Next time, I'm playing differently against likely flush draw

[M.B.E.]

From a one-table satellite (winner gets $215 seat; second place gets a bit of cash, I think $28). Down to four players and I am a very close second. Rounded to the nearest hundred, standings are:
me (UTG): 4500
button: 4700
SB: 2700
BB: 1500

Blinds 75/150. I raise to 450 with Q/images/graemlins/club.gifQ/images/graemlins/heart.gif. The button (chip leader) calls. Blinds both fold.

I have been raising a fair bit. By way of background to this hand, four players were eliminated during the first ten minutes (two of them by me). So most of the tourney we've been playing shorthanded. I was the chip leader for most of the time, but the other player (now the button) just eliminated someone.

Flop comes 9/images/graemlins/heart.gif2/images/graemlins/heart.gif2/images/graemlins/club.gif. I bet 550 into the 1125 pot. My opponent minraises to 1100. Frequently a minraise on the flop indicates a very big hand, but I don't think this player would have called me preflop with A2s or something like that. 99 and 22 are unlikely. AA or KK are possible, I guess, but I think the player would have reraised preflop (and raised more on the flop). So that leaves hands like JJ, TT, 88, A9s, and a heart draw. All in all, I'm pretty sure that I'm currently in the lead, and I think my opponent's most likely hand is a heart draw. The flop minraise is designed to get a free card on the turn. Reaching this point in my thought process, I reraised all in. My opponent thought for quite a while, and called the remaining 3000. (His pot odds on this call were 2.1:1.) If my opponent lost, he would be left with only 180 in chips -- a distant fourth place.

My read was correct: my opponent had A/images/graemlins/heart.gifJ/images/graemlins/heart.gif. His flush didn't come in, but an ace hit on the river to beat my QQ.

With an 11-outer, my opponent was correct to call the flop. At that point I was only a 3:2 favourite. I hoped he'd fold of course, especially considering that he only had me barely covered and if he folded now he'd still be in second place. If it had been a normal SNG with a 50/30/20 prize structure, then he might be correct to fold to my all-in reraise (anyone want to do the math on that?) but in this satellite, a fold by my opponent would be a big mistake, unless he put me on AA, 99, or trip deuces -- and why would he?

Next time I'm in a situation like this, I won't reraise the flop, but just call, then move in on the turn if it isn't a heart. At that point if he wanted to call me he'd be making a mistake.

Comments?

[Jonathan]

MBE, I think your analysis and conclusion are spot on!
While I was reading your post, I read the hand the same way you did, and after his miniraise, I thought to myself "push".

But just calling and waiting for the turn is clearly the better play. I guess our thinking at the time of moving all-in simply didn't go far enough. Prior to going all-in, we should ask ourselves, "If my read of the opponent is correct, is it correct for him to call or not?" Then we might find that by just calling ourselves and waiting for the turn, we construct a situation where he is more likely to make a mistake.

Instructive hand.

Your post raises a related issue. You said that when you push you are hoping he will fold. I see this kind of statement all the time on this forum. But this kind of "wishful thinking" can be expensive in poker. Rather than "hope" we should analyze what the correct play is for our opponent and assume he will choose that line, much as a chess player or bridge player assumes the correct line for his opponents play. After all, if your opponent is frequently making errors, then he's not much of an opponent, is he?

Regards,
Jonathan

[PrayingMantis]

Few thougts:

You bet half the pot on the flop. Are you sure this is the best move? If you know 1) most chances are you're ahead, 2) you're the PF raiser, 3) you're first to act, why not make it a bigger bet, pot bet, to reduce his calling-odds significanly, and put him to a decision?

I'm not fully convinced about calling his mini-raise and waiting for the turn to make a move. Obviously, if there's a good option he's on 88,A9,TT,JJ - pushing on the flop is the right move. Only if it's significantly more probable that he's mini-raising with a draw, calling now and pushing on the turn, if you like it, *might* be better, although you could face a difficult situation then, and that's why:

I believe that an opponent that is capable of semi-bluffing, like here, is probably able to bluff bigger. It is possible that you have this read on him. Against an opponent like that, I wouldn't like to see another card, since if I'm going to check/fold if a flush card or A hits, he'll probably be ready to take advantage of it *even if he doesn't hit anything*, and push me out of the pot if I check, no matter what he has. And you give him a chance to do so.

So, if you're ready to call now, and check/fold to his potential bluff later, I think pushing on the flop is actually better, despite the fact that he's getting right odds (in this scenario). This is of course very dependant on your read of him.

Another point is that there are opponents who *wont* call your flop-push, with a draw, although they are getting good odds (if they knew what you have). He's obviously not one of them, but many aggressive opponents will rather look for what they might think is a better spot, especially if they are in a good stack shape. That could makes your push still the better move - although you prefer he wouldn't call.

[La Brujita]

I have to question something and I hope it doesn't show any unsophistication on my part. Since this is basically a winner take all we can assume you play this much like a cash game.

We can agree the opponent most likely won't lay down his hand due to the pot odds; that would be a serious mistake on his part.

As I read it you have about 3000 chips after his mini raise. You are a 60.1% favorite. If you knew what he was holding and knew he was going to call aren't you happy to have your money in as this much of a favorite? In other words if your choice is get the 3000 in as a 60/40 favorite shouldn't you do it. He may be getting correct odds to call (I haven't checked) but you are getting a huge cushion on correct odds. In other words you are getting 60/40 when all you needed is 40/60 or whatever. I have made this point before and haven't created as much dialogue as I hoped. Two people in a hu pot often both have correct odds to call due to leverage from other money in pot but either they will both have no cushion or one will have some cushion in being in a pot. That person in essence is in a better place or made a "better" decision.

The only reason I see to not push in is if you think you can put yourself in a position later to get those chips in as a bigger favorite, whether it be in this hand or another. This is a separate point which I am not addressing, but I will just say 60-40 is a pretty decent favorite.

I guess I want to try and say this another way. Forget about his pot odds for a second. If you know your opponent is going to call when you are a 60:40 favorite and you don't get his chips in the middle aren't you making a serious FTOP mistake in cash game with no rake? Isn't this close to analagous?

I guess what I think honestly is you played it correctly and are second guessing yourself because you were outdrawn.

Edit to say I erased my comment about two dimes because it is not working properly right now.

[M.B.E.]

[ QUOTE ]
You bet half the pot on the flop. Are you sure this is the best move? If you know 1) most chances are you're ahead, 2) you're the PF raiser, 3) you're first to act, why not make it a bigger bet, pot bet, to reduce his calling-odds significanly, and put him to a decision?

[/ QUOTE ]
I'm pretty sure that my initial flop bet of half the pot was better than a pot-size bet. At the time, I figured that not only did I have the best hand, but my opponent was very likely drawing to two outs (perhaps three). If my opponent has a medium pocket pair like 55, 66, or 77 (which I think is very possible considering he called my preflop raise), I'd expect him to fold to a pot-size bet. But if I make a smaller bet, he's likely to call (putting me on AK or the like), and he might even raise.

Since I figure he's drawing so slim, I want to get some of his chips; I don't want him to fold. (At this stage there's no reason to put him on the heart draw; only after he minraises does that become likely.) Even taking his implied odds into account (since I will pay him off if he has a medium pocket pair and hits his card for a full house), I'd rather have him call my bet of 550 (giving him implied odds of 9.4:1) than fold to a pot-sized bet of 1125 (in which case his implied odds would be 4.6:1).

On the other part of this hand, my play after the opponent's minraise, I'm going to respond a bit later. I'm trying to work out some EV calculations, based on reasonable assumptions. Thanks for all the thoughtful comments so far on this thread.

[Che]

La Brujita-

[ QUOTE ]
You are a 60.1% favorite. If you knew what he was holding and knew he was going to call aren't you happy to have your money in as this much of a favorite? In other words if your choice is get the 3000 in as a 60/40 favorite shouldn't you do it. He may be getting correct odds to call (I haven't checked) but you are getting a huge cushion on correct odds. In other words you are getting 60/40 when all you needed is 40/60 or whatever. I have made this point before and haven't created as much dialogue as I hoped.

[/ QUOTE ]

I'm guessing that the lack of dialogue is the result of there being no reasonable arguments against your position. /images/graemlins/grin.gif Calling when you could get the chips in as a big favorite is clearly an FTOP error.

In a typical tournament someone could argue that you reduce variance by waiting out the turn card (if you know opponent will call) and then pushing (assuming opponent will then fold), but even that doesn't apply here since surviving a little longer has no value in an effectively winner-take-all tournament.

As I see it, whether the opponent's play is an FTOP error or not does not matter. I can only play my cards, and my goal is to play them correctly (i.e. make no errors myself and induce errors when I can ). As it turns out, MBE was not able to make his opponent make an error, but he has no control over his opponent's play so whether the opponent's play or MBE's is "more correct" isn't relevant IMHO.

But, if you and I are both wrong, I hope there will be some dialogue this time. /images/graemlins/shocked.gif

Later,
Che

[M.B.E.]

I did some experiments with this problem, and I now agree with the
conclusion of PrayingMantis, La Brujita, and Che: on the facts of this
particular hand, the better play was to reraise all-in on the flop, not call
the flop raise with the intention of folding the turn if a heart or ace fell.

In other words, I played the hand correctly, so my initial post (saying
that I should have played it differently) was wrong.

However, I still disagree with some of the reasoning in La Brujita's and
Che's posts, which I will discuss below.

To calculate, I made the assumption that I knew exactly what my
opponent held (A/images/graemlins/heart.gifJ/images/graemlins/heart.gif) and my opponent knew exactly what I
held (Q/images/graemlins/club.gifQ/images/graemlins/heart.gif). This makes the calculation very easy of course. It
is not all that unrealistic in this particular case, because I was already
fairly certain that my opponent was on a flush draw (for reasons
explained in my initial post) and since he called my preflop raise an
ace-high flush draw was most likely. I also tried another calculation,
where I made allowances for my opponent possibly holding other hands
like K/images/graemlins/heart.gifJ/images/graemlins/heart.gif or T/images/graemlins/spade.gifT/images/graemlins/heart.gif, and reached the same result. (I
won't reproduce that other calculation here.)

Now the flop is 9/images/graemlins/heart.gif2/images/graemlins/heart.gif2/images/graemlins/club.gif and when the action gets back to
me after my opponent's minraise there is 2775 in the pot and it's 550 to
me to call. I have 3500 remaining in my stack. (I began the hand with
4500.)

So, let's calculate my chip-EV, from this point forward, of two different
strategies. Strategy One is to push all-in, assuming my opponent will
call. Strategy Two is to smoothcall, and push in on the turn unless my
opponent hits, in which case I check-fold. I'm assuming my opponent
will not call my turn all-in.

With Strategy One, 60.1% of the time I win 5725, and 39.9% of the
time I lose 3500. EV: +2044.2.

With Strategy Two, 34/45 of the time my opponent will be folding on
the turn, meaning a win for me of 2775. The other 11/45 of the time I
will fold on the turn, meaning a loss of 550. EV: +1963.7.

So the EV of Strategy One exceeds the EV of Strategy Two by 80.5
chips.

But that's just a function of my stack size in this problem. Despite this,
it is not true that I always want to get my chips in as a favourite
when I will be called.

For example, suppose I had started this hand with 700 fewer chips
(3800 rather than 4500). Now by the time the action gets back to me
after my opponent's flop minraise, I have 2800 remaining (not 3500).

Now with Strategy One, 60.1% of the time I win 5025, and 39.9% of the
time I lose 2800. EV: +1902.8.

With Strategy Two, even with my reduced stack size my opponent still
won't be able to call my all-in on the turn with his nut flush draw (he'd
be a 3:1 underdog getting only 2.48:1 pot odds). So just as before, 34/45
of the time I will win 2775, and 11/45 of the time I will lose 550. EV:
+1963.7.

Thus with my stack at this level, Strategy Two has greater EV. It would
not be an FTOP-mistake to call the flop minraise as a favourite
rather than reraising. This is because now, the amount I can earn by
reraising as a favourite on the flop is not enough to offset the amount I
can save by folding on the turn when I am beat.

[PrayingMantis]

This is quite funny, but after you agreed with my opinion about the all-in on the flop, I now do not agree with *your* specific analysis and conclusions, but still insist on my first position. I will explain.

You build your analysis upon these 3 assumptions:

A. He has the flush draw.

B. He'll call a reraise all-in on the flop.

C. He'll fold to an an-all in on the turn, if you only call his mini-raise on the flop, AND a flush card doesn't hit.

With these 3 assumptions, it's true that you'll gain 80.5T more with strategy 1 (raising all-in) than with strategy 2 (calling on the flop).

HOWEVER, you gain these chips for a a 40% risk of busting! While in strategy 2, your risk of busting is practiacally 0.

So, I believe that depite the fact that it's a winner-takes-all, taking 40% risk of busting for merely 80.5 more chips, is a mistake. That's why I think that if you are sure of your 3 assumptions above, strategy 2 (calling on the flop, etc) is much much better.


BUT, looking at this situation from a wider and more realistic perspective (IMO) I still think strategy 1 is better. That's because of THESE 3 assumptions:

A. You cannot be completely sure he's on the flush draw, or rather on hands like 88, TT, JJ, 9A, other low pairs or unimproved Ax with one heart.

B. You cannot be sure he'll call your re-raise all-in on the flop, EVEN if he holds the flush draw.

C. You cannot be sure he wont push you out of the pot (without the flush) if you call on the flop, and then check to him if a flush card hits on the turn.

With these assumptions (that I mentioned in my original reply) pushing on the flop is the best move, with best chances of taking the pot right there, without giving any hand he holds a chance to see the turn cheaply. If he *does* call you with the flush draw (as he probably should), it is still significantly +EV for you, as you showed in your analysis.

It is also important to remember that an over-pair like QQ here is a nice hand, but very vulnerable. I believe you should play it strong and simple.

To conclude: we agree now about the better strategy, but apparently from rather different reasons. Well, that's why it's an interesting hand.

[M.B.E.]

Praying Mantis -- this is getting to be a peculiar thread. (But in a good way!)

[ QUOTE ]
HOWEVER, you gain these chips for a a 40% risk of busting! While in strategy 2, your risk of busting is practiacally 0.

So, I believe that depite the fact that it's a winner-takes-all, taking 40% risk of busting for merely 80.5 more chips, is a mistake. That's why I think that if you are sure of your 3 assumptions above, strategy 2 (calling on the flop, etc) is much much better.

[/ QUOTE ]
Hmmm, if this were a 50/30/20 SNG you have a point. But I don't think it's relevant in a 90/10 SNG (essentially winner-take-all) like this.

[ QUOTE ]
BUT, looking at this situation from a wider and more realistic perspective (IMO) I still think strategy 1 is better. That's because of THESE 3 assumptions:

A. You cannot be completely sure he's on the flush draw, or rather on hands like 88, TT, JJ, 9A, other low pairs or unimproved Ax with one heart.

B. You cannot be sure he'll call your re-raise all-in on the flop, EVEN if he holds the flush draw.

C. You cannot be sure he wont push you out of the pot (without the flush) if you call on the flop, and then check to him if a flush card hits on the turn.

[/ QUOTE ]
Yes -- these are good points. I think I agree with you that in a cash game or winner-take-all tourney these factors are usually going to be decisive in favour of Strategy One, even though under certain stack sizes, Strategy Two would yield higher CEV if we are 100% certain of our read and our estimation of how the opponent will play.

But we're never 100% certain of what our opponent holds and how your opponent plays. In this particular case, I think I could be about 80% sure that he was on the flush draw (because minraising would be such an odd way of playing TT or the other hands you mentioned). Also, if he was not on the flush draw, I think it's pretty unlikely he would bluff on the turn if a flush card hits (say less than 20% chance he would try this). However, even though there's let's say just a 4% chance of my getting bluffed out if I play Strategy Two and a flush card hits (20% of 20%), the result of that when it does happen is very bad. In short, I'd really have to be over 90% sure of both my read and my estimation of how my opponent will play the rest of the hand before I start thinking about Strategy Two in a case like this.

(On the other hand, if it were a 50/30/20 SNG on the bubble, it might frequently be correct to employ Strategy Two -- but only for variance/survival reasons, not for the reasons in my earlier posts in this thread which were about maximizing CEV.)

[La Brujita]

I have to admit manipulating pot odds is what I consider probably the most complex part of HE right now. I see MBE's point in a situation where you "know" what the other player is holding.

I have to mull over the math and the comments but this darn sure is an interesting thread.

[M.B.E.]

Just one more point. Up to now we've been assuming (or at least I have) that on the turn my opponent will fold a flush draw if I still have enough chips to make a large bet.

If that assumption is wrong, i.e. if there's a good chance that my opponent will still call my bet on the turn with only a flush draw, then Strategy Two increases its attractiveness.

There are some players online who just do not like to fold the nut flush draw, even on the turn, even to an all-in bet. If I've observed this particular opponent to play like that, and I'm quite confident (85% or more) that he does now have a flush draw, surely then Strategy Two is going to be better. Because the times when my hand turns out to be good, I still get all his chips (with either strategy). The times he hits on the river, he still gets all my chips (with either strategy). But the times he hits on the turn, with Strategy One he'll get all of my chips but with Strategy Two he gets only a few of them.

[PrayingMantis]

[ QUOTE ]
Just one more point. Up to now we've been assuming (or at least I have) that on the turn my opponent will fold a flush draw if I still have enough chips to make a large bet.

If that assumption is wrong, i.e. if there's a good chance that my opponent will still call my bet on the turn with only a flush draw, then Strategy Two increases its attractiveness.

There are some players online who just do not like to fold the nut flush draw, even on the turn, even to an all-in bet. If I've observed this particular opponent to play like that, and I'm quite confident (85% or more) that he does now have a flush draw, surely then Strategy Two is going to be better. Because the times when my hand turns out to be good, I still get all his chips (with either strategy). The times he hits on the river, he still gets all my chips (with either strategy). But the times he hits on the turn, with Strategy One he'll get all of my chips but with Strategy Two he gets only a few of them.

[/ QUOTE ]

Yes, this is a good point. Only that there's one tricky problem with this too: a player like you describe here, is obviously a pretty bad player. Many bad players will mini-raise you on the flop with a variety of hands. I'd say that a player who's willing to call all-in with his flush draw on the turn, is a player who can mini-raise you with 77 on this flop: he just believes it's a reasonable move. I think that against bad players strategy one is better, if only for one reason: strategy two is more sophisticated than strategy one. I don't think you should play sophisticated poker against bad (and many times unpredicatable) players, who can hit their set of sevens (say), on the turn, and you'll never know it until your whole stack is in the middle...

However, if you do think he's on a flush draw, and capable of calling all-in on the turn with only 4-to-a-flush, then strategy two is surely the best way to go.

[Che]

[ QUOTE ]
This is because now, the amount I can earn by
reraising as a favourite on the flop is not enough to offset the amount I
can save by folding on the turn when I am beat.

[/ QUOTE ]

Excellent work, MBE. You identified an error in my thinking and I appreciate that. Thank you.

Even after seeing your analysis, I still didn't believe that calling *could ever* be the best play. Just doesn't make sense. /images/graemlins/crazy.gif Calling and pushing had already been covered (and folding would be moronic) so the only remaining option is to raise less than allin.

As it turns out, there is a range of reraises that yield an EV greater than the EV of call-flop/push-turn IF two assumptions hold (in addition to MBE's original EV calculation assumptions):

1. Opponent will fold the turn anytime he does not have odds to call.
2. Opponent will only call our reraise. He will never push.

So to start, I calculate the maximum amount we can raise which will leave us enough chips to push the opponent off his hand when the turn misses. If the turn is a blank (I used 7 /images/graemlins/diamond.gif), we are a 75/25 favorite. So, the upper limit for R, the amount we raise above the 550 call, is the solution to this equation:

(2950-R)/(3325+2R)=1/2 and R=643.75

I use 640 in my EV calculations for the sake of round numbers and the EV is:

(34/45)*3415-(11/45)*1190=2289.33 which is 215.13 greater than MBE's allin calculation.

Unfortunately, the opponent will never call this raise since pushing then becomes +EV relative to calling (from his perspective).

Opponent EV (call)=11/45*3965-34/45*640=485.66
Opponent EV (push)=.399*6275-.601*2950=730.78

So, the opponent's EV loss from calling rather than pushing(215.12) offsets my EV gain (215.13) as it should (I assume the difference is rounding error).

As a result, the opponent should push rather than call my allin. I would rather call his minraise than get allin on the flop so I can't raise. Thus, I end up with the original result: call the minraise rather than reraise (when the numbers indicate calling is better than pushing as in MBE's "700 fewer chips" example).

I haven't done the math, but I'm guessing that a smaller reraise amount exists that would be slightly more +EV than calling yet would not be so large that the opponent would just push in. My assumption is that the dead money would allow this result to occur. However, I'm also guessing that this amount is less than 550 so the theoretically correct reraise would be illegal.

My conclusion is that IF I *know*:

1. Opponent has a draw he will not fold on the turn.
2. I know what his outs are and can fold when he hits the turn.
3. He will always fold the turn if he misses and I kill his odds.
4. He will call a raise less than allin rather than pushing.

THEN I should raise by amount R as determined above.

Will I ever *know* all these things? No, but at least I understand which plays are better than others so that I can make good decisions relative to my expectations of my opponent's likely responses.

Thanks again, MBE.

Che

BTW- Is this (calling with a made hand rather than raising) an example of the "reverse implied odds" of a made hand? I've never understood that concept so I thought that might be what's happening here. Just wondering...