See my open letter, “On sellbot field offices (SBFOs), especially concerning the problem of ‘difficulty’ in Toontown” if you are interested in more abstract considerations of the kind mentioned in this paragraph.
Balancing toonup gags is a bit of a different beast from balancing any of the other six tracks; rather than balancing the tracks against themselves, toonup is fundamentally different and can only be balanced from a perspective that includes not just the HP of cogs, but also the attacking power of cogs and the cog-elimating powers of the other six gag tracks.
In TTR, toonup is very much overpowered, because lure and sound are powerful enough to make laff loss a relatively rare event. Whenever toonup is eventually used, it does the job with just 1 or 2 gags. With nerfs to lure and sound making cogs generally (buffs to trap notwithstanding) more difficult to defeat, this effect should be counteracted a bit and thus make toonup not so overpowered.
That being said, there are still two outstanding problems with toonup:
- Organic toonup is still probably underpowered even after nerfing lure and sound.
- Bamboo Cane does worryingly little damage compared to the other toonup gags.
It’s well-known that in TTR, organic toonup is incredibly underpowered. It is, by far, the least powerful out of the seven possible tracks that can be made organic. Partly, this is due to toonup just being overpowered; since each toonup gag already does so much healing, adding just a tad extra is not going to be useful or even noticable. As a result, nerfing lure and sound, and thus making toonup less overpowered (relatively), already ameliorates the problem of organic toonup.
But there is another reason why organic toonup is so underpowered, and it has to do with the fact that toonup is fundamentally different from the other six gag tracks. With other gag tracks (particularly trap, sound, throw, squirt, and drop), very slight differences in a gag’s damage can make the difference between a cog not dying and a cog dying, simply because a cog with 1 HP is alive and one with 0 HP is not. Strictly speaking, a similar principle does apply to toonup: very slight differences in healing amounts can make the difference between a toon being fully healed and a toon not being fully healed, simply because a toon with 1 laff fewer than their max-laff is not fully healed, and a toon with laff equal to their max-laff is. But the difference between a toon with laff 1 fewer than their maximum and a cog with 1 HP is very stark. Getting a toon to the point of having 1 laff missing is “good enough”; it is not the same, but is nevertheless nearly indistinguishible from, having full laff. But a cog with 1 HP is not “good enough”; that cog still occupies a space in battle and can attack the toons at will, very unlike one with 0 HP.
This fundamental difference is not something to get rid of. It is just an inherent part of toonup affecting toons rather than cogs (unlike the other six gag tracks). Rather than trying to smooth over this fundamental difference, the best thing to do is to treat it like it is rather than trying to make organic toonup look like organic versions of other gag tracks. Making organic toonup gags do (roughly) 10% more healing than their inorganic counterparts is appealing because of its perceived consistency; but as is explained here, this consistency is only perceived, not actual.
In TTR, the Bamboo Cane gag just intuitively feels a bit underwhelming compared to the other six toonup gags. But is that feeling justified? Here are a few quantitative ways of looking at this question.
How much healing does the Bamboo Cane do in proportion to the other multi-target toonup gags (not including the level 7 gag)? The Bamboo Cane does 45/18 = 2.5 times more than the Megaphone, but the Juggling Balls do 120/45 = 2.666… times more than the Bamboo Cane. From this perspective, the Bamboo Cane is slightly underpowered.
How much single-target healing does the Bamboo Cane do compared to its corresponding single-target toonup gag? The Megaphone does 6/10 = 0.6 times as much healing as the Feather. The Juggling Balls do 40/70 = 0.571428… times as much healing as the Pixie Dust. The Bamboo Cane does 15/30 = 0.5 times as much healing as the Lipstick. From this perspective, the Bamboo Cane is underpowered.
How does the Bamboo Cane’s healing amount look in the context of toonup-gag-healing-amount-growth as a function of gag level?
Feather | Megaphone | Lipstick | Bamboo Cane | Pixie Dust | Juggling Balls | High Dive | |
---|---|---|---|---|---|---|---|
TTR | 10 | 18 | 30 | 45 | 70 | 120 | 210 |
δTTR | +8 | +12 | +15 | +25 | +50 | +90 | |
δδTTR | +4 | +3 | +10 | +25 | +40 |
As can be seen from the δδTTR row, the Bamboo Cane is the only toonup gag where the toonup-gag-healing-amount-growth takes a noticable “dip” or “decline” (in particular, note the dip from +4 down to +3). From this perspective, the Bamboo Cane is underpowered.
So it seems that the feeling that the Bamboo Cane gag is a bit underwhelming is actually justified. What, then, should be the healing amount of the Bamboo Cane? The most obvious choice is 48, since it’s the next multiple of 3 after 45. 48 actually has the advantage over 45 that it is also divisible by 2 (that is, it’s divisible by 6). Divisibility by 6 is particularly nice for the healing amounts of multi-target toonup gags because, when using a toonup gag, you are always healing either 1, 2, or 3 toons (since you cannot toonup yourself). The integers that are divisible by all three of those numbers are just those integers that are divisible by 6.
With the Bamboo Cane doing 48 laff of healing, it does 48/18 = 2.666… times more than the Megaphone, and the Juggling Balls do 120/48 = 2.5 times more than the Bamboo Cane. So they have been swapped in this respect and are as balanced as before (but now the Juggling Balls are the ones that look slightly underpowered). The Bamboo Cane would do 16/30 = 0.5333… times as much healing as the Lipstick. This is still less than both 0.6 and 0.571428…, but still an improvement. And the toonup gag healing amount table would look like this:
Feather | Megaphone | Lipstick | Bamboo Cane | Pixie Dust | Juggling Balls | High Dive | |
---|---|---|---|---|---|---|---|
A | 10 | 18 | 30 | 48 | 70 | 120 | 210 |
δA | +8 | +12 | +18 | +22 | +50 | +90 | |
δδA | +4 | +6 | +4 | +28 | +40 |
There is no longer a “dip” where the Bamboo Cane is, although it seems that the dip has just moved up by 1 gag level. While this is possibly unfortunate, it really shouldn’t matter, since the Pixie Dust is a single-target toonup gag; 70 points of healing is a lot for a single toon. However, it should be noted that bumping Pixie Dust up from 70 to 75 solves this problem (if it is a problem), and also makes Juggling Balls do 40/75 = 0.5333… times as much single-target healing as the Pixie Dust, which is exactly on par with Bamboo Cane doing 16/30 = 0.5333… times as much single-target healing as Lipstick.
Format:
healing (organic healing); propAcc
(organic propAcc
)
Feather | Megaphone | Lipstick | Bamboo Cane | Pixie Dust | Juggling Balls | High Dive | |
---|---|---|---|---|---|---|---|
TTR | 10 (11); 70 (70) | 18 (19); 70 (70) | 30 (33); 70 (70) | 45 (49); 70 (70) | 70 (77); 70 (70) | 120 (132); 70 (70) | 210 (231); 70 (70) |
A | 10 (12); 70 (70) | 18 (22); 70 (70) | 30 (36); 70 (70) | 48 (58); 70 (70) | 70 (84); 70 (70) | 120 (144); 70 (70) | 210 (252); 70 (70) |
B | 10 (12); 70 (70) | 18 (22); 70 (70) | 30 (36); 70 (70) | 48 (58); 70 (70) | 75 (90); 70 (70) | 120 (144); 70 (70) | 210 (252); 70 (70) |
C | 10 (12); 70 (90) | 18 (22); 70 (90) | 30 (36); 70 (90) | 48 (58); 70 (90) | 70 (84); 70 (90) | 120 (144); 70 (90) | 210 (252); 70 (90) |
D | 10 (12); 70 (90) | 18 (22); 70 (90) | 30 (36); 70 (90) | 48 (58); 70 (90) | 75 (90); 70 (90) | 120 (144); 70 (90) | 210 (252); 70 (90) |
First note that all proposals (A, B, C, and D) raise the healing amount of Bamboo Cane from 45 to 48, as per the section on the Bamboo Cane.
A raises the bonus healing amount for having organic toonup by a factor of 2,
from 10% extra (rounded up to the nearest integer, to a minimum of 1 extra) to
20% (ditto). This is mathematically convenient not just because it doubles the
benefit, but because division by 5 is still fairly easy (although not as easy
as by 10) in decimal, and also because this same mechanic of 20%-rounded-up is
also used for yellow damage (hpBonus
) against cogs, so players are used to
doing this kind of calculation already. Here we can see what impact this has on
per-target healing from multi-target gags:
Megaphone (2 targets) | Megaphone (3 targets) | Bamboo Cane (2 targets) | Bamboo Cane (3 targets) | Juggling Balls (2 targets) | Juggling Balls (3 targets) |
---|---|---|---|---|---|
[9, 9] ([11, 11]) | [6, 6, 6] ([8, 7, 7]) | [24, 24] ([29, 29]) | [16, 16, 16] ([20, 19, 19]) | [60, 60] ([72, 72]) | [40, 40, 40] ([48, 48, 48]) |
The intent of simply increasing the benefits of organic toonup is to make it more useful and appealing, rather than hoping that the same thing that works for other tracks (drop, squirt, throw, sound, and trap) will work for a track that doesn’t do any damage to cogs.
B is identical to A except that the base healing of Pixie Dust is raised from 70 to 75, for the reasons explained at the end of the section on the Bamboo Cane gag.
C is similar to A except that there is even further incentive for organic
toonup, by raising propAcc
by 20 across the board when the gags are organic.
The number 20 is chosen in particular for two reasons: one is that it puts
toonup gags at 97.5% accuracy when the track is maxed (97.5% being exactly
halfway between 95% and 100%), and the other reason is that 20 is conveniently
also the number used for a single stun’s contribution to gag accuracy.
The 97.5% figure is obtained because trackExp
for toonup gags is halved
compared to its calculation for all other gag tracks; with maxed toonup, that
is a trackExp
of 30. So we have an atkAcc
of
90 + 30 = 120 (because tgtDef
and bonus
don’t
apply to toonup). With the clamp_acc
function defined in the section on
clamping atkAcc
, we have:
if atkAcc <= 95:
return max(atkAcc, 0)
else:
return min(95 + (atkAcc - 95) / 10, 100)
if 120 <= 95:
return max(120, 0)
else:
return min(95 + (120 - 95) / 10, 100)
return min(95 + (120 - 95) / 10, 100)
return min(95 + 25 / 10, 100)
return min(95 + 2.5, 100)
return 97.5
And finally, D is to C as B is to A.
Choosing between these 4 options is a matter of how much organic toonup really needs to be buffed to be roughly as viable as any of the other 6 possible organic tracks, as well as a matter of how powerful one is willing to make toonup. It is clear that out of these 4 options, A is the weakest and D is the strongest. But hopefully, the case that I have made is convincing, and 75-laff Pixie Dust really does create better intra-track balance. If this is the case, then it seems that whether or not 75-laff Pixie Dust is the right choice is a matter of the sheer power of the toonup gag track overall.
Does trap need to undergo some kind of balancing of its own? Maybe. If sound is
nerfed (see the section on sound), then that might be enough to incidentally
make trap balanced; the fact remains that without any balancing efforts
(read: TTR), trap remains the least popular gag (at least amongst 6-trackers).
Notice that this doesn’t necessarily imply that trap is underpowered
(and thus unbalanced); perhaps people just routinely underestimate the utility
of a guaranteed stun for lure, and/or fail to imagine sufficiently clever ways
of using trap gags in combination with other gags to kill cogs. Also note that,
besides nerfing sound (and thus making lure more of an option… putting
aside lure nerfs) making trap more appealing, it also makes sense to consider
accuracy changes: see the section on atkAcc
clamping that explains how stuns
(and thus trap) would be more useful with a different method of clamping
atkAcc
.
Nevertheless, it’s at least worth exploring a few conservative balancing options.
Banana Peel | Rake | Marbles | Quicksand | Trapdoor | TNT | Railroad | |
---|---|---|---|---|---|---|---|
TTR | 12 (13) | 20 (22) | 35 (38) | 50 (55) | 70 (77) | 180 (198) | 195 (214) |
δTTR | +8 | +15 | +15 | +20 | +110 | +15 | |
A | 12 (13) | 20 (22) | 36 (39) | 60 (66) | 92 (101) | 156 (171) | 188 (206) |
δA | +8 | +16 | +24 | +32 | +64 | +32 | |
B | 12 (13) | 20 (22) | 36 (39) | 60 (66) | 92 (101) | 180 (198) | 196 (215) |
δB | +8 | +16 | +24 | +32 | +88 | +16 | |
C | 12 (13) | 20 (22) | 35 (38) | 56 (61) | 82 (90) | 172 (189) | 198 (217) |
δC | +8 | +15 | +21 | +26 | +90 | +26 | |
D | 19 (20) | 28 (30) | 39 (42) | 51 (56) | 82 (90) | 172 (189) | 183 (201) |
δD | +9 | +11 | +12 | +31 | +90 | +11 |
level 5 cog | level 6 cog | level 7 cog | level 8 cog | level 9 cog | level 10 cog | level 11 cog | level 12 cog | |
---|---|---|---|---|---|---|---|---|
TTR | 1 Quicksand | 1 Trapdoor |
|
1 TNT | 1 TNT | 1 TNT | 1 TNT | 1 Railroad |
" | " |
|
1 Trapdoor & 1 Seltzer | 1 Trapdoor & 1 Cream |
|
|
1 TNT & 1 Seltzer | |
A | 1 Quicksand | 1 Quicksand | 1 Trapdoor | 1 Trapdoor | 1 TNT | 1 TNT | 1 TNT | 1 Railroad |
" | " | " | " | 1 Trapdoor & 1 Seltzer | 1 Trapdoor & 1 Cream |
|
|
|
B | 1 Quicksand | 1 Quicksand | 1 Trapdoor | 1 Trapdoor | 1 TNT | 1 TNT | 1 TNT | 1 Railroad |
" | " | " | " | 1 Trapdoor & 1 Seltzer | 1 Trapdoor & 1 Cream |
|
1 TNT & 1 Seltzer | |
C | 1 Quicksand | 1 Quicksand | 1 Trapdoor |
|
1 TNT | 1 TNT | 1 TNT | 1 Railroad |
" | " | " |
|
1 Trapdoor & 1 Hose |
|
|
1 TNT & 1 Hose | |
D |
|
|
1 Trapdoor |
|
1 TNT | 1 TNT | 1 TNT | 1 Railroad |
" | " | " |
|
1 Trapdoor & 1 Hose |
|
|
1 TNT & 1 Seltzer |
A is a sort of “rationalization” of TTR, so that the deltas mostly go up by 8 each time, and otherwise are powers of 2. This has the effect of making Marbles, Quicksand, and especially Trapdoor considerably more powerful, but TNT considerably weaker (although still just enough to one-shot a level 11 v1.0 cog). Compared to TTR this actually seems to lower the use for organic trap by a bit, although having a 171-damage TNT instead of 156 actually is quite a bit better than going from 180 to 198 damage. Still, having to use a Big Weight on a level 12 cog that is being TNT’d is not really so bad considering that the Big Weight has two stuns counting towards its accuracy, and it’s only a level 4 gag.
B is just A but adjusted to have the same TNT damage as TTR. This may look fine for balancing trap but doesn’t look quite so good for balancing organic trap.
C is similar to A in that it makes Quicksand and Trapdoor considerably stronger, but nerfs TNT a bit, and also in that it is a sort of “rationalization” of TTR: the deltas make use of the sequence [(0, )8, 15, 21, 26], which follows the recurrence relation an = 2an−1 − an−2 − 1. This is, in some sense, related to the method used in the sound section, where the deltas have a slope of +1 (thus creating a triangular sequence), rather than the deltas-of-the-deltas having a slope of −1 (thus creating a tetrahedral sequence, but negative) as they do here.
D is similar to C but without the “rationalization”; the only important changes are to the damage of level ≤4 trap gags. Rather than choosing nice damage numbers or deltas, the damage figures are increased just enough to make it so that organic versions of the gags are able to one-shot a cog that is one higher level. In my opinion, if trap were to be balanced, this is probably the way to do it. The numbers are not as nice, but the effect is to make both inorganic and organic trap more useful at all stages of the game. In any case, the deltas still have the interesting(?) property that they alternate between being prime and composite, and this pattern is maintained even if you introduce a fictional “level 0” trap gag that does 0 damage. The damage for Railroad is chosen by selecting a delta value that has already been used (viz. +11).
Format:
rounds; propAcc
(organic propAcc
)
$1 Bill | Small Magnet | $5 Bill | Big Magnet | $10 Bill | Hypno | Presentation | |
---|---|---|---|---|---|---|---|
TTR | 2; 50 (60) | 2; 50 (60) | 3; 60 (70) | 3; 60 (70) | 4; 70 (80) | 4; 70 (80) | 15; 90 (100) |
A | 1; 50 (60) | 1; 50 (60) | 1; 60 (70) | 1; 60 (70) | 1; 70 (80) | 1; 70 (80) | 15; 90 (100) |
B | 1; 50 (70) | 1; 50 (70) | 1; 60 (80) | 1; 60 (80) | 1; 70 (90) | 1; 70 (90) | 15; 90 (110) |
C | 2; 50 (60) | 1; 50 (60) | 2; 60 (70) | 1; 60 (70) | 2; 70 (80) | 1; 70 (80) | 15; 90 (100) |
BC | 2; 50 (70) | 1; 50 (70) | 2; 60 (80) | 1; 60 (80) | 2; 70 (90) | 1; 70 (90) | 15; 90 (110) |
D | 2; 40 (70) | 1; 40 (70) | 2; 50 (80) | 1; 50 (80) | 2; 60 (90) | 1; 60 (90) | 15; 90 (120) |
Excluding TTR, BC is the strongest here, and A is the weakest. It’s not immediately clear how B, C, and D stack up in terms of strength because they make some orthogonal changes.
The purpose of A is clear: the accuracy of lure tends to not really be a problem, so we might as well be conservative there and instead drop the number of rounds for all lure gags (except level 7) to 1. This makes A the sort of “minimal change” needed to “balance” lure, because it means that lure is still quite useful for one round of damage mitigation and effectively two rounds of extra cog-fighting power (although you can only use trap gags in the first such round, and the first round only has up to a maximum of 3 attackers), but not so powerful that you can effectively mitigate all damage (assuming the best case of lure never missing).
One problem with A might be that since it doesn’t touch propAcc
values,
organic lure still kind of sucks. With the accuracy changes mentioned in the
accuracy section it’s not so bad (compared to in TTR), but still probably
not worth it. B is a stronger version of A that is only stronger if you have
organic lure; propAcc
gets increased by 20 instead of the usual 10, making
organic lure effectively worth an entire stun rather than half of a stun.
Another problem with A is that it makes single-target lure look even less appealing than in TTR. Single-target lures are already not used nearly as often as their multi-target counterparts in TTR, but with only 1 round of effectiveness for multi- and single-target lures, you can’t e.g. “lure left kill right” one-by-one and use single-target lures to get the left-side cogs, &c. To remedy this, C doubles the number of rounds of effectiveness for single-target lures only.
BC combines B and C and thus is more powerful than either, and also attempts to fix both problems with A.
In case BC was too powerful and/or the problem of organic lure wasn’t
solved enough, D is a weakened version of BC that makes organic lure
effectively worth a stun and a half, but lowers inorganic lure propAcc
by 10
at all levels (except 7). This may seem harsh (and indeed, it is perhaps
unfortunate that making organic lure last for more rounds isn’t really
viable and so D relies on making inorganic lure more fickle), but all is not
lost for inorganic lure users so long as their friends stun dilligently enough
(one half of an extra stun on average to reach TTR levels of accuracy) to reach
an acceptable rate of success.
In my opinion, as far as balancing “numerically” goes, either BC or D is probably the best bet.
Possibly, in addition to balancing numerically, it may be feasible to alter the actual mechanics of lure gags somewhat. In particular, the idea is to permanently mark any cog that is successfully lured, and make it so that cogs that are so marked cannot be lured. Marked (read: successfully-lured-at-some-point) cogs cannot be lured in much the same way that cogs currently in a lured state cannot be lured in TTR.
The intent here is to prevent low lure durations from encouraging a continuous stream of luring, killing, luring, cogs getting automatically unlured, re-luring, & so on & so forth. Rather than effectively giving players free (except for taking a hit during the initial unlure) re-lures in many cases because they would be luring anyways to lure fresh new cogs, this ideally encourages actual lureless play alongside lure-based play. In addition, there’s the incidental consequence of making single-target lure more useful since it’s generally less likely that there’s more than one cog that can be lured at any given time.
This change could be done in tandem with not just a numeric balance of lure (like BC), but also possibly a change that causes cogs not to immediately attack when they are automatically unlured. Also, making this mechanical change could justify using one of the numerical balancing schemes but with an extra round of duration added to all lure gags.
- Actually helps to balance lure-based and lureless play and allow them to coexist.
- Allows other changes to make more sense, like: numeric balancing, not-quite-as-nerfed lure gag durations, and cogs not immediately attacking when automatically unlured.
- Makes single-target lure gags more useful, relatively speaking.
- Requires an actual mechanical change, not just a numeric one.
- Not actually useful if you don’t care about making lure-based and lureless play balanced and coexistent.
Bikehorn | Whistle | Bugle | Aoogah | Trunk | Fog | Opera | |
---|---|---|---|---|---|---|---|
TTR | 4 (5) | 7 (8) | 11 (12) | 16 (17) | 21 (23) | 50 (55) | 90 (99) |
δTTR | +3 | +4 | +5 | +5 | +29 | +40 | |
A | 4 (5) | 7 (8) | 11 (12) | 16 (17) | 22 (24) | 29 (31) | 40 (44) |
δA | +3 | +4 | +5 | +6 | +7 | +11 | |
B | 4 (5) | 7 (8) | 11 (12) | 16 (17) | 22 (24) | 30 (33) | 50 (55) |
δB | +3 | +4 | +5 | +6 | +8 | +20 |
level 8 cog | level 9 cog | level 10 cog | level 11 cog | level 12 cog | |
---|---|---|---|---|---|
TTR |
|
|
1 Fog & 3 Trunk | 2 Fog & 2 Aoogah | 3 Fog & 1 Aoogah |
A | 2 Trunk & 2 Aoogah |
|
|
|
2 Opera & 2 Opera |
B | " | " |
|
|
|
Note that the primary difference (assuming 4 toons with max sound) between A and B is that in cases where no toons have organic sound, you need one fewer Fog to defeat a row of level 10 cogs with B. Assuming no restocks (and 3 Fogs per toon), you have enough Fogs to do this 3 times (4 with ≥1 organic sound) with A and 4 times (6 with 4 organic sound) with B. Note that this also makes organic sound actually quite a bit more useful with A, since you only need 1 organic sound to make a difference.
Taking out a row of level 10s is pretty powerful, and probably still viable if you have, say, a row of two ≤10s and two 11s, or three ≤10s and one level ≥11… One of the reasons that it’s best to err on the side of less power is that if sound is nerfed (relative to TTR), having even one soundless (or simply unmaxed sound) toon dampens the possibilities considerably, in contrast to TTR, where sound is so powerful that only 3 toons with sound gets you quite far in terms of sound spam. Removing a bit more power obviously has the potential to make this worse, but mostly just makes it more clear that sound spam is typically not an option (at least, not in the way that a TTR player is accustomed). Ideally, all possible 6-track builds should be roughly equally viable (toonupless, trapless, lureless, soundless, throwless, squirtless, dropless). In this case, we expect that — again, roughly — 1 out of every 7 “normal” (read: 6-track) toons is soundless.
Assuming that all toons have their gag tracks maxed, luredRatio
only makes a
difference in TTR in one situation: using throw gag(s) on an unlured level
≥10 cog, with no stun, while at least one other cog is lured. This is
unfortunate, because:
- This is a very niche use-case that doesn’t need to be serviced,
especially considering that even a single stun will negate the usefulness of
luredRatio
. - Having a mechanic like
luredRatio
unnecessarily complicates accuracy calculation.
The proposal is to get rid of luredRatio
wholesale, and collapse bonus
and
prevHits
so that they are the same by definition.
- Simplifies accuracy calculation.
None.
In TTR, we have:
atkAcc = propAcc + trackExp + tgtDef + bonus
Where propAcc
is fixed for each particular gag, trackExp
depends on the
level of the highest gag that the gag user has access to within the relevant
track, tgtDef
depends on the cog’s level, and bonus
is mostly just
stun (and luredRatio
, but luredRatio
is to be gotten rid of, and only ever
applied in very niche situations anyways).
Again, also in TTR, atkAcc
is actually clamped to a maximum of 95 and a
minimum of 0 (implementation-wise there is no minimum, but it’s easier to
think of it as having a minimum of 0 so that you can think of atkAcc
as just
being the percentage chance; there’s no material difference either way).
Below are (Python) functions ttr_clamp_acc
and clamp_acc
[1],
which correspond to TTR’s method of clamping atkAcc
after it is
calculated as the sum of the four aforementioned numbers, and the proposed
method of doing the same thing, respectively.
def ttr_clamp_acc(atkAcc):
return max(min(atkAcc, 95), 0)
def clamp_acc(atkAcc):
if atkAcc <= 95:
return max(atkAcc, 0)
else:
return min(95 + (atkAcc - 95) / 10, 100)
The tweakable parameter of clamp_acc
is the 10
that shows up in its
definition. 10
is a convenient number because it gives what I think is a
reasonable slope while also being very easy to calculate in your head (assuming
you’re working in base 10, you just move the decimal point one place to
the left). The magic number here is thus 145, because it is exactly when
atkAcc >= 145
that the gag is guaranteed to hit.
- Makes accuracy more meaningful in general, because its value changes more freely to reflect the strategy being used.
- Particularly, makes stunning more meaningful because it creates a nonzero benefit in many more cases.
- Makes the game less reliant on dumb luck by allowing accuracy to be pushed beyond the 95% mark (with the right gag picks), even up to 100%.
- Complicates accuracy calculation.
- This change would raise the overall power level of toons, since in some cases it improves accuracy, and in all other cases it keeps accuracy the same. This could be a con.
- Allowing accuracy to possibly reach 100% could be seen as overly rigid or deterministic(‽)
footnotes for “atkAcc
clamping”
- Python doesn’t have types, but it does have data types, and these two
functions have different return (data) types, namely
ttr_clamp_acc: int -> int
andclamp_acc: int -> (int | float)
. But really, it doesn’t matter… just be sure not to somehow trip on that.
This section deals with “facilities”, which is a term here construed broadly to refer to factories (and the corresponding factory-like facilities in the other three cog HQs) and bosses (all four in all four cog HQs). Note that this does not, however, include cog buildings.
In TTR, the barriers to entry for the various facilities are as follows:
facility | requirement(s) |
---|---|
factory, front | |
factory, side | 31 laff |
coin mint | |
dollar mint | 66 laff |
bullion mint | 71 laff |
DA office A | |
DA office B | 81 laff |
DA office C | 86 laff |
DA office D | 96 laff |
front 3 | |
middle 6 | 95 laff |
back 9 | 100 laff |
VP | 10 factories completed |
CFO | cashbot suit tasks (76 laff and access to 6 gag tracks) |
CJ | lawbot suit tasks (100 laff and access to 6 gag tracks) |
CEO | bossbot suit tasks (100 laff and access to 6 gag tracks) |
It is nice that some facilities that are considered to be designed for toons that are at least middling in terms of power are nevertheless available to anyone: coin mints, DA offices A, front 3s, and VPs. However, the requirements for facilities that have them range from “unfortunate, at least for certain styles of play” to “completely unnecessary”. There are basically three groups of requirements that we want to talk about here, and so we will tackle them one-by-one in descending order of silliness.
Really, no amount of argumentation should be required to establish that the side entrance of the sellbot factory should have no barriers to entry. The front entrance already has no such barriers, and they are both just different entrances to the same facility anyways. Presumably, the reasoning for giving the side entrance a laff limit is that if you go left after completing the first cog battle, you have to go through stompers and then complete another cog battle; if you keep going, you have to do the most dangerous walk across the catwalk area of the factory before finally getting to the part of the factory that every factory run has to do anyways. However, there are a few problems with this reasoning for instituting a laff limit:
- Most front entrance runs are long runs anyways, so they do all of this stuff as well (but with extra cog fights and catwalking).
- Just because you enter via the side entrance does not mean that you have to go left after the first cog fight; you can just go straight.
- Obstacles like stompers and goons can indeed shave off significant amounts of laff if you aren’t at least somewhat careful, but typically the real danger is in the last 4 cog fights (a.k.a. the only part of the factory that literally every factory run must complete anyways).
It’s clear that the way to go is lowering the laff limit on the factory side entrance to 1, to match the front entrance.
Some other factory-like facilities besides factories themselves have laff requirements (viz. dollar & bullion mints, DA offices B, C, & D, middle 6s, and back 9s). Is this warranted? At first it seems like it might indeed be warranted, considering that all of these factory-like facilities are significantly more difficult than factories. In addition, it seems unlikely that a toon that doesn’t meet the laff requirements for one of these facilities will actually have a reason to do the facility. For the most part, these facilities are frequented for the purpose of getting ready for cog suit promotions; but anyone who is doing this must have the corresponding suit and thus enough laff to do the corresponding facilities.
However, I want to argue that these laff limitations should be removed entirely (or weakened, at the very least), for a number of reasons:
- Shortened versions of all of these factory-like facilities are available to anyone, with no limitations. Coin mints, DA offices A, and front 3s all lack requirements, but they present identical challenges to their laff-gated counterparts! Therefore, for the sake of consistency, the requirements should be lifted (or at least, can reasonably be lifted). It could be argued that said laff-gated facilities do in fact present more challenge (enough to warrant locking them off to those with insufficient laff) because of their length. This is a kind of difficulty by attrition, where the danger is that you run out of (sufficiently powerful) gags too quickly to complete the whole facility. While this is indeed more of a problem for lower-laff toons (because they generally have fewer gag tracks and smaller gag pouches), in general this is not a good enough excuse. Imposing laff requirements means disallowing a party that has just one member who does not meet the requirement, so even a party with (to give the most extreme example) 3 maxed toons and 1 toon who just barely does not meet the requirement is barred, despite easily being able to handle the attrition. In addition, the facilities themselves provide ways of restocking gags, and players can even restock their own gags (and laff) through the use of SOS cards and unites.
- It is common for toontasks to require things that can be easily (sometimes, most easily) be found within these factory-like facilities. Of course, a toon who doesn’t meet the requirements can instead do the corresponding facilities that don’t have such requirements; however, because these facilities are meant for teamwork (generally being done in parties of 4 toons), being more flexible in which ones you can do is a big help. Being able to do a DA office D is a blessing when you need a DA office, and there are already 3 people waiting for a 4th toon to join them so that they can do a DA office D.
- Completely preventing players from being able to do something just because you have a vague idea that they probably aren’t capable of it is often poor design. It is needlessly frustrating for players who intentionally play in an unexpected way, or intentionally seek extra challenge, or just feel a bit more “railroaded” due to the game deciding for them that they “aren’t ready yet”. The possible punishment in Toontown is never very great (going sad is not exactly the end of the world), so allowing players to decide for themselves should be unproblematic.
- Removing these laff requirements is even more important when you consider implementing changes to the max-laff mechanic in general. Particularly, laff caps that allow players to limit how high their toon’s laff will ever be able to reach, and changes to max-laff acquisition (viz. toontasks & minigames that award max-laff), are both the kinds of changes that demand at least the reconsideration of laff requirements. In many of these cases, max-laff levels are being lowered overall, and so laff requirements must be removed (or at least lowered) in tandem.
- If changes to requirements for doing bosses (see the next section) are considered, then removing these laff requirements for factory-like facilities may be necessary anyways.
After considering the requirements for entry into the side entrance of factories, and for entry into other factory-like facilities, we are now brought to the same issue w.r.t. boss fights (excluding VP, because it effectively has no requirements due to only requiring 10 factories). It may seem particularly obvious that there are no significant changes to be made to the requirements for entry into non-VP bosses, considering that all three require (and only require) some part of the main taskline to be completed. However, there are, per se, ways of loosening these requirements without making changes to the main taskline — and there are reasons to do so, as well.
What would be the motivation for loosening the requirements for entry into non-VP bosses? There are basically two motivations: first & foremost is the desire to open up non-VP bosses to challenge runs. Basically, we are talking about uber toons here. In TTR, ubers are limited to maxing out their sellbot suit by defeating the VP many times (and doing many factories). Allowing them other activities (viz. 3 other possible bosses to run) to do would be a huge gameplay improvement for uber/challenge-run styles of play. The second (lesser) motivation is that as TTR’s “Crash Sellbot HQ” and “Crash Cashbot HQ” events have shown, opening up bosses to more toons is a good bit of fun. Of course, this is not to actually propose what these two events do in particular, which involves nerfing said bosses to make them more accessible to toons of all power levels.
There are two (not mutually exclusive) ways that the loosening of non-VP boss entry requirements can be achieved without any changes to the main taskline:
- Allow a similar mechanic to “Crash Sellbot/Cashbot HQ”, with so-called “rental suits” that allow entry into bosses without having the proper suit. This would be limited to just CFO, CJ, and CEO. Rental suits are permanently level 1; that is, they cannot level up and are always “ready for promotion”. The use of rental suits should probably have requirements; requiring the completion of some factory-like facilities (e.g. mirroring Sellbot HQ by requiring the completion of at least 10 factory-like facilities within the corresponding cog HQ) and/or requiring at least 1 maxed gag track should suffice (I would favor having both requirements).
- Make the tasklines that grant cog HQ suits available to any toon that meets one or both of the following requirements: (2a) the toon has completed the normal pre-requisite tasks (as in TTR), (2b) the toon has a level 50 Mr. Hollywood suit.
While these two options are not mutually exclusive, it probably suffices to just pick one. The second option seems more appealing, because it offers more of a sense of progression. The sense of progression is both because you have to progress all the way through Sellbot HQ before getting to the other HQs, and because it allows leveling up the suits that you get for the other HQs, unlike rental suits. However, this is a higher barrier to entry than the first option (which may be a bad thing), and means more laff for ubers due to the other 3 cog HQ suits also granting 5 laff each (this is not so much a problem with laff capping). The second option also probably means that factory-like facilities will be more popular (even if defeating some of them is required to get a rental suit) because bossing on low-laff toons will still mean having to earn promotions.
This section discusses cog behavior (read: attacks) and maximum laff for toons. These topics are inherently related for the obvious reason that higher maximum laff is primarily useful for being able to handle more punishment from cogs before going sad. However, make no mistake: these topics are also intimately tied to the ability of toons to defeat & otherwise control cogs, that is, gag balance. These topics are difficult because they are so entangled in the overall power level of toons in the game, including toons doing challenge runs like e.g. uber toons and/or semi toons.
That being said, even if these topics are difficult to treat due to being entangled in… well, everything else, there is one thing that we can clearly state: toons have an excess of laff points in vanilla TTO/TTR. In many ways this is just saying that the balance between toons and cogs is skewed heavily in favor of toons, hence the existence of challenge run styles of play. In addition, it is also clear that multi-target attacks that cogs can perform are disproportionately powerful when compared to their single-target counterparts. This makes cogs that have no multi-target attacks (e.g. any bossbots of higher rank than Yesman) disproportionately weak, and also makes cogs that do have multi-target attacks fickle in terms of damage output because it is a matter of chance whether or not they choose their attacks to be multi-target ones.
The concern about multi-target cog attacks is not necessarily a sign of an actual problem; it might be perfectly okay for some cogs to be limited to single-target attacks, and for it to be a rather significant matter of chance whether or not certain cogs choose to use their multi-target attack(s). But it does give something to consider when trying to balance cogs’ attacks.
There are three main ways of limiting laff that we want to consider here: laff capping, nerfing toontasks (and other max-laff-awarding activities), and lowering starting laff.
Laff capping means allowing players to choose a laff limit for their toon at the point of toon creation. A laff-capped toon starts with the same max-laff as a normal toon (15, unless modifications are made), but any max-laff increases that they receive can raise their max-laff to, at most, their laff cap. This is a pretty mechanically straightforward change and has already been implemented at least once before. Laff capping is an appealing proposition, but its usefulness is essentially limited to the creation of uber toons; it does only a little to help balance the game generally, because it is entirely voluntary.
- It’s voluntary, which has the benefit of encouraging a wider range of playstyles and allowing players to choose for themselves.
- It’s flexible enough to limit max-laff to any arbitrary degree.
- It’s voluntary, which has the drawback of doing only a little to help balance the game generally, and
- Being voluntary also tends to cause greater power disparities between toons that are trying to cooperate.
It is possible to limit maximum laff in general (and thus take a step towards balancing the game more generally) by nerfing the max-laff rewards given by toontasks and other relevant activites. The following list chronicles the progression of the mainline toontasks that award max-laff, including the max-laff that a toon has after completing each particular neighborhood/playground (assuming all max-laff boosts are from toontasks, and not other sources like e.g. fishing):
max-laff-awarding toontasks
- Toontown Central (25 laff)
- +1 laff (Professor Pete)
- +2 laff (Honey Haha)
- +2 laff (Professor Wiggle)
- +3 laff (Sticky Lou)
- +1 laff (generic)
- +1 laff (generic)
- Donald’s Dock (34 laff)
- +2 laff (Billy Budd)
- +1 laff (generic)
- +1 laff (generic)
- +2 laff (generic)
- +3 laff (generic)
- Daisy Gardens (43 laff)
- +1 laff (Sofie Squirt/Artie/Inkeeper Janet/Barber Bjorn)
- +1 laff (Sofie Squirt/Artie/Inkeeper Janet/Barber Bjorn)
- +2 laff (Postman Felipe)
- +2 laff (sprocket recovery/generic)
- +3 laff (Uncle Mud/Uncle Spud/Aunt Hill)
- Minnie’s Melodyland (52 laff)
- +1 laff (generic)
- +1 laff (generic)
- +2 laff (generic)
- +2 laff (generic)
- +3 laff (generic)
- The Brrrgh (61 laff)
- +1 laff (Chicken Boy/Wynne Chill/Eddie the Yeti)
- +1 laff (Chicken Boy/Wynne Chill/Eddie the Yeti)
- +2 laff (Sweaty Pete)
- +3 laff (Lounge Lassard)
- +2 laff (generic)
- Donald’s Dreamland I (71 laff)
- +1 laff (Powers Erge)
- +2 laff (Lawful Linda)
- +3 laff (Rocco)
- +4 laff (Zari)
- Donald’s Dreamland II (81 laff)
- +1 laff (generic)
- +2 laff (generic)
- +3 laff (generic)
- +4 laff (generic)
- Donald’s Dreamland III (91 laff)
- +1 laff (generic)
- +2 laff (generic)
- +3 laff (generic)
- +4 laff (generic)
- Donald’s Dreamland IV (100 laff)
- +1 laff (generic)
- +2 laff (generic)
- +3 laff (generic)
- +3 laff (generic)
An easy way to lower the overall amount of max-laff awarded by tasks is by lowering only the rewards of tasks that award ≥2 max-laff. This way, the entire structure of the taskline is totally intact. Here is what this approach looks like if taken to the extreme (all max-laff-awarding tasks giving exactly +1 max-laff):
all max-laff-awarding toontasks awarding +1
- Toontown Central (21 laff)
- +1 laff (Professor Pete)
- +1 laff (Honey Haha)
- +1 laff (Professor Wiggle)
- +1 laff (Sticky Lou)
- +1 laff (generic)
- +1 laff (generic)
- Donald’s Dock (26 laff)
- +1 laff (Billy Budd)
- +1 laff (generic)
- +1 laff (generic)
- +1 laff (generic)
- +1 laff (generic)
- Daisy Gardens (31 laff)
- +1 laff (Sofie Squirt/Artie/Inkeeper Janet/Barber Bjorn)
- +1 laff (Sofie Squirt/Artie/Inkeeper Janet/Barber Bjorn)
- +1 laff (Postman Felipe)
- +1 laff (sprocket recovery/generic)
- +1 laff (Uncle Mud/Uncle Spud/Aunt Hill)
- Minnie’s Melodyland (36 laff)
- +1 laff (generic)
- +1 laff (generic)
- +1 laff (generic)
- +1 laff (generic)
- +1 laff (generic)
- The Brrrgh (41 laff)
- +1 laff (Chicken Boy/Wynne Chill/Eddie the Yeti)
- +1 laff (Chicken Boy/Wynne Chill/Eddie the Yeti)
- +1 laff (Sweaty Pete)
- +1 laff (Lounge Lassard)
- +1 laff (generic)
- Donald’s Dreamland I (45 laff)
- +1 laff (Powers Erge)
- +1 laff (Lawful Linda)
- +1 laff (Rocco)
- +1 laff (Zari)
- Donald’s Dreamland II (49 laff)
- +1 laff (generic)
- +1 laff (generic)
- +1 laff (generic)
- +1 laff (generic)
- Donald’s Dreamland III (53 laff)
- +1 laff (generic)
- +1 laff (generic)
- +1 laff (generic)
- +1 laff (generic)
- Donald’s Dreamland IV (57 laff)
- +1 laff (generic)
- +1 laff (generic)
- +1 laff (generic)
- +1 laff (generic)
In combination with other propositions, e.g. the nerfing of gag tracks like sound & lure, the buffing of cog attacks, &c., this particular approach is probably too steep of a nerf; starting The Brrrgh with 36 laff is tough. Also, without changes to other sources of max-laff increases, the amount of max-laff available from toontasks alone is worryingly similar to the amount of laff available from other sources: 42 from tasks, and 37 from elsewhere (plus 15 starting laff, for an absolute maximum laff of 94).
The good news is that exactly how much laff you distribute via toontasks, and exactly when in the taskline you distribute it, is very tweakable: setting all max-laff-awarding tasks to award +1 laff works as one extreme, and some middle ground between that and something much more high-powered like vanilla TTO/TTR can easily be achieved by tweaking the rewards within any of the various neighborhoods/playgrounds.
Ultimately, to what degree it makes sense to cripple (or perhaps boost?) max-laff accumulation is wildly dependent on what other changes are made (or not made) to the game.
The quantity of max-laff increases available from non-task sources can be tweaked to better match the max-laff increases available from tasks, which brings us to the topic of how to nerf these non-task sources if necessary.
When it comes to non-task sources of max-laff increases, all of them are given +1 at a time, so the technique used for toontasks doesn’t apply. Instead, we have two (not mutually exclusive) options: some (or all) of the increases granted by a given activity have to be removed outright, and/or an overall cap can be placed on the total number of max-laff points that can be obtained from non-task sources. If we break down the non-task sources of max-laff increases, we see that most of them are from leveling up cog HQ suits:
- +7 (fishing)
- +5 (Sellbot HQ)
- +5 (Cashbot HQ)
- +5 (Lawbot HQ)
- +5 (Bossbot HQ)
- +4 (gardening)
- +3 (golfing)
- +3 (racing)
Besides the obvious options that we have for each non-task source of max-laff increases — viz. leaving it alone, or removing all of its max-laff increases — we’ll quickly consider ways of partly removing the max-laff increases from each source listed above. Note that one invariant that we maintain is always giving one of the max-laff increases as an award for totally maxing out the activity.
- Fishing gives an odd number of max-laff increases (7), so it may make sense to only award a max-laff increase every other time, starting with the first. This would mean getting an increase for catching a total of 10, 30, 50, & 70 species, for a total of +4 max-laff from maxing fishing. Another possibility is just removing one or more of the earliest max-laff rewards, e.g. removing two of them to end up with awards at 30, 40, 50, 60, & 70 species caught, for a total of +5 max-laff from maxing fishing. Yet another possibility comes from 70 being divisible by 14: awarding a max-laff increase at each multiple of 14 species caught. Awards at 14, 28, 42, 56, & 70 would place the total max-laff awarded for maxing fishing at +5.
- Lowering the number of max-laff increases awarded by cog HQs is more difficult, mostly because it is engrained in the way that cog suit promotions work: at the highest-rank suit (e.g. Mr. Hollywood for Sellbot HQ), the number of merits (using the term “merit” generically) needed to achieve each successive promotion follows a vaguely sawtooth pattern, with an increase in max-laff accompanying each fall. The purpose of having this kind of payoff is obvious, and it only makes sense to replace any removed max-laff increase with some other suitable reward. Short of changing the merits-for-promotion progression itself, removing some max-laff increases from cog HQs will require more creativity than is warranted in this document; you might try, for example, awarding things like a single extra tree spot at the toon’s estate.
- There are basically 4 viable options for gardening: max-laff increases at 20, 30, & 40 species, or at 20 & 40 species, or at 30 & 40 species, or only at 40 species.
- Golfing and racing are similar to gardening, but with just 3 trophies to choose from. Generally, you can grant max-laff increases for the 1st & 3rd, or the 2nd & 3rd, or the 3rd only.
Note that in the case that you use one or more of these methods, it may make sense to replace the now-missing max-laff increases with something else to fill in the missing space; even just a cosmetic reward.
Capping the amount of max-laff that can be obtained from non-task sources is fairly straightforward, except that you have to choose a particular number to cap at. The considerations for choosing such a number (if any) must heed the following considerations, in descending order of importance:
- Balance against the rest of the game, including the amount of max-laff available from toontasks, and general toon-cog balancing considerations.
- Not capping so low as to make too many sources of laff useless, but also not so high as to make the viability of toons in combat very reliant on doing unrelated minigames like racing.
- Aesthetic concerns about the regularity of numbers used (including the capping number itself, as well as the highest max-laff achievable in the game).
Here’s a summary of the pros & cons of stripping (some) max-laff awards from non-task sources, as well as the pros & cons of putting an overall cap on the number of max-laff available from such sources (note that these two things are not mutually exclusive):
- On its own, this technique lowers the max-laff-granting power of non-task activities while keeping some incentive for each toon to do each such activity.
- On its own, this technique is not very invasive; it doesn’t require a new mechanic to be added. At its core, all that really needs to be done is changing some max-laff awards from +1 to +0, or something very similar.
- Using this technique can allow the balancing of these various activities. For example, in vanilla TTO/TTR, fishing alone grants more max-laff than racing & golfing combined.
- In some cases, now-missing max-laff rewards will create a void that will either have to be left open, or filled with some other sort of reward.
- In some cases, this may skew non-task activities towards power players that are willing to invest the time into maxing these activities, because only investing some time into a given activity won’t give much (if any) up-front rewards.
- Skews away from excessively rewarding power players by allowing players to more easily get to the laff cap without having to completely max out every possible source of max-laff. It should be noted that this is much more important than in TTR, because in TTR getting from 100 max-laff (no non-task activities completed) to 137 max-laff (all non-task activities completed; maximum possible laff) is not much of an advantage gameplay-wise. 100 laff is more than enough to do even the most challenging things in TTR.
- Allows putting a hard limit on attainable max-laff without tampering with the various ways of getting it.
- There is some pressure to add extra rewards (cosmetic or otherwise) to fill in the “gaps” that are here created by refusing to award max-laff to players that have already hit the cap, even though they continue to achieve further progress in non-task activities. This pressure is eased by “completionist” urges, which urge players to complete things even if there is no reward per se for doing so. The pressure is eased as well by the activities themselves being high quality, and thus motivating players to do the activities for the sake of the activities themselves being enjoyable.
- Related to the previous point: placing the cap too low could make excessively many sources of laff useless (or at least, less useful). A player that has already reached max-laff now has less incentive to do non-task activities that they haven’t maxed out; this is trivially not the case if the cap is removed, since anyone whose laff is maxed out has already maxed out all activities anyways.
- It’s somewhat invasive; a new mechanic needs to be introduced that caps the overall max-laff gain attainable from non-task sources.
When it comes to tweaking cogs’ attacks, there are two main concerns that we want to address here: disparity between single-target and multi-target attacks, and the overall power level of cogs.
It’s well understood in TTR that multi-target cog attacks are more formidable than their single-target counterparts. The reason is that multi-target attacks tend to do the same per-toon damage as single-target ones; furthermore, multi-target attacks are never “duds”. By “dud”, we mean attacks like the Glad Hander’s Fountain Pen or the Yesman’s Razzle Dazzle. With a multi-target attack’s damage effectively being multiplied by the number of toons that it hits, the total damage output is much more impressive.
Of course, this impressive damage being spread evenly between two or more toons does dilute its impact to some extent. But the amount of resources (in terms of both gag expenditure, and the number of gags in each round that have to be dedicated just to toonup rather than other gags) required to patch up the damage is definitely much higher. This imbalance may be acceptable to some extent; it all depends on how neck-breakingly sharp you want the variation to be between cog attacks, including the variation between the attacks of a given cog species, as well as the variation between the arsenal of different cog species. It also makes sense to want some single-target attacks to really “sting”; it really stings when a level 11 Mingler uses Paradigm Shift and (assuming no misses; the accuracy is 90%) dishes out a whopping 24 × 4 = 96 damage! There are a number of approaches that can be taken here. So here are a few of them (note that none of them are mutually exclusive with any of the others), each accompanied by at least some discussion/deliberation:
The most obvious choice is to directly nerf multi-target attacks by tweaking the numbers. This further splits our options into nerfing damage and/or nerfing accuracy.
Nerfing accuracy might sound like a somewhat poor idea given that it introduces more of the “dumb luck” and “fickleness” that is lamented elsewhere in this document. However, it actually turns out to not be that bad of an idea, for at least two reasons. First is that, unlike making toons’ gags more fickle, making cog attacks more fickle doesn’t have the same negative impact on gameplay. When adding dumb luck to player-controlled elements of gameplay, it can quickly become frustrating, because players feel (often justifiably) that their planning & actions have been in vain just because of some dumb luck that seems to have no purpose, no way of being predicted, and no way of being controlled. But when adding luck reliance to NPCs (adversaries, in this case), the effect is much more mild. The player already sees the NPCs as being somewhat opaque, so leaving their decisions — and the effects that they have on the game world — totally or partly up to chance doesn’t necessarily make the opacity worse or make anything seem fickle. Indeed, the methodic use of chance is often to great effect whenever it makes NPCs seem more organic (although that particular use is not directly relevant here).
The other reason that adding dumb luck to cog attacks is less of a problem here is that we are talking specifically about multi-target attacks, which perform independent rolls for each toon in the battle. This causes the number of toons that are hit (or, conversely, the number of toons that are missed) to be selected from a binomial distribution B(n, p), where n is the number of toons in the battle and p is the probability of the attack hitting a given toon. Because of the central limit theorem, this looks very similar to a normal distribution; particularly, with the binomial distribution it looks quite normal even with an n as low as n = 4. The variance of said distribution is np(1 − p), so with 4 toons in the case of maximum variance[1]:
np(1 − p)
= 4 × 0.5(1 − 0.5)
= 2 × 0.5
= 1
So the expected number of toons struck is obviously 2, but with a variance (and therefore also standard deviation, in this case) of ±1. This interval, [1, 3], contains 87.5% of all result values.[2] And this is only the “worst” case; higher or lower values of p will yield smaller variances/standard deviations.
As a result of this central tendency, high levels of “dumb luck” in cogs’ multi-target attacks actually doesn’t make cogs terribly fickle in terms of raw damage output. That being said, there is, of course, more significant variation to be expected when it comes to exactly which toons are struck, as p decreases.
When deciding to nerf multi-target attacks directly, the following pros and cons of nerfing the accuracy (by itself) should be taken into consideration, although do note that nerfing damage and nerfing accuracy are not mutually exclusive:
Pros of nerfing accuracy in particular:
- Less revisionist in some sense, because you can nerf accuracy without actually changing the observed effects of toons being struck by said attacks.
- Makes multi-target attacks less disruptive without actually changing the particular number of laff points required to survive a particular barrage of cog attack(s). This makes cog attacks basically just as deadly overall, while being significantly less of a drain on toon resources.
Cons of nerfing accuracy in particular:
- Less deterministic. This is mostly a problem in dire situations, where the lowered likelihood of a cog making you go sad might be nice — but hardly a guarantee or reassurance. However, it should be noted that although multi-target attacks are generally more powerful than the corresponding multi-target ones in TTR, the per-toon damage is typically roughly the same, so this point applies to essentially any cog attack with an accuracy that is “significantly” less than 100%.
- Because only nerfing accuracy keeps cog attacks at basically the same level of deadliness, this form of nerf is not as effective when reduced access to laff points is what demands the nerf.
And now, we will take a look at the pros and cons of nerfing multi-target attacks directly, on its own:
- Weakens cogs. This can be a good thing when it is warranted due to cogs being too powerful in comparison to toons.
- Not very invasive; all that need to be tweaked are damage and/or accuracy numbers.
- Can be an effective way of balancing single-target and multi-target cog attacks.
- Weakens cogs. This can be a bad thing when it is unwarranted due to cogs already being as powerful as, or less powerful than, toons.
- Revisionist towards the set of cog attacks that exist in TTR, because the attacks themselves have to be modified.
- This method requires some meticulousness, because all cogs with multi-target attacks have to be balanced, against themselves and against all other cogs of a similar level. This includes balancing the damage/accuracy values for the attack at each cog level (of which there are 5 for a given cog species), for each such attack.
footnotes for “nerfing multi-target attacks directly”
- Maximum variance occurs at p = 0.5, because for a given n, the
variance is proportional to
p(1 − p) = p − p2.
Since this is quadratic
in p, the global maximum is just where the derivative is
zero. The
derivative w.r.t. p is then
1 − 2p, by the power
rule. Finding where the
derivative is zero:
1 − 2p = 0
−2p = −1
p = ½ = 0.5. - The probability mass function for a particular outcome k over a binomial distribution X ∼ B(n, p) is Pr(X = k) = nCk × pk(1 − p)n−k. The sum of this PMF over k ∈ {1, 2, 3}, where n = 4 and p = 0.5, is 0.875 = ⅞. Note that this is not representative of approximating the binomial distribution with the normal distribution (where ±σ contains roughly 68.3% of the result values) because there is no continuity correction (and n is small).
The considerations for buffing single-target cog attacks directly are largely similar to (or rather, inverse to) the considerations for nerfing multi-target ones. One notable difference is that changing accuracy values is no longer really an option here; most cog attacks (especially those from cogs that are a high level for their species[1]) have accuracy reasonably close to 100% anyways, and single-target attacks are not affected by accuracy changes in the same way as their multi-target counterparts.
It should also be noted that buffing single-target cog attacks has the effect of raising the threshold for how much laff a toon needs to have left to withstand a given number of cog attacks. This is not necessarily the case when tampering with multi-target attacks.
So let’s take a look at the pros and cons of buffing single-target cog attacks directly, on its own:
- Strengthens cogs. This can be a good thing when it is warranted due to cogs being less powerful than toons.
- Not very invasive; all that need to be tweaked are damage numbers.
- Can be an effective way of balancing single-target and multi-target cog attacks.
- Strengthens cogs. This can be a bad thing when it is unwarranted due to cogs already being as powerful as, or more powerful than, toons.
- Revisionist towards the set of cog attacks that exist in TTR, because the attacks themselves have to be modified.
- This method requires some meticulousness, because all cogs have to be balanced, against themselves and against all other cogs of a similar level. This includes balancing the damage values for the attack at each cog level (of which there are 5 for a given cog species), for each such single-target attack.
footnotes for “buffing single-target attacks directly”
- “Species” here refers to particular named species, like Telemarketer or Corporate Raider.
Now we are brought to options more “radical” than just changing some of the numbers, by introducing additional mechanic(s): first, attack limiting. The basic idea is to treat cogs slightly more like toons, in that they will have some attacks in limited quantity, just like toons can only carry e.g. 3 Birthday Cakes. It should be noted that attack limiting is discussed here as a way of balancing multi-target cog attacks against single-target ones, but is actually potentially more useful. Attack limiting can act as a general strategy to be used for game-balancing, as well as a way to simply make cog fights more interesting.
It doesn’t really make sense to allow cogs to run out of attacks entirely, lest we fall into the trap of making cogs mirror toons for no other reason than “symmetry” for its own sake; some asymmetry between toons and cogs is intentional and beneficial to the game’s design. Rather, when applying attack limiting, it is expected that each cog species will possess at least one attack that is unlimited in use.
Attack limiting should not break in the presence of v2.0 cogs or similar. If, for example, a v2.0 cog has a one-time-only attack and uses it, and then has their first shell defeated, their skelecog form should not be able to use that attack.
The main issue left to hash out here (short of actually proposing a particular attack limiting scheme for every cog species) is the issue of which numbers it makes sense to use. Generally, it is expected that any limit greater than 3 or so is not going to be very useful, because cogs are not expected to last very long on the battlefield and perform a lot of attacks. Limits greater than 3 are going to make the attack effectively unlimited most of the time in practice, and even a limit of 3 on a single attack is probably pretty high. Again, it depends on how you choose to balance the rest of the game as to how many attacks you expect cogs to actually be able to perform in various situations. Additionally, the effects of an attack being limited are dependent on the probability of that attack occuring in general; an attack that is limited to one time only, but has a 10% usage chance, is only somewhat limited by the one-time-only bit. In general, most attacks that are limited should be limited to either 1 or 2.
It is possible to slightly extend this concept by pooling some of the attacks that a given cog species has into a group, and then putting a limit on that group, rather than a separate limit for each attack in the group. However, this complicates things considerably for no apparent benefit. Potentially, this is useful when a cog has many attacks in their arsenal and you want to limit the use of any of those attacks that aren’t a particular one or two of them, but that seems like a niche usecase. Any other benefits are a mystery as of yet.
- Weakens cogs (usually). This can be a good thing when it is warranted due to cogs being too powerful in comparison to toons.
- Doesn’t require revising attacks; the attacks stay the same, and cogs’ use of them is what changes.
- Makes cog behavior somewhat more deterministic, in a way that toons can strategize around (viz. it is possible to know that a cog has exhausted some attack and cannot use it again).
- Flexible & multi-purpose.
- Weakens cogs (usually). This can be a bad thing when it is unwarranted due to cogs already being as powerful as, or less powerful than, toons.
- Invasive, because a new mechanic must be added into the game, which complicates the game’s mechanics.
- This method requires some meticulousness, because any cog attack by any cog species is potentially the subject of attack limiting.
Toying around with exactly how many toons can be hit by a given multi-target cog attack is a possibility. The obvious way to do this is to set the (maximum) number of targets for each of a species’s multi-target attacks on a per-attack-per-level basis, or just on the somewhat coarser per-attack basis.
Balancing on the level of target count could allow for more fine-grained balance adjustments. It also could make the cog aggro mechanic more relevant and a bigger part of toon strategy, because normally multi-target attacks ignore aggro since they don’t have to make any choice as to which toon to target. On the other hand, while this is interesting, it is the most invasive of all suggestions here in lieu of balancing multi-target vs. single-target attacks.
- Weakens cogs. This can be a good thing when it is warranted due to cogs being too powerful in comparison to toons.
- Can be an effective way of balancing single-target and multi-target cog attacks.
- Makes the cog aggro mechanic more relevant by making multi-target attacks possibly rely on aggro in terms of which toons they target.
- Weakens cogs. This can be a bad thing when it is unwarranted due to cogs already being as powerful as, or less powerful than, toons.
- By far the most invasive method listed in this section. This is presumably very difficult to implement properly, since it requires messing with cog attack animations.
- This method requires some meticulousness, because all cogs with multi-target attacks have to be balanced, against themselves and against all other cogs of a similar level. This includes balancing the target count values for the attack at each cog level (of which there are 5 for a given cog species), for each such attack. Less demanding is setting the target count values on a per-species-per-attack basis rather than also varying by cog level.
Many “generic”/“randomly generated” toontasks are excessively difficult in comparison to other tasks available in their stead, for the same reward. The infamous tasks like “defeat 200 5+ story cog buildings” obviously fall into this category, but so do many others. In many cases it’s unclear why the tasks exist in the first place, other than perhaps as a cruel practical joke played on those who accidentally accept them. In other cases, the tasks just need to be a bit more balanced to be in line with other comparable tasks with the same reward.
This calls for some adjustment of the random task generation algorithm. Toontasks need not be extremely balanced; balancing them with one another is not a huge part of improving the game, and some variation in balance is acceptable, because different toons will have different preferences and priorities at different times. Additionally, super difficult tasks (like our infamous example of “defeat 200 5+ story cog buildings”) need not be removed entirely, so long as one is willing to invent new rewards, e.g. 15000 jellybeans, or perhaps a permanent goofy cosmetic effect to add to one’s cosmetic effect inventory, to be used at any time (like clothing in one’s wardrobe).
Probably the easiest way to come up with a reasonably fair algorithm is by assigning point values to objectives (and modifiers thereof). Also, point values have to be determined for particular rewards; not just assigned to the reward itself, but to the combination of the reward with the chapter that it belongs to (e.g. Daisy Gardens, or Donald’s Dreamland II). Then, when randomly generating a task for a particular reward, one simply chooses some sensical assortment (for example, it doesn’t make sense to have to defeat 5 4+ story coin mints…) and adds together their point values. Then, one divides the point value of the reward by this sum, and rounds to the nearest integer (to a minimum of 1). This will give the number of times that the objective has to be acquired/completed/defeated/&c. It is probably worth checking the “error” term here, that is, how much you have to round off as a portion of the rounded-off value (or the original value). Error terms too large may be unacceptable, in which case such results should be discarded/avoided.
Here, we will try to devise one such point-based scheme, although this is only a suggestion/demonstration. We will start first with assigning point values to requirements. Also note that the code snippets are in Python:
requirement | points |
---|---|
cog defeated | 10 |
cog is at least level n , in no particular area[1] and/or facility[2] |
32 // (13 - n) - 1 [3] |
cog is at least level n , in a particular area[1] and/or facility[2] |
4 |
cog must be at least level 11 | 5 |
cog must be at least level 12 | 14 |
cog must be v2.0, in no particular area[1] and/or facility[2] | 35 |
cog must be v2.0, in a particular area[1] and/or facility[2] | 15 |
cog must be a skelecog | 9 |
cog of a particular type[4], in no particular facility[2] | 12 |
cog defeated within a particular area[1] | 6 |
cog defeated in a particular facility[2] | 15 |
cog must be of a particular species[5], in no particular facility[2] | 40 |
cog must be of a particular species[5], in a particular facility[2] | 15 |
building defeated | 60 |
building must be at least n floors |
15 * (n**2 + n - 2) |
building of a particular type[4] (minimum floor count of n ) |
75 * n |
building within a particular area[1] (minimum floor count of n ) |
45 * n |
factory defeated | 350 |
mint (no particular type) defeated | 470 |
coin mint defeated | 500 |
dollar mint defeated | 750 |
bullion mint defeated | 950 |
DA office (no particular type) defeated | 510 |
DA office A defeated | 540 |
DA office B defeated | 760 |
DA office C defeated | 1000 |
DA office D defeated | 1200 |
cog golf course (no particular type) defeated | 550 |
front 3 defeated | 580 |
middle 6 defeated | 1200 |
back 9 defeated | 1800 |
… | … |
There are some subtleties here, many of which are captured by the expectations already set forth by TTO. For example, tasks requiring cogs of a particular species and a certain minimum level should be avoided so as to not effectively double-count points; this kind of task doesn’t exist in TTO/TTR anyways. One of the nice things is that this kind of scheme is quite flexible; it can be tweaked as needed to generate better tasks, as one sees fit.
footnotes for “toontasks”
- “Area” here refers to a playground (Toontown Central, Donald’s Dock, &c.) or cog HQ (Sellbot HQ, Cashbot HQ, &c.). It does not refer to more specific areas like a cog building or bullion mint.
- “Facility” here refers to a location that is more specific than an “area”, like a cog building or bullion mint.
- This formula only works if the maximum cog level is 12.
- “Type” here refers to the four types sellbot, cashbot, lawbot, & bossbot.
- “Species” here refers to particular named species, like Telemarketer or Corporate Raider.
Just to recap how the acquisition of gag tracks works in TTO/TTR:
- You start the game with throw AND squirt. (2 tracks total)
- After completing Toontown Central, you gain toonup XOR sound, at your choice. (3 tracks total)
- After completing Donald’s Dock, you gain drop XOR lure, at your choice. (4 tracks total)
- After completing Daisy Gardens and Minnie’s Melodyland, you gain trap XOR (whichever track you didn’t get from Toontown Central), at your choice. (5 tracks total)
- After completing The Brrrgh, you gain exactly one of the two tracks that you are still missing, at your choice. (6 tracks total)
So, by the time that you’ve finished The Brrrgh, you have all of the gag tracks except for one, and you have some choice of which one you are missing. Notice the wording of “some choice”, since you cannot choose to have all 6 tracks except for throw, and you also cannot choose to have all 6 tracks except for squirt.
Also somewhat strange is the way that you start the game. Having completed exactly 0 playgrounds, you have access to 2 gag tracks; having completed 5 playgrounds, you have access to 6 gag tracks. This isn’t “strange” per se, since it just means giving players a big head-start on their gag tracks: if completing 5 playgrounds gets you from 2 tracks to 6 tracks, then each playground is worth something like (6 − 2)/5 = 0.8 gag tracks, and yet you start the game (0 playgrounds completed) with 2 gag tracks.
Furthermore, needing to complete Daisy Gardens and Minnie’s Melodyland to get your fifth track is also a bit odd, when otherwise only a single playground is needed to acquire a new track. In TTO/TTR, the excuse here is basically that Daisy Gardens has the first cog HQ taskline. Each of the four cog HQs has its own taskline, and the two “easiest” HQs (sellbot and cashbot) have tasklines associated with the playgrounds that they are connected to (Daisy Gardens and Donald’s Dreamland, respectively). By “associated with”, it is meant that these cog HQ tasklines are encountered in those particular associated playgrounds at the places within the main taskline’s progression that those playgrounds occupy. The taskline progresses from Toontown Central, to Donald’s Dock, to Daisy Gardens, to Minnie’s Melodyland, to The Brrrgh, to Donald’s Dreamland. This progression is reflected by the increasing difficulty of the streets (that is, the cogs roaming the streets and the cog buildings on those streets) in those playgrounds. This works out nicely, since you have 4 total gag tracks to gain (going from 2 to 6 gag tracks), and 6 playgrounds. So, 4 of those playgrounds give gag tracks, and the other 2 are each dedicated to a cog HQ taskline. The odd bit is that the sellbot HQ taskline is actually not necessary; it exists only to introduce players to the concept of cog HQs, rather than to grant a cog HQ disguise/suit (since a sellbot disguise is obtained by just completing 10 factories, no task needed). Furthermore, it interrupts the flow of gaining gag tracks. Neither of these idiosyncrasies apply to Donald’s Dreamland, which also does not award any gag tracks.
While these idiosyncrasies of how TTO/TTR handles gag track acquisition make some good sense within the context of TTO/TTR, they also leave room for changes that make good sense as well. And, within the context of some of the changes proposed in this document (e.g. gag track balancing and changes to laff limits on facilities), it may make even more sense to change gag track acquisition away from the TTO/TTR approach.
When thinking about changes to gag track acquisition, we want to consider how we can improve the players’ freedom. Like with considerations on the subject of laff limits, we have to accept that allowing more flexibility in the game (rather than less) is ultimately beneficial, and that any time that players are clearly stripped of that flexibility, they will often (when possible) find ways around it. Hence, the existence of uber toons, semi toons, reverse uber toons, cashbot uber toons, boss runs with artificially limited numbers of participants, &c.
Clearly, toons must have the option to start the game with at least one gag track. But should the maximum number of starting tracks be 2, like in TTO/TTR? And, regardless of that number, how much freedom should toons have in choosing the gag track(s) that they start with?
When it comes to how much freedom that toons should have in the gag track(s) that they start with, it is pretty straightforward to argue that the freedom should be total, given the principles under consideration here. That is, any time a gag track is chosen at toon creation, it should be chosen from a selection of all 7 gag tracks (barring any tracks already chosen, of course). If you go the TTO/TTR route and restrict what gag tracks toons start with, they will find ways to circumvent that anyways — there are plenty of throwless and/or squirtless toons in TTR. So, not only is this a pretty unnecessary restriction of the freedom that players have in building their toons, but it also is a pointless restriction, in that it can be somewhat easily bypassed.
The same reasoning suggests that, regardless of what maximum we place on the number of tracks that a toon can start the game with, the minimum number of such tracks should be zero. This may seem a little odd in the sense that it may seem to encourage (or at least, sanction) reverse ubers (or at least, toons that are completely helpless for the entire time until they get a gag track). The reality, however, is that reverses (and 0-track ubers, for that matter) can be made anyways; there are even several of them in TTR that have the maximum attainable laff (137 at the time of writing). All that is really needed is a warning sign that shows up any time that a player is electing to start with zero gag tracks, saying something to the effect of: “Are you sure you want to start the game with no gag tracks? It’s going to be a rough life! The Cogs don’t respond well to the persuasive words of Toons…”. Not terribly relevant, but also worth noting, is that not all gagless toons are necessarily either reverses or 0-track ubers — they could just be toons that never plan on fighting Cogs! There are other activities to do in Toontown, of course.
When it comes to a maximum number of starting gag tracks, things get a little messier. It depends pretty strongly on how the rest of the game is designed (c.f. the other sections of this document). For the most part, the numbers that we are here deciding between are probably just 1 and 2. Obviously anything lower than 1 is nonsense, and while 3, 4, 5, and 6 are all probably viable in some cases, in most cases that likely makes toons excessively powerful from the get-go and pretty severely dampens the sense of progression as the number grows. The sense of progression being dampened by a higher maximum number of starting gag tracks is due to the obvious reason that you have less milestones along the main taskline (since the main taskline gives you fewer gag tracks), especially because said milestones would represent spikes in power.
- If sticking with an endgame maximum of 6 gag tracks: Fits more nicely with the existence of the Daisy Gardens main taskline, in the sense that Daisy Gardens could then award a gag track (like Toontown Central, Donald’s Dock, Minnie’s Melodyland, & The Brrrgh). This can also be used to help with the concerns discussed in the “limiting laff” section if one sees fit, since some or all of the Daisy Gardens tasks that award max-laff in TTO/TTR could be replaced with gag track frame tasks instead.
- If setting an endgame maximum of 5 gag tracks: Allows you to preserve the TTR/TTO taskline and still end up with 5 gag tracks (maximum) after finishing The Brrrgh.
- Makes toons overall less powerful, up until they get their last gag track. This could be a pro.
- Given a certain endgame maximum number of gag tracks, starting with 1 track increases the sense of progression over starting with 2 tracks.
- Kind of a rough start to not be able to pair your chosen gag track with at least one other track! For some choices (viz. toonup, lure, & trap), it becomes straight-up impossible to solo even the easiest cog battle. Also notable is that trap is doubly useless, simply because it relies on someone else having picked lure, which is somewhat unlikely if you are mostly playing with other toons who also have just 1 gag track. Starting with a maximum of 1 gag track at toon creation is nigh begging players to shoot themselves in the feet.
- If sticking with an endgame maximum of 6 gag tracks: More revisionist because you have to implement another gag track acquisition taskline (presumably in Daisy Gardens), which is not only revisionist but also potentially more difficult to implement (depending on how you go about it).
- Makes toons overall less powerful, up until they get their last gag track. This could be a con.
- Much more flexible than starting with a maximum of 1 track, since for any chosen track you can freely pick from 6 other tracks in a way that makes your first choice more viable. As mentioned above, trap is a clear loser in the “pick exactly one track to start the game with” game, but here some sense can be made by also choosing lure (so that you can lure for your own traps) or choosing something particularly versatile, like throw, or squirt, or drop, or sound.
- If sticking with an endgame maximum of 6 gag tracks: This is the least revisionist option, since neither the maximum number of starting tracks nor the maximum possible number of gag tracks are changed from TTO/TTR (although you get more choice, of course).
- Makes toons overall more powerful, up until they get their last gag track. This could be a pro.
- Makes toons overall more powerful, up until they get their last gag track. This could be a con.
- If setting an endgame maximum of 5 gag tracks: This is very difficult taskline-wise and progression-wise. In the main taskline you will have to remove the gag track reward from one of the playgrounds (so that Daisy Gardens, Donald’s Dreamland, and 1 other playground all do not award gag tracks). Progression-wise, this guts the sense of progression to be even less rewarding than your standard TTO/TTR. From toon creation to the very end of the main taskline, you have only acquired a total of 3 gag tracks.
- If setting an endgame maximum of 6 gag tracks: Continues to leave the Daisy Gardens taskline barren of gag track acquisition.
In TTO/TTR, every time that you start the main taskline of a playground that awards a gag track, you are given a choice between two gag tracks, and must choose which one you want to acquire by the end of that playground’s taskline. In keeping with the logic of the previous section on the subject of starting gag tracks, this should be expanded to give the toon a choice between all of the gag tracks that they don’t already have, plus the additional choice of not acquiring a gag track at all.
There is also the possibility of changing how many gag tracks that a toon ultimately ends up with after the completion of the main taskline (assuming that they always choose to acquire as many gag tracks as they can get). In TTO/TTR, this number is 6, which is one less than the total number of gag tracks that exist in the game. So, in TTO/TTR, “normal” builds are defined by the single gag track that they are missing (typically “trapless” or “dropless”). However, the possibility exists to reduce (or even increase) this number. Here we only consider two possibilities:
- Keeping this number at 6.
- Lowering this number to 5.
Increasing this number to 7 means that every (non-uber & non-semi, henceforth NUNS) toon has every gag, something that already effectively feels true in TTO/TTR in a way that is problematic and not conducive to creativity nor teamwork. Particularly, the total number of possible NUNS builds is reduced from 7 to just 1. We also don’t consider lowering this number below 5, because it quickly starts to hamper the sense of progression through the game, and doesn’t actually increase the number of possible NUNS builds by as much as one would hope.[1]
At the moment, I am leaning towards a maximum number of starting gag tracks of 2, paired with a maximum number of ending gag tracks of 6 (that is, similar to TTO/TTR). Not only is this basically the least revisionist option, but also the option that starts the game in a more flexible and less brutal way (starting with a maximum of one track being essentially an invitation for players to shoot themselves in the feet). The maximum of 6 ending gag tracks follows from the maximum of 2 starting gag tracks, since ending with a maximum of 5 would hamper progression too much and require breaking the main taskline very seriously.
- Dramatically increases the total number of possible NUNS builds, from 7C6 = 7 to 7C5 = 21. This implies a greater variety of toons, increasing the overall creativity that goes into even NUNS toon creation.
- Further curtails the feeling of “basically every toon has basically every gag” that TTO/TTR has, even moreso than just balancing the 7 gag tracks alone. This implies a greater focus on meaningful teamwork & cooperation, as well as more engaging and varied gameplay.
- If limiting starting gag tracks to a maximum of 1: Allows keeping the main taskline totally intact, since the number of gag tracks acquired between toon creation and completion of the main taskline is still 4.
- Makes toons weaker overall. This could be a pro.
- If limiting starting gag tracks to a maximum of 2: The number of gag tracks acquired between toon creation and completion of the main taskline is now less than it is in TTO/TTR. Another playground (in addition to Daisy Gardens & Donald’s Dreamland) will have to be stripped of its gag track taskline. The game’s sense of progression is dampened by the lack of gag track advancement.
- Makes toons weaker overall. This could be a con.
- This is the more revisionist option, insofar as the maximum number of tracks that a toon can have in TTO/TTR is 6, not 5.
- If limiting starting gag tracks to a maximum of 2: Does not hamper the game’s sense of progression like a final maximum of 5 gag tracks would. Also unlike ending with 5 tracks maximum, this does not break the main taskline; indeed, the original main taskline from TTO/TTR can be preserved.
- If limiting starting gag tracks to a maximum of 1: The game’s sense of progression can even be improved over that of TTO/TTR, since a NUNS toon acquires 5 (rather than 4) gag tracks in between toon creation and completion of the main taskline. Daisy Gardens (or Donald’s Dreamland, for that matter) can be fitted with a gag track acquisition taskline.
- Makes toons stronger overall. This could be a pro.
- Dramatically reduces the total number of possible NUNS builds, from 7C5 = 21 to 7C6 = 7. This implies a smaller variety of toons, decreasing the overall creativity that goes into NUNS toon creation.
- Makes toons stronger overall. This could be a con.
footnotes for “gaining gag tracks”
- Setting the maximum number of possible gag tracks at 3 or 4 gets you 7C4 = 7C3 = 35 possible NUNS builds; not quite as much of an improvement over 7C5 = 21 as 21 is over 7C6 = 7. And of course, going below 3 possible gag tracks only gets worse (not that you would want to do that anyways).
This section is dedicated to other thoughts and proposed changes that have little or nothing to do with the core mechanics of the game.
It could safely be said that TTO/TTR had/has some of the more poorly-designed player communication systems seen in any MMO. This fact achieves nearly comical levels of irony, considering that Toontown is a game prided on its focus towards teamwork and friendship. There is a variety of factors making communication less-than-optimal and even less-than-useful within TTO/TTR that are addressed here.
Speedchat is actually one of the strengths of the TTO/TTR communication system. A method of communication through pre-ordained phrases has a number of benefits, but these benefits are watered down — and sometimes not even taken advantage of — in TTO/TTR.
One of the truly wonderful benefits of communication through pre-ordained phrases is that it invites localization (l10n). Players can set their language preferences locally, and all Speedchat-related text (including Speedchat phrases uttered by other players) then appear to the player in their preferred language. Of course, this involves all the messiness and difficulty of natural language translation, but certainly nothing extraordinary that hasn’t been tackled before. Idiomatic phrases in particular pose a challenge, but this is generally solved by either sense-for-sense translation (possibly, in certain cases, even substituting in an analogous idiomatic phrase in the target language) or even leaving the Speedchat phrase fully or partially untranslated.
As far as I know, no versions of Toontown exist that implement this feature.
If you had never used Speedchat before, and you heard that it functioned via pre-ordained phrases, you might think it’s a convenience. In some ways, it is, but as soon as you first attempt to navigate the hellish gauntlet of super-human dexterity, patience, and recall that is the Speedchat interface, you might be disappointed. The GUI provided by Speedchat is consistently frustrating to use for everything but the most easily accessed phrases, makes certain phrases nigh impossible to utter at all, requires the exclusive use of the mouse (or other pointing device) to the exclusion of keyboards, and requires the user to remember and repeatedly navigate a jumbled hierarchy of phrases.
As a result, it is in no way difficult to imagine a better Speedchat interface.
A good start that avoids the pitfalls of Speedchat as it exists in TTO/TTR is
to allow users to bind a
“Speedchat key” to some arbitrary key on their keyboard, and then
initiate a fuzzy string
search over the set
of Speedchat phrases when the user starts typing. The user can then press
Enter/Return
to confirm the closest match and thus utter the phrase and leave the
“Speedchat mode”, or even use arrow
keys after having typed a nonempty
string to select from the top n
matches. An extension to this would allow arrow
key navigation when the user has
only typed the empty string
(either by not having typed anything yet, or through the use of
Backspace); in this
context, arrow key navigation would
allow selection from the previous n
Speedchat phrases uttered by that player.
The old interface could stay; instead of being used via the
keyboard, you would have to
click on the Speedchat icon as normal.
You can even go further and allow players to bind arbitrary Speedchat phrases to arbitrary key combinations.
It should be noted that the accessibility (a11y) of these kinds of things improves significantly once you allow the choice between pointing device and keyboard. Pointing devices are inherently clumsy because they are designed specifically for tracking motions in two-dimensional space; keyboards, by contrast, only deal in discrete events that can easily be mapped to various devices (some of which may not even feature push-buttons as a primary interface). The current Speedchat interface may be difficult or impossible to use for those with motor impairments like paralysis, various RSIs, &c., as well as those with conditions that affect eye–hand coordination (particularly in the elderly).
It would ideally go without saying that within the context of the traditional Speedchat GUI, phrases should be in a fixed configuration (same hierarchy, and same order within each list), with the obvious exception of adding or removing phrases. And, in addition, phrases that are more closely thematically/semantically related should be closer together, as a general rule. Unfortunately, Speedchat systems like the one in TTR violate these principles.
A Speedchat “feature” that attempts to mitigate the automation of doodle training by randomizing the order of the Speedchat phrases in the “Pets” category violates the first principle here, compounding the confusion and difficulty already associated with Speedchat. While this is presumably justified in the context of mitigating said automation, this justification is weak at best, since the underlying problem (players regularly desiring the use of automation for something that is supposed to be an enjoyable game) lies with doodle mechanics and not with the ease of automation of Speedchat (something that often inevitably comes alongside good a11y, anyways). In any case, a very naïve mitigation such as this one is easily bypassed in countless ways, including, but not limited to: OCR (even of the most rudimentary kind, since the text will always be pixel-for-pixel the same), sniffing the process’s memory, compiling a custom client binary, &c.
Phrases that are thematically/semantically closely related, e.g. nearly equivalent phrases like “No problem.”, “You’re welcome!”, and “Any time!”, should be adjacent. It is acceptable (and sometimes desirable) to have the same phrase in more than one location, for example, putting “Good luck!” in both the “Goodbye” and “Friendly” sections.
One of the great weaknesses of the TTO/TTR Speedchat+ system is the lack of any kind of user-facing chat log whatsoever. This is in stark contrast to the textual communication systems of essentially every other multiplayer game; in TTO/TTR, if you don’t see the speech bubble, the message was never sent as far as you’re concerned. Having an ordinary chat log like every other game has long been a requested feature from TTO/TTR players.
Not much needs to be said about the exact implementation of such a chat log, in the sense that it is a commonplace technology nowadays and should therefore be more straightforward to design. Some nice features might include, among other things:
- The ability to easily control how much screen real estate is occupied by the chat log; for example, a way of toggling its visibility, or even a way of resizing it and/or repositioning it.
- The ability to scroll up and down the log.
- A horizontal rule inserted into the log every time that a change of map occurs, with any sequence of multiple such rules in a row collapsed into a single rule.
- The ability to click on a toon’s name in the chat log and have it function like clicking on the toon’s nametag.
- Only showing a message-sending toon’s name once in the case that they send 1 or more messages in a row; the name need not take up space by being duplicated multiple times in a row.
- The ability to show or hide message
timestamps. The
timestamps themselves can probably
just be in the format
%T
, or%X
if locale-dependence is desired (usingstrftime
syntax), in the user’s local time zone.
Because the purpose of Speedchat+ is to allow more-or-less free-form textual input, and Toontown is intentionally designed for persons of all ages (particularly w.r.t. children), some form(s) of automatic bowdlerization are typically employed to filter out unwanted words, or even unwanted phrases. Although such methods of automatic bowdlerization are usually useful to some degree, their actual effectiveness is dubious, and the particular implementations found in the wild are sometimes victims of poor design.
When it comes to effectiveness, even with very strict filters (like e.g. that of TTR), players still easily and widely find ways to talk about the intentionally blocked words/phrases/topics anyways (and, furthermore, the strategies employed for this purpose are diverse). This is usually not even the fault of the filter being poorly designed or having a narrow purview; this is just a reality of purely automated (that is, no actual persons being involved) modes of censorship. This inherent limitation should be taken into account when considering how overbearing to make a particular filter. While it is not true that this inherent limitation implies that all such filters are totally useless, it does limit how much it makes sense to accept a certain level of inadvertent censorship of ordinary (that is, unproblematic) language.
When it comes to the design of these
filters, there exists a variety of
strategies, and this document will not go over the various merits and pitfalls
of each one. Instead, it is illustrative to look at some existing design
failures. One failure is the mixing of incompatible strategies, like employing
blacklisting in addition to
whitelisting. Such a scheme is
inherently confused, and
serves also to unnecessarily confuse players and create an overbearing filter
that makes normal speech more error-prone and more of a challenge.
Neither blacklisting nor
whitelisting are poor strategies
in and of themselves, but the combination of the two (as well as certain other
incoherent combinations) should be avoided. Another failure is one of
punctuation; depending on the
exact implementation (usually implicitly or explicitly relying on regular
expressions),
punctuation is sometimes made more
difficult to use because of a
filter, even if the
punctuation is being used
ordinarily. One possible clean way
of preventing this failure (in
English and in many other
languages) is simply ignoring
punctuation when doing the
filtering (counting
hyphenation as a form of
spelling rather than of
punctuation when not surrounded by
spaces). For example, They said, "no."
would be
filtered identically to
They said no
(note that the space between the
comma and the first double quotation
mark is preserved).
The last glaring issue w.r.t. Speedchat+ bowdlerization methods is concern for localization (l10n). In an ideal fantasy world, l10n for Speedchat+ would be just as straightforward as Speedchat. Of course, Speedchat+ is much more free-form, so compatibility between languages is difficult or impossible. A few options present themselves:
- Speedchat+ is monolingual. Particularly, this usually means English-only, both because Toontown originated in the United States and because English acts as a kind of de facto global lingua franca. However, Toontown servers that are based in and/or target regions that are not primarily English-speaking and would not use English as a lingua franca would choose some other language (e.g. Portuguese for Brazil and other Lusophonic regions).
- Speedchat+ for a given user at a given time only allows input from a particular language, namely whatever their current language preference is. Speedchat+ messages then appear to other users untranslated. It behooves such an implementation to allow users to switch their language preference easily and on-the-fly (viz. without having to re-login), to allow multilingual users to communicate freely. Players should easily be able to view the language preference of other toons, e.g. via their nametags or by clicking on their nametags.
- Speedchat+ ignores language preference, but still allows communication in two or more languages (including two or more languages freely mixed within a single message). Players should still easily be able to view the language preference of other toons, e.g. via their nametags or by clicking on their nametags.
- Each district in the game is associated with a language. Within each district, Speedchat+ acts like (2.), but as if every player in the district had their language preference forcibly set to whatever the district’s language is.
(1.) has the advantage of being the easiest to implement, and of having the best chance of mitigating unwanted language. However, it also is the most restrictive and is socially exclusive towards those who do not speak the particular language of choice.
(2.) has the advantage of allowing multiple languages and is thus less exclusive. The need to switch language preference to speak a different language (and thus also the, for the most part, inability to use multiple languages within a single message) is clunky, but in exchange, this strategy has a better chance of mitigating unwanted language than (3.). This also requires having a filter for each language, and so takes more implementation and maintenance work than (1.).
(3.) has the advantage of allowing multiple languages and is thus less exclusive. Being able to freely use various languages is a great boon, but comes with the downside of making it more difficult to mitigate unwanted language than it is in, say, (2). Whether this requires more or less implementation/maintenance effort than (1.) depends wildly on the filtering strategy, but in both cases there is just one filter (rather than a filter for each language).
(4.) has the advantage of allowing multiple languages, but unfortunately does so by walling these language communities off from one another, which is probably needlessly divisive, as well as stifling to multilingual users. This also does not play nicely with the ordinary use of districts, which are simply identical copies of the same world that may be switched between for game-mechanical reasons (e.g. to find a building). This also takes about as much effort as, and has about as much chance of mitigating unwanted language as, (2.), since a separate filter for each language is needed.
Speedchat+ messages are ordinarily restricted to a very short length. This hobbles communication to a great extent. It incentivises very abbreviated language to the point of being difficult to decipher. It also requires splitting thoughts and sentences across multiple messages, which leads to unnecessary interruption/incoherence, and forces listeners to wait for later installments of the same sentence or thought and then piece them together. Obviously, some reasonably short limit needs to be put on the length of Speedchat+ messages, but there is no reason they cannot have a much larger limit than they do in TTO/TTR in order to allow fluid communication.
Also, in TTO/TTR, Speedchat+ message “length” is determined in a rather obscure way that doesn’t correspond well to the actual textual content. Better is to simply judge length on the basis of extended grapheme clusters, grapheme clusters, codepoints, or even the number of bytes in the UTF-8 encoding of the text.
If things like toon names and Speedchat are too removed from the actual “mechanics” of the game per se to have a place in this document outside of this “odds and ends” section, then governance of the game is a whole new level of unrelatedness.
Nevertheless, for what it’s worth, in light of how essentially every Toontown private server is currently being run at the time of this writing, here are just a few quick points. I won’t attempt to justify any of these points, as that would be more than is warranted for something that is this off-topic, and the justifications should hopefully be quite obvious anyways:
- The game should be open source: the code comprising the server, and especially that comprising the client, should be publicly available and licensed under the terms of some free software license.
- The development of the game should be done publicly, that is, transparently, rather than internally/privately. This includes, among other things, public distributed version control, a public pull request system, a public bug/issue tracker, &c.
- The administration of the game should be done publicly, that is, transparently, rather than internally/privately. This implies some relatively “open” form of governance over the game — one that keeps public anything that can reasonably be public (i.e. transparency) and encourages end-user participation, contributions, and discussions. This means “open” in the relevant ordinary sense, not necessarily so-called “open(-source) governance”, which is a much more robust concept that itself takes on countless different forms. In effect, the differences between “staff member” and “end-user” should be minimized, and the process of staff recruitment — and analogously, attrition — should be public/transparent.
- The primary communication channel(s) used for the game (outside of the game proper) should be, as much as is possible/practical, free (as in, free software) and decentralized (as in the internet). Effectively, this means completely avoiding the use of Discord™ and other similar centralized and proprietary services like e.g. Slack™. See “Discord Is Not An Acceptable Choice For Free Software Projects” for clear reasoning on this topic (which also applies to many other things besides Discord™ and besides free software projects).
WARNING: this section contains a larger-than-usual dose of personal opinion. Caveat lector.
There is a part of lure gags as a mechanic that will always be a little bit
broken regardless of how you tweak the parameters. By parameters I mean the
number of rounds that various lure gags (organic, and inorganic) last,
propAcc
for various lure gags (organic, and inorganic), and how many cogs
that various lure gags (organic, and inorganic) target (1, up to 2, up to 3, or
just however many cogs are unlured at the time).
Imagine, as a simple example, that you have a party with an even spread of lureless toons and lure toons: 2 of each. The party is fighting a row of level ≥10 cogs, and they decide to lure all the cogs (with Hypno, let’s say). In TTR, they can probably continue fighting just using lure (and some assortment of trap, throw, squirt, and maybe some drop) even if more cogs join the battle (like in one of the boss battles). When tweaking the parameters of lure gags to balance this, there are broadly two scenarios that come up:
Scenario 1: you, in some way, nerf the accuracy of lure in an attempt to balance it. This has the very clear and unfortunate disadvantage that this makes the game (particularly lure, of course) even more reliant on dumb luck. Generally, you (a) get lucky and have the lures mostly work and arrive back at what looks like TTR lure-based gameplay, (b) put a lot of effort into stunning in order to make the accuracy hover around 85~95% again and thus arrive back at what still looks like TTR lure-based gameplay (but with more stunning, which doesn’t matter too much considering how long the lure gags’ effects last), or (c) you don’t stun well enough and/or don’t get particularly lucky and thus (seemingly inexplicably) keep face-tanking cog hits while performing an unusual number of lureless kills (viz. unusual for a lure-based context). In addition to the aforementioned disadvantage of making things more reliant on dumb luck in many cases, none of these 3 outcomes seem particularly appealing; none of them actually seem to “balance” lure.
An additional thing to note is that in outcome (c), the “lureless kills” in question are not as desirable as they might sound at first: because they are effectively “accidental” in some sense (because they are the result of dumb luck making lure gags miss), they don’t look like actual lureless play, and generally take one of two forms: (c0) some gag(s) are used on a cog as it is lured, but the lure misses and so the gag(s) do not kill the cog; the cog is killed at some point, but it takes more rounds than expected, or (c1) the gag strategy chosen is a so-called “insurance strategy”, and so the cog dies in one round regardless or whether or not lure hits (assuming the other gag(s) hit). In particular, (c0) is obviously not what anyone really wants from (quasi-)lureless play, and (c1) is nicer but makes especially clear the problem of lure being entirely dumb luck; you hope that the lure hits, but don’t (read: cannot) expect it to, and when does hit, it just acts as a way of saving some resources. But making this lure’s niche as a gag track is probably a recipe of making lure unappealing in general and thus unbalanced. Also, the fact remains that you can never rely on lure for anything remotely critical in either (c0) or (c1).
Scenario 2: you, in some way, nerf the duration of lure gags in general in an attempt to balance the gag track. At first, this seems nicer than Scenario 1, since it doesn’t make lure any more fickle than it has to be. But there is a sneakier, more subtle reason as to why this scenario is still “kind of broken”, regardless of whether or not it is combined with Scenario 1. The reason that I use the phrase “kind of broken” is because it depends on your perspective as to whether this is a real problem, a minor problem, or a complete non-issue.
To illustrate, imagine a simple example where Hypno is a gag that works identically to how it does in TTR, with the exception that it lasts for just 1 round (rather than 4). Choosing 1 round in particular is somewhat arbitrary (2 rounds could be a possibility as well), but much more than 1 round, and you aren’t really nerfing it, and 1 round is the simplest to work with anyways. When one toon chooses Hypno, the 3 other toons have the opportunity to pick other non-lure gags; in many cases this will be somewhere between 0 and 2 toonup gags and the rest being attack gags like trap, throw, and squirt. Assume, for simplicity, that the lure hits. The attack gag(s) are used, and some cog(s) still remain both alive and lured. The damage that said cogs would otherwise do immediately after the toons finished using their attack gag(s) is prevented because they are lured; this is part of the point of using lure in the first place. If no new cogs join between the time of the attack gag(s) being used and the next time that the toons get to make gag choices, and all of the remaining (lured) cogs can be taken out with just the gags that are chosen in that very next round, then great. The cogs are defeated, and the toons never had to deal with any pesky cogs becoming unlured. But in the case that these two conditions are not both met, some cog(s) will automatically be unlured after the toons use their attack gags for the second time (or they do actually get defeated, and it is not until the stream of incoming cogs starts to overwhelm the toons’ resources & luck that this automatic unluring occurs; either way, it happens).
While this seems fine and dandy — after all, this is basically the point of nerfing lure gag durations — the problem comes from the awkward ordering in which things occur when a cog is automatically unlured. When the cog(s) automatically unlure after the toons have used their gags, the immediate next thing that happens is that the automatically unlured cog(s) attack the toons. In the context of TTR, this makes a great deal of sense: unluring a cog can make sense for various reasons (including killing it outright), but allowing a cog’s lured-state duration to elapse and thus automatically get unlured is basically always a mistake, and thus gets swiftly punished. When nerfing (and rightly so, I might add) lure via what is here called Scenario 2 (or any combination of Scenarios 1 and 2), this makes considerably less sense, for two related reasons: (a) swiftly punishing toons for allowing lure to expire seems to make less sense when the lure is designed to expire quickly, and (b) this mechanic (so far as I can tell) seems to inherently make it more difficult to reconcile lure-based play and lureless play; despite the fact that the cogs get automatically unlured, there is little chance for toons to plan ahead and take advantage of their knowledge of when cogs will automatically unlure, because the cog will take a whiff at them anyways before the next round comes and they have the chance to re-lure the cog anyways.
To give a (very) simple example that hopefully makes (b) more clear, imagine a lured level 11 cog that is going to get unlured as soon as the next round of toon attacks ends. In anticipation of it getting unlured, I might use a Piano gag on it (Piano, just because it does enough damage to one-shot a level 11 v1.0 cog). But if I actually did that, the Piano would just miss (since the cog is lured and drop gags have a fixed 0% accuracy on lured cogs), and then the cog would immediately unlure and attack me. Again, depending on your perspective, this may or may not seem wonky in a bad way.
There are, of course, some reassurances that can be afforded to both (a) and (b), which I will address here. One could reframe (a) as not being a swift punishment for allowing a cog’s lured-state to expire, but rather an occasional punishment that is inherent to using lure gags at all; that is, it is just part of lure itself that you have to be wise to only use lure when you know that you can face-tank the requisite number of cog attacks (or have a way to dispose of the cogs swiftly enough to avoid said face-tanking). As a rationale, I think that this is actually fine and good, but I would note that this attitude could possibly have the effect of balancing lure too much into the realm of “only marginal usefulness”. For (b), as far as I can tell, there are two (not mutually exclusive) ways of assuaging its concern: you can either not care and be fine with this kind of friction between lure-based play and lureless play, or you can insist that sound gags “solving” the problem in certain edge cases is sufficient to make things work.
When I say that sound gags “solve” the problem in certain edge cases, I basically just mean that the use of drop gags can be afforded on certain cogs that are about to be automatically unlured, with the use of one or more sound gag(s). Of course, this has two problems: one is that this only actually makes sense in cases where the cogs not being dropped on have low enough HP (either from being low level, or from being hit in some previous round(s)) to all (or at the least, all except one) die to the sound gag(s). If this condition isn’t met, then it is almost certainly better to do something else (probably just aggressively luring). The other problem is that what is really going on here is a shortening of the lure gag’s effect duration by 1 round via the use of sound gag(s). Yes, this technically allows you to take shots at the cog(s) before they get their free attack(s) on you (viz. the attack(s) that they get for being automatically unlured), but part of the price that you pay is that you effectively shorten by ½ the number of rounds that the lure gag’s effects last. The effective shortening is because sound gags have low damage-per-(lured-)cog, come before all the other gag tracks that can be used to attempt to damage lured cogs (viz. throw, squirt, & drop), and get no bonus damage (read: orange damage) on lured cogs. The sound thus effectively acts as a low-damage wakeup. The side-effect of effectively shortening lure effect durations by ½ rounds is somewhat troubling when we are generally talking about lure gags whose effects last for around 1 round each. So, again, the coexistence of lure-based play and lureless play still has the same friction and the problem doesn’t really seem to be solved (excepting perhaps some edge cases).
See the section on balancing lure mechanically for a possible solution.