Poker: Tournament Idea

[SCavenger]

Anyone familar with duplicate bridge? It's where teams rotate around to different tables and play exactly the same randomly predetermined hands. After the teams have played each table, they are ranked according to their performance relative to other teams that played the same hands. The best aggregate performance is the winner.

Would you think that a similar format might be a better measure of a poker player's relative skill as opposed to current tournament formats? My sense is that success would be more due to skill than cards.

For example, a tournament could consist of 10 tables with 10 seats at each. Players would sit at a different table, in a different seat for ten table sessions. Players would not move as a group but be scattered amonst the tables. Each table session would consist of a predetermined # of hands and each player allocated a predetermined # of chips. The hands and the board would be the same for each session at a given table.

At the end of the 10 sessions, players would be ranked (according to their chip count at the end of each session) against the other players that played the same table/seat (ie cards).

So, assume I had the following sessions outcomes (in chips):
1000
0
10000
6000
500
2000
4000
0
3500
1000

These outcomes rank relative to other players who played them:
8
10
1
2
8
4
3
9
2
10

Some function of these rankings would determine the overall winner(s). Perhaps the lowest sum or some point mapping (1st = 15pts, 10th = 0pts).

Bottom line, players would be measured against other players using the exact same hands (player and opponent), same board, same number of hands, the same chip position. The only difference would be the players faced. However, even that would be mitigated by the shuffling at table changes.

What do you think?

[MNBridge]

I've always thought the duplicate bridge strategy should be used for other tournaments. (i.e. Cribbage, Bridge, whatever.)

I think you would have to be careful about how to set it up but it should work.

[Expunge]

It’s an interesting idea. You would still need many hands at each station to determine real skill. However so much of poker depends on who you are playing against at that time so there is still luck of the draw. Maybe if you could set up a computer program where the opponents have set rules on when to bet raise or fold. Again you’ll need a ton of hands. Once I crank out my first true AI for my poker program I think I would be able to set this up. I might not even need to know the computers cards, just as long as I can program the computer to play with a hard set of rules. I would have some advantage knowing what these hard and fast rules are though. Anyways this won’t happen until after the exam as the programming project is on hold until then.

[Bama Gambler]

The other problem is a tournament can't a set number of hands, because everyone that is low goes all-in on the last hand. Then the tournament turns into a one hand tournament. If you want to measure skill look over the course of a year. The best tournament players are up $$$ and the wannabes are down $$$.

[SCavenger]

The other problem is a tournament can't a set number of hands, because everyone that is low goes all-in on the last hand. Then the tournament turns into a one hand tournament. If you want to measure skill look over the course of a year. The best tournament players are up $$$ and the wannabes are down $$$.

Not really, the idea is that you want to have the most chips relative to the other players that have had the same exact hands. So you probably wouldn't want to throw away even a small stack that may be the best that anyone had done with that series of hands.

However, let me ask this - do you play to maximize value for each hand? Or are there some plays where you might "throw away" chips in order to alter an opponent's perception of your play for future purposes? If the former, then limiting the # of hands would not seem a problem.

[O. Hannah]

I see this system failing for holdem poker and here is why:

1) Bridge rates EACH hand.

2a) Given the nature of NL poker....if you rate by EACH hand...it's sort of crazy because losing hands will actually all rank LAST and all the people that sat out will tie for 2nd. :crazy:

2b) In for a penny, in for a pound....no reason to fold as finishing the hand even down 1 chip (from all the folders) is the same as being all-in. Basically, it turns into a lottery situation involving the SB, BB and anyone catching a hand.

[SCavenger]

I didn't know that bridge hands were rated (I only have a rough notion of duplicate bridge). I don't think it would be necessary here. In any case, the unit equivalent (to a bridge hand) would be the table session of hands not each hand.

Maybe I didn't explain well enough:

I sit down and play at seat 1, table 1, with 1500 in chips and play 30 hands. Somebody notes the number of chips I have at the end.

Now you come to play seat 1, table 1 with 1500 in chips for 30 hands playing exactly the same sequence of hands I was dealt, the same boards and facing the same cards in the opponent's hands (but different opponents).

The idea is to fix some of the variables, mainly the cards dealt, in order to isolate the effect of the play. The only thing that is unfixable is the opponents. However between playing against most opponents (over the course of different table sessions) and being measured directly against some (those that have had the same table/seat), the effect of opponent's differing plays should be considered a wash.

I agree with Bama's assertion that a truer measure of skill would be a long-term sample, this is an attempt to measure it in one event where the "luck of the draw" at least in cards (maybe not opponents) is controlled for.

[11pecans]

so each table has 30 decks of cards?

pick up your hand 1 play it put it back in envelope for next rotation
pick up envelope #2, play hand 2, put it back

similarly you would have the dealer envelopes

I'm not saying it's a bad idea from a probability stand point, but I don't see it ever beeing popular, poker tournaments already have a way of measuring success and eliminating players and determining a champion, its not going to change

[MNBridge]

How about the following changes:

Play the same cards until there is only one player remaining.
1st is worth some value
2nd same and so on.

Could just be 10 to 1

You would play Y tournaments and have a total score compared to all the other players that played that seat. Those are your real opponents.

This would take forever and you would need a lot of players. 100 people would be 10 matches with 10 people in each match.

I think you would need a computer to play this. (Easier to store hands and deal same cards)

[SCavenger]

This would take forever and you would need a lot of players. 100 people would be 10 matches with 10 people

That was the rationale for limiting the # of hands played at each table session. So you play, say 30 hands at ten table sessions.

I think you would need a computer to play this. (Easier to store hands and deal same cards)

Yeah, need some kind of technology beyond envelopes.

tournaments already have a way of measuring success and eliminating players and determining a champion

True, and the current method is much more dramatic than this proposal. On that alone, it would probably never fly. But there seems to be a lot of complaining about lucky cards, lucky streaks, lucky draws and so on. This is an attempt to remove the "luckys" so one can only blame themselves. Guess you guys won't be contributing venture capital to my new start-up. :D