Hey everyone,

I had a few questions regarding ICM on the stone money bubble with a short stack, as well as 3 particular spots. What range do you jam in these spots? I know risk premium is very high for very short stacks on the stone bubble in theory (because the chance of getting top payouts is so low so the value of the min cash is quite high) but does that thinking change for massive multi-flight tourneys with huge payouts up top? It's pretty hard to run this type of spot in ICMizer given there's going to be 600+ payouts and stacks to input.

Tourney Info
Seminole Hard Rock Poker Open Event #1 - $400 buy-in, 8 starting flights, $1M prize pool guarantee (will likely end up near $2.5M range)
10% of players paid - Each flight plays down to 10% on Day 1 and all players who make it to Day 2 are in the money
Just under 50% of prize pool paid to top 10 spots
Min Cash - similar event earlier this year paid $767 for a min cash
Flight 1D - 825 entrants, 83 paid/return for Day 2

Hand 1:
Blinds 3k/5k/5k BB ante, 86 players left (3 off the bubble)
Hero in HJ has 49k (9.8 BB), all stacks cover
Folds to Hero in HJ with Ac5c

Hand 2:
Hand for Hand play, stone bubble (84 left)
Blinds 3k/6k/6k
Hero on the BTN has 36k (6 BB), SB and BB are <12 BB but all cover
Folds to Hero on the BTN with KhQh

Hand 3:
Hand for Hand play, stone bubble (84 left)
Blinds 3k/6k/6k
Hero in the LJ has 43k (7.2 BB), all stacks cover
Folds to Hero in the LJ with AcQh



Thanks!

As always, 2 crucial pieces of information are missing:

1. How many stacks are shorter than you? I understand you might not know this, but a sense of this matters.
2. What is the average stack? Doubling to an average stack has a lot more utility than doubling to a 1/2 average stack. In the former case, you will have meaningfully increased your ev, in the latter, you might be adding some relatively small ev, since you are still likely to min cash and add to that whatever equity your stack has going forward.

In addition the players behind you matter; the guy who snapped with KQs was probably not making a good call, given how tight your range should be. Based on your range he is ahead of exactly 3 combos. If my math is correct he is behind 120 combos, and exactly even with 3 combos. He isn't even closing the action; or getting the full BB discount. If BB does fold, he is risking 5.5 to win 14. He is not ~40% against your range, I think. Add to that the ICM factor and his call is lighting money on fire.

Importantly, his call cost both of you EV.

As always, 2 crucial pieces of information are missing:

1. How many stacks are shorter than you? I understand you might not know this, but a sense of this matters.
2. What is the average stack? Doubling to an average stack has a lot more utility than doubling to a 1/2 average stack. In the former case, you will have meaningfully increased your ev, in the latter, you might be adding some relatively small ev, since you are still likely to min cash and add to that whatever equity your stack has going forward.

In addition the players behind you matter; the guy who snapped with KQs was probably not making a good call, given how tight your range should be. Based on your range he is ahead of exactly 3 combos. If my math is correct he is behind 120 combos, and exactly even with 3 combos. He isn't even closing the action; or getting the full BB discount. If BB does fold, he is risking 5.5 to win 14. He is not ~40% against your range, I think. Add to that the ICM factor and his call is lighting money on fire.

Importantly, his call cost both of you EV.

Thanks for the response!

My bad there, the board didn't show average stack but I think it should be easy to figure out - starting stack was 20,000 chips and 10% of the field made the money so I think the average stack would be right around 200k. So a full double for me would've put me still at less than half the average stack. One interesting consideration here I forgot to mention is that Day 2 would actually start at the level that the earliest flight finished at, and and a flight had already finished at the prior 3k/5k/5k blind level, so the blinds are going back a level on Day 2. Would having a bit more depth on Day 2 change the EV of a jam at all?

As for the number of stacks shorter than me, I had no clue, but my thoughts at the time were I wasn't going to be the first player to blind out if I folded everything given I had just gone through the blinds when we started hand for hand. I checked the chip totals and there are 6 players that bagged less than the 36k I had when we started hand for hand, so 1-2 per flight. So I'm thinking I would've been one of the shortest stacks in the room but likely not the shortest. FWIW the hand where I jammed AQo ended up being the last hand of the bubble, two 20 BB stacks got QQ in vs AA so there's probably value in waiting even if I was the shortest stack to allow coolers to happen.

The players behind me is an interesting point - both players to the left were definitely willing to gamble and were definitely punting given ICM. The SB had snapped me off with KQs, and the BB had jammed 9 BB with 96o from the BTN the orbit before and showed the table. I think the blinds thought I was wider than was actually planning on being too though because I was playing more aggressively before losing most of my stack with AK vs AA a level before. So my shove definitely had less fold equity than it would have against more ICM conscious opponents, but I definitely would have had around 45% equity when called.

Given this info do you think the KQs was ok or do you think it was likely too thin? How about the AQo jam?

Didn’t read the spoilers yet. I think 1 is a fold, 2 is a jam & 3 is pretty dependent on the stacks/players behind you.

Being very short on stone bubble in a big live event is definitely a lot harder to navigate then online, since you are missing a lot of information and have to make your decisions based on assumptions. There could be a 2bb or similar stack at another table without you knowing, there might be a bunch of people shorter then you or you could be the shortest stack left.

I think a lot of the decisions come down to how much you value the min cash. The prospect of having to come back the next day with a short stack and most likely have to get it in pretty quickly probably would also make me a bit more willing to play more aggressively in these spots.

I see that phrase a lot. For any tournament that you have a reasonable bankroll to be playing, the $ev should be linear through the area of the min cash. Exceptions can be made for something you sattied into, or a once a year 'shot' (like the ME). Other than that, $ev should reign supreme.

I haven't taken a pen to paper, but these spots when you are doubling to something that is only 1/2 average are very bad for you. So much of your equity is the min cash, and doubling doesn't really change that much. There are spots where you need to be greater than 80%, which means even AA is a fold.

For more on this and the tools to do the math, buy Dara O'Kearney's excellent book: Endgame Poker Strategy.

Seconded. Good book.

One way I've started thinking about ICM and why the money bubble is important is-- look at the pay jump from not cashing to the min cash, and see how much longer in the tournament it takes to have a similar pay jump from the min cash, or how long it takes for one spot to be a similar jump in cash.

Since I could find these payouts easily, let me use the WSOP Main Event this year as an example.

The jump from 1301 to 1300 is a $15,000 increase ($0 to $15,000).

To make another $15,000 on top of the minimum, you'd have to make it from 1300 to 476 ($31,900).

The next single-spot pay jump of $15,000 is from 81 to 80 ($86,000 to $101,700).

So when you look at it like that, you can see why bubble play is so important, because that pay jump to the min-cash is the kind that's not going to show up again in the tournament until much later. (Probably even more emphasized in a smaller field-- some of the small-field ACR tournaments I play will have a mincash as high as 2.4x the buyin.)

Anyway, looking at the hands I think 2 is probably a shove; fold equity is paramount in these spots and only having two people behind you is important plus you block a lot of the better calling hands and should have decent equity when called. 1 is probably a fold, and I'm less certain on 3-- both cEV shoves but the ICM question is really a math question centered around how many stacks in the tournament are shorter than you.

I think, all those 3 hands are fine jamming spots. With high ICM pressure calling ranges are affected much more than jamming ranges, as long as we still have fold equity. And yes the payout structure does affect the risk premium. The relevant number is, how much of the price pool is distributed in min-cashes. If as an example we have a 10$ tournament, where the min-cash is 30$, and 10% of the field gets paid, then 30% of the price pool is distributed on the bubble. And since there is now 30% less money in play, each chip is worth 30% less, than it was before the bubble. This is essentially, where the risk premium comes from, and its an important concept to understand. 30% is a typical number for online MTTs, but if the number was higher or lower for this event, you should make either bigger or smaller adjustments for the risk premium.