22/33/44

[willyc]

I've been dealt these hands about 250 times each, and they are by far my biggest losers. It seems that when I hit my set and manage to get all my chips in, I'm losing to set-over-set more often than I'm ahead.

Are these winning hands for anybody? Do people think it is worth set mining with them?

[Albert Moulton]

Depending you your position, effective stack depth, and skill level of the other players, you can always choose to fold these hands in EP instead of play them for either an open raise or limp/call.

In POT-LIMIT & NO-LIMIT POKER by Ciaffone & Reuben, Chapter 16, "How Deep The Money Is Determines Your Play," Ciaffone writes, "Even though set over set is rare, it is so expensive to be on the lower end that small pairs are a double-edged proposition. I recommend playing deuces through fives only in late position and unraised pots."

That advice might not apply at your level, but it might.

Feel free to follow it if you think it will improve your win rate.

Alternatively, just call when a raise is less than 5% of the effective stacks. The 5/10 rule for playing small pairs for set value is "call" if it is less than 5% of effective stacks, "fold" if it is over 10%, and make a judgement call in between. With low pairs, error on the side of caution and fold if the raise is over 5% and/or you are out of position with deep stacks. That might help, too. I suspect you are probably calling raises that are too big relative to effective stacks, and playing them out of position too often. The result is that you pay too much when you miss. And you win too little when you hit.

[pzhon]

It's a common misconception that low pairs are a gold mine. One joke I tell my poker students is, "Small pairs are better than big pairs, since small pairs can flop a set and stack someone."

I believe these playing low pairs for set value when it is not appropriate is a significant leak for many winning players.

It helps to recognize that small pairs are not very profitable on average. After all, TT isn't hugely profitable, and it is much better. Most winning SSNL players are close to breaking even with low pairs (within 2 big blinds of breaking even). This comes from playing them in some situations which are clearly +EV, and erroneously playing them in some situations that are clearly -EV.

22 is much weaker than 66. It may be hard to see the difference at first, but 66 is better in many ways.

[img]/images/graemlins/diamond.gif[/img] 66 flops more viable 6-10 out straight draws.
[img]/images/graemlins/diamond.gif[/img] 66 is less likely to be counterfeited, with or without a set. Consider the difference between 66 and 22 on a board of TT553, or 62444, or when the board is a flush.
[img]/images/graemlins/diamond.gif[/img] 66 can win with set over set, not just lose. Consider what happens someone else also calls for set value with 44, as many players do. The main time you will play a big pot involving the player with 44 is when you both have sets, and you will win with 66, but lose with 22. The difference between winning a stack and losing a stack 1% of the time is 2 big blinds, which is quite significant.
[img]/images/graemlins/diamond.gif[/img] When you have 66, it is less likely that your opponent has a suited connector like 65s or 76s or 86s. (32s looks like a suited connector, but it really is a suited 2-gapper since it makes no more straights than 52s, and far fewer people play 32s than play 65s.) These are poison for low pairs, since you will be bluffed out on most flops, but they will very rarely pay off a set. They tend to play close to perfectly against a low pair played for set value, so it is very bad to run into them.

In order to play 22, you need a middle pair like 66 to show a clear profit, and it usually doesn't.

When you are playing for set value alone, it is best to have a tight opponent who will more frequently have a hand with which he can pay you off, and it is best to have position so that it is easier to get paid off. If you are out of position, you will get much less money out of someone with KK on an ace-high flop on average, and you might not be able to get much money in before something like a 4-flush or 4-straight on the board kills your action.

It's good to play someone who will often give you a free card on the flop. Sometimes this is to let you catch up when they flop top set and are vary confident that you don't have top pair, but it is pretty valuable to get a free card from someone who is trying to be tricky with an overpair, and who will stack off because he didn't represent his full strength on the flop. Such a free card can be worth 8 big blinds with roughly 100 BB stacks. If you get a free card 25% of the time, a small pair can be worth an extra 2 big blinds, which is quite significant.

While people like to talk about the 7.5:1 odds against flopping a set, even if you can get all of the money in when you do, you don't always win. Against an overpair, you lose about 20% of the time, roughly 8% when you flop set under set, 8% from your opponent spiking his 2 outs, and 4% from 1 card flushes and straights, and the times you get counterfeited. This means that even if you are sure that you will get paid off, you still can't profitably call more than about 1/12 of the effective stack size. That you won't always get paid off, such as when your opponent raises with hands other than AA, means you need a much larger effective stack, and it might not be right to call for set value even with extremely deep stacks.

That said, sets can be a lot of fun. Just don't try to have sets at every opportunity.

[SmokeyRidesAgain]

Someone please to be post the top set mining thread.

[DiamondDog]

I think it's worth mentioning, for anybody playing at the lowest micro stakes (like I am) that at our level you can shove with a set of deuces and still expect to be way ahead of your opponents' calling range. Many opponents will call with all manner of junk. That's been my experience, anyway.

[helter skelter]

[ QUOTE ]
Someone please to be post the top set mining thread.

[/ QUOTE ]

?

[ QUOTE ]
The difference between winning a stack and losing a stack 1% of the time is 2 big blinds, which is quite significant.


[/ QUOTE ]

Could you clarify? I don't understand what you are saying here.

[ QUOTE ]
In order to play 22, you need a middle pair like 66 to show a clear profit, and it usually doesn't.

[/ QUOTE ]

or here.

[Dagaco]

I agree with Diamond Dog. If you shove low sets like those generally the caller has ace rag -at low levels anyhow.

[pzhon]

[ QUOTE ]

[ QUOTE ]
The difference between winning a stack and losing a stack 1% of the time is 2 big blinds, which is quite significant.


[/ QUOTE ]

Could you clarify? I don't understand what you are saying here.


[/ QUOTE ]
If you (66 or 22) and a player with 44 call, you will both flop sets or better at the same time about 1% of the time, on 180 out of 17296 flops.

When that happens with 100 big blind stacks, it is very likely that the lower hand will stack off to the higher hand. If you have a set of 6s over a set of 4s, then you expect to gain about 100 big blinds. If you have a set of 2s under a set of 4s, you expect to lose about 100 big blinds. That's a difference of 200 big blinds. If 66 will play exactly the same 99% of the time, but will be about 200 big blinds better 1% of the time, then 66 is about 2 big blinds better for this reason alone.

Of course, you will not always find yourself in a pot with another set-miner, and such a player might have 77 instead, and as I pointed out, there are many other differences between 66 and 22.

Here is another difference: When you flop a set or better with 22 against AA and get all of the money in, you win about 81.0% When you flop a set or better with 66 against AA, you win about 82.1%. You can see these values using the free and extremely useful tool PokerStove by setting just one community card. This will slightly overweight the possibility that you flop quads, so both values are high by about the same fraction of a percent. You can search my past posts for 44 to see how to correct this.

That is a much more common 1% difference between winning a stack and losing a stack. It is worth about 2 big blinds when you are up against AA with 100 big blind effective stacks. The difference against KK is much smaller, closer to 0.2% or 0.4 big blinds.

[ QUOTE ]

[ QUOTE ]
In order to play 22, you need a middle pair like 66 to show a clear profit, and it usually doesn't.

[/ QUOTE ]

or here.

[/ QUOTE ]
If 66 is a call by 1 big blind, and the differences between 22 and 66 add up to a total of 1.5 big blinds, then calling with 22 is wrong by half of a big blind.

A lot of players don't see the differences between 66 and 22. (I hope reading my post helps.) They might say that overall, calling with 22-88 primarily for set value roughly breaks even or shows a tiny profit with some combination of stack sizes, position, and raise size. If that is true, then it is probably right to call with 66-88, and wrong to call with 22-44.

[helter skelter]

I think I understand now. So it would be better not to play the small pairs if certain opponents are in the pot (like setminers) or use some form of pot control? Or not play them at all?

[gedanken]

edit: pzhon is both smarter AND faster than me...


[ QUOTE ]
[ QUOTE ]

The difference between winning a stack and losing a stack 1% of the time is 2 big blinds, which is quite significant.

[/ QUOTE ]
Could you clarify? I don't understand what you are saying here.

[/ QUOTE ]


look at all the hands where you get all in with either a set of 2's or a set of 6's. Now say that 1% of the time 6's would win but 2's wouldn't. The difference between 6's and 2's would be 2% of the average stack, or 2BB assuming 100BB stacks.

Now say pocket 6's are returning about 1 or 2BB per hand for you. Those pocket 2's won't be very profitable. Note that profit from 22 is going to come almost exclusively from sets (if you're not playing all-in pre against morons, that is).