Poker-AI.org

Re: Raising pre-flop

2014-03-31T23:46:49+00:00

Quote:

Never calling on the SB is a mistake btw.

Yes you are right.

NEO Calls on pre-flop IF:

(Card1 = 2) and (Card2 < K)
(Card1 = 3) and (Card2 < Q)
(Card1 = 4) and (Card2 < J)

else

raise

Tom

Statistics: Posted by Tom — Mon Mar 31, 2014 11:46 pm

Re: Raising pre-flop

2014-03-31T22:24:14+00:00

Quote:

I think Selby is Pre-Flop standard for HU Bots.

Well I have never heard of it but that does not mean much. To me having an optimal strategy that does not consider post flop play is not of much use. Many hands like suited connectors, gappers, broadways etc are better off being called when 3 betted which represent a big chunk of combos due to their post flop playability. The hands to 4B would be your big hands, some blockers (like a wheel Ax hand), small pairs, and the very worst of your range. I will have to look up exactly what I did but I think that somewhat close. I get confused on my bots 3B and 4B ranges unless I look at the code. They are actually quite a bit different.

Quote:

But some Bots like NEO play Pre-Flop more aggressively. On SMALL BLIND PLAY they ever Raise or Fold, never call.

I played a few thousand hands against NEO and it has some major leaks preflop. Against NEO all you have to do is 3B every hand on the SB and raise 100% of buttons. NEO never 3 or 4B folds so its very predictable if I remember correctly. Never calling on the SB is a mistake btw. You can beat that bot without ever playing post flop although they have probably fixed that by now.

Statistics: Posted by shalako — Mon Mar 31, 2014 10:24 pm

Re: Raising pre-flop

2014-03-31T18:36:09+00:00

Ok..so about 50 hands are called vs raised when the bot is on the button which are a ton of combos. That is nearly a third of all playable hands that you would call and fold to a potential 3B. That is a ton of fold equity you are losing by not raising as well. So..you not only losing money by not raising but it also looks like your also going to have a high fold to 3B% which will hemorrhage a ton of money as well.

Your bot should be raising at least 80% of buttons until the villain adapts. Mine will raise 100% of buttons with no player history. You must use position and fold equity to your advantage in HU.

So it looks like your calling 50 hands and folding 10..which would be a 65% button raise. From your chart it does not say what % of those raising hands are called to a 3B but only raised (R2)..or is that the CR1 and CR2 hands? If so that will only be maybe 25% of your raising range. So..just from a glance at this chart it looks like your fold to 3B % will definitely be way way too high.

If the guy is 3 betting the hell out of you then your bot must adapt by increasing your 4B or calling frequency or your gonna get crushed by fold equity alone.

Statistics: Posted by shalako — Mon Mar 31, 2014 6:36 pm

Re: Raising pre-flop

2014-03-31T18:46:30+00:00

Quote:

I have never heard of the Alex Selby optimal strategy for HU. Does C mean call? I have never heard of calling being correct on the button so I am curious about this. If my bot saw it call it would probably 3B bluff with a wide range of hands so I see this strategy as highly exploitable.

I think Selby is Pre-Flop standard for HU Bots.

Yes C means call and

F => Fold
C => Call
R1 => Raise
R2 => Raise, and reraise if raised back.
R3 => Raise, reraise and re-re-raise if raised back.
CR1 => Call-raise
CR2 => Call-raise and reraise if raised back.

And Star * means a random

e.g. in SMALL BLIND PLAY

*1 (KK) 60.1% R3, 39.9% CR2

Code:

If random >= 0.61 
  R3
else
  CR2

Check Selby here: http://poker.cs.ualberta.ca/publications/davidson.msc.pdf

Chapter 3.2 Pre-Flop Betting Strategy Page 19 - 20

or here https://groups.google.com/forum/#!msg/rec.gambling.poker/EBCQDWwHyGQ/TagTiDKCsuEJ

But some Bots like NEO play Pre-Flop more aggressively. On SMALL BLIND PLAY they ever Raise or Fold, never call.

So you could not use Preflop Game for Opponet modeling.

Tom

Statistics: Posted by Tom — Mon Mar 31, 2014 5:44 pm

Re: Raising pre-flop

2014-03-31T17:29:24+00:00

http://www.archduke.org/simplex/index.html

Alex Selby wrote:

These pages describe a computer program that finds the (or a) perfect strategy for two-player Texas Holdem if there is no betting after the flop

Statistics: Posted by spears — Mon Mar 31, 2014 5:29 pm

Re: Raising pre-flop

2014-03-31T16:40:06+00:00

Statistics: Posted by shalako — Mon Mar 31, 2014 4:40 pm

Re: Raising pre-flop

2014-03-31T18:19:24+00:00

Hello,

I think betting on the preflop is a very important Part of Poker.

in NL and FL my Bot use a Hutchison strategy:

http://erh2.homestead.com/hem.html

in HU my Bot playes the optimal Alex Selby Start Table : ->

Code:

  F  => Fold
  C  => Call
  R1 => Raise
  R2 => Raise, and reraise if raised back.
  R3 => Raise, reraise and re-re-raise if raised back.
 CR1 => Call-raise
 CR2 => Call-raise and reraise if raised back.

 SMALL BLIND PLAY    // Dealer

                               suited
      |  A   K   Q   J   T   9   8   7   6   5   4   3   2
    -------------------------------------------------------
   A |  R3  R2  R2  R2  R2  R1  R1  R1  R1  R1  R1  R1  R1
   K |  R2  *1  CR1 CR1 CR1 R1  R1  R1  R1  R1  R1  R1  R1
   Q |  CR1 CR1 R2  R1  R1  R1  R1  R1  R1  R1  R1  R1  R1
 u J |  R2  *2  R1  R2  R1  R1  R1  R1  R1  R1  R1  R1  R1
 n T |  *3  R1  R1  R1  R2  R1  R1  R1  R1  R1  R1  C   C
 s 9 |  R1  R1  R1  R1  R1  R2  R1  R1  R1  R1  C   C   C
 u 8 |  R1  R1  R1  R1  R1  R1  R2  R1  R1  C   C   C   C
 i 7 |  R1  R1  R1  R1  R1  R1  C   R2  C   C   C   C   C
 t 6 |  R1  R1  R1  R1  R1  C   C   C   R1  C   C   C   C
 e 5 |  R1  R1  R1  *4  C   C   C   C   C   R1  C   C   C
 d 4 |  R1  R1  C   C   C   C   C   C   C   C   R1  C   C
   3 |  R1  C   C   C   C   C   F   F   F   C   F   R1  C
   2 |  R1  C   C   C   C   C   F   F   F   F   F   F   C

 [*1] (KK)  60.1% R3, 39.9% CR2
 [*2] (KJo) 62.7% R1, 37.3% CR1
 [*3] (ATo) 73.0% R2, 27.0% CR1
 [*4] (J5o) 64.3% C , 35.7% R1



IF BLIND PLAY    // Not Dealer

 When the small blind has called

                               suited
     |  A   K   Q   J   T   9   8   7   6   5   4   3   2
   -------------------------------------------------------
   A |  R3  R2  R2  R2  R2  R1  R1  R1  R1  R1  R1  R1  R1
   K |  R2  *1  R1  R1  R1  R1  R1  R1  R1  R1  R1  R1  R1
   Q |  R2  R1  R2  R1  R1  R1  R1  R1  R1  R1  R1  R1  R1
   J |  R2  R1  R1  R2  R1  R1  R1  R1  R1  R1  R1  R1  R1
 n T |  R1  R1  R1  R1  R2  R1  R1  R1  R1  R1  R1  R1  C
 u 9 |  R1  R1  R1  R1  R1  R2  R1  R1  R1  R1  C   C   C
 t 8 |  R1  R1  R1  R1  R1  R1  R2  R1  R1  C   C   C   C
 d 7 |  R1  R1  R1  R1  R1  R1  *2  *3  C   C   C   C   C
   6 |  R1  R1  R1  R1  R1  R1  C   C   R1  C   C   C   C
   5 |  R1  R1  R1  R1  R1  C   C   C   C   R1  C   C   C
   4 |  R1  R1  R1  R1  C   C   C   C   C   C   R1  C   C
   3 |  R1  R1  R1  R1  C   C   C   C   C   C   C   R1  C
   2 |  R1  R1  R1  C   C   C   C   C   C   C   C   C   R1


 [*1] (KK)  98.1% R2,  1.9% R3
 [*2] (87o) 26.0% C,  74.0% R1
 [*3] (77)  99.8% R1,  0.2% R2



 When the small blind has raised  // Not Dealer

                               suited
     |  A   K   Q   J   T   9   8   7   6   5   4   3   2
   -------------------------------------------------------
   A |  R3  R2  R1  R1  R1  R1  R1  R1  R1  R1  R1  R1  R1
   K |  *1  R2  R1  R1  R1  R1  R1  C   C   C   C   C   C
   Q |  R1  R1  R2  R1  R1  C   C   C   C   C   C   C   C
   J |  R1  R1  R1  R2  C   C   C   C   C   C   C   C   C
 n T |  R1  R1  C   C   R1  C   C   C   C   C   C   C   C
 u 9 |  R1  R1  C   C   C   R1  C   C   C   C   C   C   C
 t 8 |  R1  C   C   C   C   C   R1  C   C   C   C   C   C
 d 7 |  R1  C   C   C   C   C   C   R1  C   C   C   C   C
   6 |  R1  C   C   C   C   C   C   C   R1  C   C   C   C
   5 |  *2  C   C   C   C   C   C   C   C   R1  C   C   C
   4 |  C   C   C   C   C   C   C   C   C   C   C   C   C
   3 |  C   C   C   C   C   C   C   C   C   C   C   C   C
   2 |  C   C   C   C   C   C   C   C   C   C   C   C   C


 [*1] (AKo)  14.2% R1, 85.8% R2
 [*2] (A5o)   5.4% C,  94.6% R1

Statistics: Posted by Tom — Tue Mar 25, 2014 6:47 pm

Re: Raising pre-flop

2013-12-18T04:34:31+00:00

Quote:

I'm not even considering my position yet, but I'll work on that this week.

Any thoughts? Do you have any reference for a good strategy I could implement for FL?

Position is critical so get that working then lets see how it compares to what I have. I never used a LUT so I am not familiar with any of those numbers as I just hard coded a bunch of IF THEN, CASE statements etc.

Statistics: Posted by shalako — Wed Dec 18, 2013 4:34 am

Re: Raising pre-flop

2013-12-16T16:19:37+00:00

shalako wrote:

Quote:

I think I didn't get this one? I always follow the a prefixed strategy (based on tables like this: http://www.pokerstrategy.com/download/c ... _fl_en.pdf , so if I get 10 premium hands in a row, I'll bet 10 times

Ok I looked at that FL chart its wrong all over the place. For one thing it needs to get more specific then just Early, Middle and Late positions. There is a big difference in hand selection even between UTG and UTG+1

So right now I'm not following that specific strategy. What I'm doing is using a LUT for pre-flop hand strengths, based on #opponents. And then I select my ranges like this:

PREMIUM: 1.5,
RERAISE: 5,
RAISE: 17,
CALL: 30,
LIMP: 35,

Meaning the strategy to use (PREMIUM=always raise, RERAISE=raise, and re-raise if 3betted, etc...) for the top Nth% hands (from my LUT table). I also make some adjusts like, if a pair will be folded based on this strategy, change it to CALL.

I'm not even considering my position yet, but I'll work on that this week.

Any thoughts? Do you have any reference for a good strategy I could implement for FL?

Thanks!

Statistics: Posted by corintio — Mon Dec 16, 2013 4:19 pm

Re: Raising pre-flop

2013-12-16T16:09:48+00:00

At least a little tighter. 32 VPIP would be on the looser side at an SH table, but at an FR table it's really loose.

Are you limping a lot? It can be good, when there are many limpers, with a good multiway hand, but it's rarelly good first in. You should also respect raises on the micros and not defend any 2.

I always used the pokerstrategy approx charts mixed with the pokerstrategy silver charts as a basis for my preflop game. I loosened up a little in the later positions.

Statistics: Posted by HontoNiBaka — Mon Dec 16, 2013 4:09 pm

Re: Raising pre-flop

2013-12-16T16:03:56+00:00

HontoNiBaka wrote:

They are stupid. They will also tell you, that raising with an open ended flushdraw is stupid, even though you are the favourite with such a strong draw, because "lol, stupid raising with no hand"

You stats look a little strange IMO. Pretty big gap between VPIP and PFR and high PFR for 10 max tables. What preflop recommendations do you follow?

Thanks for your reply HontoNiBaka.

So you think I should lower my PFR? Regarding the gap between VPIP and PFR, that's because I'm trying to force the bot to be a little more loose. Do you think I should keep it tight?

Statistics: Posted by corintio — Mon Dec 16, 2013 4:03 pm

Re: Raising pre-flop

2013-12-16T16:00:08+00:00

shalako wrote:

I have a FL 10 player AI that I used back in 2004 on PartyPoker that did very well at 2/4 and below. I can dig up the preflop strategy if you need more help. I was rather complicated as I recall. At those limits and the wealth of loose FL tables during that time I found opponent modeling was not needed until 3/6+

I had a rather sophisticated player modeling database running but found it didnt deviate from the baseline strategy as often as I thought it would even at 3/6.

Hey shalako, it would be awsome if you could send your preflop strategy to me, as I'm using FL to tune my bot (correct table reading, automatic play with pauses, etc..) and don't want to loose too much

Regarding opponent modeling, one thing that I'm experimenting right now is to detect the "always raise" players and respond accordingly, or else my bot will keep folding two pairs and even small three-of-a-kinds

Thanks!

Statistics: Posted by corintio — Mon Dec 16, 2013 4:00 pm

Re: Raising pre-flop

2013-12-14T20:07:14+00:00

Quote:

Statistics: Posted by shalako — Sat Dec 14, 2013 8:07 pm

Re: Raising pre-flop

2013-12-14T13:27:00+00:00

Statistics: Posted by HontoNiBaka — Sat Dec 14, 2013 1:27 pm

Re: Raising pre-flop

2013-12-14T05:24:40+00:00

Statistics: Posted by shalako — Sat Dec 14, 2013 5:24 am

Re: Raising pre-flop

2013-12-14T04:31:52+00:00

Thanks for the reply.

Nose wrote:

Ok, first of all you should consider your opponents. They are playing on play-money tables so probably they only have a limited clue about the game.

Yeah, that's what I thought.

Quote:

Next there are many factors to consider when preflop-raising:

a) bet-/ raising size

I'm playing on fixed limit tables, so I don't think this matters in the pre-flop.

Quote:

b) betting frequency

Quote:

c) how many of your bets are you defending? ie: you play 32/17 but only defend {KK+, AK} against a min-3bet then your preflop-play can be considered poor

Yeah, right now my bot is weak in this point, it folds a lot for 3bets... I'm working on that right now.

Quote:

but yes, I do bet preflop

Thanks again for the insights
-Corintio

Statistics: Posted by corintio — Sat Dec 14, 2013 4:31 am

Re: Raising pre-flop

2013-12-14T04:17:20+00:00

Ok, first of all you should consider your opponents. They are playing on play-money tables so probably they only have a limited clue about the game.

Next there are many factors to consider when preflop-raising:

a) bet-/ raising size
b) betting frequency
c) how many of your bets are you defending? ie: you play 32/17 but only defend {KK+, AK} against a min-3bet then your preflop-play can be considered poor

...

but yes, I do bet preflop

Statistics: Posted by Nose — Sat Dec 14, 2013 4:17 am

Raising pre-flop

2013-12-14T02:17:02+00:00

Hi guys,

Finally I got my bot (alpha version) running, and playing against the high stakes limit play-money. And it's winning, but it's far from a real winner.

I'm using some standard pre-flop strategies, raising the top 17% (+/-) hands, but in general my bot is very tight. So my vpip/pfr is about 32/17. And my raises are bothering the players at the table, who start complaining about my raises, and saying that raising on the pre-flop is just plain stupid.

Well, what I would like to know from you is: have I got this wrong? All strategy guides, books, etc... I find on the Net say we should raise at pre-flop, we have even metrics for that (vpip/pfr). Do you (and your bots) raise pre-flop? If yes, is facing this issue (with the players) normal?

Thanks for any advice

Statistics: Posted by corintio — Sat Dec 14, 2013 2:17 am