ICM Calculations in Poker Tournaments

Thoughts from the forum in calculating ICM and what to do about it in sit and go tournaments and multi-table tournaments. Phadraeus writes:

Firstly, I have learned that push/fold is not an easy formula. Rather it is a question of judgement and feel, based on the following factors:

M;
Q;
ICM and tournament equity;
Position;
Profiling
tournament type (MTT v STT)
and blind speed (turbo or regular)

A yellow M and very low Q with fast blinds might justify pushing from an M of 12. However mostly, I would think that pushing with that large a stack will be premature.

Mzone is basically a system of risk management. The lower your M, the wider your risk window needs to be opened...that's pretty straight forward...

Now, in a STT (S&G) you also need to balance Mzone with ICM (Independent Chip Model). *Most* STTs involve most or all players being in orange and red Mzone at the end of the tournament. If you are pushing a wide range from yellow M when most of the other players have similar stack sizes, then IMHO you are probably not playing a correct ICM strategy.

In STTs, the push fold mode is partly about M, and partly about what leverage you will have for flop play (and the two are very much inter-related). However, a low M does not necessarily mean that your tournament equity is low. Low M and high tournament equity (i.e on the bubble with 4 players having similar stack sizes) is a time for more conservative play than pure Mzone strategy would suggest. In this case there are two ways of cashing. One by doubling up with the best hand and, Two, by allowing someone else to make a fatal mistake.

Remember that part of ICM strategy is that you profit by staying out of hands where your opponents are playing too loose. This is just another calculation in poker you must keep in mind.

The other functional aspect of the M strategy is post flop play flexibility. With an M of 12, you mostly have enough chips to make a 3xBB raise, manuevre post flop, make meaningful plays and/or get out of the hand and wait for a better spot.

However, with an M of say 5, any pre-flop raise is mostly going to leave you pot committed and/or having not enough chips after you pre-flop raise to put to any meaningful use. Accordingly, you are better off shoving and getting the maximum leverage possible from your stack.

My personal preference for this range in a STT (normal blind speed) is when my M gets to 6.

If blinds are say 100/200 then a 6M stack is 1800. If I raise pre-flop I will be raising 600 (3XBB). If there is a caller to put me into flop play, then the pot will be 1500 compared to my remaining stack of 1200. I therefore do not have enough chips to manoeuvre with relative to the pot size and since the pot is larger than my stack I am pretty well committed to it. Thirdly, if I fold post flop I will be down to an M of 4, where after I have considerable less fold equity for future shoves.

If I double up from an M of 6, I am mostly in the top 3 in a STT (unless I got to an M of 6 due to an early large loss). If I double up from an M of 4 comfort levels are not nearly so great.

So for all those reasons, an M of 6 is my preferred trigger point for STT normal blind speed.

Now, between and M of 6 (my pushing trigger) and an M of 9, there are more than 2 blind levels of folding time. That means if instead of pushing at an M of 9, if you played very tight and folded, waiting for a big hand, you would see between 20 and 30 additional hands before you pulled the trigger. Given that there are 1326 possible hands, seeing an additional 25 hands gives you opportunity to improve your pushing range by around 2%. Might not sound like much, but since a double up from an M of 6 will still put you back in contention for a win, it is pretty well a free 2% (2% is also the margin by which a PP beats over-cards - so it is frequently a percentage that we rely on).

However if you are playing turbos, then instead of 20 to 30 additional hands, you may only see 10 to 15. Now your percentage edge gained by folding is not as great. Also, because turbos force looser play and more frequent all-in moves, you need a larger stack than normal to retain any meaningful fold equity.

Accordingly, pushing with a much higher Mzone strategy seems appropriate for turbo STTs (M=9 to 12) would seem appropriate to me.

Now, for MTTs....

Here, things change quite alot, mostly due to the antes. Before the antes, I tends to adopte the same push/fold approach as Sit and gos (i.e M of 6 = trigger time). (and again, this is also Q dependent - very low table Q an M of 8 may mean you may have very little fold equity depending on stack sizes around you)

However after the antes start, your M drops considerably without a corresponding drop in stack size. This means that you retain significant fold equity well into the red Mzone. Consequently, you can afford to be more selective with your pushes in MTTs and I generally start pulling the trigger with an M of 5 or less. A higher M than that actually gives you opportunity for a significantly greater range of plays, including re-shoves, stop and goes and all-in continuation bets.

I also mentioned position above. This is a difficult and curly part of the whole equation. According to Harrington HOH v2 (which is really mostly applicable to MTTs) position does not matter much for inflection point play (i.e pushing from red Mzone). The reason is that in EP, you are under more deparation to make a move before the blinds swallow a big chunk of your stack. So you are pushing a similar range from EP and LP, just for different reasons.

However in STTs, position is a much more significant factor. First in Vigorish is still king, however ICM considerations demand that position is properly played. Again, because chip values are non-linear near the bubble, ICM is critical. Fundamental to ICM is the fact that if your oponent plays too loose, they hurt themselves, but they hurt you too. The players who benefit (with +ve EV) are the ones sitting back and watching the all-in clash.

In my experience, most players do not play the bubble correctly. When blind pressure is high, most people call a shove with way too wide a range. The sklansky gap concept seems to go out the window and all-in shoves frequently get called by Ax suited, A9+ off-suit, KQ, KJ, QT and any pocket pair. Because of all of that loose play, I tend to prefer to push through as few oponents as possible, meaning I still want genuine strength to push from EP on the bubble, irrespective of M (provided I have meaningful tournament equity in ICM terms).

Lastly, I mentioned profiling. The nightmare situation here is when you have a moster stack loose player on your left. Now you have very little fold equity and pushing from an M of 6 or 12 is not the issue. Rather this issue is that you need solid hand strength to push because you player profiling means you have negligable fold equity.

Similarly, pushing and re-shoving into medium sized stacks is much more important than M trigger points alone. For example, I might re-shove a marginal hand into a medium stack from an M of 12, but I would not push the same hand into a monster stack from an M of 7.

So, that turned into a pretty long post. That is just my thoughts on it anyway. Hope it assists and I'm sure others will have clearer insights...

(Calling out for a post from Raziel on this because his turbo ICM play is probably the best of the forum)