Requesting Feedback - A Potential Alternate Rule Set

[Kal]

 

[Grim Tuesday]

Minna, miteite kure!

 

[Scar]

Bones, I'm not impressed.

@daKID, The point of more games is NOT so you can play more stages. The point of more matches in a set (like when we go from 2-3 to 3-5 for semis and finals) is to get more points of data. It's a well-known concept in statistics that as you increase the number of trials, you continually decrease the margin of error of whatever it is you're measuring.

The point of science is to eliminate our own biases and to run experiments to find out how things really are. If this is a scientific process to determine a better ruleset, we check our personal ideas and biases at the door, and we run tests as objectively as possible. If you introduce 2 variables you're immediately tainting the results of any test. How can you be sure that people enjoying the ruleset is due to more stages? To less stock? To a smaller time limit?

If we isolate one variable, in this case the timer, and the hypothesis is (as Cactuar has stated) that stalling will be eliminated as a viable strategy in contests with a 3-minute timer, please let me know how you plan to prove it one way or another if matches / sets are played on 20 different stages? It's plainly obvious from even what little testing has been done that more strategies are created by adding new stages. Now there's no way to attribute change in strategy to *only* timer, or *only* stage, but instead we have some vague idea of what will happen if we change everything and compare apples to oranges.

 

[Kal]

I do think that players should stick to the stage list they are already familiar with. It's not going to be easy to assess whether you like the change in stock if you include with it a change in your usual stage list. Scar is right on this point. I think the point da K.I.D. is making, though, is that, with only starter stages, we are playing more matches than we have available stages, which goes against one of the reasons for having the counterpick system in the first place.

 

[Kink-Link5]

I hate when threads get overinflated in size because I can't respond directly to the original post.


Well **** that, I'm doing it anyway.


This ruleset is amazing, but I would suggest making the stage strike done from all available stages instead of a predefined "Starter" list that benefits a handful of characters indirectly.

This would solve the issue of so-called neutral stages, the only problem being that we might be stage striking for 10 minutes

 

[Kal]

I've run a tournament with the No Johns ruleset that Kish Prime wrote, and we struck from the entire stage list. It's a good approach in terms of fairness, but it has some pragmatic issues. People found it doable when provided with small printouts of all available stages, but it felt cumbersome and many people just ignored it altogether for the first stage and simply struck from a smaller list of nine stages.

Personally, I see nothing wrong with striking from the entire list. I think people need to just complain less.

 

[Bl@ckChris]

lol just imagining one person striking all the neutrals and just ending up like "how you gonna act?" scares me. the idea of a last choice being "great bay, onnett, or flatzone?" intrigues me, but i still can't imagine ever staking my money/tournament life on it lol.

 

[~Twitch~]

lol just imagining one person striking all the neutrals and just ending up like "how you gonna act?" scares me. the idea of a last choice being "great bay, onnett, or flatzone?" intrigues me, but i still can't imagine ever staking my money/tournament life on it lol.

me and kevin call that the karn strat. LOOOOL

 

[Bl@ckChris]

lmao karn would SOOOO do that.

twitch, have you and kevin tried this ruleset at all? i'd like to hear how gay kevin is on flatzone platforms lol.

 

[Cactuar]

I just love going Falco and doing momentum stuff on the oil slick.

 

[stelzig]

I hate when threads get overinflated in size because I can't respond directly to the original post.


Well **** that, I'm doing it anyway.


This ruleset is amazing, but I would suggest making the stage strike done from all available stages instead of a predefined "Starter" list that benefits a handful of characters indirectly.

This would solve the issue of so-called neutral stages, the only problem being that we might be stage striking for 10 minutes

Yes, brilliant. Let's reduce playtime even further, this ruleset doesn't do it enough yet!

 

[Kal]

Stelzig, are you sure you quoted the right post there? Or is there some in-between-the-lines thing I'm missing? Also the total playtime is roughly the same in this ruleset as the standard one. It's just that individual matches are shorter. This has been explained before.

 

[twizzlerj]

this thread has gotten out of hand cactuar's popularity and ability to debate can really get a thread going i guess.

 

[Kal]

In Cactuar's defense, I, as usual, have the lead in post count.

 

[Cactuar]

Variety is the spice of life.

*shrug*

@stelzig: The playtime to set time ratio is lower. Actual playtime is roughly the same.

A bigger issue is that people feel like there isn't enough... I guess we could call it plot development... in 2 stock matches. They feel like each match should tell a story? I'm not sure how to take that criticism as it seems to be more of a "I am super used to playing 4 stock matches" response than anything else.

 

[twizzlerj]

I want this thread to die so some noob can wait a few years then post this rule set again and instantly get shot down.

 

[Screech]

I think the stock and timer cuts are a little extreme. I love the direction that you're going and think that the ruleset change is very much on the right track. More matches (and quicker matches) in a single set really allow a player to demonstrate and execute his skillset in a lot more ways.

3 stock, 5 minute games with 5-7 matches per set is a good middle ground. 3 stock matches still penalize SD's without making them an almost guaranteed game ending error. Slip ups happen and while you don't want to give advantage to the better player it's just as bad to force a loss on somebody who tripped up. Your idea is good but probably needs some more consideration.

 

[Kal]

I am starting to think that three stock is a better choice, but I have not playtested it enough to know for sure. However, best of seven with three stock matches best of five with four stock in terms of maximum possible duration and probability of the better player winning (with some assumptions; see this post). I think it's a good idea. I will try and run my next tournament as best of seven with three stock while using only the starter stages.

 

[Penguin_ftw]

This is a massive thread, so it is possible this has been addressed already, but how would crew battles be handled?

 

[Kal]

I think crews would be identical, though you could just allocate half as many stock to each crew. Not sure what Cactus thinks.

 

[KrIsP!]

Crews work differently, I see no reason to shotren the stock count seeing as we would end up with 2 stock vs 1 stock after the first match and comebacks would be hard to pull off. Their not a very big issue so I don't see much point to including them in this rule set wouldn't be necessary. Singles and double sI see as testworthy with this rule set butI doubt crews would gain much from it and I don't think it's as big an issue when it comes to what Cactuar wants to change.

 

[Kal]

You make a good point about the one stock remainders. Crews with two stock could be somewhat awful.

 

[Cactuar]

Yeah, crews is way out of scope. I haven't considered it at all and I won't waste my time considering it in the near future.

 

[Kal]

We'll see about that, Cactuar.

*gets on google chat*

 

[BlueStar03]

I watch tournament videos on youtube to improve my game and I would love to see more stages used. Are there any ideas on how this ruleset would affect character selection/tier lists? Would we see a wider variety of characters used? If so that would make it more entertaining from this viewer's perspective.

 

[KirbyKaze]

I want to just state I would be far more conservative with the stage list and probably just add Rainbow Cruise, Mute City, Brinstar, and two other levels (probably MK2 and something else; probably Corneria) back into the current 6 we have and then gradually test for more and more levels. Bo7 in normal sets, Bo11 in finals.

You need to get 8 kills split between 2 games in order to win a set against someone under the current ruleset. You need 12 in a finals (between 3). All this ruleset really does, aside from forcing the use of new stages, is change how you can go about getting those kills because it further breaks up the stocks-games relationship. Kind of promotes some gimmicky characters for specific levels, but I don't really view this as a bad thing. I'm not even convinced a lot of these strategies are really all that overpowering... I think you can very easily take preventive measures against a character gun-camping on Corneria, for instance. Destroying the guns, or simply not letting them reach the guns. Get a stock lead.

I dunno. I feel it opens up opportunities for a variety of strategy styles to be used... ones that basically don't exist in this current environment just because the current ruleset emphasizes some really specific criteria for character viability. There also reaches a point with a lot of defensive characters where you've basically won once you've got a certain lead (3-1 and 4-2 are pretty much GG, unless you're on death's doorstep or kill yourself). I mean, comebacks do happen, but I feel in most games once I get a certain lead I pretty much wind up grinding my opponent to death. This ruleset prevents matches from devolving into the grind of death, which I think is interesting.

That said, I kind of like the current ruleset because it benefits me greatly for all the reasons stated above. I would still probably enter a tournament under this ruleset and have a lot of fun.

Our current ruleset promotes going solo in a tournament because all the stages are very similar, promote similar strategies on the whole, and reward you for consistency above all else. This ruleset promotes having a secondary and gives a buff to raw power and gimmicks. I'm not sure those are intrinsically horrible things.

 

[KAOSTAR]

I want to just state I would be far more conservative with the stage list and probably just add Rainbow Cruise, Mute City, Brinstar, and two other levels (probably MK2 and something else; probably Corneria) back into the current 6 we have and then gradually test for more and more levels. Bo7 in normal sets, Bo11 in finals.

You need to get 8 kills split between 2 games in order to win a set against someone under the current ruleset. You need 12 in a finals (between 3). All this ruleset really does, aside from forcing the use of new stages, is change how you can go about getting those kills because it further breaks up the stocks-games relationship. Kind of promotes some gimmicky characters for specific levels, but I don't really view this as a bad thing. I'm not even convinced a lot of these strategies are really all that overpowering... I think you can very easily take preventive measures against a character gun-camping on Corneria, for instance. Destroying the guns, or simply not letting them reach the guns. Get a stock lead.

I dunno. I feel it opens up opportunities for a variety of strategy styles to be used... ones that basically don't exist in this current environment just because the current ruleset emphasizes some really specific criteria for character viability. There also reaches a point with a lot of defensive characters where you've basically won once you've got a certain lead (3-1 and 4-2 are pretty much GG, unless you're on death's doorstep or kill yourself). I mean, comebacks do happen, but I feel in most games once I get a certain lead I pretty much wind up grinding my opponent to death. This ruleset prevents matches from devolving into the grind of death, which I think is interesting.

That said, I kind of like the current ruleset because it benefits me greatly for all the reasons stated above. I would still probably enter a tournament under this ruleset and have a lot of fun.

Our current ruleset promotes going solo in a tournament because all the stages are very similar, promote similar strategies on the whole, and reward you for consistency above all else. This ruleset promotes having a secondary and gives a buff to raw power and gimmicks. I'm not sure those are intrinsically horrible things.

trying to skip the bull **** so I don't want to go back and read. what stock count and timer are you potentially supporting?

 

[KirbyKaze]

2 stock
3 minutes

 

[KAOSTAR]

originally I didn't like the idea of two stocks. but when you think about it, there is potential to make characters a little more viable. similarly to the longer the match the more opportunities for the better player to outplay the lesser one. maybe it works with characters to some degree, as in you only have to win two disadvantaged stocks vs 4.

also not that this part matters as much but it is similar to other fighting game formats, and shorter matches are easier to watch, record upload commentate etc. it may help the scene.

I also think melee being a technical game, that shorter bursts could lead to more consistency overall.

idk if this is the best format but I feel the current one needs a change.

I wish people were more open minded about trying new things.

 

[KirbyKaze]

Current one is conducive to me being good which is likely a flaw of the system.

 

[KAOSTAR]

yea umm, I just like shorter games lol. 3 stocks 5 minutes ain't half bad either.

 

[Blistering Speed]

2 stock, 4 minutes. I don't see a reason to change the current timer:stock ratio.

 

[Kal]

4 stock, 7 minutes is pretty common. The idea was to allocate one minute per stock. However, you realize that the match is over once seven stocks are lost, so you need to only allocate one per stock up to seven. The same is done here: one minute per stock, but whenever three are lost, the match is over.

 

[onionchowder]

I like this idea, but I think it's a bit overkill (3-stock, 5 mins, sets of 5, maybe? 3 is the default number of stocks anyway, and people like the number 5). Since sets are bigger, stage lists ought to be expanded as well, but I don't think the entire stage pool should have more than ~10 stages (in Bo5, ~12 in Bo7).

From the perspective of maintaining consistent in win outcomes (i.e. reducing volatility), large counterpick lists are actually very bad. Since large counterpick lists offer the counter-picker a larger advantage, the probability of an upset is actually larger with large counterpick lists, since the first match matters more and the other 4 matches (in Bo5) split 2-2 more consistently with more biased counterpicks.


Now that I think about it, long games probably hurts technically demanding characters (e.g. Fox), since they have to maintain a high level of technical precision for the entire match. Shorter matches and larger sets means less sustained intensity and more breaks.

 

[Kal]

We expect 3-stock, best of five to closely match 4-stock, best of three and 2-stock, best of seven. Keep in mind that, with a longer set-length, individual matches become less important, hence large counterpick lists aren't actually necessarily worse. It's not hard to come up with a scenario where a single loss on best of three guarantees the loss of the set, but that does not apply to best of five.

I also have some qualms with this idea that counterpick stages increase variance beyond what is normal. It's circular reasoning, because it assumes that these counterpicks are not legal in the first place. The counterpicker's advantage is also greatly exaggerated.

 

[onionchowder]

My point is that a 2-stock Bo7 with a large counterpick list is more volatile than a 2-stock Bo7 with a small counterpick list. We can modify counterpick list size independently of #stocks and Bo#.

On counterpicking; The idea is that the counterpicker's advantage is increased by a larger stage list relative to a smaller stage list, all else equal. It doesn't matter how big the absolute counterpicker's advantage is, all we care about as ruleset designers is how to get a relatively low counterpicker's advantage, since we have to work within our constraints.

I don't see where you think my reasoning is circular. Counterpicker's advantage increasing with larger stage lists is independent of what the current state of stages is.

 

[Kal]

My point is that a 2-stock Bo7 with a large counterpick list is more volatile than a 2-stock Bo7 with a small counterpick list. We can modify counterpick list size independently of #stocks and Bo#.

On counterpicking; The idea is that the counterpicker's advantage is increased by a larger stage list relative to a smaller stage list, all else equal. It doesn't matter how big the absolute counterpicker's advantage is, all we care about as ruleset designers is how to get a relatively low counterpicker's advantage, since we have to work within our constraints.

What advantage do I get by having Brinstar, Rainbow Cruise, etc., available for counterpicking? Maybe I'm missing the point, but it's not as though I have strategies available to me that are unavailable to you. What's stopping you from making the best character choice on any stage I pick?

I don't see where you think my reasoning is circular. Counterpicker's advantage increasing with larger stage lists is independent of what the current state of stages is.

The above part is not circular. I was referring more to the idea that there is increased variance if, for example, Brinstar is legal. "Peach should normally lose to Fox, but wins on Brinstar," is the type of circular logic I was referring to, which is what I assumed you meant when you mentioned "volatility." I perhaps misunderstood what you meant by volatility (I assumed you meant "variance"), so it would help if you explain what you mean by the term.

However, I think the above part is just wrong. The counterpicks are banned currently because they are either "too random" or give certain characters too strong an advantage. Since Cactuar explicitly mentions that he sees nothing wrong with a player being competent with a broad range of characters in this ruleset, it's silly to suggest that it gives the counterpicking player an advantage. In virtually every case, you can just use one of the characters who is advantageous on the stage your opponent chooses.

 

[onionchowder]

What advantage do I get by having Brinstar, Rainbow Cruise, etc., available for counterpicking? Maybe I'm missing the point, but it's not as though I have strategies available to me that are unavailable to you. What's stopping you from making the best character choice on any stage I pick?

I was operating under the assumption that players don't switch characters mid-set. The argument still holds if players only play a few characters and their opponent is aware of what secondaries they have, which I think is a realistic assumption.

The above part is not circular. I was referring more to the idea that there is increased variance if, for example, Brinstar is legal. "Peach should normally lose to Fox, but wins on Brinstar," is the type of circular logic I was referring to, which is what I assumed you meant when you mentioned "volatility." I perhaps misunderstood what you meant by volatility (I assumed you meant "variance"), so it would help if you explain what you mean by the term.

However, I think the above part is just wrong. The counterpicks are banned currently because they are either "too random" or give certain characters too strong an advantage. Since Cactuar explicitly mentions that he sees nothing wrong with a player being competent with a broad range of characters in this ruleset, it's silly to suggest that it gives the counterpicking player an advantage. In virtually every case, you can just use one of the characters who is advantageous on the stage your opponent chooses.

You're right, when I say "volatility" I basically just mean "variance". I see now what you meant by circular reasoning, but that issue is eliminated when you change the way you view things.

The idea is that the "better" player ought to win more consistently. When the rules are changed, the game is fundamentally different, and the metric of "better" is changed slightly (e.g. Peach players are relatively better in a ruleset where Mute City is legal than in the current ruleset). Assuming that for any given ruleset, there is an objective, well-ordered metric of "better" (that is, we can theoretically take all players and rank them), then our measurement of volatility is how consistently the "better" player wins. The "better" ranking can vary by ruleset, but as long as it exists, we can analyze the volatility of the ruleset.

One issue with the theoretical objective ranking is that it is basically impossible to measure due to matchup ratios. Also, it probably doesn't exist, since playstyles can behave in a Rock-Paper-Scissors fashion. However, I think the idea of volatility it presents is still robust.

 

[Kal]

I was operating under the assumption that players don't switch characters mid-set. The argument still holds if players only play a few characters and their opponent is aware of what secondaries they have, which I think is a realistic assumption.

Yeah, the assumption underlying the MBR ruleset is basically that your character choice is fixed for a set, and so they select stage to minimize the sort of variance you would see in matchups like Fox vs. Marth and Peach vs. Falco. However, top players generally adapt to their environment, even with regards to the characters they play: Ken picked up Fox to play against Azen, and Armada picked up Young Link to play against Mango. I don't think it's unreasonable to expect that, with a ruleset with several stages, good players would learn to have enough characters to account for all of them.

The idea is that the "better" player ought to win more consistently. When the rules are changed, the game is fundamentally different, and the metric of "better" is changed slightly (e.g. Peach players are relatively better in a ruleset where Mute City is legal than in the current ruleset). Assuming that for any given ruleset, there is an objective, well-ordered metric of "better" (that is, we can theoretically take all players and rank them), then our measurement of volatility is how consistently the "better" player wins. The "better" ranking can vary by ruleset, but as long as it exists, we can analyze the volatility of the ruleset.

The problem here is that the measurement of "better" is usually based on the probability of winning in the first place. What does it mean to suggest that I am better than my opponent if to suggest anything other than that my chances of beating him are greater than 50%?

You could try and analyze variance, but it becomes quite muddled. However, sans randomness, I don't think there can be any disagreement that consistency is unaffected. So, you might be able to make a case that Brinstar, Mute City, etc., can impact variance. But Rainbow Cruise, Pokéfloats, etc., seem to decidedly not impact variance at all.

I would further (though separately) argue that Brinstar, Mute City, etc., have so little randomness as to negligibly impact variance. Really, only in extreme cases do we see randomness have a significant impact on the consistency of results. Things like items-on, Icicle Mountain, and Flatzone have impact consistency quite a bit. Things like Brinstar, Mute City and Corneria, noticeably less so.

A final remark: the consistency of results alone do not justify anything. We could have an incredibly consistent ruleset by playing only Marth dittos on Final Destination. However, we often allow for increases in variance in the interest of increased depth. In general, to suggest that something should be banned solely because it increases consistency, without even arguing how much, will not suffice.

 

[da K.I.D.]

Current one is conducive to me being good which is likely a flaw of the system.

I love you KK, im tempted to make this my sig.