wtf, these arguments against 3 or 5 stock matches are just ********.
"4 stocks is good because everything else doesn't feel right"
Why do you THINK anything else doesn't feel right? A few of you are just leaving it at that, so that makes me think you aren't even considering that 4 stocks feels right because that's the way you usually play.
"if it weren't for 4 stocks, amazing combos/combacks that we have seen would have never happened"
Lol, first of all, so what? we're trying to find the best way of testing our skills against other players in tournament, and are speculating as to why 4 stocks was chosen and why it is the best way. If we can't come up with a good reason to choose 4 stocks over any other number, it's safe to conclude that we just decided on the number without any more consideration than "it just feels right." Of course, that's not enough justification when we're dealing with srs business (money's on the line). Amazing combos/comebacks are completely irrelevant.
Now, obviously, in all likelihood, you'd have other amazing combos, but one would never think that a small change in stock count which doesn't have anything to do with combos or comebacks (I'll explain in a sec) would cause there to be nothing similar. That would be like saying if any one factor that is irrelevant to the cause is different, then the effect is different. Like saying any small change in the past automatically causes negative effects cuz lolz the present is the best possible.
"you don't get amazing comebacks with anything other than 4 stocks"
or
"it would be harder/easier to comeback with anything other than 4 stocks" (this is actually correct but not for any reason that anyone's discussed)
you gotta be ****tin me. Obviously you would still have comebacks. Consider a ruleset where a set consists of two BO3s with 2 stocks. Consider any ****ing ruleset, do I even have to explain this? Whenever you have a lead, you can have a comeback. A change in stock count would not change this fact, so this argument is also irrelevant. More stocks would allow for larger leads, therefore less but more amazing comebacks. Less stocks would result in more but less amazing combacks.
"longer matches allow for a more accurate test of skill."
This is true. But we're not talking about match length, we're talking about stock count. To isolate the stock count variable and keep match length controlled you'd just have more rounds with fewer stocks or less rounds with more stocks. This also isolates what we're discussing from the time constraints issue (well, almost).
Okay, now those are out the way we can finally discuss what for some reason nobody has mentioned, which is what the effect the stock count has on the match. I thought that was the original question! The stock count controls round length. Since we play in sets under time constraints, the round length controls how many times large leads are reset. Whether the first match is a JV5 or a close match, the score is still 1-0.
This has serious implications. The rules of the tournament are affecting the gravity of any lead. As a TO, you control the degree of slippery slope in the game! That's probably better left to the game designers, who know more about the game than you, and if you add too much or too little slippery slope, you could **** up the game! Too little slippery slope means there's no advantage to the guy in the lead, and a comeback could occur at any moment. This is Brawl's problem. There is very little slippery slope in that a hit in Brawl doesn't increase your chances of another hit as often as in Melee. Low slippery slope makes decisions feel less meaningful as there will be no long term consequences for successes or failures. It makes you feel like you always have a fighting a chance -- but never the security of a lead. High slippery slope helps keep the guy in the lead in the lead. Comebacks are more rare and more amazing as a result. Decisions at first are very meaningful, as the any mistake will have dire consequences later on. This is evident in Starcraft, where losing any amount of mining workers puts you way behind in economy later on. Any "random" occurrence or mechanical error may make players feel cheated, as those errors will probably make them lose, and they feel that their opponent did not defeat them, but rather it was random or they only defeated themselves. Near the end of the game, decisions are meaningless, as the game has already been decided by past occurrences. This is much like many games of Chess, where only a few pieces on one side remain, yet the game continues with the king hopelessly getting in and out of check until a player concedes. While decisions are more meaningful with high slippery slope, the endgame is meaningless to play out, as the losing player has no fighting chance.
So the real question we're trying answer is "how far should the TO adjust the game's slippery slope in order to best test our skills?" With the above paragraph in mind it seems ideal to adjust the slippery slope so that decisions and leads always feel meaningful, but the losing player still always feels like he has a fighting chance. Fortunately, Melee already does this. With a high "tech skill" ceiling and deep nuances, and therefore heavy emphasis on these skills, differences in skill are usually significant enough for the players better at these skills to keep the lead no problem. However, those skills' emphasis is not so overpowering that players can afford lapses in thought or predicting the opponent correctly (well, in most cases); a player with superior (enough) mindreading and kill efficiency can still take the lead. Also, the game allows for low percentage deaths, which can cause leads to shift dynamically (negative slippery slope). The ability to gimp or kill efficiently, the high mechanical skill ceiling, and the meaningfulness of correct reads resulting in combos but eventually allowing a reset in the form of a new stock are what gives Melee it's perfect balance of positive and negative slippery slope. I'd say it's actually on the high side, but that's where the TO comes in...
As I explained before, playing in sets automatically causes negative slippery slope. We all like this slippery slope because it allows us to make up for errors due to "randomness" or self-inflicted errors. This is partly why long matches are preferred; it allows players to make up for their errors. Playing in sets is an easy fix. Any abnormal error like a pro suiciding 4 times in a round against an average player doesn't matter with the current ruleset. Sets also make the few correct decisions by the losing player not matter, however, which decreases the accuracy of result of the set.
The real reason we play 4 stocks then is that most fighting game communities and practically any competitive community and everyone that likes the idea of playing two outta three, i.e. everyone that likes a normalization of the lead after a certain point believes that you should have one mulligan (a mistake you decide just shouldn't count because it's not representative). It just so happens that the first round also allows you to figure out your opponent's skill level and play style. Since we like playing two outta three, someone decided that 4 stocks per match was all we had time for.
But if you wanted to change what's long been accepted as the norm, here the things to consider, as a summary:
-How long do you want the average set to be? Tournaments must be completed on time, so the time it takes must be predictable and short enough.
-How many times do you want to take the specific length of the lead and normalize it, causing negative slippery slope?
-How much do you want decisions (whether they be SDs or amazing combos) to matter?
-Do you want to allow things (counterpicks, stage selects, breathers) to happen more often that can only happen if you play more sets?
optional
-Do you want the game to be fun to watch? With high slippery slope your audience will leave before the game ends and with low slippery slope you won't have much of an audience.
-Do you want your game to be fun to play? High slippery slope causes random errors to be more frustrating while low slippery slope is boring.