ICM Adjustments
Why I've Changed Gears Early in Final Tables
For most of my tournament poker career, I’ve paid little attention to sliding up the pay scale. My approach was to live without fear of busting, keeping my eyes on the top prize. I slept well at night, even if I busted in a favorable chip position, “knowing” that I was doing the right thing in the long run.
But was this approach wrong? “Wrong” might be overstating the case a bit. Here are my 23 final table finishes based on the Hendon Mob data:
1st - 4
2nd - 2
3rd - 4
4th - 0
5th - 3
6th - 2
7th - 2
8th - 2
9th - 4
We don’t want to draw too many hard and fast conclusions from a sample size of 23, but these results seem to be on the right track. The main goal is always to maximize top three finishes and victories, and I’ve definitely achieved both at a strong rate in these 23 appearances. But could I have done better?
Notice that 8 of my 23 results are in the bottom three. This is roughly the amount you would expect from random chance, so it’s not exactly terrible. But also notice that I have as many 9th place finishes as firsts. Some of that is coincidence, but some of that is because I have played without concern for sliding into 8th or 7th, always going for the win or at least the top three.
We know now that this approach, while correct in spirit, in incorrect in the details. I had the overall goals right, but ICM analysis shows that 9-handed at the final table should be one of your most risk-averse moments. We should keep the goal of winning in mind, but we should pass on a lot of our smaller edges at the start of the final table. The simple, intuitive explanation for this strategy? It’s a great chance to move up significantly without winning any chips, and we’re still further away from the top three than it seems.
What might my results have looked like if I’d been a bit more conservative at the start of final tables? Would I still have four wins and ten top-three finishes? Probably not, as the whole point of my earlier strategy was to maximize those outcomes. But I’m thinking that I could’ve still reached the top three quite often, with significantly fewer bottom finishes.
Let’s say, hypothetically, that by playing a bit more conservatively, I could’ve changed my results to this:
1st - 3
2nd - 2
3rd - 4
4th - 3
5th - 4
6th - 2
7th - 2
8th - 2
9th - 1
Would I have preferred these outcomes? My first instinct was, “no way—first place is worth so much more than the others! How can it better to have fewer wins and fewer top-three finishes, replacing them with some measly fourths, fifths, and sixths?”
Plugging in the payouts from the most recent final table in my Hendon Mob, it turns out that my actual results would’ve won $179,502. The hypothetical results? $182,842. To be sure, it’s very close, and to be sure not all tournament payouts are the same. The more top-heavy the pay structure, the less conservative you should be at any point, and vice versa. Still, I think all the ICM models are onto something. You can and should sacrifice a small increase to your winning chances for a moderate increase in avoiding 8th and 9th place finishes.
In a future post, we’ll look at some specific adjustments I’ve made to start doing just that—including in a recent online tournament where I finished fifth. :)