Multiway Deals

The Tourist

Can someone explain how to do a multiway FT deal (ie more than two people) based on chip counts, along the lines of this two way deal:

1st prize: 1000
2nd prize: 600

Player A Chips: 100,000
Player B Chips: 50,000

Player A gets : 600 + (100,000/150,000)*400
Player B gets : 600 + (50,000/150,000)*400.

Is this expanable to any number of people? If not, what is the usual method?

The_Chopper

The Tourist wrote:

Can someone explain how to do a multiway FT deal (ie more than two people) based on chip counts, along the lines of this two way deal:

1st prize: 1000
2nd prize: 600

Player A Chips: 100,000
Player B Chips: 50,000

Player A gets : 600 + (100,000/150,000)*400
Player B gets : 600 + (50,000/150,000)*400.

Is this expanable to any number of people? If not, what is the usual method?

This is the usual method so player A gets 600 plus 2/3 of 400 and Player B gets the rest.

If there are 7 players which I saw happen at a tournament lasy year - everyone takes 7th and chip chop the rest as you've done above

The Tourist

That's what I thought, but the way I figure it that is a bit off. The two player deal is based on the probability of player A finishing first being = the proportion of player A's chips to all the chips in play. If you add one more player but assume the same thing for first place you get the chance of each player being first. If you then assume player A is first and calculate the chance of player B or C coming second (as being in the same proportion of their initial chips) and then do the same for players B and C being first you get a different answer to the above method. For example in a very quick calculation where 1st prize is 50k I found the differences were of the order of 2k. Has anyone come across this kind of thing before. Or a good method you can do in your head at a poker table?

RoundTower

The Tourist wrote:

Has anyone come across this kind of thing before. Or a good method you can do in your head at a poker table?

Short answer: there is no quick method for working this out.

Long answer: google ICM.

CoolBoardr

The Tourist wrote:

That's what I thought, but the way I figure it that is a bit off. The two player deal is based on the probability of player A finishing first being = the proportion of player A's chips to all the chips in play. If you add one more player but assume the same thing for first place you get the chance of each player being first. If you then assume player A is first and calculate the chance of player B or C coming second (as being in the same proportion of their initial chips) and then do the same for players B and C being first you get a different answer to the above method. For example in a very quick calculation where 1st prize is 50k I found the differences were of the order of 2k. Has anyone come across this kind of thing before. Or a good method you can do in your head at a poker table?

??

1st prize: 1000
2nd prize: 600
3rd Prize: 200
Total: 1800

Player A Chips: 100,000
Player B Chips: 50,000
Player C Chips: 40,000
Total: 190,000

Player A gets : 200 + (100,000/190,000)*1200 = 831.57
Player B gets : 200 + (50,000/190,000)*1200 = 515.78
Player C gets : 200 + (40,000/190,000)*1200 = 452.63
Total: 1799.98 (close enough!)

RoundTower

if you do it like the chip leader always gets a very good deal. This is useful to know, because many people will agree to a deal like this.

Hitman Actual

The Tourist wrote:

That's what I thought, but the way I figure it that is a bit off. The two player deal is based on the probability of player A finishing first being = the proportion of player A's chips to all the chips in play. If you add one more player but assume the same thing for first place you get the chance of each player being first. If you then assume player A is first and calculate the chance of player B or C coming second (as being in the same proportion of their initial chips) and then do the same for players B and C being first you get a different answer to the above method. For example in a very quick calculation where 1st prize is 50k I found the differences were of the order of 2k. Has anyone come across this kind of thing before. Or a good method you can do in your head at a poker table?

The maths for this can get quite complex. Have a read of this and this.

I can run through the maths of the three-way ICM deal if you really want, which is handy to know if you play 6-seater STTs as you can just plug the equations into a spreadsheet and look through loads of bubble situations very quickly. For more players than three, I wouldn't touch the maths with a ten-foot clown pole.

The Tourist

Thanks for the links. It looks like I was on the right track. One word of caution: one of the links talks about a random walk model as if it were the exact answer, when it is just a model (ie with some simplifying assumptions) as far as I can see.

Just to clarify what I was saying above, here's an example I worked out:

Prize1 50,000
Prize2 25,000
Prize3 15,000

Chips A 100,000
Chips B 50,000
Chips C 35,000

This leads to (see links above for mathematical details):
PlayerA PlayerB PlayerC
Prob(1st) 54% 27% 19%
Prob(2nd) 33% 38% 29%
Prob(3rd) 13% 35% 52%

Commonly Used Deal
A: 39,324.32
B: 27,162.16
C: 23,513.51

ICM Style (Fair?) Deal
A: 37,182.18
B: 28,269.74
C: 24,548.08

In this case the commonly used deal is as per CoolBoardr and The_Chopper's explanations above and the ICM deal is done as follows:
PlayerA = Prob(A 1st)*Prize1 + Prob(A 2nd)*Prize2 +Prob(A 3rd)*Prize3

What I'm trying to get is a feel for when it is right to accept the commonly used deal.

RoundTower wrote:

if you do it like [Ed: the commonly used deal] the chip leader always gets a very good deal. This is useful to know, because many people will agree to a deal like this.

Sounds good.

Marq

just don't do deals.
then you don't have to think about any of this silly maths stuff.

TommyGunne

Never deal. If you are doing a deal you are playing above your bankroll or somehow you have lost your necessary gamble!
2 situations where you might possibly deal:
1) You are against vastly superior players.
2) You are offered an incredibly favourable deal.

Mr.Plough

TommyGunne wrote:

Never deal. If you are doing a deal you are playing above your bankroll or somehow you have lost your necessary gamble!
2 situations where you might possibly deal:
1) You are against vastly superior players.
2) You are offered an incredibly favourable deal.

dont buy that, as has been said before, id happily deal when play is crapshooty