ICM, pot odds and minraising

[daviddem]

Hello,

OK, so I know cash games and SNG are very different in terms of odds because of the different $ value of chips.

Example:
- we have a nut flush draw on the flop
- there is 1000 in the pot
- villain moves all-in with his last 1000

Call or fold?

In a cash game, this is a call: we get the pot odds to call (break even is calling for approximately 1166). In a SNG, the situation is very different and ICM needs to intervene here: the general situation, number of remaining players, blind sizes, stack sizes, etc all have a major effect on whether this call should be made or not.

My question is, knowing the above, does minraising sometimes make sense? I see a lot of the high stakes SNG players and pros do it regularly (including preflop). I assume the logic behind this is that since the chips you have are worth more than the ones you are thriving to earn, then you should tend to risk less (and so, raise less). However, I am still confused:
- I see a lot in this forum that we should never minraise (taipan168 low buy-in's SNG do's and don't's)
- if it does make sense to sometimes minraise preflop or postflop, when and why?

Thanks for the help. I have only just started playing the $1 SNG on Full Tilt, and damn, they are not the fish fest I thought they were going to be! Out of 9 players, you usually get one or two playing badly, but the rest is quite tight early and in the later stages seem to understand blind stealing and ICM pretty decently.

[Lance]

General rule: don't min-raise anybody at low buy-in's SNG especially at limits as $1,
you can't bluff a donk.

preflop min-raising = only in appropriate situation
postflop min-raising = never


Personally i use preflop min-raising only in those cases
1. when i am 2nd deep stack on the bubble and have not so strong hand to push when chipleader is on my left (most of chip leaders have no balls to 3bet with marginal hand). By this i can steal blinds from 2 short stacks. It is really efficient because your min-raise is respected from UTG.
2. in situation blind vs blind
3. in heads up (in fact it is covered in 2th case)
4. "slow playing" AA, KK when i want to give a chance to reraise me. It is player dependent.

But again, this stuff i can use effectively at the limits as 6.5$ and 16$, dunno if it works at higher limits.


Maybe later when you become an experienced player and will play higher limits, you can make this move against particular player where you have reads on him and know his reaction.

[taipan168]

The SNG dos and don'ts really applies to lower buyin SNGs (say $16 and below) and against bad players at any buyin level. Bad players are generally calling stations who call too loose, and if you minraise preflop you're getting the worst of both worlds:

- If you have a weak hand and are stealing, you reduce your fold equity by minraising
- If you have a strong hand, you reduce the amount of chips that you can get into the pot by minraising.

Once you get to high buyins, I think a minraise can be quite an effective tool against the right players, but even at the $27/$38/$60 level that I have played, most raise 2.5x BB once the blinds get to 50/100.

[daviddem]

OK thanks for the advice guys. For the heads-up though, like Lance, I do minraise preflop on the button because that is what Harrington advises to do. I have not finished Collin Moshman's book and I am not sure whether he advocates the same. I also see different people play it differently, for some 2XBB is the standard, for others 3XBB. What do you guys do, and why?

[taipan168]

OK thanks for the advice guys. For the heads-up though, like Lance, I do minraise preflop on the button because that is what Harrington advises to do.

You should do things because there are good reasons to do them, not because that's what a book says! I do not like minraising heads up because opps' calling ranges tend to open up when heads up, and they get big odds (at least 3:1) to call your raise anyway.

I have not finished Collin Moshman's book and I am not sure whether he advocates the same. I also see different people play it differently, for some 2XBB is the standard, for others 3XBB. What do you guys do, and why?

If stacks are <15x BB when it gets to heads up I usually just shove whatever I want to play unless I have a monster hand, if stacks are deeper than this I usually make it 2.5x BB to go.

[Kijjo]

Heads up it's very important that you raise in position if you're playing the hand - as much or more than your opponent's standard raise - this way if you both raise an equal number of times, you're playing for the same level or larger pots while you're in position. Since the button is a position of strength, you want to play for more chips when you're in that spot.
So, depending on the opponent, if he always calls when he's on the button, I'm ok with just min-raising every time because I'm playing for pots that are twice as large when I'm in a strong position. Similarly, if he always min-raises, I'll try to stick to 3xBB.

[daviddem]

OK. Do you always raise or fold on the button? Again Harrington makes the argument that you always pretty much have the odds for a least calling on the button with a2c. So his straightforward line seems to be to call with trash, raise the good hands 2xBB (to encourage action), and raise the medium hands 3xBB (to try and terminate the action preflop). Of course at the same time blending in some variation in your game. He says if you get raised too much when you just call, then just dump the trash.

Seems to make sense to me. Although I will also dump the trash if stack sizes/blind sizes dictate. I realise this may be too basic a strategy at the higher stakes, but for the small games I play it seems to do the job.

Taipan, don't you want your opp range to open up when you bet your very good hands? In situations where my opp always attacks my calls, then I like occasionally to call with a very good hand and either call his raise or reraise to build a bigger pot. And after a good slapping, he will respect my calls much more...

In turbos, heads up, you are generally in push/fold mode, but usually not at the start of the heads up in the standards.

[Kijjo]

Do you always raise or fold on the button?

I hope you NEVER do something always.
I think heads-up is more like a dance than a chess match. You gotta learn your opponent. Hopefully playing the entire game or final table you've gotten a handle on them - but then alot of people really flip a switch when it gets to heads-up. Figure out their mindset, watch patterns and counter punch with a strategy that will cost them more in the long run.

[Extremophile]

I think heads-up is more like a dance than a chess match. You gotta learn your opponent.

A little off topic but in chess you've got to learn your opponent, too and try to play in the opposite of his style in order to make the things difficult for him. That is why it's difficult to win against Kasparov, who plays almost every style in perfection.

[RockyMoose]

Let me also add that the $1 tourneys on FullTilt are a bad idea. The fee is $.25, which is a 25% rake! At least play the $2.25 tournaments, where the rake is half that, at 12.5%, and it's just a dollar more. By moving to the $2.25s you've instantly improved your ROI by a considerable amount.

[taipan168]

OK. Do you always raise or fold on the button? Again Harrington makes the argument that you always pretty much have the odds for a least calling on the button with a2c. So his straightforward line seems to be to call with trash, raise the good hands 2xBB (to encourage action), and raise the medium hands 3xBB (to try and terminate the action preflop). Of course at the same time blending in some variation in your game. He says if you get raised too much when you just call, then just dump the trash.

I think that Harrington's advice works better in deeper stacked situations. Even in non-turbo SNGs, by the time it gets to heads up, stacks are usually relatively short (say 30x BB or less). The other thing is that at low buyin SNGs if opps will call a 2x BB raise they'll call a 3x BB raise, so if I'm not shoving preflop (with less than 12-15x BB) then I'm raising 2.5x BB whether I'm raising a good hand or a weaker one.

Taipan, don't you want your opp range to open up when you bet your very good hands? In situations where my opp always attacks my calls, then I like occasionally to call with a very good hand and either call his raise or reraise to build a bigger pot. And after a good slapping, he will respect my calls much more...

Yes, but if you are consistently raising 2.5x BB then opp will open up his range anyway and you're far better off if he is going to raise/shove over if you have a monster hand once you've raised by 2.5x BB than if you'd just called.

[birdman417]

could someone show us the calculations for the example in the OP?

[daviddem]

You mean the pot odds calcs for the flush draw?

Once villain has moved all in, there is 2000 in the pot, and it costs us 1000 to call. So pot odds are 2:1, and so we have to win more than 33% of the time for this call to be profitable. Question is: is this the case?

To simplify, let's say our flush, should we make it, wins all the time (not exactly true, but close enough for an estimate). There are 9 cards in the 47 hidden cards that give us our flush. So there are 38 cards that do NOT give us the flush on the turn. So the probability that we do NOT make the flush on the turn is 38/47=80.85%. If we do NOT make the flush on the turn, the probability that we also do NOT make the flush on the river is 37/46=80.43%. So the probability that we do NOT make the flush on both the turn and river cards is 0.8085*0.8043=0.65=65%. So the probability that we DO make the flush on either the turn OR the river is 100%-65%=35%, which is more than 33%, so the call is profitable in the long run.

To estimate these probabilities in play, there is the rule of 2 and 4 that gives you the approximate odds of making your hand on the turn or on the river. Basically, multiply your outs by 2 to find how much chance you have to make your hand on the turn and by 4 to find out how much chance you have to make it on the turn or river. So for the above example with 9 outs, multiply 9 by 4=36%, which is pretty close from the actual 35%. This rule becomes less precise beyond 9 outs and you can refine it as follows: calculate your rule of 4, then subtract the number of outs beyond 8 to get the "adjusted" rule of 4. Then divide by 2 to get the "adjusted" rule of 2. So with 12 outs, adjusted rule of 4 is 4*12=48, subtract 4 (number of outs beyond 8) -> 44%, and divide by 2 to get the adjusted rule of 2: 22%.

If you don't like calculations, you can also get a real time odds calculator like Texas Calulatem.

The above calculations are true in a cash game, where chips have a value proportional to the amount you paid for them. This is not true in a SNG, because the chips $ value is NOT proportional to the amount you paid for them, because of the pay structure of SNG's. This is the "ICM" effect (much more maths and probabilities involved here!). Basically, the less chips you have, the more each chip is worth in $ equivalent. The more you have, the less each chip is worth in $ equivalent.

[daviddem]

Let me give a quick example of the ICM effect. I know this has been done before, but it is always good to be reminded how dramatic the effect can be. Say we are at the very start of a 9 handed SNG, 1500 starting stacks. Let's say the winner takes $500, 2nd $300 and 3rd $200. Ignore the rake for simplicity. According to the ICM, each player expected value (EV), based on the size of his chip stack, is $111. Now let's say at the first hand, you stack another player (ignore the blinds for simplicity). In a cash game, your worth has just doubled. Not so in a SNG. In a SNG, according to the ICM, your EV is now $203, not $222. What happened to the missing $19? They have been split evenly between the other players. So you risked your entire $111 of EV to earn only $92 of EV... This tells you how much less the chips you earned are worth than the ones you already had: the 1500 chips you earned are worth 92/111=82.9% of the 1500 you already had.

The effect is even more dramatic on the bubble. Same example, 4 players and all have the same number of chips. Everyone's EV is $250. Now you stack a player. How much is your EV now? $383. The remaining $ have again been distributed evenly to the other remaining players (you do the job and take the risks, and they benefit! No wonder the bubble lasts a long time!). The chips you earned are worth only 133/250=53.2% of the ones you already had!

[birdman417]

thanks for the in depth examples david... this helps a lot
my biggest leak is folding to raises thinking i DONT have the odds to call but in reality i do.

[taipan168]

My classic ICM example is the following hand:

Stack sizes:
UTG 2000
Button 2500
SB 5000
BB (Hero) 4000

Standard 50/30/20 payout 1-table SNG. Blinds are 200/400, UTG and Button fold and SB shoves. SB has shoved the last 5 hands which have been folded to him.

You look down at AKo.

Due to the ICM effect, this is one of those situations where it doesn't matter whether SB is shoving only 1% or 100% of hands, it is not correct to call. This is a fold 100% of the time. In a cash game, this would be a profitable call over the long run if opp is shoving very wide, but in this case you always lose money by calling.

[daviddem]

My biggest leak is folding to raises thinking i DONT have the odds to call but in reality i do.

Don't forget thatb you also have implied odds: if you make your hand you may possibly win more chips from your opponents than he already put in the pot.

Estimating your pot odds is an important part of the game. Read all you can find about it in the articles section of FTR. Sklansky's and Harrington's books also detail all that very well (my personal preference goes to Harrington's books).

Yes this is a very good example taipan. I think if I remember that in that situation it is only correct to call with AA or KK (or QQ if opp range is really wide).

[andyjok]

If you like your books and the maths may i suggest you read kill everyone david if you havnt already.

[DrStrangeShove]

I would consider ICM and the stacks behind me. At some tables min raises mean something, some they dont. If youve got big stacks behind you that you know dont mind reraising you, I would flat call. But if you dont feel there's an aggressive big behind you, then I cant see min raising as a good thing. Usually a 3x gets the message across to those behind you that they should fold. And ICM would be a big part of the thought process. Is it early? Late? As an average stack mid early maybe a min raise gets action, and blank hits turn, youve got one or two pot commited and youve got a good shot at a big pot spot. Lots to consider here. ICM and other stacks would be my main focus.