Okay everyone, firstly I'd like to say that I am not a gambler. I love to hang out at casinos but I know that every game there is -ev so i play table minimum for fun because i don't mind losing a little for a good time so this isn't a gambler trying to convince himself he can beat the casino. But I've been really studying this ultimate texas holdem game and i'm convinced that if you play an optimal strategy, it may be beat. I'm basing this pretty loosely that most players are playing a very bad strategy versus optimal, yet they can't possibly have less then a 10% disadvantage in the game or else losing runs would be extreme and no one would play. But there are so many ways you can improve yourself in that game that I believe with an optimal strategy you should be able to play at least 8-10% more efficiently then a player playing poorly.

For instance, most players don't realize that you are playing against a random hand, so they wait and wait until they hit a pair before they bet the 4x allowed preflop. What they should be doing is playing a random hand's range and anytime you are over 50% to a random hand, its 4x jam time.

What I would like to do is break down my strategy preflop:

any ace
any k5+, andy king suited
Q5s+, Q8+
J8s+, JT+
33+ for pocket pairs.

Now again, most players don't jam with these hands unless they have big A hands or medium pockets plus. so imagine how much of an edge we have versus them by taking these huge bets.

Post flop everyone plays pretty standard, assuming they didn't go all in for 4x preflop, they pretty much bet when they pair anything which is fine. But alot of equity is lost by most players by not knowing when to call on the river with high cards. I've actually pokerstoved certain cards against certain boards to determine when you should call or not. Its important to note that when the board pairs and doesn't pair makes a huge different.

In the game, when the board pairs you get paid 4-1 if you win by calling river. Not going to explain how you get 4-1, but anyone who has played the game and really thinks about it will see you get 4 to 1 by calling because when board pairs you get paid on the ante bet.

When the board doesn't pair is interesting, so lets say you call the river for 1 bet, well if the deal doesn't have a pair then you get your ante bet back. which means if the dealer doesn't pair you have a freeroll to win the hand since you put out a bet but get a bet back. the chances of the dealer not having a pair is 35%. Again if the dealer doesn't pair, you get 1 bet back.

So i did some pokerstoving and came out with a list of when you should call. Its kind of messy but hopefully it can be interpreted. Note that when I say a situation is close, that basically means you have to use your discretion on how drawy the board is. Anyone who plays the game I'd like to get some feed back and let me know what you think.

============================================

Non paired:
Any King
All queens except 1 cards to flush is close, OE is close if no over card. One over card is 10%

Any jack one over, no over is close. 1c to flush with 1 over is close.

Any ten 1 over. 2c to flush with straight possibility is close.

Nine 2 overs is close kicker above lower cards is better, go if no 2c to flush + straight possibilities. 3 overs is definite.

8 needs 3 overs.

7 needs 4 overs


IF BOARD PAIRS:

all kings unless 1 card to flush

all queens except 1c to flush unless 2 overs, 2c to flush is close. No overs is close 3%, with 3c to flush 2%. No overs kicker needs to be greater then one on board.

Jack 1 over is close 4%, no jack without an over. no jack if 1c to flush

Ten needs 2 overs, no go if 1c to flush.

==============================================

One final note. I'm very aware casino games for the most part can't be beat. So even if i'm wrong about this game it still would be pretty cool knowing you are extremely close to ev against the casino and with beers and comps might even come out ahead. Any insight would be appreciated.

http://wizardofodds.com/ultimatetexasholdem

Your basic strategy seems a lot like his observations. Beyond that, I'm not familiar enough with the game.

preflop yes our strategies mirror, but no one has ever talked river hero call strategies which are spots that comes up alot.

if you are talking about Texas Holdem Bonus that is my weakness at casino's. although believe it or not I am prob up at that game. I hit the bonus for a few grand all together. I usually start my casino's sessions doubling up $100 real quick at a few diff casino's and add about $600 to my BR than hit the Poker room with those winnings. "So far" this strat has worked for me.

Ultimate Texas Hold'em ≠ Texas Hold'em Bonus

See this thread for more info on Texas Hold'em Bonus.

why would one not put the 4x bet in preflop with 22? it is 50.334% vs random cards. there are way more than just the one mentioned on WoO that have an edge vs a random hand.

is it because high equity spots come up frequently enough on flop or at the turn/river stage as to the medium or small bet have even higher ev than making the +ev large bet preflop?

WoO claims that he cant come up with a optimal strategy for all streets since his computer "solved" the game using brute force going over all possible outcomes. however i feel as if it should be possible to come up with a strat.

im also pretty sure that the game is indeed -ev even with perfect strategy. on the other hand im kinda hoping that there might be games where the inventors havent thought it all the way through.

I'm pretty certain that the game is -ev as well, and i'm just basing this off the fact that the casinos know how to make money, but it isn't inconceivable to think an edge is possible. Keep in mind blackjack can be +ev as well but the casinos keep the game around mainly because of popularity but also because players who do not play optimally usually lose way more then the few who do.

As far as 22, I'm not sure why the pf strategy excludes 22 and some other hands that slightly beat out random, I can only imagine that its because of what you said with the +ev bet on river negates the equity you lose by not calling flop. I've thought about it too, and I reasoned even if 22 is more plus ev to make a bet its very slight and I wouldn't mind trading in the .3% edge for reduced variance. Even on a 20 dollar bet, by not taking the 4x pf and calling any river you lose like 5-6 cents equity maybe if my math is correct.

I reworked the math, turns out I was completely wrong on what you should be calling nonpaired boards with, the correct way of playing assuming my calculations are correct are as follows.

all kings unless one c to flush + str8

Queen 1 over is close

Jack high 2 overs is close .5 percent.

Ten needs 3 overs 3%. 3c to flush okay, no 1c to flush




Common post flop situation strategies to be continued....

Hi. I have worked out a lot of the odds for the river bet a long time ago, but lost the calculations, and now am not sure if they are 100% correct. But I may get back to that soon. There are more factors that I did not consider before (like ties on the river). I really love the game, and almost always win big at it following my strategy (which really surprises me!). I have written down the results though, and have these to share with you....(in the 'high card call' column, 0 is don't call, 1 is call with highest card, etc.)

But the key to the game really is what to do on the river without a made hand.

Preflop: same as Wizard of Odds
The reason 22 is not included is because it loses to any over pair, but also will lose the ante bet, whereas there is about a 9% chance that the 33 will win the additional ante bet.

POST FLOP

DRAWS:

No chance of high card and UNPAIRED board:

Actually need >63.8% chance to win in order to bet, and higher in case Dealer hits straight or flush. This leaves ONLY:

Top two of straight flush draw 8 or higher
Split straight flush draw Ten or higher

EVERYTHING ELSE ON THE FLOP IS YOU NEED >50% IN ORDER TO BET:

No chance of high card and PAIRED board:

Top two of 9 high straight
Top card of 10 High straight
4-flush 9 high

Chance of high card UNPAIRED board

Flush draw, down to 5 (Jd 5c / Ac Qc 6c)
Double backdoor straight WITH back door flush (Jd Tc / Ad 8d 7h )


MADE HANDS:

Any top pair
Any middle pair
Bottom Pair, EXCEPT:
There is a 3-flush AND you have none AND pair is 2 or 3
There is a 3-flush, and pair is:
6 with lower-than 6 kicker
5 and kicker plays or is ten high
4 and kicker plays or is king high

RIVER

Paired Board

Board High Card Call
Open-Ended Straight 0
AKQJ Fold the 9
4-Flush 1
Gut-shot straight 1 if straight is 9 or lower
2 if straight is 10 or higher
No straight or flush 3
Backdoor straight 3
Backdoor flush 3
Backdoor 2-straight 3
Backdoor straight and flush 3 if have one of the outs
2 if have none of the outs
3 of a kind 4
2 pair 4 if one or both pair under 8
5 if both pair 8 or higher
4 of a kind Any (because blind pays 10-1)
ANY TIE (2 pair, 4 of a kind) 2 possible over cards only

Unpaired Board

Board High Card Call
4-flush 0
Open-ended straight 0
Gut-shot straight w/ backdoor flush 0
Backdoor flush 1
Gut-shot straight 1
Backdoor straight AND flush 1
Backdoor straight (3) 1 / 2 if have big second card or dealer outs
Backdoor straights (1, 2) 2
AKQJ Call with the 9
No Draws 2


I would LOVE feedback, and when I get back from Vegas next week, I will work on these some more (and hopefully have tales of good results!)

I'm having alot of trouble reading some of your notes but very interesting and thank you.

It is very close to being an even money game, but a majority of the house edge comes from the fact that the house only pays the blind when the qualify with a pair or better and takes it when you lose.
Hypothetically lets say I gave you a hand where you are a perfect 50/50 flip and you bet 4x with it.
Then 50% of the time you lose 6 bets and 50% of the time you win. However, since the dealer will not qualify every time you will not win 6 bets every time thus giving the house an edge. The bonus payouts on the blind do not make up for this either. I don't know how to show this, maybe someone can help me out.

I know of one way to beat it but you have to be careful. If you can pull it off, you can have a bigger edge then a card counter; only catch is your max value bets are limited. So anyone who plays this game knows that the house edge really comes from players not playing optimally. they don't 4x bet ever unless they have a premium hand, and they look at me like i'm crazy when i go "all in" with a king high or some queen highs. So what I did last time was make friends with the table and the dealers. Then I jokingly suggested that hey if you don't want to bet it let me bet it for you since i'm crazy like that (try to make the deal while the dealer isn't paying attention). So basically every time they have a 4x hand, i'd do it and whenever they like the board and say they'll take the rest of the action, we just work out the difference. For example if i 4x when they have A7, but the flop pairs one of their cards, they give me the word and i automatically take 2x the action. or if they decide to not call until the river, i take 3x the action they take the 1x. It works out for them as well because even if they decide to not call the river, they still have the off chance of winning their ante bet or blind bet without having to make any call at all. and it works out for me because i just take an even money bet whenever I see them pick up a hand that is greater than 50% to win against a random hand. Dealers didn't say anything because they are clueless and don't realize i'm only doing it for the huge edge as they are oblivious to the math of the game. In their eyes i'm just some degenerate that likes to bet with an ace or king high, which in their minds is a dumb thing to do. One dealer said to me "pit boss might say something if he sees you doing side action, so don't do it too loud".

Theres another obvious way to gain edges in the game, which is that exposed cards from other players will help you make decisions. Whether this can be used to actually *beat* the game, who knows. If it is true though, then maybe a complicated collusion team can beat the game for a lot.

I think this is true also. While dealers are not supposed to allow players to show each other their cards, very few of them actually enforce this rule in my experience. I think with a full table of people showing cards to each other, playing optimally it is very beatable. However, to be able to adjust betting ranges that quickly seems very difficult. However there has been situations where I had a Q on a board with like AK223 and was going to bet it, but saw that 5 other players didnt make a single pair and felt that a bet was no longer optimal. Theoretically i think if cards are shown the player COULD gain an edge but how should the betting ranges change? Any insights?

Knowing a few more cards is such a slight overall edge. Maybe on some boards where players are holding important cards, but for the most part I don't think it is enough. If you work out the math to the game it goes something like you need 23% equity with your high card to call on a non paired board, and 19% on a paired board. Knowing an important card might take 1.5% off but most decisions are usually not that close. Even if players aren't showing their cards you can still put them on a pair when they bet out. In fact, keeping track of that information might even be more valuable then a neighbor with random cards that probably won't affect your decision most of the time.

I finally worked out a basic strategy table for Ultimate Texas Hold'em, and will work out the collusion strategy sometime in the future. I'm not expecting much EV improvement for collusion, because its basically an even-money game, and the dealer has to qualify for the ante bet, etc. (Unlike Mississippi stud, where there's no dealer hand, and the pay table is much like Jacks-or-Better; collusion there improves the game from -4.91% to +1.5% EV).

Anyways, here's the Ultimate Texas Hold'em basic strategy:
http://discountgambling.net/ultimate-texas-holdem/
http://discountgambling.net/mississippi-stud-ev-barona/

Steve

I logged a week of grinding out minimum bets and basic strategy UTH, and won about $2000. My friend played higher limit with me for 2 sessions, and won $2250. I post some reflections on the week:
http://discountgambling.net/2009/10/...-texas-holdem/

Steve

I finally got around to a 2-player collusion strategy for UTH, not that it helps much. FYI:

http://discountgambling.net/2009/10/...-texas-holdem/

So from your study do you know mathematically if you have a definitive edge yet?

I didn't try to quantify the collusion edge, because it looked too small relative to the inherent house edge. If you want my program, I'll give it to you, and you can look further into possible river exploits. I'm sure you'll get +EV for a sufficient number of confederates and appropriate strategy. If you want the program, you'll need to know how to compile C++.

Or, if you have a Mac, I can give you the binary executable.

What casinos in Vegas have UTH...preferably on the Strip?

I have been working on strategy for this game for quite a while. Play testing the strategy for over 2000 hands and recording the results of each hand on an excel spreadsheet has produced a +EV result. I am interested in discussing theories with like minded people and I would love a program that could produce random hands and results for this game if anyone is up to the challenge. Please email me if you would like to discuss strategy as I am opposed to divulging details on a public forum at AndyContra@aol.com

If you really want to playtest a strategy you need to write a computer program that will playtest millions of bets for you.

I wanted to bump this and ask.. Is this beatable, and I played it but its was 3x raise. Does this change the odds at all?