Preflop Aggression: More or Less Luck?

[Mempho]

[ QUOTE ]
[ QUOTE ]
Depending on the situation, raising can either lower or increase the 'luck' factor. Like Andyfox says, if it's used to thin the field and increase your chance of winning a pot (albeit a smaller pot), it can reduce your variance.

But if you're raising AJs in the BB after 7 people have limped in, then it will have the opposite effect on your variance.

[/ QUOTE ]

I challenge this assertion. When you raise AJs in the BB after 7 people limp in you have an equity edge. As your equity increases, your variance isn't as negative. For example win rate 5, variance 10 = swings from -5 to 15. win rate 15, variance 10 = swings from 5 to 25. This is just an example of how increasing your win rate makes your variance less noticable, these numbers do not represent anything real.

Certainly by raising more often there are some effects which mean higher variance, but there are other effects that make that variance not as "painful".

[/ QUOTE ]

I understand the assertion you are making. Have there been any studies on how marginal preflop raises affect variance and expectation....certainly they increase expectation but do they do so to the point where they actually reduce the variance from not raising?

[sledghammer]

[ QUOTE ]
[ QUOTE ]
Depending on the situation, raising can either lower or increase the 'luck' factor. Like Andyfox says, if it's used to thin the field and increase your chance of winning a pot (albeit a smaller pot), it can reduce your variance.

But if you're raising AJs in the BB after 7 people have limped in, then it will have the opposite effect on your variance.

[/ QUOTE ]

I challenge this assertion. When you raise AJs in the BB after 7 people limp in you have an equity edge. As your equity increases, your variance isn't as negative. For example win rate 5, variance 10 = swings from -5 to 15. win rate 15, variance 10 = swings from 5 to 25. This is just an example of how increasing your win rate makes your variance less noticable, these numbers do not represent anything real.

[/ QUOTE ]

This is incorrect, because pot equity is a proportion of the pot, not a fixed number. For example, if your equity is .4, the variance will be .4*(1-.4), or .24. Playing an 8SB pot preflop, where you have .4 equity, has EV of 2.2SBs[3.2 pot share -1sb invested], with a Variance of 1.92SBs. Doubling the pot by raising doubles your expected value, to 4.4SBs, but also doubles your variance, to 2.84SBs.

[mmcd]

But if you're raising AJs in the BB after 7 people have limped in, then it will have the opposite effect on your variance.

I assume you meant AJo here. Not raising a decent suited A after 7 limpers would be borderline criminal.

[Mempho]

[ QUOTE ]
One thing left out of your analysis is that the raising got the blinds out of the pot. So fewer competitors and thus a diminished chance that a "lucky" flop will hit an inferior hand. And, in general, players in a game where they suspect the pot will be raised are less likely to limp in from early positions, also cutting down on the luck factor. The fewer players involved, the less chance a high quality pre-flop hand will be drawn out on.

[/ QUOTE ]




I certainly don't disagree...the blinds are an X factor in the equation but its not quite as simple as it may first seem. Noone doubts that raising gives you a better chance of winning and many times creates some dead money in the pot from the blinds or from the occasional extra poster. These are things that argue in favor of raising, but there is another issue that lots of preflop raising creates:

Let's look at a hypothetical deal in an online game. Let's pretend that you're a very technologically intelligent cheat [img]/images/graemlins/mad.gif[/img] and that, as a result, all of the cards on your screen are dealt face up. You are a mastermind and have virtually no chance of getting caught and therefore you are going to be able to play in a way that perfectly coincides with the fundamental theory of poker. You are in MP3 in this situation...here is what you see...

SB Q [img]/images/graemlins/club.gif[/img] 3 [img]/images/graemlins/spade.gif[/img]
BB J [img]/images/graemlins/diamond.gif[/img]9 [img]/images/graemlins/diamond.gif[/img]
UTG 5 [img]/images/graemlins/spade.gif[/img] 8[img]/images/graemlins/diamond.gif[/img] (He correctly folds immediately)
UTG+1A [img]/images/graemlins/heart.gif[/img]J [img]/images/graemlins/heart.gif[/img] (He is passive preflop and limps)
UTG+29 [img]/images/graemlins/spade.gif[/img]6 [img]/images/graemlins/club.gif[/img] (Correctly folds)
MP1 Q [img]/images/graemlins/diamond.gif[/img]T [img]/images/graemlins/spade.gif[/img] (Correctly folds)
MP2 A [img]/images/graemlins/spade.gif[/img] 7 [img]/images/graemlins/diamond.gif[/img] (Correctly folds)
Hero A [img]/images/graemlins/diamond.gif[/img]K [img]/images/graemlins/spade.gif[/img]

Well, since we're cheating here we should label ourselves villain [img]/images/graemlins/grin.gif[/img], but now we've got to decide what to do. We look behind us and see:

CO 9 [img]/images/graemlins/heart.gif[/img]8 [img]/images/graemlins/heart.gif[/img] (he'll call a bet but not cold call)
Button K [img]/images/graemlins/diamond.gif[/img]J [img]/images/graemlins/spade.gif[/img] (hmmm, he would play that unraised from the button)


And then the blinds...SB is folding no matter what and the BB is calling no matter what.

So...do you want to fold, call, or raise now that we've seen the stone cold truth? We have the best hand, so we're not going to fold. Let's see how this works. If we raise, we are going to drive out button and the cutoff, the SB has already checked the fold box and the big blind is calling anyway.

So first we look at what a raise accomplishes. We raise and it gets folded to the BB, who calls, then UTG+1 calls. So, its 3 to the flop for 2 bets a piece and 6.5 SB are pulled to the middle.

What is our equity in this situation?

Our hero wins 56.21% against these 2 opponents.


So, we have 56.21% of a 6.5SB pot or a 3.65 SB slice...less the 2 bets of our own and we have a positive expectation of 1.65 SB.


Now, if we don't raise, the button and the cutoff will come in...we will have 5 people to the flop for 5.5 SB, thus:

UTG+1 22.19%
Hero 41.28%
CO 19.22%
BT 4.79%
BB 12.51%

So you have a 41.28% slice of 5.5SB, or a 2.27 SB piece. If you subtract your original bet, you have a positive expectation of 1.27SB.

So your profit from raising is only an additional 0.38SB in this situation because one of the byproduct of raising is to discourage dominated hands from making calls in position.

In this particular scenario, you would probably be better off not raising because you have some implied customers if you hit your hand...meaning that the pots you do win (which will be fewer) should be larger due to the implied payoff from not raising...specifically when one of our limpers we let in flops top or middle pair.

[mike l.]

.

[andyfox]

Good stuff there.

Your analysis indicates what percentage of the time you will win with the best cards. It doesn't indicate how many times you can win without the best cards; whatever that percentage is, it will be a much higher, one would think, without the two players behind you. The flop will miss everybody more often with just three of you and you'll be more confident betting it when they check to you than you will be with two others still to act behind you. So it's not just raising the blinds out, but raising the players out behind you that should be considered. This is especially crucial as one move up in stakes and a larger percentage of the pots are won not by having the best hand on the river but by having the only hand before the river.

BTW, when I plugged in the cards into Twodimes, it showed Ad-Ks winning 50.92% against Jd-9d and Ah-Jh.

[mike l.]

"This is especially crucial as one move up in stakes and a larger percentage of the pots are won not by having the best hand on the river but by having the only hand before the river."

maybe it was a misperception on my part, but it seemed like there was a dip in the number of showdowns around the 20-40/40-80 level but then that table showdown percentage went back up as i played higher. does that make sense?

[andyfox]

I've played minimally at 80-160 and 100-200. You might be correct for So. Calif., especially Commerce. I would suspect not for Las Vegas, but hopefully others can chime in here, and for online too. Regardless, I think when you lose you lose more with the two players behind you, so when we say we win 56.21% (or whatever) of the time, I think we lose more $ with the two behind than we do, proportionately, when our raise buys us position. That, I would think, is especially true as you move up and the players are trickier and more aggressive, especially with position on the turn and river.

[Noo Yawk]

[ QUOTE ]
"This is especially crucial as one move up in stakes and a larger percentage of the pots are won not by having the best hand on the river but by having the only hand before the river."

maybe it was a misperception on my part, but it seemed like there was a dip in the number of showdowns around the 20-40/40-80 level but then that table showdown percentage went back up as i played higher. does that make sense?

[/ QUOTE ]

I actually have a theory about this. I found that the 20-40 to 40-80 level games are the first level of pyschololgical barrier games, where people can lose thousands, rather than hundreds fairly easy, causing them to play weak-tight. The ones that get used to it, take advantage and beat the crud out of these players. When they move up, they refuse to fold as they believe everyone's trying to bully them the same way. Just an opinion I've formed while I was bored. Maybe it has some merit, maybe not.

[Mempho]

First of all, sorry I took so long to get back to you. I had a busy, busy weekend.

[ QUOTE ]


BTW, when I plugged in the cards into Twodimes, it showed Ad-Ks winning 50.92% against Jd-9d and Ah-Jh.

[/ QUOTE ]

I ran it throught PokerStove...I included the dead cards since we knew the entire deal here.



[ QUOTE ]
Your analysis indicates what percentage of the time you will win with the best cards. It doesn't indicate how many times you can win without the best cards; whatever that percentage is, it will be a much higher, one would think, without the two players behind you. The flop will miss everybody more often with just three of you and you'll be more confident betting it when they check to you than you will be with two others still to act behind you.

[/ QUOTE ]

Granted...there is no doubt that the PFR gives you postflop leverage to steal some pots that are not mathematically yours. I do not know the average win% on a continuation bet (TAG stats) with AK. Our win percentage on the bluff against two opponents is considerable, but it is certainly not going to be a huge percentage. Of course, our bluff percentage on a 5-handed flop drops exponentially...we have virtually no chance of winning uncontested. I'm sure that you will agree that like everything in poker, there is a tradeoff here. You know that AK has a lot more postflop options unraised, of course. Also, when you are outflopped, it is very easy to get away from.

Some basic flops with AK are:

You flop TPTK and are not extremely vulnerable such as a flop of A 9 2 rainbow

You flop TPTK and are vulnerable such as a 2-flush of A 9 2

You flop TPTK and are extremely vulnerable such as a 2-flush of AJT

You flop a monster such as AKK or QJT

You whiff completely such as 9 8 5

You whiff but you've got a nice draw such as Q J x rainbow or J x x monotone where you have the ace of trump.

You flop TPTK and are losing with little hope of winning (up against a set or something like that)

You flop TPTK and are losing but have a decent draw such as someone flopping bottom 2 pair (where you have the 5 outer plus some counterfeit out potential that can be picked up on the turn).

If you flop a hand with AK in an unraised pot, you can cause opponents to make some huge flop and turn mistakes even if they are on good draws. It is more difficult to do this if you are the PFR (maniacs are an exception).

If you miss the flop with AK, you can muck it quickly in an unraised pot if it becomes immediately clear that you are beaten.

If you're up against a set with TPTK...you're going to lose either way.


Anyway, I'm not advocating not raising with AK at all. I'm just stating that not everything is as clear cut as it may seem. Poker can be played from varying angles of attack and there are times when you are better off taking the opposite of the textbook approach. I nearly always raise with AK...but there are situations where it may be clearly better to call, especially if the following conditions are met:

a) You have almost no fold equity...i.e., players have already entered the pot and one of them is almost sure to see you to the showdown
b) The players yet to act behind you will play hands that you will beat to a pulp when you both connect. This turns AK into more of an implied odds hand.
c) It could have metagame benfits.

That said, these situations are difficult to judge and you could almost never be wrong in making the opening raise with AK.

Of course, the point of this post was in relation to preflop aggression. I think that AK is pretty much an automatic raising hand preflop. Of course, we know that many of the hands raised in today's midlimit online games are not nearly as good as AK.

The main hypothesis of my post was this:

If you get into a game that is so aggressive that it requires you to put in a lot of money with marginal hands preflop, you are going to have higher variance outcomes and if you are a good postflop player, it will lower your expectation. The only leverage you have is to get your opponents to lay down incorrectly. This is often expensive and high variance. It does work, however. More often, though, your postflop reading skills are neutralized in the same way that a no-limit expert's postflop skill is nullified by "sliders" in tournament play.

So, how important is preflop passivity in relation to the following variables?

1) VPIP
2) Avg. Pot Size
3) Passivity