Playing as a "Team" to reduce variance ?

[skipperbob]

Has anybody thought about (or done it, for that matter) of pooling results with two or three other players with whom trust is not an issue and who all feel that their play is roughly equivalent from a skill standpoint.

The most difficult thing for me, playing $215 SnG's, is dealing with the humongous swings (variance); (+/-) 20 buy-ins is not uncommon . If I were part of a trustworthy partnership (not cheating partners, just pooling results), it seems to me that our mutual results would more closly approximate the average win-rate of the group.

????

[Jman28]

I've thought about this.

In theory, I think it makes perfect sense, although the incentives to get better aren't as strong once you have this group formed.

However, finding a group of people with almost exactly the same skill level and with whom trust is a non issue might be pretty difficult (especially if you want to improve your game... you would have to improve uniformly).


Edited because I can't spell incentives

[the shadow]

Interesting idea. In case you didn't see the thread, Jerrod Ankenman (http://forumserver.twoplustwo.com/showthreaded.php?Cat=&Number=2421419&page=&view=&s b=5&o=&vc=1) and Bill Chen are doing this in MTTs.

The Shadow

[the shadow]

[ QUOTE ]
finding a group of people with almost exactly the same skill level and with whom trust is a non issue might be pretty difficult

[/ QUOTE ]

The team would not need to have everyone at the same skill level. Imagine two players -- one with a higher ROI but a high variance and the other with a lower ROI and a lower variance. It could be mutually advantageous for two players to team up, for much the same reasons that it makes sense to add bonds to a stock portfolio.

The Shadow

[UMTerp]

I agree it could get pretty messy to find the team you trust and whatnot, but one solution would be to pay each team member his "salary", then withhold a percentage of profits that are distributed on some incentive basis (results-based).

Another big hurdle to overcome would be when one team member couldn't play his fair share over some length of time due to personal reasons.

[UMTerp]

[ QUOTE ]
Imagine two players -- one with a higher ROI but a high variance and the other with a lower ROI and a lower variance.

[/ QUOTE ]

That's pretty irrelevant in SNGs. It'd be hard to be "high-varience" if you tried...

[the shadow]

I'm not to sure about that. My guess is that a player who single-tables $200 SNGs has a higher variance than someone like Lori who multi-tables $10s, $20s, etc.

[UMTerp]

I'm not convinced of that. His maximum theoretical downswing (within a certain confidence interval) would certainly be bigger than Lori's (percentage-wise), but that's only because she would have a higher winrate from playing the lower buy-in SNGs.

Because of the "step" (for lack of a better word) nature of SNGs (the fact that there are only 10 distinct results), it's hard to create a lot of extra variance and still be a winning player. The only way I can think to do it would be to start pushing a lot more marginal hands (and calling marginal ones) on the bubble. Even that wouldn't do a whole lot of damage I wouldn't think though.

That's why Aleo used to say that it was a pretty safe bet for just about all SNG players to enter a "1.4" (I forget the exact number, but I think that was it) in the spot for standard deviation in his spreadsheet when playing around with confidence intervals.

Now I'm not saying that there's NO way to increase variance - of course there is - but the mathematical nature of SNGs tends to give a more concrete right/wrong answer moreso than other forms of poker, hence making variance pretty standard for most players that carry a simliar winrate.

[RedManPlus]

Variance for each player is reduced...
ONLY if you pool the profits close to 100%.

Since this type of team has no legal standing...
It would only work with full-time Pro Players...
That have a reputation to protect...
Because that's your only protection.

That's why Pros sell a piece of themselves...
And buy a piece of someone else...
To reduce multi-table tournament variance.

That said...
If you are a good online SNG player...
Why are you worrying about variance?

rm+


/images/graemlins/cool.gif /images/graemlins/cool.gif /images/graemlins/cool.gif

[PickyTooth]

because constant income is fun and not being sure what can of money youll be making next month is not.

[Slim Pickens]

This is a very bad idea. It works from a mathematical standpoint, but absolutely fails from a social one. I highly doubt you could ever find a group of poker players able to agree upon (and stick to) an objective evaluation of talent for distribution of profit. Think about it.

[Sponger15SB]

[ QUOTE ]
Has anybody thought about (or done it, for that matter) of pooling results with two or three other players with whom trust is not an issue and who all feel that their play is roughly equivalent from a skill standpoint.

[/ QUOTE ]

Yeah me and like 2-3 other guys do this, we all join the same SNG and then we create a chat room through AIM and tell eachother our hole cards and when to fold so we can build up our chips around the bubble if one is short stacked.

We usually lock up at least 2 ITMs this way. Its a great way to reduce variance.

[Quicksilvre]

As long as the cash is dealt out equally, and no one gets greedy, it could work. I predict that it all ends in tears.

[checksplay]

There are a lot of teams that do this for blackjack. On www.bj21.com (http://www.bj21.com), a poster named DD' has described a good system that takes into account playing time, results, and a host of other things. The math is easy, it's the people part that is hard. Should work in theory, though.

[microbet]

I think to really work something out you would have to have some close relatives or family friends that play for similar stakes, but who has anything like that?