Sunday, July 22, 2007

There Is No Correct Play Style

UTG+1 limped.
You are sitting on the CO with Jack of Spades Ten of Spades.
What do you do?

Answer: It depends…

In NL Holdem there are many different play styles to choose from, all with their own strengths and weaknesses.

A solid TAG will be making money because of having the best hand quite often. In a large % of the times where they are continuing postflop with a top pair hand, they will have the better kicker. Further a TAG will be trying to take a fair share of pots where they miss with continuation bets.

A smart LAG will be trying to take more than his share of the pots pre and postflop by trying to gauge the strength of their opponents hands and taking down pots when they have good folding equity. The second strength comes from getting paid off very well when they actually have a strong hand and someone with a hand like TPTK just doesn’t believe them anymore paying off for up to multiple buyins.

Tight-passives, also known as rocks, generally have a strong preflop hand when they play, and silly enough, a large part they lose compared to other play styles, which is taking down pots when they missed, they make up for by getting bluffed at too often. Even with an unimproved AK or AQ, they don’t seem to mind check-calling the flop, turn and sometimes even river.

These are the most common play styles, though there are many more like ultra-aggressive, loose-passive calling stations, ultra-nits and others.

I estimate most novice and intermediate poker players will try to figure out which play style works out the best for them, and stick to it as default play style.

But I think sticking to a default play style is a mistake. Even if one works out very well for you, it has certain disadvantages. Observant players can adjust accordingly after figuring you out, which is pretty easy for them if they use a HUD. Further, using a single play style can make you play hands in a sub-optimal way.

In my eyes someone who wants to become an advanced player should look at each situation individually and try to estimate the most EV way to play it. Think outside the box, and don’t even use logic such as “I am going to establish a TAG/LAG image and then change gears in a bit.” That logic may work for live tournaments/cash games, but has its flaws for online cash games. Image building doesn’t work on every player, since some people do not use HUD’s, and some people only look at their own cards and their TV.

Starting out with a default play style may have its merits when joining a new table without any info on the players yet, or if you multitable without a HUD. But if you want to be more flexible, you can avoid even this with some preparation, at least if you use a HUD and play on a PT-enabled site.

Before starting to play, open up a bunch of tables from your favorite site and let it datamine for about 30 minutes. In the meanwhile just browse some forums, check some hand histories from last session, watch some porn or at the very least do something other than sitting down at a table.

Even while you won’t have many hands on the players when you do sit down at a table, every bit of information is better than nothing, so use it properly.

Now take a look at the hand at the beginning of the article. You hold JTs in the CO, UTG+1 limped, what do you do?

Let’s say your HUD shows UTG+1 is relatively loose preflop, passive postflop, and folds to over 50% of all c-bets. The people to your left are not overly loose, and the BB doesn’t seem to be someone defending his blinds to steals all the time. The most EV move here would probably be to make a raise that would get called by UTG+1 most of the time and hopefully not by the blinds, so to something like 4xBB total, and c-bet any flop. Size of c-bet depends on flop texture.

If UTG+1 folds to c-bets less than 40% of the time or if the blinds seem loose, the correct move would probably be to just call and hope for a multiway pot with that holding.

Some people to your left with exceptional high PFR ratings? Fold unless you are effectively deep stacked.

Good table selection may be a way to maximize winnings when sticking to a certain play style, with basics such as people who like playing LAG looking for tight people to their right, TAG’s looking for loose people to their right.

But with a more flexible approach you will be dominating more tables by adjusting correctly. You will try to play every situation for maximum EV. A nice side effect is that your own stats don’t give dataminers at your table enough info to exploit you. At some tables you may look like a complete maniac, at another the biggest nit ever, but hopefully on both tables sitting behind a nice big stack.

More important, playing like this should increase your pre-, postflop and hand reading skills at a faster rate than sticking to any standard play.

Yours truly,

"Keep your friends close, and your enemies to your right."

Monday, July 16, 2007

Poker is a Game of Chance

Bean-counters from the tax department state this. It has to be correct, since the Law says so (the Dutch Supreme Court ruled it in 1998.) Many other people also agree with this, like for instance the group of people that buy state lottery tickets and have an estimated IQ of over 120. Those with a lower IQ would probably call it a game of luck instead.

“Poker is a game of incomplete information.”

Interesting statement, or at least, it must be, since it’s been written in every article on poker which is over one page long.

Let’s see if any of both of these statements are true, by taking this single hand as example.

Remember though, this is not about the actual hand, it's about the methods used to analyze other factors in a poker game than basic odds and hand strength. Feel free to let me hear your thoughts on this.

I also posted this on Internet Texas Holdem, click link to see the original post:

Available information:
I only played a few hands at this table, about 1 orbit, so I don’t have any reads yet on the other players.
I have slightly over a full stack, and the big stack is sitting two seats to my left, the small stack two to my right. Lousy table selection by me.

Full Tilt Poker - No Limit Hold'em Cash Game - $0.50/$1 Blinds - 6 Players - (LegoPoker HH Converter)

Soultwister (SB): $103.90
BB: $180.40
UTG: $81.70
MP: $100.00
CO: $58.50
BTN: $100.00

Preflop: Soultwister is dealt
Ace of Clubs Queen of Clubs (6 Players)
UTG calls $1.00, MP raises to $5.00, 2 folds,

What happened so far? UTG limped, which is a weak play in 6-max in my eyes. A player behind him raised, which doesn’t say much about his hand. He could be raising here for value with a strong hand, or just raising a decent hand to isolate against the limper.

The size of the current pot tells me I should fold or reraise, getting 1.66:1 pot odds out of position. But if I call, the big blind is getting good pot odds so may call with any two cards, and the player UTG will most likely call for the same reason. So my estimated pot odds of calling here would be between 2.66:1 and 3.44:1. Good enough for a call. Folding here isn’t bad either. Reraising is just too risky without more information, or at least would be a high variance play OOP without info. Not my style.

Soultwister calls $4.50, BB calls $4.00, UTG calls $4.00

Flop: ($20)
King of Spades Jack of Clubs 4 of Clubs (4 Players)
Soultwister bets $12.00, BB raises to $34.00, 2 folds, Soultwister raises all-in to $98.90,

Very nice flop for me, even while I just have ace high now. How many outs do I have to hit the stone-cold nuts if that is what I need to win? 9 clubs and 3 10’s, so 12 outs to the nuts. That gives me about 45% chance to hit the nut flush or nut straight by the river.

I can check here and decide what my best options are when the action gets back to me, most likely checkraising. I would have chosen that option if UTG had been the initial raiser, since that increases the chance to get more players/money into the pot. It’s unlikely that this flop will get checked through, so my relative position is great right here.

The other option is betting out small into the PF raiser. This gives UTG good odds to tag along if he has any broadway cards and add some more money into the pot, while MP may interpret this bet as an informational bet, semi-bluff, value bet or pot sweetener. And he’ll be correct.

After my donkbet the BB reraised me, which generally means he must have a very strong hand, since he doesn’t seem to be afraid of preflop raiser.

I made a mistake here, I figured my 12 outs and current pot size would probably give me good odds to be in a profitable all-in situation, so put my chips in without realizing that I was already getting 3:1 pot odds. With 12 outs, I only need to get 2.92:1 to make my hand on the turn. So there was no need to make a move here at all.

When I go all-in and get called, the total pot will be $217.8 while I invest $86,9 here. Which is 40% of the total pot, while my outs tell me I’m winning 45% of the time. So with that information readily available, I’m making money here?

If I just look at my outs I win (0.45*217.8 )-86.9=$11.11. Folding earns me $0. So with that info I’d be losing $11.11 by folding compared to pushing all-in if villain never folds.

Pot odds can be deceptive though since there are more hidden factors. Is this situation really as profitable as it seems?

First, the rake was not substracted from the total pot. The rake is $3 maximum on full tilt at this limit, I get about a third back thanks to rakeback, so $2 total rake which lowers my winnings from earlier calculation to (0.45 * 215.8 )-86,9=$10.21. Small difference in large pots, but can have quite some impact on small pots where the odds seem marginally +EV.

Second, I won’t be winning 45% of the time at all. Against villains hand range, his sets/two pairs could improve to a full house or quads. But how likely is that to happen?
cards win %win lose %lose tie %tie EV
Ac Qc 325 32.83 665 67.17 0 0.00 0.328
Kc Kd 665 67.17 325 32.83 0 0.00 0.672

EV: -$16.11

cards win %win lose %lose tie %tie EV
Ac Qc 423 42.73 558 56.36 9 0.91 0.432
Kc Ad 558 56.36 423 42.73 9 0.91 0.568

EV: $6.33

cards win %win lose %lose tie %tie EV
Ac Qc 409 41.31 581 58.69 0 0.00 0.413
Js Kc 581 58.69 409 41.31 0 0.00 0.587

EV: $2.23

There’s always a chance here that villain is making a move here with a draw or bluff, actually making me a favourite, but to calculate my odds I’m going to keep it simple and not take other hands into consideration, so let’s say he has at least top pair, two pair, or a set, and has one of these hands: JJ, AK, 44 or KJ. AA/KK look quite unlikely here because of his preflop play. And if I’m already making a worst case scenario, let’s scrap KQ/K10 from his range as well.

With the pot size I needed to win 40% of the time here to make this move profitable. There are 6 possible combinations for him to have a set if I take KK out of his hand range, 6 for the other hands mentioned. If each combination is just as likely that leads to (32.8+32.8+43.2+41.3)/4= 37,5% chance to win here.

So instead of earning $11.11 here like the first rough estimation I’ll be losing $5.98 here if I’m called all the time.

Does that make pushing all-in a –EV move? That depends on how often he needs to fold to make this profitable. If I move all-in here, I bet $86.9 (1.32 times pot) to win $64 (current pot – rake) if he folds. He will be getting (152.9/64.9) = 1 in 2.36 to call. What hands could he possibly fold here? Perhaps AK or KJ.

Time for some real math using the Fimbulwinter-method.

A=Current pot size – rake
X=Villain’s folding percentage / 100.
B= Winning chance
C= Total pot if called – rake
D= Your bet/raise

EV = Ax + (-x ) ((B*C )-D )

EV = 64x +(-x)((.37,5 )(215.8 )-86.9 )
EV= 64x +(-x )(-5.98 )
-x(-5.98 )=5.98x
EV = 64x +5.98x -5.98x = 0
5.98 = 64x +5.98x
X = 5.98/69.98 = 0.085

He only has to fold 8.5% of the time here to make this profitable. A random villain would lay down his hand more often than that for his remaining stack. So without trying to guess how often villain is actually reraising me here with a hand like KQ, K10, Q10, aces or a pair of deuces, this move is +EV right away.

BB calls $64.90

Turn: ($217.80)
Ace of Diamonds (2 Players - 1 All-In)

River: ($217.80)
2 of Spades (2 Players - 1 All-In)

Pot Size: $217.80 ($3 Rake)

BB had
King of Clubs Jack of Diamonds (two pair, Kings and Jacks) and WON (+$110.90 )
Soultwister had
Ace of Clubs Queen of Clubs (a pair of Aces) and LOST (-$103.90)

So is there any chance or luck involved? Short term indeed, but who cares what cards show up after you put your money in, poker is all about long term results. And in the long term, luck is a neglectable factor. And the chances are known.

Incomplete information? With the information available in this hand I could make no mistake that would make me lose more money than folding, even when having the worst hand
. I should have just called on the flop against an unknown, and reraise against some regulars when I know more about their post-flop tendencies. More information would just allow me to make slightly better estimates of expected value.

So what IS poker? I think it's just a game in which you have to minimize your mistakes and try to induce your opponents to make their mistakes larger. If you just do that, and also keep taking the dead money on the tables often enough, it should be a profitable game. And dead money on poker tables is quite common if you realize we live in a world where people who are considered to have an over-average intelligence buy state lottery tickets.

Yours truly,

"Keep your friends close, and your enemies to your right."

Return Of The Blog

It has been nearly a year since my last update. Where has Soultwister been?

Busting the tables still, or busting his bankroll?

A bit of both, but in my favor.

Around the time of my last update I switched from playing limit to No Limit. Why the switch?
As the saying goes, the key to limit poker is about trying to put your opponent to a decision for some of his chips. That was not good enough for me, I want them to make bad decisions for a large part of their chips.

Further, skill matters even more in NL than in limit, and it is a hundred times more fun. Correct bet sizing, position play, player reads, trying to make equity calculations on the fly, inducing bluffs for value and more factors like those make No Limit Hold'em the best game for me.

What limits did I play in NL?

I started out playing on $.10/.25 NL 10-max, working my bankroll up towards $.5/1 and a bit of $1/2. , though the later limits were all played shorthanded on 6-max tables.

6-max is where the real money in poker cash games is. First you will be playing a lot more hands per hour, secondly the average winning rate in BB/100 hands seems to be better. The combination of those makes it much more profitable then playing full-ring.

So how did I do?

From august up till march I withdrew a total profit of about €6000, so around $8200. I made a big mistake though, which is withdrawing too much money constantly spending it on RL things and leaving an online bankroll too small to adapt to big swings.

During april I noticed the error of my ways. At the end of april I had $2900 less online compared to the 10th of that month, even while the total loss of that month was only around $300, and actually still slightly profitable because of rakeback.

In the beginning of June I lost my last bankroll on Full Tilt Poker and leaving myself with an online bankroll of only $1200.

But was this because of variance, bad play or bad bankroll management?

It's a combination of all three factors, and this is also the reason why I returned to Blogging.

The last few months I've spent most my poker time analyzing my play and was shocked at the conclusions I could draw from it. I was terrible at making correct value bets, did not give other players enough room to make mistakes, did not look good enough at the pot odds I were offering my opponents and picked some of the worst spots for bluffing.

When analyzing a hand history the best way for me seemed to be to put it on paper, look at all factors involved in the hand and try to extract every bit of useful information you can find in it. Results of the played hand do not matter, even the most basic hand can be pretty interesting and give useful information to use at later times.

I also returned to the poker forum I used to post in when I played limit before, . It's a great place to post your thoughts, ask for other people's thoughts about played hands, or comment on their played hands yourself. Take a look at the forum if you haven't been there before, it is a great forum for beginners, novices and advanced players.

I will use a different approach here with my new posts than I used to. I'm no longer going to worry about the results of hands/situations posted here, or just worry about the optimal play, but rather am going to use these hands as a means to work out or explain concepts which can be used as a means to improve my game. And hopefully this will also help readers to improve parts of their game.

There will also be a lot more math involved in these posts. Even while the different calculations used to calculate the equity of a hand, required fold equity and other factors of the game will be very hard to calculate on the fly while in a hand, but after making many calculations like these, it becomes easier to estimate the expected value of different options or situations better.

Yours truly,

"Keep your friends close, and your enemies to your right."