Following my reading on an exchange posted on the forum GrenoblePoker , I wanted to deconstruct the principles outlined in order to be as accessible as possible.
The Independent Chip Model will allow us to calculate based on our carpet our probability of winning, finished second and finished third. By approximating the distribution of prices, we obtain euros (a kind of image of the value of our carpet).
Here's an example (you can take the form of a question, try to answer before reading the following): SNG
9 players, buy-in € 9 + 1, 4 left. Distribution of classical
price: 50%, 30%, 20%
Prize pool: 9 x 9 = 81 € €
Blinds 100/200
UTG
1000 Button 1400 SB 2600
We (66) 4000
To moment of the game, we have a broken ICM as follows: Player Chips
Prob Prob 1st 2nd 3rd Prob Equity
UTG 1000 0.1111 0.1545 0.2521 12.34 €
Button 1400 0.1556 0.2071 0.3118 16.38 €
SB 2600 0.2889 0.3204 0.2586 23.68 €
BB ( us) 4000 0.4444 0.3180 0.1775 € 28.60
Moving to Action: Preflop
: UTG folds, push button, SB fold.
Action (a purely mathematical point of view (MHI% price allocation))
1 / folder You?
2 / You follow?
____________________________
Before I answer, analyze different situations in case of fold or call.
1) One fold, carpets become:
UTG 1000, Button 1700, SB 2500, 3800 We either € 34.20 (42.22% of prize pool) from the ICM
2) We call and we lose :
UTG 1000, Button 2900, SB 2500, 2600 We either € 23.40 (28.89% of the prize pool)
2b) One call and you win:
UTG 1000, Button 0, SB 2500, 5500 We either € 49.50 (61.11% of the prize pool)
2c) We call and we split:
UTG 1000, Button 1450, SB 2500, We either 4050 € 40.95 (45% of prize pool)
We noted that the button (3rd belt) is a pretty tight player, and he chose instead to wait until the shortstack jumps. Suppose we put on: 22 +, A2s +, A8o + + and KTs KQo
By using PokerStove (http://www.pokerstove.com/), our 66 win, lose or splitter in the following proportion:
lost: 33.4%
wins: 65.5%
split: 1.1%
Thus, by combining these data with those of the ICM, we get: For a
fold: 42.22% of the prize pool worth
Caller: (28.89% x 33.4%) + (61.11% x 65.5% ) + (45% x 1.1%) = 50.15%
If you call it will end up with 50.15% of prize pool (note that this is theoretical, since the maximum gain is 50% hoped the prize pool) If
we fold it will end up with 42.22% of the prize pool.
Caller is gaining 7.97% of prize pool, or 6.45 €.
In the long term, paid with 66 wins in this situation! Ie with the hand on which range we put our opponent.
Conclusion:
First, we note that these calculations are dependent on the profile of players and ranks as one attribute.
Then the value of understanding the principle of ICM will be in time. Over the NSE, we will find similar situations and, as and when our mastery of this concept, some decisions may seem we are much clearer.
Finally, it is not a fixed model. Other factors may help in the decision: one estimate have an edge and avoid situations slightly + EV the bubble to allow us to glean chips for the period ITM ...
The Independent Chip Model will allow us to calculate based on our carpet our probability of winning, finished second and finished third. By approximating the distribution of prices, we obtain euros (a kind of image of the value of our carpet).
Here's an example (you can take the form of a question, try to answer before reading the following): SNG
9 players, buy-in € 9 + 1, 4 left. Distribution of classical
price: 50%, 30%, 20%
Prize pool: 9 x 9 = 81 € €
Blinds 100/200
UTG
1000 Button 1400 SB 2600
We (66) 4000
To moment of the game, we have a broken ICM as follows: Player Chips
Prob Prob 1st 2nd 3rd Prob Equity
UTG 1000 0.1111 0.1545 0.2521 12.34 €
Button 1400 0.1556 0.2071 0.3118 16.38 €
SB 2600 0.2889 0.3204 0.2586 23.68 €
BB ( us) 4000 0.4444 0.3180 0.1775 € 28.60
Moving to Action: Preflop
: UTG folds, push button, SB fold.
Action (a purely mathematical point of view (MHI% price allocation))
1 / folder You?
2 / You follow?
____________________________
Before I answer, analyze different situations in case of fold or call.
1) One fold, carpets become:
UTG 1000, Button 1700, SB 2500, 3800 We either € 34.20 (42.22% of prize pool) from the ICM
2) We call and we lose :
UTG 1000, Button 2900, SB 2500, 2600 We either € 23.40 (28.89% of the prize pool)
2b) One call and you win:
UTG 1000, Button 0, SB 2500, 5500 We either € 49.50 (61.11% of the prize pool)
2c) We call and we split:
UTG 1000, Button 1450, SB 2500, We either 4050 € 40.95 (45% of prize pool)
We noted that the button (3rd belt) is a pretty tight player, and he chose instead to wait until the shortstack jumps. Suppose we put on: 22 +, A2s +, A8o + + and KTs KQo
By using PokerStove (http://www.pokerstove.com/), our 66 win, lose or splitter in the following proportion:
lost: 33.4%
wins: 65.5%
split: 1.1%
Thus, by combining these data with those of the ICM, we get: For a
fold: 42.22% of the prize pool worth
Caller: (28.89% x 33.4%) + (61.11% x 65.5% ) + (45% x 1.1%) = 50.15%
If you call it will end up with 50.15% of prize pool (note that this is theoretical, since the maximum gain is 50% hoped the prize pool) If
we fold it will end up with 42.22% of the prize pool.
Caller is gaining 7.97% of prize pool, or 6.45 €.
In the long term, paid with 66 wins in this situation! Ie with the hand on which range we put our opponent.
Conclusion:
First, we note that these calculations are dependent on the profile of players and ranks as one attribute.
Then the value of understanding the principle of ICM will be in time. Over the NSE, we will find similar situations and, as and when our mastery of this concept, some decisions may seem we are much clearer.
Finally, it is not a fixed model. Other factors may help in the decision: one estimate have an edge and avoid situations slightly + EV the bubble to allow us to glean chips for the period ITM ...
