Is there any data proving that 2 stock is actually faster than 3 stock?

[Thundering TNT]

I know that in theory, 3 stock should take longer, but a 2 stock game with a lower timer seem more likely to time out more, due to camping being easier. Is there any data showing if either of these are true?

 

[Amazing Ampharos]

Thinkaman did a good study into this early in the game's lifespan. I believe his conclusion was that 3 stock games take less time per stock but a bit more time overall than 2 stock games, but this was quite a while back and the meta has developed since then. Unfortunately, I can't find the thread since smashboard's search feature has the critical flaw of making it oddly hard to find all threads in a particular sub-forum made by one user. Either way, in practice most tournaments that run late do so for non-game related reasons (poor set-up utilization, insufficient set-up count, long between game delays, inefficient pool sequencing, allowing bracket bottlenecks, etc.); the argument of 2 stock tournaments to save time is and has always been a strange one.

 

[RedNova]

Thinkaman did a good study into this early in the game's lifespan. I believe his conclusion was that 3 stock games take less time per stock but a bit more time overall than 2 stock games, but this was quite a while back and the meta has developed since then. Unfortunately, I can't find the thread since smashboard's search feature has the critical flaw of making it oddly hard to find all threads in a particular sub-forum made by one user. Either way, in practice most tournaments that run late do so for non-game related reasons (poor set-up utilization, insufficient set-up count, long between game delays, inefficient pool sequencing, allowing bracket bottlenecks, etc.); the argument of 2 stock tournaments to save time is and has always been a strange one.

And with that said, do you think it would be plausible to do tournaments with 3-stocks?

 

[cardboardowl]

And with that said, do you think it would be plausible to do tournaments with 3-stocks?

Given that tons of tournies do 3 stock, yes

 

[BestTeaMaker]

I'm still collecting match data for 2 vs 3 stock. It's stopped for a while because of life, but hopefully, I should be able to get back to watching matches and timing how long it takes.

The early analysis of 2 vs 3 stock is that, at most, it takes about 40 seconds-1 minute more on average, which honestly is not a lot of time.

 

[HeroMystic]

And with that said, do you think it would be plausible to do tournaments with 3-stocks?

Tons of tournaments still do three stock, especially when you account for non USA regions.

 

[Thundering TNT]

I'm still collecting match data for 2 vs 3 stock. It's stopped for a while because of life, but hopefully, I should be able to get back to watching matches and timing how long it takes.

The early analysis of 2 vs 3 stock is that, at most, it takes about 40 seconds-1 minute more on average, which honestly is not a lot of time.

I'd love to hear where you got your data from. What tournaments were these results taken from?

 

[RedNova]

Tons of tournaments still do three stock, especially when you account for non USA regions.

That's one of the reason I asked. It seems like there's a lot of tournaments also running 3 stock, alongside 2 stock. I was wondering if it would be possible to decide in one universal ruleset? At least stock wise

 

[clydeaker]

After doing a few quick tests I found for every one stalk there should be three minuets. Two minuets per stalk like Melee and Project M tournaments seems to be slightly to short for smash 4 in my opinion. I think smash 4 battles should take about as long as Melee, Brawl, and Project M tournament matches are, witch is about 8 min per match. Over all I think 1v1 Smash 4 matches should be set to 3 stock for 8-9 minuets.

 

[LiteralGrill]

I'm still collecting match data for 2 vs 3 stock. It's stopped for a while because of life, but hopefully, I should be able to get back to watching matches and timing how long it takes.

The early analysis of 2 vs 3 stock is that, at most, it takes about 40 seconds-1 minute more on average, which honestly is not a lot of time.

Over the coarse of a tournament, 1 extra minute per match can easily add up to extra HOURS of tournament time. It looks small at first, but larger in practice.

 

[chipndip]

The rule of thumb is rather simple:

Since a two stock match should usually never take longer than a three stock match (I mean, we can act like number crunchers all we want, but at the end of the day, less chances is less chances), a larger tournament should obviously invest in a two stock format, while smaller ones have the option of a three stock format. It's not complex, really. If we're going for an eventual "golden standard", two stock is more applicable because you run less of a chance of over-extending into the wee hours of the night...or morning.

Not trying to sound ignorant or arrogant, but the meta of "I have less stocks. I gotta play safe." doesn't just disappear when you're on 3 stocks. It only keeps itself under wraps until the first KO, so it's not like people get more reckless out of nowhere, save for outlier situations.

 

[FooltheFlames]

I think we should switch back over to 3 stock, everyone knows Brawl is more slower than Smash 4 and it had 3 stock going almost exclusively. It would still be faster than that atleast.
And I think that the first player to get a kill in a two stock match gains too much a lead when the skill levels are about the same for each player, regardless of the Rage effect. So tourney matches are becoming who can get the first kill instead of what it could really be.

 

[Oracle]

The main reason ppl use 2 stock is that logistically, its nearly impossible to have multi game events including 3 stock smash 4. 3 stock s4 takes longer than melee/pm, which affects how you can budget your time in a major. If I run waves of pools, then the smash 4 pools will take longer than melee/pm, which heavily hurts my ability to run the tournament on time.

For solo s4 events, or events focused on s4, 3 stock isnt that big of a deal, but most majors have s4 as well as other games, which makes 3 stock unfeasable. I think post evo it will be worthwhile to test 3 stock and see how long it lasts, but for now there isnt anything wrong with 2 stock

 

[Raijinken]

Over the coarse of a tournament, 1 extra minute per match can easily add up to extra HOURS of tournament time. It looks small at first, but larger in practice.

Adding total hours to a tournament can be helped tremendously by having extra setups. Numbercrunching to follow assumes: Everyone shows up to their round on time, each round starts sequentially (I add 1 minute to the total round time to make the round length reflect setup), each round goes to time but doesn't go to tiebreaker, each set is played to its maximum length.

2s6m with one minute downtime
A 30-person tournament takes either 58 or 59 sets (depending on bracket reset in grand finals), of which 3-4 are Bo5.
A Bo3 set takes 21 minutes, and a bo5 set takes 35 minutes. Total tournament hours: Roughly 21 hours 40 minutes

Simple rulechanges at a worst-case scenario:
Bo3 takes 27 minutes, Bo5 takes 45 minutes. 28% increase in time required.
Estimated 27 hours 45 minutes total tournament hours.

Let's assume something slightly less than a worst case scenario. Let's assume that no match ever goes to time, but all sets go to maximum. Thus, new assumptions are:
2s6m match lasts for 4 minutes plus setup, 5 minutes total. Each Bo3 lasts 15 minutes, Bo5 lasts 25 minutes.
3s8m match lasts for 5 minutes (this is the higher end of the estimated time difference), plus setup, 6 minutes total. Each Bo3 lasts 18 minutes, Bo5 lasts 30 minutes.
(Please note that I think both these estimates are high, but since certain matchups go long and others don't, we'll use this for calculation)
Now a 2s6m tournament takes 15 hours, and 3s8m takes 18.5 hours.

However, this assumes we're running a single string (or setup, queue, thread, core, however your studies prefer to grasp a process). This number is divided across the number of setups to estimate the total time a tournament would take.

A 2s6m tournament under the "long but not worst-case" assumptions above, on three setups, takes approximately 5 hours to run to completion. The 3s8m takes a bit over 6 hours.

If you add just one more setup to the 3/6 tournament, and everything else is run exactly as assumed, you're looking at around 4.5 hours. One setup.

The key is to multithread (or multitask or multiqueue) to the appropriate level to make as many different processes run as possible. Of course, this is harder when setups are limited due to a small or unwilling community.

However, this also results in the tournament being longer because more Smash is being played. Assuming your tournament exists so that people can play Smash, this is what we call "value-added time." Contrast people being late to matches, coaching, fingerwarmers, and so many other things that are "non-value-added" time.

In other words, if you adjust setup count appropriately, you can have slightly longer (and in my opinion, for other reasons, better) matches, for comparable runtime.

Don't blame stock count for things running over when it's the easiest thing to fix.

 

[LiteralGrill]

In other words, if you adjust setup count appropriately, you can have slightly longer (and in my opinion, for other reasons, better) matches, for comparable runtime.

Don't blame stock count for things running over when it's the easiest thing to fix.

Yes adding more setups as well as running tournaments more efficiently can help, I realize that, I just wasn't happy with someone trying to trivialize that 1-2 hours added to the tournament by presenting it as 1-2 miinutes a match (as one makes it look much smaller than it really is). Not all events have access to lots of setups though, and even some majors don't manage to really have enough setups sometimes so I can't always blame them for just going to 2 stocks. Both work in the end and neither have presented major issues so if a tournament picks either it won't really bother me.

 

[Shouxiao]

If an extra min was not added to the current ruleset(basically For Glory + 1min) many matches would simply time out or be close to timing out. One could argue that with the current format if one stock goes on for awhile then all the person with the stock lead has to do is camp/stall the rest of the match.

 

[Raijinken]

If an extra min was not added to the current ruleset(basically For Glory + 1min) many matches would simply time out or be close to timing out. One could argue that with the current format if one stock goes on for awhile then all the person with the stock lead has to do is camp/stall the rest of the match.

To be fair, that's exactly how it always is. But that's been discussed to death elsewhere. The only way to prevent that is to remove the option as a win condition. It's a bit less likely to happen with 3/8 (but as or more likely with 4/8). The issue is that the match has to be forced to end some time, but that time has to be long enough to discourage stalling.

 

[Oracle]

@ Raijinken that argument is pretty fallacious because stock count isnt related in any way to how many setups you have at a venue. One ruleset wont give you any extra setups, so saying "the extra hour will go away if we have more setups" is silly. Where are those extra setups gonna come from?

Youre gonna always have the same number of setups at a tournament regardless of stock count; there isnt a whole lot a TO can do about that.

 

[Raijinken]

@ Raijinken that argument is pretty fallacious because stock count isnt related in any way to how many setups you have at a venue. One ruleset wont give you any extra setups, so saying "the extra hour will go away if we have more setups" is silly. Where are those extra setups gonna come from?

Youre gonna always have the same number of setups at a tournament regardless of stock count; there isnt a whole lot a TO can do about that.

I'd think it clear that if the tournament can't handle the extra time, they should be more than welcome to cut the time down in any way they choose.

Obviously there's no way that, match per match, 2 stock would be longer than 3 stock on average. It's simply not how the math works, and from that stance it's a pointless debate.

But there are ways to address the additional time in a tournament, which is what I was meaning to address. I've got no clue how TourneyLocator is run logistically, but obviously if you've got no way to increase your setup count, going with the shorter rounds is a quick time fix that sacrifices relatively little.

On the other hand, if a community or venue is able to come by extra consoles (whether from players or a slice of venue fee), it seems reasonable to me to do so in order to more efficiently run what is, at least in my opinion, a better way to play the game.

We've hopped pretty far from the wording of the OP, though. 3-stock matches are, on average, longer. Tournaments are an entirely different matter. I suppose I inadvertently set up a bit of a straw man since I was attempting to explain a means of solving tournament length, not match length. And for that, I apologize.

 

[FooltheFlames]

"Proof" that when both players know both their own main inside and out, and are extremely familiar with the match up-
3 stock can be faster than 2 stock!

I say lets just give the Smash 4 metagame a few years when everyone knows everything and we may be returning to 4 stock matches... who knows?

 

[Shouxiao]

"Proof" that when both players know both their own main inside and out, and are extremely familiar with the match up-
3 stock can be faster than 2 stock!

I say lets just give the Smash 4 metagame a few years when everyone knows everything and we may be returning to 4 stock matches... who knows?

I like fast pace matches like that. I ultimately see 3 stocks being use later.

 

[JoeR]

I agree. 3 stock may come in later once the meta develops more, but for now 2 stock keeps things a tad faster and more exciting for viewers. 2 stock prevents very lopsided matches I feel that get boring quick. Also, who doesn't like quick tournments that then leave time for more play? Friendlies go quicker too so everyone plays.

 

[Khao]

The one thing I generally don't understand why 8 minutes is the standard for Melee, but somemtimes seen as plain unnacceptable for WiiU/3DS.

 

[Pyr]

The one thing I generally don't understand why 8 minutes is the standard for Melee, but somemtimes seen as plain unnacceptable for WiiU/3DS.

Because this isn't Melee and has different requirements?

Having a hard cap on match time keeps the overall average game time lower. Game time is a very real issue that needs to be dealt with. It was much less so in Melee's growing days.

 

[Jehtt]

Guys, remember that most tournaments have more than one match going on at a time. If a minute is added to every match in a 30 man tournament, it's not going to add exactly thirty minutes.

 

[Sixfortyfive]

The one thing I generally don't understand why 8 minutes is the standard for Melee, but somemtimes seen as plain unnacceptable for WiiU/3DS.

Old habits die hard.

It's easier to force new rules onto a new game than it is to do so for an entrenched, decade-old game.

 

[Shouxiao]

A bit of a side note but when a 2 stock 6min match last nearly to time it does get the ResidentSleeper(twitch.tv face) feeling. From a viewers point seeing more aggressive play from a 3stock match would make things more exciting. Once one stock is lost generally players have to play defensively.

Not saying defensive play is bad. When things are too defensive for a good portion of the match it can be dull. Well I expect things to be defensive if both players are on last stock at high %. I know some characters need more defensive play but it feels like after the first stock is lost the amount of defensive play shoots up too much.