22 The Newbie Chronicles: Turbo Texas Hold 'Em

[mdlm]

When I started this poker project one of my initial goals was to beat the best lineup in Turbo Texas Hold’Em (TTH). I had to abandon this goal because I could not get TTH to run correctly on my two-monitor system.

I finally dug out an old 100mhz machine and got TTH to work. It is far, far better than Acespade, its main competitor. In fact, with one exception, I cannot find a single feature in Acespade that is superior to the corresponding feature in TTH. (The one thing I can’t find in TTH is the ability to set up a specific hand and ask one of TTH’s personalities what it would do in that situation. It’s possible that this feature exists and I just haven’t found it).

I set up an all-Conan table and played against it for 100 hands in a 20/40 game with rake and toke. My understanding is that Conan is one of the two best TTH personalities. I was able to beat it for about $750 in these 100 hands. I paid about $50 in rake and toke.

Like Acespade, the main weakness that TTH’s personalities seem to have is that they regularly fold better hands in the face of strong betting. In the 100 hands I played I won 13.5 pots and three of those pots were won on the river when TTH folded a superior hand.

In addition to these defects, the Conan personality appears to grossly overvalue certain hands. If I recall correctly in one hand it went 6 or 7 bets with AK against my 44. It lost that hand when the highest card that appeared was a Q.

Compared to Jones, Conan open-raises far more often. For example, Conan open-raises with A7s in MP while Jones calls. I do not know which is better but it is interesting to note the differences. I should do a thorough analysis of this and maybe program a Jones personality into TTH.

After I played 100 hands against Conan I looked at TTH’s analysis of my play. Preflop I scored 89 on the hands I played and 80 on the way I played them. Overall TTH said that I was a little bit too loose and much too aggressive.

One of the best features of TTH is Sidewinder Sid, a cartoon character who walks through your play and tells you where your play differs from the play of the advisor personality. What I like about this analysis is that it makes it very easy to identify large leaks.

What I am hoping to do is to create a TTH personality that is superior to all of the built-in personalities and then use it to drill. I don’t know if I can create something that plays mid-limit poker, but creating a Jones personality is certainly possible.

==>
Goal Update

Last week I spent 23 hours on poker: 10 hours in PokerPages tournaments, 4 hours on 2+2, and 9 hours playing with Turbo Texas Hold ‘Em.

I did not spend any additional money last week. I have spent a total of $476.43 out of my $1000 budget.

An update on each of the four goals (which are to be accomplished by 3/30/03):

1. Read and study Jones’ “Winning Low Limit Hold ’Em”
I have confirmed 2 1/3 out of the three points I need to achieve this goal. A point (flush draw value bet) is pending an analysis of 10,000 hands.

2. Beat Acespade
Goal Completed on 11/5/02.
Over a period of 100 hours (3600 hands) I beat Acespade’s best lineup at the rate of over 4 BB/hr.

3. Beat Masque World Series of Poker
Goal Completed on 11/17/02

4. PokerPages 85% rating in one calendar month playing 20 tournaments
My current PokerPages rating is 78.45%. I played four tournaments in the past week and finished #20 out of 125, #38 out of 121, #6 out of 97, #22 out of 131.

[Easy E]

"...The one thing I can’t find in TTH is the ability to set up a specific hand and ask one of TTH’s personalities what it would do in that situation"

I'm not sure I understand what Acespade is doing here. What is this "ask one of the personalities" thing? Can you be more specific?

In Turbo Texas Holdem (what version?), if you want to get advice from the Advisor about a particular hand/situation, I think you'll get what you want by doing this:

1) Stack the deck, setting the hands for yourself (seat #10), any opponents you choose to set hands for, and any of the board cards that you want (and save it, in case you want to reuse it, or use for high-speed testing)

2) Set a Repeatable Deal number, so you know the cards dealt to the random spots will always be the same for a particular hand and the non-stacked Profile(s) actions should usually remain the same (unless you have some of the randomness/adjust play settings in place)

3) Set your other game settings as you'd like, as you learn them. Then start a new game.

4) Each hand that is dealt, check the advice for your Stacked hand, based on the actions that occur on that hand. Then Replay the hand, try different actions and see what changes about the Advice on different streets, after different actions.

5) You can substitute in different lineups, with different player mixes, then restart the game (the repeatable deal code comes into play here) and try the same hands with different mixes of players.

6) And, of course, with SID enabled, you can get further analysis on what you did.

Is that what you meant by 'asking'?

By the way- if you don't like the Advisor in place, make your own! You can create a profile and substitute it as the Advisor, at least in ver 5- I'm not sure how many releases prior to 5 that this was in place.
That could address your "ask personalities" goal? You could sub various Profiles in and then hit Advice (this would affect SID's analysis too, I believe- I haven't tried that yet)

EXTRA CREDIT- If you have the time, which you seem to, why don't you do a detailed side-by-side comparison challange- Acespade vs. Turbo Holdem? People have been asking about it lately and the last detailed one that I've seen is almost a year old.

[Clarkmeister]

"What I am hoping to do is to create a TTH personality that is superior to all of the built-in personalities and then use it to drill."

"9 hours playing with Turbo Texas Hold ‘Em."


Anything to avoid playing in a real game, eh?

[Easy E]

The man is obviously committed to his original plan, come hell or high water.

Now, if he bails on Stage 2, THEN we can lay into him [img]/forums/images/icons/wink.gif[/img]

[Nick B.]

<<In addition to these defects, the Conan personality appears to grossly overvalue certain hands. If I recall correctly in one hand it went 6 or 7 bets with AK against my 44. It lost that hand when the highest card that appeared was a Q.>>

Speaking of overvaluing certain hands, who goes 6 or 7 bets with 44. Maybe you should try playing against a random table. I think the game would be fairly easy to beat with all the people playing the same. The diversity of the table is what makes some games tough.

[marbles]

"I think the game would be fairly easy to beat with all the people playing the same. The diversity of the table is what makes some games tough."

--I'll go you one further... The diversity of the table is what makes poker poker.

[Louie Landale]

The computer does NOT play well, mostly because it cannot use what's already happened to influence its play. Once you can beat it, then you are in fact investing your time detrimentally if you continue to play it, since you are learning bad strategy and tactics that MAY work against the personalities that have no chance against real folk. Just because certain "good" computer personalities NEVER call the river with 2nd pair, doesn't mean real opponents don't.

Computer personalities are litterally "disciplined brain-dead folk".

Use it as a research tool, possibly creating profiles and letting the computer play 1mil hands at a time. But take what you "learn" with a grain of salt.

- Louie

[Easy E]

Pretty good post, Mr Landale :0

However, if you THINK about what is going on and why, rather than just accepting "i should fold second pair on the river to a bet, like the 'expert' Profiles do", I think that even the "brain-dead" scenarios can teach something... and doing it in the software, rather than a live game, lets you think about it without distraction/pressure.

I mean, if you bluffed the river and made the winning hand fold, what does that say about how YOU should play the 2nd pair when faced with a bet on the river? Could it be a bluff? What type of player is that (somewhat easy to analyze your own play and motivations, right?) who made the bet, and what could it mean? etc etc....

And, as you said, you can then follow up with some high-speed simulations, and Profile adjustments, and get an idea (NOT a 'carved-in-stone" rule, but as a possible EV on the play) on the value of never calling with second pair... always calling with second pair.... second pair against various boards and types of opponents...
.. as well as the flip side- always bluffing on the river with 3rd pair, never doing it, doing it with certain boards... blah blah blah.

At least, IMO, that's how I try to use the tool... though I'm still trying to fully grasp the 'bad habits' that I might be ingraining... and what failings the TTH software has in encompassing all variables that could affect my decision process down the road....

And I take EVERYTHING I read, hear or learn with various-sized grains of salt (some are boulders!)

[pudley4]

</font><blockquote><font class="small">In reply to:</font><hr />
What I am hoping to do is to create a TTH personality that is superior to all of the built-in personalities and then use it to drill.

[/ QUOTE ]

Define "superior".

I took the "Maverick" profile and put it up against the low limit lineup, using the game conditions at my local card room ($2-$4, 10% rake, max $4, $1 jackpot drop at $15, $1 toke on pots above $20). After 1,000,000 hands, Maverick was a significant loser (although far ahead of the competition). I created a "Jones-style" profile that was a significant winner in this game. However, when I put the tight profile up against an all-Maverick lineup it got crushed. You need to always adjust to the current game conditions.

</font><blockquote><font class="small">In reply to:</font><hr />
Compared to Jones, Conan open-raises far more often. For example, Conan open-raises with A7s in MP while Jones calls. I do not know which is better but it is interesting to note the differences

[/ QUOTE ]

After 5 months of study you should be able to tell us why "Conan" would open-raise here and why Jones would open-limp here.

[Easy E]

"I took the "Maverick" profile and put it up against the low limit lineup, using the game conditions at my local card room ($2-$4, 10% rake, max $4, $1 jackpot drop at $15, $1 toke on pots above $20). After 1,000,000 hands, Maverick was a significant loser (although far ahead of the competition)."

** If you took the rake out, was Maverick a gross winner?
Also, does your statment that mean that every other Profile lost even more than Maverick did (I assume so)?

" I created a "Jones-style" profile that was a significant winner in this game. However, when I put the tight profile up against an all-Maverick lineup it got crushed. "
The "Jones-style" profile being the "tight" one?