Cash game pro Luke ‘desultory’ Hatfield looks at the theoretical differences between cash games and tournaments and explains the ways to get an edge in tournament play
For a very long time I have purely been focused on the theory of cash games. I used to be an 18-man and nine-man sit-and-go pro when I started my poker career but I have since forgotten all of my sit-and-go theory knowledge.
But during the Spring Championship of Online Poker (SCOOP) I thought it would be a good time to refresh my tourney skills and I thought I would share various things I have re-learnt in the process. So, I will start from the beginning; Given that tournaments have even more variance than cash games, one can conclude that one will need a larger bankroll for playing a $100 tournament than a $100 buy-in cash game.
This is assuming you are of similar relative proficiency at both types of game. I haven’t bashed the numbers, but there is a consensus you need 100 buy-ins to be ‘safe’ from busting from playing tournaments. A fairly conservative estimate for cash games would be about 50 buyins. So that’s at least twice as much!
If you are competent, it seems you will be ‘in the money’ (cash prizes) around 20% on average for sit-and-gos and maybe slightly less for MTTs. So on average, you will only win cash once every five tournaments you play. Please don’t be short-termist and think you are competent because you won one or because you cashed in three straight tournaments.
You will need a much bigger sample sizes than this. It is said that you will need to play 1,000 tournaments to discount variance, but I think this is an underestimate.
Measuring skill levels
Tournament players measure their competence by a term called ‘Return On Investment (ROI). ROI is calculated by the following sum: (total cash prizes – total money spent on buyins)/total money spent on buyins.
The best sit-and-go players earn 20% ROIs and general consensus says that the best MTT players can earn 100% ROI. This is completely different to the BB/100 win rate measure that cash game players use.
By far the most important difference to understand when transitioning to tournaments is that in cash games, any +EV move should be executed because +ChipEV is directly related to +$EV. But in tournaments, as your stack doesn’t have a fixed value, +ChipEV doesn’t always equal +$EV. In other words, you can make a move that will win you chips, which will lose you money in the long term.
Sometimes +ChipEV can be –$EV and therefore in order to be successful one has to relinquish some +ChipEV spots in tournaments that one never should in a cash game. This is all assuming that your goal in poker is to make the most money. If your goal is to come fi rst more than your average share in tournaments then your tactics may be slightly different.
There is a mathematical model called the Independent Chip Model (ICM) which calculates how much a player’s current stack is ‘worth’ at a given time in a tournament. It produces your percentage share of the prize pool when considering your stack and the amount of players left in the tournament. To help you understand this I’ll give you an example from a typical nine-handed sit-and-go.
There is a sit-and-go with a standard payout structure of 50% for first, 30% for second and 20% for third and there are four players left. We will assume that the four players have exactly the same stack size, and therefore their share in the prize pool is equal to their share of the chip total.
Two players go all-in and one gets knocked out. Then on the next hand, one of the smaller stacks goes all-in against the bigger stack and gets knocked out. The respective ICM values are as follows in the table you can see on the right of the page. When there are three players, the big stack has 50% share of the total chips but it’s wrong to assume his stack is worth 50% of the prize pool on average, because that assumes he will come first every time.
It is true that the big stack will have more chance to come first than the small stacks but the ICM attempts to give a more reasonable approximation of stack worth. Obviously none of them will win exactly what the ICM %EV suggests, it is just the average long term winnings they can expect to gain with that specific stack. If you want to calculate ‘stack worth’ in $EV you multiply the ICM stack percentage by the total prize pool. For example: $100 prize pool * 25% ICM = $25 stack worth.
The meaning of ICM
So what are the implications of this? As you can see from the table, when two players go all-in and one player gets knocked out the other two players not involved in the action have had their stacks increase in value by 5.8% of the prize pools. By sitting tight, they have won money. In cash games, you have to play to win money.
Another thing to note is the player who doubled his stack has only gained an extra 13.3% of EV, so by going all-in he risked 25% of his EV to gain 13.3% EV. This can create some strange dynamics where players should fold hands as good as AK even if it is +ChipEV to play it! Say you are in the late stages of a sit-and-go with a payout structure 40%, 30%, 20%, 10% for 1st, 2nd, 3rd, 4th.
There are three players left so you are already in the money. The blinds are now 50/100. The big stack of 2,000 chips goes all-in on the button and the small blind calls for 1,001 chips. You have 1,000 chips and you have AKo in the big blind.
You know that the button is shoving 100% of hands here and you know that small blind is calling 9.5% of hands (A9s+, ATo+ & 66+). If you call you will win a pot of 3,000 about 38% of the time and you will lose your stack 62% of the time. So doing an ChipEV calculation: You call & win: (1,000+1,000+100) * 38% = 798. You call and you lose: (900) * 62% = 558. So 798 – 558 = 240. As a fold is 0 EV, and 240>0 you should call if not considering ICM.
You can expect to increase your stack by 240 chips, so in a cash game you would snap call. However, ChipEV doesn’t take into the cash value of your tournament stack in $EV so let’s do an ICM calculation now. You currently have 28.33% of ICM%EV. If you lose your stack you have 0%ICM EV and if you win the pot you now have an ICM EV% of 37.5%. You lose 62% and win 38% so 62% * 0% + 37.5% * 38% = 14.25%.
By calling you have just lost 28.33% – 14.25% = 14.08% of the prize pool. If the prize pool is $100, you have just lost $14.08 on average by making that call. You should fold the AK if you are basing your plays off the ICM model.
The reality Of ICM
So, say you are disciplined and competent enough to fold the A-K and then the small blind flips over A-Q, the button flips over A-2 and the board comes A-3-7-5-K.
You hate yourself because you just folded when you would have won if you called. If you think like this you are destined for misery with poker. If you want to be one of the best, you should just shrug, laugh and say to yourself: ‘Oh well, I made the correct fold in the long run.’
The underlying ICM dynamic becomes very important in the later stages of a tournament, but the further away the tournament is from the payout positions, the less implication ICM should have on ones decision making. Early in a tournament, ICM has no impact at all and therefore the game play should be similar to a cash game.
ICM requires too many calculations to run on more than 10 players, so when playing in an MTT, it is difficult to know at what point you should start to consider ICM factors. Some say it only needs consideration when you reach the final table, but I think this is incorrect. It is hard to say exactly when, but I think you should definitely be tighter the closer you get to the bubble.
Because of the implications of ICM, a contrasting theory in MTTs, is that one should play more aggressively around the bubble because everyone else should be tightening up. This is especially true if you have a large enough stack that you will not be knocked out if you lose an all-in and therefore the bigger stacks become very powerful in these positions.
You should open and three-bet much wider ranges and in theory you should get many more folds. Say you are on the bubble and you have the big stack of 3,000 chips and the three other players at your table have 1,000 chips each. Let’s say the big blind is 100 chips, so the short stacks have 10BB each. If the other three players are competent and therefore tightening up, then you can min-raise every unopened pot and you should win the blinds.
If the players are fishier and like to call min-raises a lot then you can usually continuation bet half the pot to take it down. Please be wary that if you are opening 100% then you are definitely folding to three-bets too much. When competent players realise this, they will three-bet all-in against you quite a lot, even on the bubble, and therefore your 100% min-raising becomes unprofitable.
In more competent games where your opponents have less than 10BBs, open shipping all-in is better. In this scenario, you can open ship 100% of hands from the button and 50% of hands from the cutoff. When trying these tactics it will result in two scenarios: you get called by a good hand and drop your stack down to 2,000 at which point you should start playing your usual game again, or you will suck out and have 4,000 and can continue your shipping all-in tactics. Usually before a call happens you will have decimated the other stacks so much that it should be a victory on average from here on.
Another important thing to consider in tournaments is stack sizes. When blinds are small relative to stacks (stacks of 100 big blinds or more) the game play will be mostly postflop. But as blinds increase and stacks become relatively smaller in relation to the blinds (12-40 big blind stacks) the game play becomes much more of a preflop game.
Relative to 100BB cash games, you will see wider value three-bet ranges and consequently wider three-bet calling ranges (with 7-7+, A-Ts+, A-Jo+ as standard) because of the smaller stacks. The smaller stacks will lead to a game which has less opening, less calling preflop, more three-betting and less folding to three-bets.
Tournament play is much more dynamic than cash game play for all of the above reasons. This can be very confusing for the specialised cash game player, who is used to opening a lot and playing a certain way with 100 big blinds. Cash game players will tend to open too much and call three-bets tighter than is optimal.
On the opposite side of the coin, because of a lack of postflop experience tournament players will struggle in the deeper postflop cash games. That is why there are not many players whom are successful long term in both variants. The modern day 40BB short stackers come closest to being competent at both games, but I know of none who crush tournaments and cash games.
Some say cash games are more skilled because of the deeper stacks, and to a certain extent it is true that deeper stacks mean more decision making options and therefore more complications, whereas smaller stacks can often lead to a perfect mathematical solution (going all-in), which in essence is much simpler to execute and understand. But still, any cash game player that disrespects a tourney pro simply doesn’t fully understand the complexities of tournament play.
PokerPlayer magazine is now free on your phone or tablet!