*** Warning: Absurdly long. Please don't quote the whole thing in a reply! ***
This hand has some unique qualities that, in my view, make the "standard" play of a turn bet incorrect.
But some people whose posts I respect told me I'm badly wrong, so I delved further to see how this hand compares to another seemingly similar situation, where I believe the "standard" play IS correct.
(Note that this discussion applies to reasonable stack sizes like in the original post. If you're playing an uncapped cash game with 5000 bb's that's different because there's a bunch more room for river play.)
Recall our Hand: J
J
And the board we are against in this hand:
Board 1: A T 7 J
Compared to this very similar board:
Board 2: A Q 6 J
Both boards have an ace, two diamonds, and a possible broadway, and we spiked our set at the turn.
The changes in board 2 are subtle... the Ace is suited and the broadway card is higher, and I notched the 7 down to a 6. But these subtle changes have a HUGE impact.
I believe the typical player (myself included until a few days ago) blithely bets his Jacks on either board, assuming the most likely villain hand is a big ace and he's owning.
But in fact our set of Jacks is MUCH weaker on board #1. On board #1 there are many more hands that have us crushed, or have big drawing equity.
Just as importantly, board #1 gives our opponent fewer big hands that we crush.
First note that in a vacuum (i.e. villain doesn't look at his cards until after the flop), the big hand / big draw combinations are nearly identical on the two boards.
But as we'll see, the villain's hands that are required to create those combinations are FAR more likely in board #1 considering the villain's "reasonable" early position preflop raising hands.
I'm assuming a wide raising range (he made a small raise, early position but vs very small blinds, i.e. say a good LAG player raising some good and speculative hands alike, but good speculative hands, not offsuit clunkers), adjust as you wish...
Pairs, AQ+, A9s+, A5s, KQ, KJs, suited connectors 56s+
Comparing this range on board #1 and board #2...
HANDS THAT CRUSH US AND BIG COMBO DRAWS
Broadway
Board 1: KQ (16 hands)
Board 2: KT (0 hands in assumed raising range)
Broadway is of course our nightmare scenario.
Note that on board #1 broadway is a
significant risk, while on #2 it can be virtually ignored!
(No, I didn't save the best for last. If you want to stop reading now, you've gotten one of the most important points already.)
2nd nut straight
Board 1: 89s (4 hands, one with flush redraw)
Board 2: Not possible
A fairly small part of board #1 range, but a threat that doesn't exist at all on board #2.
Broadway/nut flush combo draws
Board 1: A
K
, A
Q
Board 2: K
Q
Rare but twice as likely on board #1. Note there would be a bunch more of these draws on both boards if we didn't have the J
blocker (I screwed this up in a previous post I think).
Without the blocker, both boards get significantly more dangerous, with board #1 still nearly twice as dangerous in this category.
HANDS THAT WILL CORRECTLY CALL REASONABLE BETS
Naked nut flushdraw
Board 1: A
9
, A
5
(2 hands)
Board 2: K
X
(0 in raising range)
Some concern on board #1 and virtually NO risk on board #2.
Lower flushdraws
Board 1: 5
6
Board 2: T
9
, 8
9
, 7
8
Finally board #2 fares a little better! Of course, 8
9
already had a made straight on board #1 which is why it doesn't show here, so getting that out of this range is a Pyrrhic victory.
On board #2, a couple of these have weak straightdraws as well.
On balance these draws are less dangerous than the NFDs in that our opponent can't feel real confident when they hit, so we may get away cheap if they do. They also can't confidently semi-bluff as the NFDs can with top pair/flushdraw.
GOOD HANDS / WEAK DRAWS THAT WILL INCORRECTLY CALL REASONABLE BETS
Top pair/broadway draw
Ok, now THESE are the hands that everyone was worried we weren't getting value from if we don't bet the turn.
These are the hands that we "put the villain on". We naturally (and, I submit, too narrowly) think like this: Villain raise = big ace, ace high board, therefore set = jackpot.
Let's see how many there are.
Board 1: AK (12), AQ (12) less the 2 combo flushdraws, for 22 hands
Board 2: AK (12), ATs (3) for 15 hands
Well, quite a few. And in fact even more on Board #1, which should make all the value-betters happy (in fact, contrary to my theory about checking behind on board 1, it would seem, but read on).
WEAK HANDS / WEAK DRAWS THAT *MAY* INCORRECTLY CALL A BET
Underpair/broadway draw
Board 1: K
J
, Q
J
(2 hands)
Board 2: KQ (11), K
J
, J
T
(13 hands)
Villain is a whopping 6x as likely to have this draw on board #2. This is because KQ is a big part of villain's range, and KQ makes the nuts on board 1 and a weak draw on board 2.
We want these hands to call, because they have only 3 clean outs to the nuts (which we then barely pay off), 1 out to broadway which puts a flush on board as well (which we may not pay off at all), 3 outs to 2-pair (which they crying call at river)... and then the paydays... the 2-outs to trips which fill us up.
Of course a competent villain should realize much of this (even if he puts us on a hand no better than two pair and figures trips is HIS payday), so he shouldn't really call much more than a small bet here.
TWO PAIR OR BETTER HANDS THAT WILL GIVE ACTION
Finally we get to the hands we hope our villain has... a made hand that is hopefully too strong for him to fold, and usually drawing thin. In descending order of strength:
Sets
Board 1: AA, TT, 77 (3 over sets, 6 under sets)
Board 2: AA, QQ, 66 (6 over sets, 3 under sets)
We are twice as likely to be beaten by a set on board #2, another rare victory for board #2, and its last.
Naked top two
The (hopefully) can't-fold scenario. This is the bread-and-butter hand payoff hand for sets.
Board 1: None possible (0 hands)
Board 2: AQ (9 hands)
Board #1 has NONE of these hands to pay us off with, there are no AJs hands possible (bummer!) while Board #2 has 9.
This is a big incentive to bet the turn on board #2 in an attempt to induce a raise/shove from villain "protecting" his top-two hand.
Top and middle pair
Actually not very strong on this board because villain must fear better aces-up as well as broadway, but will certainly at least call.
Board 1: A
T
, A
T
(2 hands)
Board 2: A
J
(1 hand)
Only one such hand on board #2, but maybe more likely to spazz all-in because psychologically it looks stronger.
Middle two pair
A very weak made hand on this board, almost more correctly should be considered a drawing hand. Competent opponents would muck it vs any real action, and certainly not raise. But it probably looks too pretty to fold to most reasonable bets.
Board 1: J
T
(1 hand)
Board 2: Q
J
(1 hand)
TOP PAIR OR OTHER HANDS THAT WILL GIVE LITTLE TO NO ACTION
Naked top pair
Suited aces that missed the flushdraw or two-pair. Note that AT+ all have some sort of draw so are not included here.
Board 1: A
9
, A
5
, A
9
, A
5
(4 hands)
Board 2: A
9
, A
5
, A
9
, A
5
, A
9
, A
5
(6 hands)
Simply awful on this board. Competent opponents fold to virtually any turn bet. If we knew opponent had this, we would probably check and hope he improved to trips at river or called unimproved as a bluff-catcher.
Pocket pair/broadway gutshot
Board 1: KK, QQ (12 hands)
Board 2: KK, TT (12 hands)
Horrible hands for villain to call. Only 3 clean outs to broadway, and if he spikes his set it puts 4broadway on board so he has no trips payday. Really these are only very thin bluff-catchers.
Small Pairs
Pocket pairs drawing dead, or nearly so. Obviously those drawing dead we'd like to give a free river to hit their "miracle", because they have to fold turn.
Board 1: 88, 99, 66, 55, 44, 33, 22 (42 hands)
Board 2: 99, 88, 77, 55, 44, 33, 22 (42 hands)
These are quite a few hands. Checking is clearly best on board 2 with zero risk. Against board 1 there is a small risk of 88 or 99 hitting a gutshot, but it's small payoff if they do, so again checking is best vs most players.
Airballs
I'm not sure how many hands and don't want to do the math.
Obviously, the only way to get money from them is to let them bluff. And letting them get to river with first-right-to-bluff is probably the best way to let them do so. Only the most gutsy or foolhardy villains would bluff-raise a turn bet on this board.
SHUT UP ALREADY WHAT'S YER POINT
Ok, ok... the point is we want to see how specific hands in villain's range react to a value-sized turn bet, to see if betting is correct.
ALL-IN MATCHUPS
Let's say at turn, both players have a hand good enough to go with, and our turn bet lights the fireworks.
The good... hands that we crush.
Board 1: 6 hands (6 sets)
Board 2: 12 hands (9 top two, 3 sets)
The bad... hands that crush us or have a draw big enough to make us sweat profusely.
Board 1: 25 hands (20 straights, 3 over sets, 2 big combos)
Board 2: 7 hands (6 over sets, 1 big combo)
The ugly... look at the ratio of good/bad all-in matchups on the two boards:
Board 1: 6 out of 31 good (happy at turn only 20% of the time)
Board 2: 12 out of 19 good (happy at turn 63% of the time)
That's a huge difference in terms of number of hands we are happy to see. Of course one side or the other still has outs, so putting math on the allin ranges, we are:
39% on Board 1 vs: AA,TT,77,AdKd,AdQd,KQs,98s,KQo
63% on Board 2 vs: AA,QQ,66,AQs,KdQd,AQo
So it's a huge EV difference as well.
VALUE TOWN
Hands that will call a decent turn bet every time, and sometimes a river bet, while drawing very thin. This is the much-vaunted value that we are missing if we (gasp!) check behind at turn.
Includes: Top pair drawing thin to broadway (with little implied odds). Top and middle pair. Top and bottom pair.
Board 1: 24 hands (AK and AQ top pair, ATs top/middle)
Board 2: 16 hands (AK and ATs top pair, AJs top/middle)
Board 1 has 50% more of these hands, so at first blush we should be pounding more often.
But wait... look what happens when we compare to the previous all-in confrontations. If we go auto value betting...
Board 1: 31 allins vs 26 value towns
Board 2: 19 allins vs 16 value towns
I thought this was VERY interesting. I earlier incorrectly assumed the big matchups were less likely than the mundane ones (and some others agreed with me).
Presumably the "standard" turn action goes "the set bets" and "the big ace calls".
But in fact it turns out that on EITHER board we are MORE likely to find "the set bets" and "the big ass hand raises".
If you think villain spazzes in with all aces-up, then there are even fewer "big ace calls" and more pushes (good for us on board 2).
Of course there are some other check/calls in here, continuing...
FLUSHDRAWS
Board 1: 3 (2 nut flushdraws, 1 low flushdraw)
Board 2: 3 (3 low flushdraws)
While we can't reasonably price any of these draws outs, it's still desirable to make them pay, and/or set up a big river if we both hit (but that happens very rarely here, i.e. 1 outer on board #1).
However, on board 1, the two top pair/NFD hands have a potential semibluff-shove. And as we now know, we fare badly vs his legitimate shove range. Do we so badly want to try to make a NFD pay and possibly have to contend with those semi-bluffs in his range further confusing us? I'd say no.
WEAK HANDS/DRAWS THAT MAY CALL TURN
These shouldn't call turn, but might depending on our bet sizing.
Board 1: 3 hands (2 underpair/broadway, 1 middle two)
Board 2: 14 hands (13 underpair/broadway, 1 middle two)
Far more hands can make a bad call on board 2, again leading towards value betting board 2.
OTHER HANDS
I'm assuming all the other hands will fold if we bet turn.
ENTIRE CALLABLE RANGE
In an attempt to make betting the turn look as good as possible, I added up all the hands above that at least call a turn bet (including those that clearly shouldn't vs a set).
Board 1: 63 hands (of which 33 are possible all-ins)
Board 2: 52 hands (of which 19 are possible all-ins)
DUCY?
On board 1 we CANNOT value bet with impunity, because over half the time villain can CALL, he has a hand he can legitimately SHOVE if he decides to.
Over half the time!
Even if he doesn't shove but instead puts in a healthy raise, at reasonable stack sizes (i.e. PSB left at river) we
lose money with second set no matter how coyly we play the river.
And that's just if he plays straightforward based solely on the value of his hand. If he gets a little sneakier, mixing things up, e.g. putting in some raises with big combos and top set, flatting the nizzles... well then we really get lost.
I believe the risks of betting are too great. His calling range has too many monsters lurking, and we have a potential monster-killer if we can get to the river cheaply.
Further, and I believe possibly just as importantly (though the math tedium to figure out is too daunting for me to try to prove), by concealing our strength I think we get far more value from weak hands that fold to a turn bet (fearing another river bet) or hands that couldn't even consider calling turn but improve enough at river to call.
And, of course, we let villain possibly river-bluff his airballs.
All in all, it seems to me check-behind is far superior for board #1 at these stack sizes.
In sharp contrast is board #2. On that board we should hammer away. There are too many hands we get value from to do anything else. And if the villain is stupid enough to raise, it's his -EV funeral when we shove.
Hopefully seeing the guts of the hand ranges laid bare helps give a better intuitive grasp of what's happening. Because I'm not doing it again for a different range.
Going back to boring old PokerStove math and putting in Villain's callable ranges:
Hero J
J
Board 1: A
T
7
J
Villain: AdKd,AdQd,Ad5d,KQs,KhJh,QhJh,JhTh,98s,6d5d,KQo
Hero EV: 0.39
Board 2: A
Q
6
J
Villain: AKs,AhJh,ATs,KQs,KhJh,QsJs,JhTh,Td9d,9d8d,8d7d,AKo ,KQo
Hero EV: 0.89
Well over twice the EV on a very similar looking board!
The moral of the story:
It's awfully hard to make a value bet into a calling range where you are -EV.
One final note, so you don't think i'm a checking nit...
In this situation I *would* bet this turn a bunch, just not (any longer) with JJJ in position.
I think it's perfectly fine to bet a LESSER hand like top-pair with no good draws.
A lesser hand folds some weak hands that we WANT villain to fold, either because he has decent equity vs our weak hand (and a fold is incorrect if he knew what we had) or because although he's totally crushed we won't ever get any more money in ahead anyway (e.g. he has pocket deuces and will fold turn to any bet, but call river if we let him spike a 2 outer for free).
Just as importantly, if we bet a lesser hand and get raised, we can easily fold (oh well, 1-pair drawing dead) without sacrificing much. Even if villain is semi-bluff raising, on a board like this he probably has us owned anyway.
But with a hand as powerful and hidden as a set, on a board that a big portion of villain's calling range is the nut straight... it's a big mistake to allow the villain to kill the value of our hand by making it too expensive to see the river.