8-Man Pools (New Style)

[S0FT]

This is not revolutionary and has been discussed before, but I don't think it was done in a clear and concise manner. I will give an example pool so everyone knows exactly what it would look like. Then I will list what I think the pros and cons of this system would be.

Swiss Style 8-Man Pools for Big Tournaments

8-Man pools

1st Round

Matches
P1 vs P5
P2 vs P6
P3 vs P7
P4 vs P8

Results
P1 1-0
P2 1-0
P3 1-0
P4 1-0
P5 0-1
P6 0-1
P7 0-1
P8 0-1

2nd Round

Matches
P1 vs P3
P2 vs P4
P5 vs P7
P6 vs P8

Results
P1 2-0
P2 2-0
P3 1-1
P4 1-1
P5 1-1
P6 1-1
P7 0-2
P8 0-2

3rd Round

Matches
P1 vs P2
P3 vs P5
P4 vs P6
P7 vs P8

Results
P1 3-0
P2 2-1
P3 2-1
P4 2-1
P5 1-2
P6 1-2
P7 1-2
P8 0-3


Pros
1. You have the same outcome every single time. There will always be a clear top 4 that advance to bracket.

2. It saves a lot of time because it cuts down on pointless matches (see point #3). It is also easier to plan:
----------------------------3 Rounds of 4 sets each
----------------------------4-Setups/pool
----------------------------Average set = 9-10 minutes
----------------------------Average pool = 30 Minutes
There is a clear order of whom each player plays against as opposed to a round robin where people just kind of play whoever and sometimes you wind up having to wait for one player to play 6 sets. That can never happen with this system.

3. A major problem with Round Robin pools is halfway through the pool you already have a good idea of who will make it out and who won't. So a lot of pointless matches are played and a lot of sandbagging occurs. In this system every match matters (except for 3rd round P7 vs P8). So sandbagging can end up really costing you.

4. It is really simple to write on paper. No more writing out pool sheets with the graphs.

Pro Summary: Once understood it is really simple to implement, saves time, gives you a reliable top 4, and consistent results every time (which is really important for a major, no more weird tie breaker shenanigans)

Cons
1. Although there is a clear 1st seed, 2nd - 4th seed are tied. So seeding a bracket may be harder.

2. Would need at least 1 person in each pool who understands the system. For some reason a lot of people do not understand how swiss works.


TLDR: Can we re-evaluate what swiss pools would actually look like, both the pros and cons in a clearer more concise manner.

Discuss....

 

[C.J.]

Yeah, swiss format- I'm not sure why people do RR. It's really commonly seen in MtG and other games like that.


As for tiebreakers being an issue- just do exactly what Magic does. Base it off of the records of the people you beat/lose to. There's a formula so it's objective/consistent as well.

 

[FrootLoop]

^ there's a formula for tiebreakers in RR too.

obviously RR is more consistent, as RR top 4 is more reliable than Swiss top 4.
4 setups/pool is a lot to expect.

Not saying swiss is bad, but you didn't mention these things.

 

[S0FT]

I feel like pools right now are this very muddy process. There are to many factors and when weird tie breakers occur not many people know the rules. With swiss it is very clear: If you want 1st seed then you need to win 3 sets, if you want to make it out you need to win 2 sets. There are no exceptions. Also if pools use this simpler system they can now follow a consistent formula. Which would mean TDs would have an easier time determining the amount of time its going to take, given player count and how many setups there are.

FrootLoop:
4 setups/pool is the maximum number of setups it would take per pool. The reason why that is important, is now TDs have a more reliable estimate of how long pools should take given player count and the number of setups. Right now there is no reliable way of determining how long pools will take under RR (that I know of).

The problem with tie breakers in RR is that the tie breaker can be for determining who receives 5th, 4th, or 3rd place (who will make it out and who won't), and different TDs have different rules for determining what breaks ties, and sometimes it is not even mentioned in the tournament rules. So once again most of the time it is not clear exactly what needs to happen to make it out of a pool. As opposed to a swiss where if you win two sets you will always be in the top 4. If you lose two sets you will never make it out. Which to me means that a swiss top 4 is actually a more reliable method for determining who the top 4 will be.

 

[FrootLoop]

Would swiss not factor head-to-head into tie breaks?

iirc FGC does double-elim bracket pools. This is even faster and simpler, plus is doesn't rely on having large setup : player ratios.

You also have to argue about accuracy. Half of the people in your example had their success depend on 1 set. This comes up much more infrequently in RR.

 

[TheCrimsonBlur]

^double elim bracket pools are actually much slower than round robin pools. Their advantage is that they require fewer setups (a bracket reduces to a couple setups after a few rounds).

----

I think this is a bad idea but a very interesting suggestion. As always with these type of things, it usually looks a lot better on paper until you put names down. Lets use pool 1 of Kings of Cali as an example (it was a 7 man pool but lets add a player named "?"):

P1: PP
P2: ROFL
P3: L
P4: Matt Anderson
P5: gkinfinity
P6: Saca Su Moto
P7: JC
P8: ?

1st round
PP v gkinfinity
ROFL v Saca Su Moto
L v JC
Matt Anderson v ?

2nd round
PP v L
ROFL v Matt Anderson
gkinfinity v JC
Saca Su Moto v ?

3rd round
PP v ROFL
L v gkinfinity
Matt Anderson v Saca Su Moto
JC v ?

PP 3-0 (W vs gkinfinity, L, ROFL)
ROFL 2-1 (W vs Saca, Matt; L v PP)
L 2-1 (W vs JC, gkinfinity; L v PP)
Matt Anderson 2-1 (W vs ?, Saca; L v ROFL)
gkinfinity 1-2 (W vs JC; L v PP, L)
Saca Su Moto 1-2 (W vs ?; L v ROFL, Matt Anderson)
JC 1-2 (W vs ?, L v gkinffinity, L)
? 0-3 (L v Matt Anderson, Saca Su Moto, JC)

Does Matt Anderson really deserve to be out over gkinfinity? They never played head to head, and if you look at their games played, gkinfinity would have made it out had he had Matt's schedule (?, Saca, and ROFL is significantly easier than JC, PP, and L).

Furthermore, determining 2nd-4th seed matters. Its the entire reason we do pools in the first place. You alluded to it earlier, but it is a really crippling fault. If you go into a 32 man bracket, of the 2-1 players, how do you decide who has to fight a 1st seed? In this example its probably Matt, but there are other pools where it won't be as clear cut.

As a spectator, why is it desirable to have a pool and not even have ROFL and L play? Thats the most interesting, most competitive matchup, and would not be "sandbaggy" at all, yet it is entirely avoided.

----

Round robin pools are FAST. Their only setback is that the require a lot of setups (for an 8 man pool you need 4). If you have optimal setups though, an 8 man round robin only takes 60 to 75 minutes. Thats an incredibly efficient way to take out half your competitors. The problem with round robin isn't time, its setups, and this solution doesn't really address it. You still need 4 setups per pool, and saving 30 minutes is really not that significant.

 

[MaskedMarth]

Swisses are better with larger groups. One thing that's attractive about Swisses IMO is the chance to play many people near your skill level.

Suppose we had people in groups of 20 or 30 and ran eightish rounds of Swiss?

 

[hungrybox]

This reminds me a lot of Swiss. and Swiss is really good for running an event

 

[EthereaL]

I'm still surprised that the Power-rankings aren't used to seed once the bracket is made.

Accurate Power-ranking > Pool performance

 

[clowsui]

And then you run into the problem of Accurate

GGs EthereaL

 

[EthereaL]

Touché.

With that said, I think accuracy would be easy to garner.

 

[Acryte]

@About seeding accuracy using this method...

You could use this format and then each round of pools give each player one "Challenge Match" where they can challenge someone from their pool. They can only challenge players higher seeded or tied with them who they haven't played yet. Bad players who lucked out with easy matches can be targetted to drop back down and good players can try to improve their position. It bumps the winner up to the rank of whoever they beat and the loser drops down to a position based off his Local ELO calculated from that pool and each subsequent challenge match. You can order the challenges based off seeds. Lowest seed challenges first, highest player (able to challenge someone) challenges last (giving the higher seed advantage of later pick). A player doesn't have to use their challenge, they can pass, but they do have to play if challenged. Mostly you only have the potential to move up if you win or down if you lost a challenge... or down if someone above you lost and someone is nudged higher than you. Since the higher seed player gets later pick, if you get nudged down by someone you feel is worse than you, you're still able to challenge someone above you to recover your lost rank OR challenge the person who nudged you down (it would be unfair if the people above him were too strong but he was noticeably weaker than you but you were unable to challenge him since you had already played him in pools. Kick his weak *** back down the ranks).

You could program a simple ELO pools calculator type app easily where you can keep track of multiple pools and it calculates ELOs for each pool for determining rankings. A player's General ELO (from all ranked tourney matches, aka not from pools local to this tournament) can give you the edge in who is higher for people seeded the same exiting pools... for example, If you came out of your pool 1st and so will 7 other people from their 7 respective pools then it decides who is higher ranked among those people. Your pools performance will still mostly dictate your seeding for bracket. ELO from outside the tourney just affects those who are the same seed exiting their respective pools when deciding pairings (they aren't tied in rank anymore).

A good player isn't punished because he played more poor players (and won obviously) than someone else, the ELO is first calculated upon start of challenges but remains locked in order of seeds based off wins/losses. When someone wins or loses a challenge it adjusts the ELO values of everyone dynamically but DOESN'T reorganize ALL the players. If the challenger wins he takes the rank of the other person and the loser drops to a position based off his newly calculated ELO. If the player who was challenged wins he retains his rank and the challenger instead is dropped to a position based off his newly calculated ELO. The people who shift down or up because of him DON'T CHANGE ORDER based off ELO changes due to the challenger winning or losing his match (aka it doesn't completely reseed everyone based off of that person winning or losing). The loser just shifts down shifting others up. This means a good player who beat 2 scrubs and 2 good players wouldn't be penalized for those scrubs losing challenges, and his win against that scrub being worth less now. His initial seeding based off wins (which match by match is assigned to playing the stronger players) will prove his worthiness to be in his top spot. If he is to get knocked down, someone below him must beat either him or someone above him. If he is unworthy of his current place in rank, then he'll lose a challenge and get knocked down accordingly to the appropriate ELO position for him, shifting others up.

tldr;

Players can challenge someone for their spot. Loser shifts down based off a newly calculated ELO. You can challenge anyone higher than you that you haven't played yet in pool, or someone you have played if they nudged you down by beating someone above you.