LO8 Tough river decision?.....Maybe not.

[cgkid]

bbart- I agree 100% with your anaysis. I don't think raising was the best play and you pointed most of the reasons why. nice post.

Is folding really that exploitable in this situation? The pot seems to be pretty protected by the two players behind me who have yet to act an will be getting a very good price on a call. HU it's definitely exploitable, but four handed?

[bbartlog]

I agree folding seems best. As for the exploitability, yes, with two others also in the pot it seems like that's not such a worry. Especially if they are likely to call without the nuts. I was just using the term in a technical, game-theory sense: if you fold in this spot when you don't have the nuts (100% of the time), then BB theoretically gains by promo-betting some set of high hands at least some of the time. In reality the play errors at low limit lean so much towards calling incorrectly (rather than folding incorrectly) that BB will basically never be trying that here, though.

[RobNottsUk]

Was the turn bet sound?

Folding "toughly" for 1 bet at 5.5-1 on the river, is inviting poor loose players, to throw in future bluffs on scare cards, and get tricky with you. The BB only needs 25+% chance of success to make the proposition profitable, so they'll be tempted to take an optomistic shot on the river.

By my rough reckoning, there's 15 cards out of 44, 15:29 making it only 51.7% to make the Lo. The perfect 3 gives you a wheel, but it's the 3rd nut str8 (76543,65432), and one of those players behind can easily have A2 to split with you. If the pot were smaller, then the river fold is clearer. There's no fold equity, with that turn card offering so many Hi possibilities, and opposition expecting you to be betting an A2 draw.

So was betting the draw + overpair actually a small mistake, and over-playing the hand? You could very easily be 2nd best for Hi already, passive players give too many free cards.

On a loose-passive table, it's quite likely to be check-ed behind, yet they'll still put in crying river calls. Should someone bet behind you, you would actually gain some information; and the non-pfr betting tends to look stronger, especially if they're not a "folder".

Is this clear cut, or close?

[cgkid]

I think the turn is a clear value bet and I'd bet A2Kx in this spot when checked to as well. The T a good card for my hand and my equity against their range here is easily over 33%. These are loose players whose range is much larger than wraps, two pair+ and nut low draws. My river play was bad but I don't question the turn.

[RobNottsUk]

When I first saw the hand, I thought so to. But thinking about it more, I think you have a problem against hands like A2KT, A23K etc who could easily have called your pfr, and the flop 'cbet'.

If you get raised behind and there's only heads up or 3-way action, facing a likely split, you've lost money on the turn.

If you had prospects of a scoop, it's a clear bet, but it looks like there's no hi cards you want to see on the river, no medium cards give you a scoop and the perfect '3' is unlikely to bet best against a loose field, as the K54 flop put the top end of the wheel on the flop (meaning a better str8 is very likely).

You may have 'best hand' on the turn (against their wide ranges), but if there's only a few great river cards for you, where's the value?

I think perhaps you like a smaller pot on the river (so you fold without the nut Lo), or a much bigger one, where the info gained by bet behind you, makes it more likely you can close the river betting with better odds rather than being in the middle.

I'm not saying the turn bet is a bad mistake, I just wonder about it, it seems analagous to some 'best hands' on flops that should fold without Lo potential, due to risk of outdraw and likely split.

I think if you had AA23, the extra 5 Lo cards, and the improved wheel possibility make it much clearer to bet the turn.

[Buzz]

Rob - You had me agreeing with you for a few minutes.

But if the board just pairs, Hero will scoop unless an opponent makes trips. For example, if the board pairs with kings on the river, with twelve cards between them, one of Hero's three opponents figures to make trip kings about five times out of eight.

the math set up: P=1-C(41,12)/C(44,12)

But Hero is not going to collect from anyone on the river if the board pairs and he has a winner. Indeed Hero will probably lose a bet on the river, paying off somebody who makes trips if the board pairs. (Depends of course).

Thus by betting here, Hero gets his river pay-off for the twelve unseen cards that will pair the board on the river and when Hero's two pairs win. For each of these twelve unseen cards, Hero figures to win (assuming no bias for or against these twelve cards) three times out of eight.

Assuming all three of these players are expected to call the third round bet (as they indeed all did), then when the board pairs on the river, five eighths of the time Hero loses one bet from this round, but three eighths of the time, Hero collects a total of three bets from this third betting round.

(3 bets)*3/8-(1 bet)*5/8=+4/8 bets or +0.5 bets.

And that’s for each of the 12 cards that pair the board.

But we really need to look at all 44 of the possible river cards, ascertain Hero’s chances of winning and losing, and how much Hero will win or lose, then total the result, in order to know if a bet is favorable or not.

For all of the possible river cards, I estimate as follows:<ul type="square">12 pairing cards, +6
4 treys, +8
11 low cards, +2
2 aces, +2
15 other cards, -15[/list]
I’ve already explained how I got +6 for the 12 pairing cards.

For the 4 treys, hero makes a wheel, but will get tied by another wheel approximately two times out of five and will (very approximately) split with high about another overlapping two times out of five.
It’s very approximate because we don’t know if anybody will see the flop if dealt 76XY or 62XY. For most, I’d guess if would depend on the XY component of the hand, perhaps their position, and their general philosophy of seeing flops. (And of course there’s no way to know that with unknown opponents, and even very difficult to know for certain if you’ve been sitting at the table with them for a couple of hours. After this flop, anybody who saw the flop with 76XY probably continues, but maybe not everybody who saw the flop with 62XY. Again, it depends on XY and other factors.
At any rate, very approximately for each the 4 treys, <ul type="square">1/5 scoop
2/5*3/5 split
2/5*2/5 get quartered
2/5*3/5 scoop
2/5*2/5 win 3/4
for a total of
4/25 get quartered
6/25 split
4/25 win 3/4
11/25 scoop

Then 0*4/25+1*6/25+2*4/25+3*11/25=
0+0.24+0.32+1.32=1.88

And finally 1.88*4=+7.52 (which I rounded to +8 the first time through, and which is close enough).[/list]
For the 11 other cards that make the nut low for Hero, Hero will get tied by another nut low about two times out of five (a bit less, but getting occasionally sixthed makes up the approximate difference). I didn’t figure this the first time through, but now I will.
1*11*3/5+0*11*2/5=11/5. Hero only makes about +2 for those eleven cards, rather than the +11 I assessed the first time through. Sorry.

Finally the two aces are tough. Anyone loose enough to see the flop with 23XY probably is loose enough to continue after this flop. (Of course 23AY or 234Y is not, by my standards, very loose at all). But is anyone who saw the flop with QJXY continuing after the flop? (I wouldn’t generally think so, but you never know). And you beat everyone who saw the flop, made a set and continued (but I’d have expected a raise on the flop and/or turn from any set playing the hand). Thus a set is not likely, but is still possible, but an ace on the river making you top set beats it anyway. Two pairs is very likely, but your top set beats that too. But your set of aces doesn't beat a wheel. And that is what it looks like when BB bets the river.

It occurs to me that I made a mistake on my first post in this thread, I’ll go back and correct that. (The one neat thing about being mod is I can go back and correct those errors after the time has expired, without having to post a separate correction post. And I'll at least leave an indication that I edited.)

At any rate, let's say from your perspective on the turn, you win half 2/4 of the time with aces get scooped 1/4 of the time, and scoop (you still have 27 for an "emergency" low) 1/4 of the time. That's admittedly very crude. I’m just going to give each ace +1*1/2-1*1/4+3*1/4=+1. (I originally assigned +2 each, with a question mark, but I think +1 is better).

Finally, for the 15 other cards, by betting on the third betting round, Hero loses one bet on the third betting round.

Totaling, +6+8+2+2-15 =+3. Since the total is positive, Hero should bet the turn.

Buzz

[RobNottsUk]

Hmmm well, I did consider that. But the board pairs and now you're calling hoping it's a bluff. Many loose low stakes Tricksy-Fish players have a betting reflex on any scare card, that gives them a chance of stealing a pot by a bluff.

In Hold'em it's really iffy at full ring multi-way, to rely on trips not being out, especially if it's cards ppl like to play. With much more chance of those cards being out against you, I'm really wondering how much value those 'outs' actually have.

The fact is, betting does risk running into a turn raise behind, when your draw (and you are drawing) is at it's most vulnberable.

Even loose players flop real hands sometimes...

Just folded bottom 2 pair on a 542 board on flop, from BB with no Lo draw, and the turn was a 4; even then someone else made trip 4's, and the pot was split on a str8.

More often than not, someone does have the trips (or the nut full) is my impression, and frequently some of them lose huge pots with under-fulls.

At the low stakes, do consider the impact on your Table Image, if you fold for 1 bet on the river in big pots. A bet from a less tight player on turn, acts to "Inhibit a Bluff" so you can be sure, when the BB bets into the field on your set that he's not bluffing.

Though in low stakes poker, plenty of players wake up and bet two pair, with an obvious str8 on board, so I feel real unease at folding at 5-1 for 1 bet. When passive players raise, it's a whole different ball game and tends to mean the nuts.

Buzz, I really appreciate your time, and when I'm not meant to be focussing on a game, I'll have to work through your maths (and do some of my own) to try to learn from this. The stats of card distributions helps a lot in low levels, because assuming you're up against random hands is not far off the truth.

As it is, I have real doubts about betting Nut Lo draws without backup, and without real scoop potential, in games where there's almost 0 fold equity.

[Buzz]

[ QUOTE ]
Buzz, I really appreciate your time, and when I'm not meant to be focussing on a game, I'll have to work through your maths (and do some of my own) to try to learn from this. The stats of card distributions helps a lot in low levels, because assuming you're up against random hands is not far off the truth.

[/ QUOTE ]Hi Rob - And I appreciate your time. While you evidently were composing this reply, I was also composing so as to correct my previous post. When writing the post to which you replied, I was suddenly very tired when I was in the middle of writing it (the original). The rewrite didn't change my mind about the conclusion in the post to which you responded, but I did refine the numbers and I did change my mind about my first post in this thread.

[ QUOTE ]
The fact is, betting does risk running into a turn raise behind, when your draw (and you are drawing) is at it's most vulnerable.

[/ QUOTE ]Good point. And if the raise knocks out one of the three opponents, that's bad for Hero.

[ QUOTE ]
At the low stakes, do consider the impact on your Table Image, if you fold for 1 bet on the river in big pots. A bet from a less tight player on turn, acts to "Inhibit a Bluff" so you can be sure, when the BB bets into the field on your set that he's not bluffing.

[/ QUOTE ]Another good point.

Thanks.

Buzz

[RobNottsUk]

Having watched play on paired boards with extra attention lately, I think they really benefit the AA2x in short-handed pots. Say you have AA2x, raise pre-flop isolating a limper and the BB. Then betting overpair + Nut Lo draw on a 744 pot, tends to take it there and then against reasonable players. Weaker ones, will call hoping to improve, or see free turn card.

In a hand where there's 4+ players seeing the flop, in a loose game they'll often call 1 bet, extremely light. In that situation, if you make the Nut Full house on the turn (from top set) and the river card puts runner runner set on the board, you are in trouble!

In a 3 way pot, generally you have 2-way hands against 1 Hi hand, or 1 2-way, a Lo hand and a Hi hand. The schizophrenic nature of Hi/Lo reduces the likely strength of the Hi hand more than linearly.

That's not the case, in 5 way or 6 way action, you probably have 2 good Lo hands, a trash Lo/Hi, a Hi and 2 dodgy Hi hands to beat. Each added hand, hugely increases the liklihood of trips, or sets filling up when the board pairs.

From practical experimentation, I think the intial responders are over-estimating the liklihood of str8's being out, when in a loose game where poor players only have to call 1 bet.

Whilst I haven't calculated it rigorously after experimentatoin, I suspect the consensus under-valued the cyring call with top set, against loose tables and poor players who miss str8's, as players tend to play "Hold'em Hands" like KQ45 (They called earlier with a pair + gutshot etc).

Furthermore I think that trip's or better is exceedingly likely in highly multi-way pots on any board, paired by the river.

As for the tainted outs to A's up, I think you'll find someone may well bluff, but in every hand I saw this, a player had trips, or filled up, or made quads on the river (and was attempting to trap with a check/raise).