UTH Advanced Strategy

The advanced strategy for decision point 2 does not agree with the UTH calculator.

Specifically holding 2K with QQ on board you should raise with any A,J,T kicker.

QQA - Raise average value of raising= -0.254980 average value of checking = -0.294960

QQJ - Check average value of raising= -0.475243 average value of checking = -0.450256

QQT - Check average value of raising= -0.494414 average value of checking = -0.464107

Should the rule be changed to only raise with A kicker or is the calculator in error?

Quote: zrlcsx

The advanced strategy for decision point 2 does not agree with the UTH calculator.

Specifically holding 2K with QQ on board you should raise with any A,J,T kicker.

QQA - Raise average value of raising= -0.254980 average value of checking = -0.294960

QQJ - Check average value of raising= -0.475243 average value of checking = -0.450256

QQT - Check average value of raising= -0.494414 average value of checking = -0.464107

Should the rule be changed to only raise with A kicker or is the calculator in error?

I'm not sure I'm following the full calculation you're making.

Assuming you're talking about kickers in hand, not on the board, Ax (any ace) in your hand should have been bet 4x pre-flop, so there's no decision to be made by the time you know QQ is on the board.

By Mike's kicker rule, except for some 4toa flush or straight hands, QQ and anything else on the board is 11 cards or less the dealer can have against you, plus a maximum of 8 kings and aces for 19, so the jack plays. In that specific instance, since QQ eliminates most outside or inside straight possibilities that wouldn't favor the jack, I can't think of an instance you wouldn't bet it except 4toa flush.

The 10 could play, depending on other overcards.

As they say, a little knowledge is a dangerous thing on my part. So perhaps you could explain more fully, at least for me, what your question is?

Edit. Been thinking about this some more. Perhaps decision point 2 is the 2x bet, even though you're talking about kicker strategy? (I need to look at the calculator you're referencing, I guess). If QQx is the flop,.the vast majority of the time the jack would play no matter the river, and it's an aggressive strategy, so it's probably correct to bet 2x there.

I would think.the 10 would be a close decision as well, and the math might say bet 2x. The number of times you could pair, trips, etc, and high cards on the board are good for you, since they're not then in the dealer hand, 2x might be the right aggressive bet.

The difference might lie in the 2x, even if it's not +ev in the way you're dissecting it, is worth more than a straight + or -. In part, that's because you make 2x instead of 1x, so that allows for it to be less ev to play than check, in part because you know the dealer always qualifies, and in part because you're trying to offset and minimize the cost of the Blind bet by betting 4x and 2x whenever possible. Since the Blind bet pushes below a straight, but you lose it when you lose, betting 2x on even a slightly negative hand would have the effect of adding a little value to the overall cost of losing the Blind on other hands.

Thank you for the quick response.

I am only concerned about the 2x bet (Mike calls it decision point 2 in his Advanced Strategy) after the flop. If my hole cards are K2 offsuit (would not have bet 4x) and the flop is QQT, QQJ or QQA. Mike's advanced strategy for shared pairs in all three cases cases you should raise. Mike's calculator says to check QQT and QQJ and raise QQA. All cases have negative outcome. Only the QQA has Raise greater that Check (values in original post are from Mikes Calculator).

I would think it's an exception case that factors in what I said in my last paragraph (those 3 factors). What little I know about the algorithms behind the calculator, it could be a programming oversight, and they should always match, but in general using those calculators, there are very specific (and a long list of) exceptions for most (all?) of the strategy generators for VP variations and other games, and that could be one. There's usually a simple, complex, and optimal strategy for every game, depending on how far into the thousandths you want to dig.

I'll step out because I can't give you a better answer, but hopefully one of the better math guys on here will.

My experience with the UTH calculator on WOO is that it provides mathematically correct values of EV for any case that you define. It is a wonderful calculator and very useful.

The advanced strategy for UTH on the WOO site can be improved upon (particularly, in my experience, for decision 2 -after the flop). Some of us have used the UTH calculator to develop a list of exceptions to the UTH advanced strategy that we use when we play the game. It is simply a question of how many rules you want to memorize to capture a tiny bit of extra EV.

You have started out in the right way - you are using the calculator to explore where the exceptions might be and appear to have discovered some! Just remember that no one claimed that the advanced strategy was mathematically optimal for every case.

I understand that that the advanced strategy is less than optimal in order to simpify. But in this specific case it is more comlex by adding J and T kicker. I am afaid I am missing something.

I would expect any exception would act to make the strategy more complex. Here it simplifies.

Maybe in your case but how about if you have K3
Or K4 does it change then?

The values change slightly for K4 and K3, but still indicate checking is best for the T or J. Ace kicker remains a raise.

These cards matter because of straight possibilities. There are also flush possibilities in some cases. Are those accounted for elsewhere? k2/qq10 rainbow will look much worse than the same hand with 3 suited cards.

I should have indicated in the beginning that there was no chance of a flush draw. If k2 was suited i would have already been in at 4x.

Kh2d/QhQsJh has an unsuited pocket hand and is 3 to a royal with 2 cards to come.

So, the strategy advice: "holding 2K with QQ on board you should raise with any A,J,T kicker" is made because sometimes the suits will line up such that you have a royal flush draw?

I agree with OP. That seems like a badly formulated strategy rule.

The strategy rule should probably read

"Holding 2K with QQ on board you should raise with any A kicker."

Of course, you should also raise with 2K and QQ on board if you have any 3 cards to a royal flush, but "raise with any 3 cards to a royal flush" is stated elsewhere.

I agree -

The strategy rule should probably read

"Holding 2K with QQ on board you should raise with any A kicker."

I respect the all the work Mr. S has done and it seems strange his advanced strategy would have this error, if that's what is is, for shared pairs. I have been looking at the other rules for shared pairs and am finding the same conflict with the calculator. In every case that I have looked at so far the calculator would simplify the rules.

Quote: gordonm888

Of course, you should also raise with 2K and QQ on board if you have any 3 cards to a royal flush, but "raise with any 3 cards to a royal flush" is stated elsewhere.

According to Pokerstove:

{Rainbow QQT board}:

1,070,190 games 0.004 secs 267,547,500 games/sec

Board: Qs Qh Tc
Dead:

equity win tie pots won pots tied
Hand 0: 48.492% 44.56% 03.93% 476876 42081.00 { Ks2c }
Hand 1: 51.508% 47.58% 03.93% 509152 42081.00 { random }


{Three to a RF board}:

1,070,190 games 0.010 secs 107,019,000 games/sec

Board: Qs Qh Ts
Dead:

equity win tie pots won pots tied
Hand 0: 49.990% 46.16% 03.83% 494027 40956.00 { Ks2c }
Hand 1: 50.010% 46.18% 03.83% 494251 40956.00 { random }

We're behind in both spots-- I don't see why betting 2x in either case is a good strategy.

Quote: nimbus2k

We're behind in both spots-- I don't see why betting 2x in either case is a good strategy.

IF I'm understanding the math correctly, it's because there's not a straight line correlation between the play bet. and the ev.

The Blind bet must be factored in. It is a highly -ev bet, that is offset by the inherently +ev Play bet under optimal strategy. The reason so many people lose so fast at UTH isn't because they don't bet the play. It's because they don't bet the play early enough in the process, or big enough (nothing should be 3-bet, either 4-bet or check pre-flop. A lot of people check A-little , face-face, and small.pairs, then 2 or even 1-bet them.).

If you get a simple win, the Blind only pushes. If you get a straight (more possible than usual by a little with that board), you make 1:1. But if you lose, you ALWAYS lose the Blind.

In this particular case, many possible results are eliminated, because you know the dealer will qualify with the board paired. So the Ante will play, not push, on a win or loss, which affects the math as well.

You also know your deuce is very unlikely to play, so you're depending on the king only, though at least a small part of the ev is calculated on your deuce pairing on the river. But at that decision point, your deuce, even if paired, could counterfeit to a higher 2nd pair.

Also, atm, you only have 15 cards out against you, and that number is unlikely to increase. QQTxx (1 pair is 11) and 4 aces. So the king will play in almost any scenario.

The question is 2x vs 1x. And, factoring in the Ante and Blind, which will both win/push (Blind) or both lose regardless of the size of your Play bet, the 2x will help make up for their loss on other hands (as a whole), whether this particular hand wins or loses, for only 1 more unit.

I can see the -ev going into the high 30% range before it's correct to check there.

Per the calculator on the Wizzards site:

QQT - ev of raising= -0.494414
QQT - ev of checking = -0.464107

both cases have negative ev and are very close but checking is less negative.

The wizzard gives credit to J. B. for the calculator. I am assuming all of the possible senarios are taken into account in calculating the outcome. Wish I had the math chomps to do the calculations myself but I do not, so i rely on this calculator.