The Unity Ruleset: Discussion

[Supreme Dirt]

My friend, Nysyarc, brought up how in a local tourney they tried a different way of doing bracket in which losers bracket became pools of 4 people where 1 person would advance 2 rounds of a bracket. It basically just adds 1 extra set to be played per pool done here is an example:

[collapse=Pretty large image]

[/collapse]

This would be beneficial as it lets lesser players get more tourney matches while not affecting winner's bracket at all besides maybe the incentive to not get sent into loser's bracket. It also makes results more accurate as well.

Time wise it shouldn't really be that much of a difference because its only 1 extra set per pool. Since its a pool, people should be gathered already so they don't need to be called to the TV more than once per pool while in bracket, you would have to call them twice as there are 2 rounds of losers to go through.

What do you guys think?

It's not 1 extra set per pool. lrn2math. It's 3 extra sets.

 

[Ussi]

This is what I get for just repeating the guy instead of actually counting the sets myself.


Still, what do you think of that format?

 

[Supreme Dirt]

Tin's using it at a tourney this Saturday, so I'll get to try it firsthand. So I'll save my judgments for after I've experienced it.

 

[Tesh]

That sounds pretty cool, but according to the picture, people that lose in WR1 still can be eliminated by losing in LR1 right? That would still kind of suck for lesser players right? Like the guy listed as silent swag is still 0-2 then out.

 

[Supreme Dirt]

Well, it's good for locals, not for like a national or anything, since the person coming from loser's to GFs has played (possibly significantly) more sets than the person coming from winners.

 

[Ussi]

That sounds pretty cool, but according to the picture, people that lose in WR1 still can be eliminated by losing in LR1 right? That would still kind of suck for lesser players right? Like the guy listed as silent swag is still 0-2 then out.

It counts as losing either LR1 or LR2 cause in the pool you'll get 1st, 2nd, 3rd, or 4th, 3rd and 4th is essential getting 0-2 and 2nd is going 1-2 in bracket (winning round 1 losers but losing round 2) Those who would go 0-2 in bracket still get 2 more matches in than they would have.

Each pool is 2 rounds of losers bracket

 

[Zankoku]

Actually Ussi, the GSS system did not use any pools for LR1. It starts at LR2, so going 2 and out still happened.

 

[Flayl]

Seems pretty pointless to me

 

[Ussi]

Actually Ussi, the GSS system did not use any pools for LR1. It starts at LR2, so going 2 and out still happened.

Oh wow =_=;;

Well then losers round 2/3 is in pools basically then :s

oh well, nevermind me then.

 

[infiniteV115]

Seems pretty pointless to me

The reason our TO (Tin Man) is doing it is because it leads to more accurate results.

To explain how this system works, when you are in the losers bracket, instead of playing 1 person at a time, you will be placed in a pool of 4. You will then play everyone in your pool and only 1 person will advance. I believe that this system is worth a try due to the amount of benefits it promises. The biggest advantage this system offers is the improvement in accuracy of the results. This is what the results would look like if 32 players entered the tourney.

[collapse=Placings]1.
2.
3.
4.
5.
6.
7.
7.
9.
9.
11.
11.
13.
13.
13.
13.
17.
17.
17.
17.
21.
21.
21.
21.
25.
25.
25.
25.
25.
25.
25.
25.[/collapse]

If anything, I figure it's worth a try. I believe the accuracy that these results present is extremely valuable. Now there is a top 5! Now there is a top 10! This also allows people the opportunity to play more matches in the losers bracket. I believe this would be good in developing the scene since people will have more exposure to playing others in a tournament setting. It also removes some instances of a "lucky bracket" since people will be required to play more sets than before to advance. Even though it still is possible to dodge some players, compared to the standard double elimination bracket, there's no escape!

 

[Tesh]

^ I still dont get how its good for the absolute worst players around.

 

[John12346]

Well, it does allow them to play a minimum of 4 matches(lose once in winners, then play 3 pools matches in losers), instead of the meager 2 matches that they're allotted per tourney(lose once in winners, lose once in losers), so that's something right there.

 

[Tesh]

Well, it does allow them to play a minimum of 4 matches(lose once in winners, then play 3 pools matches in losers), instead of the meager 2 matches that they're allotted per tourney(lose once in winners, lose once in losers), so that's something right there.

Actually Ussi, the GSS system did not use any pools for LR1. It starts at LR2, so going 2 and out still happened.

this is what i was talking about.

its an interesting system for small locals or even more viable if 1 or 2 stock matches were used, but...

i guess you could filter those people into an epic fail pool and the winner places.... just above last.

 

[tekkie]

is it easy to set up in tio though?

 

[Zankoku]

The reason the pools didn't start until LR2 was because it's easiest to set up from there in tio - you just make the one person who won the pool the one who advances, and manually adjust the results listing.

 

[Tesh]

Thats an excuse, not a pro.

 

[infiniteV115]

Ok so it may not benefit the absolute worst players, but it doesn't hurt them either. It's the same as losers' bracket for them, and it's a pro to everyone else.

 

[Ralph Cecil]

Is there really much you can do for the worst player that doesn't involve giving them a handicap? o-o

 

[DeLux]

In regard to the losers pools proposal and testing, I feel it's important to consider the three values of consistency, accuracy, and logistics.

I think if we're looking to support accuracy of results, a round robin tournament would indisputably provide the best results while maintaining consistent win conditions throughout. However, it is usually logistically infeasible.

Because of the logistical reasons, most tournaments opt for a double elimination style bracket. It maintains consistency by monitoring losses as a means to be eliminated from bracket, but some of the accuracy might be lost due to seeding.

The system proposed seems to attempt to strike a medium between the two. At best, it is more accurate while maintaining the same level of logistical viability. At worst, it is more accurate while being less logistically viable. However, it loses out in the consistency value in terms of criteria needed to advance as the criteria between winners bracket and losers pools are remarkably different.

As a TO, I think consistency in terms of criteria for bracket advancement is a value that deserves consideration of being maintained. However, I would think that accuracy at the lower levels is a lesser value because the point of a tournament is to determine the winner of the event, otherwise we wouldn't award prizes to the top placers only and spurn the lower placements. Moreover, I feel like the one and done nature of the losers bracket gives it some of its competitive tension which is healthy for fostering growth.

That being the case, I don't think I'd support it in favor of the current system.

 

[Supreme Dirt]

Just popping in to say that after experiencing loser's pools and watching several friends experience round after draining round of them, they're a terrible idea in practice. It's basically saying "oh, you dropped into loser's? wait like 5 hours. Then you will never get another break again."

The way tournaments are done already is fine. Pools for large tournaments where seeding is necessary (though... considering who I ended up against in first round brackets... I seem to think seeding may always be necessary :/) and then straight double elimination brackets.

Though I wouldn't mind seeing how triple elimination worked out, but that'd be terribly complex.

 

[Supermodel From Paris]

LOL, I love it when people use the word "accuracy" in tandem with the word "results." If you already know what the results will be before the tourney starts, why hold it at all? The point is the find the best player. This is part of the reason skill-based seeding is broken and why no fighting game community but Smash uses it.

"More accurate results?" Seriously? The results are what they are. You can produce a set of different results by using a different seed, but they will never be more are less "accurate."

Can the URC discuss the possibility of standardizing location-based seeding so we can do away with this broken understanding of tournaments once and for all?

 

[Supreme Dirt]

Exactly. I mean, I would've been fine had I been seeded against Suzaku with random seeding, but being seeded against him based on "skill" just feels like being given the finger. I mean seriously, this is how bad you think I am?

To have improved lots and then get seeded poorly, and non-randomly, just hurts.

Also location based seeding. Like wtf yesterday... So much BS. People having to fight way too many other people from their own crews.

 

[Flayl]

Professional sports use seeding all the time >_>

I mean yeah I get that if it's a local there might not be much spectator value so seeding might not matter much, but it's downright dumb not to seed regional/nationals because it can and does kill hype when top level matches happen early without an audience.

Also depending on the progression of skill level, not seeding means somebody might earn prize money without "earning it".

 

[Supermodel From Paris]

Sports have a basis for seeding though, as do certain smash nationals (usually regional TOs that agree to keep track of seeding points for APEX, Genesis, etc.). We have a basis to seed those events rather than just assuming that X player is bad and Y player is good and sticking potential top 10 placer into a pool with Ally and Shugo.

More importantly, spectator value isn't really the point. Finding the best player is.

Often when I make this point, I get a reply from some high-level player saying he'd feel robbed if he had to face Ally in WR2, but guess what hot shot, we all feel robbed. That's the point. Unless you have some kind of pre-existing numbers to prove that you deserve to have your "unfair" matches postponed a few rounds, you don't. Save that kind of seeding for events that are part of some kind of series in which players have a real numerical value, because otherwise from the time you walk into the venue until your first match you ain't ****.

 

[Supreme Dirt]

It's also a confidence issue.

I mean seriously, if you know matches are seeded based on Skill, and are facing Ally round 1, that means that the TO considers you the single worst player in the room. Not so bad if it's earned over a series, but...

And a TO doesn't always know how good a player is. I mean... say at a MK unbanned tourney, I'm known as a Dedede, and decide to go Metaknight that tourney, say I beat players I "shouldn't" because of Metaknight, ****ing up the seeding, since if you're seeded above the halfway point, you're expected to lose your first set. Things like that can't be accounted for. For all a TO might know, what they've seen of a player was just a secondary, or even a just-picked-up character, and then they base their seeding off of that, and then it turns out, with their main, they're actually a really good player.

tl;dr, a TO can't properly account for the skill level of every single player at a tournament.

 

[Supermodel From Paris]

And on top of that, I've seen completely disgusting negligence and nepotism in tourney seeding I was present for. My favorite is when the TO leaves the room for a minute and asks his helper to finish seeding for him. and suddenly his helper has like six of his "trusted friends" over the laptop skill seeding.

Unreal.

 

[Ussi]

Seeding by skill I would assume to just have top 4/8/16 (depends how big the bracket is) be seeded and the rest be random.. not everyone be seeded by skill basically, but having known good players not have to play each other too early. But upsets still happen.

 

[Zankoku]

I seed by TrueSkill.

 

[Supreme Dirt]

I seed by TrueSkill.

I am okay with this.

 

[Budget Player Cadet_]

Just popping in to say that after experiencing loser's pools and watching several friends experience round after draining round of them, they're a terrible idea in practice. It's basically saying "oh, you dropped into loser's? wait like 5 hours. Then you will never get another break again."

This reminds me of what happened at my tournament... I ran Bo7 GFs instead of Bo5, and had to watch 14 games of back and forth, most of them Snake vs. Samus.
Never. Again.

The way tournaments are done already is fine. Pools for large tournaments where seeding is necessary (though... considering who I ended up against in first round brackets... I seem to think seeding may always be necessary :/) and then straight double elimination brackets.

Larger tournaments only? What joker is saying that crap? Look, I'll be frank: regardless of how small your tournament is, if you have time, absolutely absolutely absolutely do pools.

I did 4 pools in about 2 hours at my tournament last weekend... We had 26 people, and we only eliminated 10 people from the bracket. I still consider it worthwhile. Why? Because without it, worse players get 4-6 games... and then they're done, and get to wait for the tournament to finish. I remember KTAR6, and I am very glad I made it fairly far in the bracket*– there was no time or space for freeplay anywhere... and this isn't that rare. Meaning that lousy players will spend a long time sitting around on their *****, waiting for something to happen. While a veteran player might be willing to put up with it because of friends in the community and whatnot, a newbie is going to look at his $15 entry fee and think, "Wait... I just played a grand total of 4 matches, and now I may have to go... What a waste of time and money."

Seriously guys – do pools.

Though I wouldn't mind seeing how triple elimination worked out, but that'd be terribly complex.

Lol 3D bracket images

LOL, I love it when people use the word "accuracy" in tandem with the word "results." If you already know what the results will be before the tourney starts, why hold it at all? The point is the find the best player. This is part of the reason skill-based seeding is broken and why no fighting game community but Smash uses it.

All right, take these results from a hypothetical 8-man tournament.

1. Nairo
2. Random1
3. M2K
4. Sphere
5. Ally
5. Anti
7. Cadet
7. Random2

Is that accurate? Let's assume that the bracket was set up so that Random1 only had to beat me and random2, and all of a sudden he was in the finals. Look, I get that there is a certain level of bias involved in seeding. But this bias is without reason. Just hypothetically, which of the following situations is worse:
1. An above-average but unknown gets seeded low in the bracket. He has to face up against a top player early on and gets his *** kicked, but then starts doing real work in losers. He then gets seeded higher at the next tournament because we've learned better.
2. R1 Winners: M2K vs. Ally, Ramin vs. Glutonny. R2 Winners: M2K vs. Glutonny. R2 Losers: Ally vs. Ramin. Or, alternatively: Ramin, Glutonny, M2K, Ally, ADHD, and Dabuz in the same pool.

Excuse me for being a contrarian here, but I'd argue that situation 2 is a lot worse for competition. The reason we run double elimination (and less often than we should, pools) is very similar: it's to avoid such randomized results. Seeding is both necessary and useful.

Also, when you say that other fighting game communities don't use seeding... You mean there's a chance of something like Daigo vs. Jwong round 1 at EVO? Because that would be ********.

 

[Supermodel From Paris]

All right, take these results from a hypothetical 8-man tournament.

1. Nairo
2. Random1
3. M2K
4. Sphere
5. Ally
5. Anti
7. Cadet
7. Random2

Is that accurate? Let's assume that the bracket was set up so that Random1 only had to beat me and random2, and all of a sudden he was in the finals. Look, I get that there is a certain level of bias involved in seeding. But this bias is without reason. Just hypothetically, which of the following situations is worse:
1. An above-average but unknown gets seeded low in the bracket. He has to face up against a top player early on and gets his *** kicked, but then starts doing real work in losers. He then gets seeded higher at the next tournament because we've learned better.
2. R1 Winners: M2K vs. Ally, Ramin vs. Glutonny. R2 Winners: M2K vs. Glutonny. R2 Losers: Ally vs. Ramin. Or, alternatively: Ramin, Glutonny, M2K, Ally, ADHD, and Dabuz in the same pool.

Excuse me for being a contrarian here, but I'd argue that situation 2 is a lot worse for competition. The reason we run double elimination (and less often than we should, pools) is very similar: it's to avoid such randomized results. Seeding is both necessary and useful.

Also, when you say that other fighting game communities don't use seeding... You mean there's a chance of something like Daigo vs. Jwong round 1 at EVO? Because that would be ********.

This scenario is statistically impossible. If Nairo placed higher than Ally using location-based seeding then he was just a better player, statistically.

See this thread:

http://www.smashboards.com/archive/index.php?t-302722.html

You can see that ankoku ran a few simulations to see what would happen. None of the results came even close to being that different from the results we "expect" (whatever that means).

The flaw in your argument is more or less that you think you already know what the results are and you're trying to get things as close as possible to those results, using our previous tournament results as a broken metric. You have got to be smart enough to understand the flaw in that logic. Please tell me you are.

 

[Zankoku]

It's not just statistically impossible, it's just simply impossible because even the worst-seeded double elimination bracket in existence should not involve a random achieving top 2. Random1 would've had to either defeat Nairo in winners' finals or defeat Mew2King in losers' finals to have gotten that result.

I believe tournaments like MLG and EVO seed by points gained through either previous circuit performance (MLG) or ranking tournaments similar to Road to Apex events (EVO). However, MLG has the second most convoluted bracket system I've seen (recently trumped by GSL's latest and greatest) and EVO's "pools" are in reality just isolated double elimination brackets where the top 2 move on.

 

[Budget Player Cadet_]

That 8-man was single elimination. But still... Random at 3rd is already pretty bad in a bracket that has 4 incredible top players.

This scenario is statistically impossible. If Nairo placed higher than Ally using location-based seeding then he was just a better player, statistically.

As was Random1. Which, as I'm trying to make clear, is the problem.

 

[Zankoku]

It's absolutely impossible if you made it single elimination because you'd then end up with four players tied at 5th, rather than two at 5th and two and 7th. There's also the issue of tiebreaking the two people tied for 3rd. Your examples suck.

 

[infiniteV115]

"Any tournaments that do not follow this ruleset will not be eligible for stickies on SWF or featured coverage on AllisBrawl."

Just making sure, this was a decision made by the administrators/site-runners (I don't know what to call them) of SWF, and not the URC, correct?

 

[[FBC] ESAM]

Correct. I mean you can think of it in a certain way considering AZ is part of the admin group, but that piece of permission was granted before the URC even existed.

 

[infiniteV115]

Alright, thanks.

 

[T0MMY]

Edit: Misread, w/e

I'd like to bring up that one story that t1mmy/t0mmy(forgot who) made where their ROB ultimately lost to a MK because he was forced onto the ledge so many times, despite having the percentage lead. Right there is a good case of a character with LESS jumps and no intent of timing out the opponent, but still exceeding the LGL. What would you say about that?

That was me (I play R.O.B., t1mmy plays the Kirby characters).

Keep in mind, I had a lead the ENTIRE game (even stock leads). Neither of us were employing dirty tactics (like planking), yet the timer ran out.

The rules did not state how a tie would be broken, nor did the posted rules have an LGL listed. By expectation I would win due to %-lead, by default I would win because I won the Sudden Death (which I always play out).
By every reason I should have won... BUT!...

Because there was an LGL rule used SOMEWHERE in the country, the WA players used that reason as to claim that I lost ( it was mob-rule there; the TO did not make any decision, they simply entered it into the bracket as my loss and refused to hear me out).

Thanks to some arbitrary out-of-game rule thrown onto the tournament scene using dirty politics I lost money to pay for my trip, I dropped in PR, and the panelists in their PR used it against me for an entire year to keep me below other WA players (is it any wonder I am against panelist-based PRs?)

So I ask how this LGL rule, which was developed to keep MK from playing dirty, helps the scene at all when it clearly has been exhibited to go against all semblance of fair play and rewards a lesser opponent a win (who happened to play... Meta Knight!)

As if it's not difficult enough to rightfully earn wins against a Meta Knight with R.O.B. as it is, but I have to contend with out-of-region politics being thrown against me? haha ([^^]


TL;DR: ledge-grab limits mess with the meta game, are arbitrary out-of-game rules implemented by dirty politics and need to be removed. I'd rather lose to a Meta Knight using dirty in-game tactics instead of losing to dirty out-of-game politics.

 

[Grim Tuesday]

lol at sudden death meaning anything

 

[Life]

LGL would be unneeded if we went to one-stock food.
>.>
<.<
<_>
*flees*