Are high cards overrated?

[Overdrive]

Almost everyone recommends playing 4 high connected cards in O8, because if the flop comes with high cards you of course will be in good shape to scoop and win high. But in Bill Boston's book where he ran 100,000 simulations using Wilson software high cards did not come out looking too good. For example, 10-J-Q-K has a negative ev in his simulations and 10-J-Q-K suited only comes out slightly ahead, winning high just 5% of the time and scooping just 9% of the time when played. So it seems the value of 10-J-Q-K comes only from being suited and not from it's high card potential.

So do you follow the common advice to play any 4 cards between a 10 and an Ace? Or is this a wrong strategy? Because it seems like most of the high hands he ran simulations on came out bad. For example even Double suited 10-J-K-K, 10-J-Q-Q, J-Q-K-K all are listed as having negative expectaions in his book.

[TGoldman]

I pretty much consider high only hands as speculative drawing hands, similar to a hold'em hand such as 78s. I think the Hutchison point count system is pretty good for evaluating the strength of high only hands:

[ QUOTE ]
To qualify, all four of your cards must be Ten or above AND include (1) two pair, or (2) a pair and two suited cards, or (3) two double suits. Eliminate any high hand containing three of the same rank.

[/ QUOTE ]

I think you need to distinguish between high hands with an ace -- particularly a suited ace -- and high hands without an ace, for a variety of reasons that are probably obvious.

[gergery]

EV calculation based on who wins the pot are a good start and very helpful, but miss some key bits from real play. Namely, they don’t measure how much you win when you win, and how much you lose when you lose.

Hands like KQJTs that hit a good flop tend to be in clearly defined positions with strong scooping potential and opponents with low hands that can’t play well. So if you were to calculate $ returned per postflop bet, they would likely do quite well

Whereas hands like 3456s might have good Wilson showdown value, but in actual game play will a) rarely be in confident positions that they can value bet in, b) rarely be more to bet aggressively to get others to fold better hands, c) often be picking up half-pots instead of scooping, which are much less profitable.

--Greg

[chaos]

High only hands do well in multiplayer unraised pots.

You want no low possible so that if your hand wins you will scoop. This will only happen about 1 out 4 times. So you want to play cheaply to see if you get a favorable flop. When the flop comes with 2 or 3 high cards there will be dead money in the pot form the low oriented hands.

Having a suited hand adds value. If you have a suited Ace it adds significant value. A suited King adds some value. Lower suited cards add little value. What value they do add comes from back door flushes.

If you are in late position you will know whther the pot is multiway and unraised. (Of course there is always the chance that a blind can raise.) If you are in early position you do not have this knowledge. I would not recommend playing high only hands from early position unless the game is loose and most pots are unraised.

[pipes]

[ QUOTE ]
EV calculation based on who wins the pot are a good start and very helpful, but miss some key bits from real play. Namely, they don’t measure how much you win when you win, and how much you lose when you lose.

Hands like KQJTs that hit a good flop tend to be in clearly defined positions with strong scooping potential and opponents with low hands that can’t play well. So if you were to calculate $ returned per postflop bet, they would likely do quite well

Whereas hands like 3456s might have good Wilson showdown value, but in actual game play will a) rarely be in confident positions that they can value bet in, b) rarely be more to bet aggressively to get others to fold better hands, c) often be picking up half-pots instead of scooping, which are much less profitable.

--Greg

[/ QUOTE ]

Greg, I'm not sure what you meant about the Wilson showdowns, but Boston's simulations are not simply showdown statistics with no betting. Its based on a 10/20 tough lineup on Wilson software playing simulated poker.

[Phat Mack]

I assume we are talking about limit

I like these hands, but I'm not sure that they are the gold mine that some people think.

As Niss pointed out, they are better with an ace--having one means that you can hit the hand hard while helping to keep the low off. Having them suited adds a LOT of value. Rainbow high cards can be difficult to play: their straights can share pots, and they can leave you strung out and vulnerable on 4th street.

We used to have wars on this forum about whether or not limit O8 was a position game, and I am not trying to revive them, but I think high hands do much better when played on, or close to, the button. Unless you are at a very predictable table, they can be hard to play and get paid off for from up front.

Having said all that, they don't win a lot of pots, but can win some big pots.

[gergery]

[ QUOTE ]
[ QUOTE ]
EV calculation based on who wins the pot are a good start and very helpful, but miss some key bits from real play. Namely, they don’t measure how much you win when you win, and how much you lose when you lose.

Hands like KQJTs that hit a good flop tend to be in clearly defined positions with strong scooping potential and opponents with low hands that can’t play well. So if you were to calculate $ returned per postflop bet, they would likely do quite well

Whereas hands like 3456s might have good Wilson showdown value, but in actual game play will a) rarely be in confident positions that they can value bet in, b) rarely be more to bet aggressively to get others to fold better hands, c) often be picking up half-pots instead of scooping, which are much less profitable.

--Greg

[/ QUOTE ]

Greg, I'm not sure what you meant about the Wilson showdowns, but Boston's simulations are not simply showdown statistics with no betting. Its based on a 10/20 tough lineup on Wilson software playing simulated poker.

[/ QUOTE ]

Thanks, I wasn’t clear on how Wilson does the simulations. I’m not convinced it can effectively model good poker play, and I have anecdotally heard there are significant flaws with it, but since I haven’t read Boston’s book or used the software I’ll reserve judgment.

-g

[toots]

Haven't tried their Omaha software, but their Hold 'Em seems to simulate someone who plays better than the average B&M fish. Not that this is a huge edge, but it's better'n nothing.

[GooperMC]

Just for reference here are my PTO numbers

Any suits:
4 cards above 10: .37 BB/100
4 cards above 9: .14 BB/100
4 cards above 8: .07 BB/100
4 cards above 9 No A: .62 BB/100
4 cards above 8 No A: .01 BB/100

Rainbow
4 cards above 10: .30 BB/100
4 cards above 9: (.10) BB/100
4 cards above 8: .02 BB/100
4 cards above 9 No A: (.33) BB/100
4 cards above 8 No A: (.56) BB/100

It would be intereting to see these number for players that have played more hands then me.