2/28/2006

02-28-06 [poker] - 2

02-28-06 [poker]

There's a good training technique in Limit Hold'em : whenever you feel tempted to call, either fold or raise. Obviously this is not optimal, but it's a good way to force yourself out of passive calling play and make you commit to either thinking your hand is best or not. It helps your game to make yourself play this way for a few sessions. I've found a new one for myself in No Limit. Whenever I feel tempted to make a small bet or a "feeler" bet, either make a full size bet, or just check. A lot of the time I find myself with a mediocre hand not really knowing what to do, so I make a small bet. That's terrible for a lot of reasions, so I'm trying to train myself out of it.

In other news I feel like my game is back on, so I'm getting read to move up to $200NL again. My bankroll is currently at $3500 and when it hits $4000 I'll start playing $200. As long as I keep playing okay that should be this week.

02-28-06 [poker] - 1

02-28-06 [poker]

I believe the experiment with my new name has been successful; I get people chatting to me about it all the time when they do nutty things. In the future all my names are going to be things like "usuck" "IbluffU" "urafish" things like that. A lot of psychologically weak players are easily affected by that. When you bet and they fold, if your name is "tightrock" they assume you have, but if your names is "Ilove2bluff" they assume you bluffed them. On a related note, I'm finding that having the chat off is great for my game. I sort of enjoyed trash talking and chatting, but it was distracting, and now I just focus on the action more. This type of name works on fishes who are looking at names and not PokerTracker stats.

Another good type of name to pick is to just make yourself look like a total douche. This type of name works on semi-pros who assume that everyone out there is a moron fish and just need an excuse to think you are. Names like "PhilIvyRulez" will have you pegged as a moron douche (note intentional mispelling of Ivey), or "luvHotchiX111", or "fittycent" or "mobenjaminz". Names with "fish" or "rock" or something in your name are a bad idea because they show that you know something about poker. On the other hand names like "str8flush" are fishy all the way.

2/27/2006

02-27-06 [poker] - 2

02-27-06 [poker]

There are two types of uncertainty (and hence gambling) in poker - the cards your opponent has and the cards that will come. For some reason gambling on the cards that will come is very easy, because you can know for sure if it's a +EV move or not, but gambling on what your opponent has is very hard for me.

02-27-06 [poker] - 1

02-27-06 [poker]

If you add up all the little edges and decisions in poker, they become significant. Just some wild guesses, but at low limit NL I think you can make 5 PTBB/100 just playing super vanilla nut-peddling (1 PTBB = 2 Big Blinds). Then there are lots of little things you can add to your game which edge that up. Playing a bit looser and attacking pots will give you 2-3 PTBB/100, not so much because you make money on those hands but because it increases your pay-off on your big hands. Catching some bluffs and trapping the super-aggros is another 2-3 PTBB/100. Good bet sizing and trapping people for their stack is another edge. Stealing the blinds and playing well against steals is another chunk (the blinds at 6-max are 12.5 PTBB/100, but good play probably just give you a 2-3 PTBB/100 edge on them). Add this all up and perfect play probably gets you 20-25 PTBB/100 , which is pretty insane. Of course just playing super-vanilla you can say you are a "winning player" and you don't need to do any of these other things, but all that value adds up. Say you play 30% hand of hands, so over 100 hands you play 30 ; if you made an extra 1/2 of a big blind on each of those, that's 7.5 PTBB/100.

2/26/2006

02-26-06 [poker] - 2

02-26-06 [poker]

I've been thinking about bet sizing and "The Price Is Right". Say you know you have the best hand on the river and he has something mediocre. You want to bet as much as possible such that he'll call. The more you bet (that he calls) the more profit you make. If you bet too much, he'll just fold and you make $0. This is like the "Price is Right" in that you want to get as close as possible to his maximum call amount, but without going over. Your profit as a function of bet size is :

P(B) = (B < B_max) ? B : 0

The problem is you don't know B_max, but you can estimate it with some error. Let's say you know it's between L and H, with an even probability of being anywhere in there. What should you bet to maximize profit? Clearly it's somewhere in the range [L,H] , but where exactly?. Maybe I'll do the math and figure out where exactly.

Well, Sean beat me to the math and sent me this :

Is it clearly in the middle? Clearly we have the boundary cases,
let N be the bet amount:

   N = L -> E = $N
   N = H -> E = $0

So you might hope that it's a curve that grows as N increases,
and then starts decreasing again. But that might be overwon by
the person's tendency to fold. Now, if L is 0, and H is non-zero,
clearly there must be some > 0 value, though. So that sounds
good. But I think in practice it's not. (Also, this totally changes
for a gaussian instead of a uniform distributution).

For a bet of N, probability of folding is (N-L)/(H-L), so probabiliity
of calling is 1-(N-L)/(H-L).

E = N - N*(N-L)/(H-L)
E = N - (N*N-N*L)/(H-L)
H-L is constant, so computing the derivative ignores it:

dE/dN = 1 - (2N - L)/(H-L)

set that to 0:

1 = (2N-L)/(H-L)
(H-L) = (2N-L)
H = 2N
N = H/2

Weirdly, L canceled out, so I probably screwed up. But it could
be true, that that's always the peak of the quadratic, and therefore
the optimal result is:

     L <= H/2: N = H/2
     L >= H/2: N = L

This makes some intuitive sense. Let's ask whether it's a good
idea to bet (H+L)/2, versus betting L.

In other words, we have a sure thing of making $L, so let's factor
that out (I think this makes it more intuitive). Now we want to know
if we should increase that to $(H+L)/2.

If we increase, our increase versus $L is:
   $(H-L)/2
and we have a 50/50 chance at it. If we increase and lose,
we lose the $L
   -$L

so E = 0.5 * (H-L)/2 + 0.5 * -L
       = 0.25*H - 0.25*L - 0.5*L
       = 0.25*H - 0.75*L
So this says if H is less than _3 times_ L, it's a losing bet
to take the 50/50 gamble of raising to (H+L)/2. So it makes
sense; when L gets close to H, the rate of extra-money-beyond
L doesn't sufficiently offset the increased chance of losing L.

That all looks right to me, and is kind of interesting. If L is significant at all, you want to bet the maximum that you're *sure* they'll call, and trying to edge it a little higher is -EV because the risk of losing them is too great. Say for example L is $20 and H is $40, you must bet just $20 !! Even $21 is worse because it makes them fold 1/20th of the time so your EV is just (19/20)*21 = $19.95. However if L is zero, eg. they have something so weak they might not call any bet, the best bet is H/2.

In reality, they have more like a Gaussian distribution of call values, though I doubt that changes the answer too much. Also you have to look at multiple models, say there's a 75% they call based on a model like this, and a 25% chance they call any bet (but your stacks are not huge compared to the pot, so this isn't insane). In that case there's a bit more reward for edging your bet up, the risk of losing them if they have something weak is hedged a bit by the EV gain when they have something they call any bet with.

02-26-06 [poker] - 1

02-26-06 [poker]

Infrequent events have a lot more noise. If someone plays 90% of hands, +/- 1 event is not going to change that much. However, if they actually raise 2.5% of hands, then a little run of hot or cold cards can throw that number way off, anywhere between 0% and 8% even on pretty big samples (like 200 hands). This actually is very significant when you're trying to figure out what very tight people raise with. Say you're playing someone who you've seen raise with 2% of cards, which roughly means JJ+. They come in for a raise and you look down at JJ. Yikes!! If your stats are actually accurate then you should probably just dump JJ. However, they could easily be someone who opens 5% of cards and have just been on a bit of a cold spell.

02-26-06 - 3

02-26-06

Things I have to do to sell my poker app :

  • Various code fixes, write up help, etc. This is the easy part

  • Make a web site for my retail presence. Have to hire someone to do this since my web = ass.

  • Advertise & spread word of mouth. Maybe get a partnership.

  • Set up an Installer & the whole Shareware registration code thing. I'd buy Armadillo so this is reasonably easy.

  • Make it work on PokerStars. Right now it only works on Party.

None of that is too hard.

02-26-06 - 2

02-26-06

I think it would be hillarious to hire a migrant worker to work my vegetable patch. It's like a 8' x 4' patch of dirt, I could sit out there and drink mint juleps and watch him hoe and such.

02-26-06 - 1

02-26-06

Win32 question : I would love my program to act like either a windows program or a console (CLI) app depending on how it's started. eg. if it's run from a DOS prompt (ok, a "cmd" prompt), then act like a console app. If it's started any other way, act like a windows app. The easiest thing is just to make it a console app, but then when I start it as a windows app I get this ugly DOS output box popping up. If I make it a windows app, there must be some way to detect I started from a DOS prompt and hook up my stdin/stdout to that console.

Right now I'm doing two different builds, like this :


#ifdef _CONSOLE
int main(int argc,char *argv[])
{
#else
int CALLBACK WinMain ( IN HINSTANCE hInstance, IN HINSTANCE hPrevInstance, IN LPSTR lpCmdLine, IN int nShowCmd )
{
	int argc = __argc;
	char ** argv = __argv;
#endif

and that's super lame.

I find this to be one of those things that's impossible to search for on google. Any kind of search for "WinMain _console" things like that just give you super-basic pages about how to make "Hello World" in Win32. I'm sure someone has done a nice solution for this before, but I've never seen it.

2/25/2006

02-25-06 [poker] - 2

02-25-06 [poker]

Two interesting non-standard hands :

Hand 1 :

$100 NL, $1 BB
Pre-flop: (5 players) Hero is Button with K K 
2 folds, Hero raises to $4, SB calls $3.5, BB folds.

Flop: T T A ($9, 2 players)
SB checks, Hero ????

The normal line here is just to c-bet as if I had the ace. That's what you should do 90% of the time. But if you are against the kind of loose villain who will pay you off later with any pair, you should check. Why? If you bet here he will fold any hand you beat, and only call with better hands. However, if you check and the turn comes an 8 and he holds Q8, now he'll call a bet with a worse hand, and might even lead with it. If he had an ace all along you're losing your bet anyway, so why not wait for him to make a hand that you beat in some cases? Also there are no draws on this flop except a gutshot (and you hold a lot of the outs for it) so it's not too risky to give a free card - eg. almost any card that comes makes villain a hand that's still worse than yours. The interesting thing here is that it's a good spot to slow play (but only against this particular kind of donkey opponent), even though in the scheme of things my hand sucks, I can't beat a T or A, but I'm just ignoring those hands since I'm losing to them no matter what I do, I just want to maximize against all the other hands he might have.

Hand 2 :

Stack sizes:
Hero: $100
CO: $59.95

Pre-flop: (6 players) Hero is UTG with 9 9 
Hero calls $1, UTG+1 raises to $2, CO raises to $7, 3 folds, Hero calls $6, UTG+1 folds.

Flop: 3 A 6 ($17.5, 2 players)
Hero checks, CO bets $13, Hero ???

The normal line here is for Hero to just fold. I called preflop trying to hit my set and missed. But wait - look how short CO's stack is, and look at the action preflop - he put in a solid *reraise*. Against most people this means he has AA-QQ. When an A hits on the flop, there are only 3 ways he can have AA and 12 ways he can have KK or QQ, which means it's 80% (12/15) likely that he has KK or QQ, and is now scared of the ace. The best move here is to push all in! Of course you can only do this if he can fold a QQ or KK here, but most people can fold to a check-raise push. Another line is just to lead at this flop and then push the turn. You have to be pretty sure that CO is only reraising QQ+ preflop, if he would also reraise AK then that hand is very likely and problematic. You have to have a pretty solid read for this move to work, villain needs to be pretty tight/weak, predictable enough to only have QQ+ preflop, and weak/smart enough to be able to fold his QQ on the flop. In that case preflop you're not only drawing to a set, you're also drawing to an Ace! This is sort of like when you draw to a straight with a flush draw on the board, you have some outs to make your hand and extra outs to a good bluff card.

02-25-06 [poker] - 1

02-25-06 [poker]

Even the worst players know that there are cases where No Pair is good and cases where One Pair is not. Obviously if you have something like AQ and a really aggressive guy check-raises your cbet, your hand may well be good. On the other hand if you have something like KT with the top pair T's, but the board is 89T and there's a bet and raise in front of you, your hand is junk. This is pretty basic poker.

On the other hand, even relatively good players can't see the same thing when the hand values are escalated. There are times when Two Pair is a monster, and times when a set or a flush are junk. Beginner fishes make the mistake of getting stuck on One Pair hands. Intermediate fishes make the mistake of getting stuck on "good" hands. The right thing to do is always adaptively evaluate your hand strength based on the situation.

A good example of this is when you have a hand like AA and he has JJ. Any decent player can get away from JJ, but when the flop comes 888, a lot of people get stuck on it. Their hand looks great now because they have a house, but of course their JJ is still beat by just the same things as before.

02-25-06 - 3

02-25-06

We saw "Born Heller" (a band) at the Monday Club last night. It was pretty cool. They make some really weird sounds with harp and bass, and the girl sings like an opera singer or something. Unfortunately it appears they don't have any mp3's on the web. Some of their songs were just beautiful, erie, haunting, but then they went into this artsy-farsty dischordance shit which was very unpleasant. Also saw "Mi and Lau" there who I thought were very boring and uninspired.

02-25-06 - 2

02-25-06

There's a stream that runs beside my house. You have to climb down an embankment, and there you'll find a little waterfall, and a pool about 5 feet wide and 3 feet deep. The water is crystal clear and the rocks smooth. I often go and sit by this little pool, and it has a secret few know. A giant old catfish lives in this pool. The creek is too small for him to swim out, so he's stuck in it. Usually he hides behind the stones and few see him, but I go to the pool and whistle a little tune that he knows, and he comes right to the surface and sticks out his head. We've spent many afternoons together, talking about water skimmers and tadpoles, and politics.

02-25-06 - 1

02-25-06

The sectarianism and divisions in America today are because of cable TV. In the good old days you only had 3 networks to choose from, and they were all the same anyway - broadcasting a homogenous inoffensive, cohesive American identity. Now each group can watch their own cable channel that caters to their small world-view.

2/24/2006

02-24-06 [poker] - 1

02-24-06 [poker]

Some entertaining poker gossip for those of you who don't follow the poker dirt rags :

Andy Beal came back to play the Corporation again after all ($50k/$100k limit hold'em). He beat the original players (Jen Harman, Tedd Forrest, Todd Brunson) out of $6 million or so to make him up +$3. Then Phil Ivey sat in. (cue ominous music). Phil played Andy for about 3 days straight, long sessions, and Phil was in fast aggressive gear the whole time. Phil destroyed him for over $16 million. As I understand Phil was playing with the Corporation bankroll, not his own, so the profit gets split between all the pros who backed the Corporation. Too bad for Phil, the rest of them kind of stunk it up and he destroyed.

In online news, Party Poker has done some serious cracking down on cheaters. Two of the top online tournament players - ZeeJustin and JJProdigy - have had their party accounts closed and all the funds seized because they were found to have entered the same tournament with multiple accounts. Most of the top pros play multiple accounts, and most have admitted to having friends play their accounts, and then which ever one gets deepest in the tournament, the top pro takes over. This is the first time the poker sites have gotten serious about cracking down on these guys. Each one of them lost roughly $200k by getting their accounts frozen.

02-24-06 - 1

02-24-06

I've been thinking about the whole GoldBullion retail thing, and there's one thing that makes me uneasy about it. If only top players buy it, then selling it is -EV for me. That is, if say only the top 1000 Party semi-pros buy it, I make maybe $10k profit, but I've just created a super breed of multi-tabling sharks that will make the tables much worse for me to play at, and that could eat into $10k of my poker play profits!

There are currently about 2 million registered online poker players, about about 200,000 play on any one day. The amount of money bet per month is perhaps somewhere around $400 M , though it's hard to get a good estimate of that. I don't have a good feeling for how much of that market I could penetrate. Certainly PokerTracker has sold a ton, but there's a ton of competitors, like PokerEdge, PokerOffice, etc. which I don't think have sold as much (and fight against my market share). The little helpers like MTH and PartyMine I don't think have actually sold much, but I have no idea. Perhaps the closest comparison to me is PokerAce HUD, I'm curious what his sales are.

The other potential big problem for me is if Party changes their software again in a big way which breaks my app. Obviously they are trying to fight helper apps somewhat to level the playing field, which I think is good in theory. I doubt they'll make another big change any time soon, since this last one caused them so much trouble. However, there are some small changes they could easily make which would cause me a lot of trouble, which I just have to pray they don't do.

It's kind of a dilemma. If I just use it and don't release it, it's an awesome +EV tool for me which improves my poker profit. If I do release it, it might hurt my poker profit, and help a lot of other people. I would make some money on sales, but hard to say how much. If it ever got real big, Party would crack down and do something about it, and then I'd be busto. It's a gamble.

2/23/2006

02-23-06 [poker] - 1

02-23-06 [poker]

Right now the online poker world is a food chain of horrible fishes and whales and more or less skilled sharks; the sharks are equipped with all kinds of helper software, books, web sites, they multi-table and get bonuses and rakeback, and the fishes just get destroyed. This is not really the community that the major poker sites want. They'd much rather have everyone be semi-skilled so that it's more a game of chance, more like blackjack or slots or normal casino games. That way they can just skim a rake, no one is a big winner or loser, so the losers don't go busto and just keep playing and pumping the rake, the way they play the lottery. As it is now, a fish might buy in for $1000 and lose $900 of that to a shark, and $100 goes to rake. The shark just cashes out that $900, the rake in games he plays is effectively paid by the fishes. In the ideal world (for the poker sites), two fishes would just play each other and trade money back and forth. Neither would cash out a signficant win and eventually all their money would be lost to the rake.

There are more and more semi-pros these days, and most of them multi-table which effectively multiplies their presence. At almost any table at party there's at least one semi-pro and often 2-3. At a 6-max table this means there are usually only 2-3 bad players and a few other semi-pros. At the $2000 NL tablesd for example, it's hard to sit without running into lolo or Liquid_Farts. Those guys are cash sinks, basically you're going to lose a few $ an hour playing with them because they're better than you, and certainly with the rake it's not profitable to play them. Thus, the other guys at the table have to be enough of a cash source to make the table +EV. Obviously it would be awesome if no one was allowed to multi-table except me. I do think that the 10-tabling that some of the pros do is pretty sick. If you have a bunch of pros 10-tabling, it's hard to find any tables without several of them! I think a limit of 3 or 4 tables would be pretty reasonable.

02-23-06 - 1

02-23-06

How do people do GUIs in Windows these days? I've only ever done manual Win32 GUIs, which I don't mind coding, but they look pretty ghetto unless you put a lot of work into sprucing them up. I worked with MFC a tiny bit and consider a giant turd. I don't really want to get into C# at the moment since my app is all C++ and I like it like that. So, is there some new MS way of doing GUIs that's better than MFC? I've always hated all that autogenerated code and so on that MS does, but I guess it's hard to avoid.

2/22/2006

02-22-06 [poker] - 4

02-22-06 [poker]

I find the term "nut flush" and "nut straight" confusing. When someone says "I flopped the nut straight". Did you flop the nuts? (which was a straight), or did you have the nut straight and the true nuts was something else (a flush, I guess).

In play news, my dry spell is officially over. I've had some negative variance recently and also made a lot of bad mistakes. Today I played great and got lucky and scored bigtime. I know this will sound trivial/stupid, but I really play better when I'm catching cards and getting lucky. I don't fuck around with marginal hands, and it also gives me a great image so I can terrorize people with preflop raises and steals. All that attacking that I do doesn't work if you can never show down a winner. When I'm running bad, I start getting impatient and trying to force things and trying to win hands that I can't, and find a profit somewhere.

02-22-06 [poker] - 3

02-22-06 [poker]

Some common poker wisdom that I think is pure BS :

1. Build a pot with your draw so that it's big if you hit. There are perhaps cases where this is true (like if you have a draw that's > 50% to hit), but it's almost always -EV.

2. Don't act in such a way that you'll have a hard decision later. eg. don't raise because if he reraises it will be a tough decision. Umm...no. Make the best decision at each point. An actual tough decision is actually not a bad thing, if a decision is very hard it means it's roughly EV neutral.

3. Don't chase a draw if you're going to fold when it hits. Or if you hit your hand and fold people will say "why did you play that in the first place?". Totally wrong. You were playing to make certain hands in certain situations. If your opponent tells you you're beat, you fold even though you hit the hand you wanted.

02-22-06 [poker] - 2

02-22-06 [poker]

I just put up a new article on playing one pair in the poker section. Enjoy. The example hands are all recent, and I think I misplayed 3 out of 5 of them in real life, so obviously this is still a big problem in my game. I think my offline poker analysis skills are very strong, but I have a lot of trouble actually doing it fast enough at the table. Yes, the cbloom poker computer needs a CPU upgrade.

I'm working on another article which I think is much cooler about advanced concepts in stack sizes. I think there's some material there I've never seen discussed anywhere else (like "Pot Sticking Implied Odds"), so hopefully I'll actually finish it some day.

02-22-06 [poker] - 1

02-22-06 [poker]

I'll make plays all the time that are not optimal based on knowing that my opponents suck. One classic case is when you know your opponent is on a hand that they can't release. If they were good, they'd be able to fold it when you show big strength, but you know your opponent can't fold a low flush, or good overpair, etc. so you can just go ahead and jam the pot. Against an optimal opponent you'd have to use a lot more deception to maximize EV, but against bad players you just jam it. I think a lot of what Sklansky talks about in the HSNL forum doesn't correctly take this into account - you're playing against humans who make mistakes, not against optimal opponents.

02-22-06 - 2

02-22-06

I went for a nice long bike ride in the Edna Valley today, my first of the year. It's so beautiful out there right now, all green from the rain. The grape vines are still bare of leaves, but the rows between them are full of grass and wild flowers.

02-22-06 - 1

02-22-06

The new Kleenex box designs are really cool; I like the brown/burgundy one that's sort of a paisley grid thing. Yeah, I spend a lot of time with Kleenex. If only mucus was a precious substance, I'd be rich!

2/21/2006

02-21-06 [poker] - 2

02-21-06 [poker]

I see a lot of people online playing this really horrible style. It goes like this : limp a lot of hands, call with draws, try to hit something really big. When you do hit something big, bet hard. If not, fold to aggression. This style should totally suck, because you can just fold when they hit a hand and start betting, but it's amazing how successful these guys sometimes are. I pay them off sometimes and I hate it when I do. The way to play these guys is raise preflop against their limps, bet hard on the flop with or without a hand, and just fold when they push at you.

02-21-06 [poker] - 1

02-21-06 [poker]

The crucial street is determined by stack size. B is your stack size in terms of BB's.

B < 10 , the crucial street is preflop. Your big moves are preflop - any pot you play is for your stack.

B in 10-30 , the crucial street is the flop. Your key move is on the flop. You have to be very careful with a continuation bet or a raise on the flop because

B in 30-80 , the crucial street is the turn. With this stack you can preflop raise and cbet or call on the flop with a lot more hands, because you will be making the key decision on the turn of whether to put in a lot of chips.

B > 80, the crucial street is the river.

Of course this varies and it's very rough. For example if there's heavy preflop action, that pushes the crucial street earlier because it swells the pot in relation to the stack sizes.

This comes up in lots of types of thinking. Say Player A is a superior postflop player vs. Player B. They play a match where Player A is forced to play every 2 cards he's dealt preflop. If the stacks are small, Player B will destroy player A, say B < 10. If the stacks are big, like B > 100, then Player A can destroy Player B despite a huge preflop disadvantage, since Player B can't press his advantage without risking his stack and moving a lot of chips postflop.

2/20/2006

02-20-06 [poker] - 2

02-20-06 [poker]

People talk about how the story of guys like Chris Moneymaker is so great for poker. Personally I never found that story very compelling. Yeah, some shmo can win the lottery. Big fucking deal, we don't do lifestyle stories about the guy who bought the lottery ticket each week. "I've just been buying lottery tickets in my home town each week. I'm just having fun. I'm not even the 4th best lottery player in my home town." Lame. The real great poker story are the 16 year old kids who are making $500k a year by being more skilled at high limit games and destroying the online game. Self-taught prodigies that built a bankroll from nothing and are crushing a game that rewards pure intellect.

02-20-06 [poker] - 1

02-20-06 [poker]

6-max : great because you'll have more hands heads up with really bad players, which gives you a bigger edge. Also the players just tend to be worse on average compared to full ring, more loose maniacs. Favors players who know how to play around the blinds over those who don't, or who are too tight or too passive. Favors TAG and LAG.

full ring : great because the mistakes that bad players make are worse in full ring. eg. really punishes them for being loose, especially in all the early positions. Lets you be really tight and not pay too much in blinds. If you sit to the left of a fish, you have more hands in position against them (8/10 instead of 4/6). Easier to multi-table because you can play very tight vanilla value poker (playing tight means fewer hands at a time and fewer tricky hands). Really favors tightness, even tight/weak and tight/passive players can make decent profit.

02-20-06 - 2

02-20-06

Dan and I went to Rinconada mine today. I'd never actually been inside the mine before, I'd heard it was closed and couldn't find the opening, so thought it was bulldozed in, but that's all wrong. We hiked up to the top of the grade, the flowers are blooming, and there's also snow on the ground! It snowed last night on the Cuesta Grade and there are patches of icy snow in the shade. The mine caverns are amazing, huge carved out holes traversed by thin bridges of rock. It really reminded me of the Blackburrow gnoll dungeon (near Qeynos), which is kind of sad. We checked out the top of the main cavern and snapped some photos and also found the side entrance where the old mine carts went in on rails. We didn't have flashlights or ropes or anything so we couldn't get too far in, so we'll have to go back sometime with gear. Perhaps the coolest site was outside. An excavated pit on the hillside above the main mine area had filled with water from the recent rains, and the water was a strange cloudy blue, filled with mineral powder from all the ores. The mine was a "Cinnabar" mine (which is some mercury rock thing), and the place is dangerously high in Mercury still. There were grates over the mine openings, but determined people had made mincemeat of them for our convenience. I put some photos up at Yahoo .

02-20-06 - 1

02-20-06

"The Return" is a sad, beautiful movie. The cinematography is really great. The little brat gets a bit too annoying after an hour of his pouting. I'm so jealous of people who know how to do things in the wild - fish, make fires, etc.

2/19/2006

02-19-06 [poker] - 1

02-19-06 [poker]

Party's increased the stacks in many of the multi-table tourneys, but also increased the blinds. The sit-n-go 2 and 3 tables used to start with 1000 chips and 10/15 blinds. Now they start with 2000 chips and 20/40 blinds, which is actually a bigger blind in proportion to your stack. The big multis now start with 3000 chips, which is roughly equivalent to 1500 before, so that's a nice change, it means you start with 75 BB so you at least start decently deep.

Cash play certainly is much more profitable than tournament play for many reasons, and I'm still a cash newby, but tournament play really is more fun. That exhileration of getting knocked out, getting all in and risking your life, making a final table, winning a first place, there's nothing like that in the cash games. After a good day of cash play, I just feel like I put in my hours at the job, I made some good plays, some bad plays. After a tournament win, I feel like the king of the world, baby! (1st place today at a 20-person)

02-19-06 - 3

02-19-06

this shit makes me furious . Really, the veins are bulging on my forehead and I think I just crapped my pants. There are so many cocky little fuck-head kids making huge money in poker these days, I just want to beat them at their own game so bad.

02-19-06 - 2

02-19-06

New slogan for "Tiger Balm" : asian kid trying to be like a hip-hop rice boy says "It's the balm!!" (sound's like it's the bomb).

I'd like to hire a Mexican migrant worker to work my vegetable garden. It's about a 4' x 8' patch.

02-19-06 - 1

02-19-06

Frozen shrimp is a handy healthy food that can be cooked without bothering to thaw it. There are two good ways :

1. Broiled. You can broil frozen shrimp without thawing. The key is to place them on a rack over a pan or baking sheet, because as they thaw they'll give off a lot of water and you don't want them to sit in that. Simple cover in olive oil and season them as you normally would (something like herbs de provence or TJ's 21 are good), and broil until crispy.

2. Boiled. The key here is to place them in the water when the water is still cold. Heat up the water to boiling, and by the time the water gets to boiling the shrimp are thawed and then will cook just like they normally would. I like to do this and make a shrimp Ramen - just put the shrimp in the cold water, heat it up to boiling, cook them for 2 minutes & skim the foam off the top, then put the ramen in the same pot and cook the 3 minutes more for the noodles. Add napa cabbage and black mushrooms to the hot broth, as well as chili garlic sauce (Sambal Olek) or whatever you like. (a little Sherry is a good trick for giving it a false flavor of "royal broth"). (btw it's strange that so much Chinese cooking calls for Sherry, when Sherry comes from Europe; I wonder if it's just a convenient stand-in for some Chinese alcohol with a similar flavor).

2/18/2006

02-18-06 [poker] - 3

02-18-06 [poker]

I've made myself a mantra. I'm going to try saying this out loud to myself each time before I starting playing:
I'm ready to play.  I'm ready to -
	be patient
	play each hand as well as possible
	stab at pots and then fold to aggression
	concentrate on the game
	lose the minimum as well as win the maximum
	don't try to win pots I can't win
	let them bluff and win pots from me, and don't let it bother me
	not worry if I'm winning or losing, just play each hand well
	be aggressive, raise with draws in good situations
	not attack obvious strength
	play solid, vanilla, don't do funny shite
	just fold marginal hands that sort of hit the flop
	not chase, not limp junk and not call postflop
	just fold when I'm beat
	take bad beats and just shrug them off
	use reads and stats to make small adjustments to basic solid play

02-18-06 [poker] - 2

02-18-06 [poker]

Flush draws suck for chasing, but they're great for pushing. When someone plays back at my c-bet, even if I suspect he's doing it with junk I don't want to risk playing back with pure air. Hence, playing back with a flush is perfect. You raise preflop, cbet, someone raises and you suspect they might not have the good, you push. Not only is this a great time to play back, it's one of the very few times to play back or to push draws. The other thing about flush draws is it really blows when the turn blanks. If possible you want to get them to fold on the flop or get all in on the flop.

02-18-06 [poker] - 1

02-18-06 [poker]

I've been thinking about when to check-raise. It's something I do so rarely, and I kind of do it randomly, which is bad. I'd like to give myself some specific criteria for when check-raising is good.

On the flop - if I'm the preflop raiser I'll almost never check-raise. I'll usually continuation bet, and rarely check/call or check/fold. If I'm not the preflop raiser, and I hit some kind of medium hand, I'll usually just lead out or check/call or check/fold. Check-raising might be useful for trapping a continuation bet. My first thought is if I suspect he's whiffed but will still c-bet, I can check-raise to win his c-bet. The problem there is if I'm called, I'm out of position in a big pot with a mediocre hand. Check-raising with a draw sucks because if he calls and you miss, you're on the turn out of position and have to lead again or check, which shows weakness and invites an attack. A good time to check-raise is when you want to build a pot but are worried your opponent might just call. For example, you flop two pair or a set and the board has a flush and/or straight draw. If you're sure the preflop raiser will continuation bet, this is a good time to check-raise because if you just lead he can call with good odds to hit his draw and fold all other hands. This is especially good with hands like bottom two pair which are very vulnerable.

On the turn - I like check-raising a lot more here when I have a big hand. If I check/called the flop, he has the betting lead and will often bet again. If I put him on a draw, I won't check for fear he checks behind, but if I put him on like top pair and I have two pair or a set, a check-raise is perfect here. Also, if I was the preflop raiser and I bet the flop, against aggressive attacking opponents an check-raise is very powerful. You raised preflop, you bet the flop, now you check the turn, it shows weakness and many villains can't resist attacking. Of course you have to be pretty sure he's an aggressive attacking type of villain.

02-18-06 - 1

02-18-06

The funniest thing happened at the Olympics. This girl Lindsey Jacobellis was way ahead in the snowboardcross race and decided to do a trick off a jump to show off, and wiped out an cost herself the gold. Silly girl, that's a rookie mistake, anyone who's played SSX knows you should do your tricks at the *beginning* of the run so that you power up with turbos, then you just blast your turbos and glide near the end. I guess she was just going for the high score, which is greedy, if she would've won she could've upgraded her board or gotten a cool afro or something. (BTW snowboardcross is a ridiculous stupid sport).

2/17/2006

02-17-06 [poker] - 6

02-17-06 [poker]

Not all second nuts are created equal. Lets look at some cases of 2nd nuts and how often the nuts can be out there. Once you've seen your hole cards and the full board, there are 990 possible holes that one opponent can have.

You have the top house, only beaten by quads (pair on board) :
You hold [ Q Q ] and the board is [ 5 5 7 9 Q ] (rainbow). Only [ 5 5 ] beats you. That's 1 hand.

You have the top house, only beaten by quads (trips on the board) :
You hold [ A A ] and the board is [ 5 5 5 7 9 ] (rainbow). Any [ 5 x ] beats you. That's 44 hands.

You have the king high flush (3 of suit on board) :
You hold [ Ks 7s ] and the board has 3 spades, no pairs, no str8flush possible. Any [ As xs ] beats you. That's 7 hands.

You have the king high flush (4 of suit on board) :
You hold [ Ks 7x ] and the board had 4 spades (no pairs, no str8flush possible). Any [ As x ] beats you. That's 44 hands.

You have the 2nd nut straight (3-straight on board) :
You hold [ 7 J ] and the board is [ 8 9 T 2 A ] (rainbow). [ Q J ] beats you. That's 12 hands. (same count with a 4-straight on the board, but the number of tying hands goes way up in that case)

The one that really surprises me is the case with trips on board. It's actually pretty likely to see quads in that case, and certainly if you have something like [ T T ] and the board is [ 5 5 5 7 9 ] , that's not a very strong hand at all. Of course if you have a hand like [ A A ] on a board of [ 5 5 5 J J ] you don't even have the 2nd nuts, you have the 3rd nuts and a huge amount of hands beat you.

02-17-06 [poker] - 5

02-17-06 [poker]

It's been a slow week for me in the poker world. I busted out 1st hand in the live tournament. Online I've made a small profit, but I haven't made any of those big hands that bust someone and give you a nice win. I've found a few "whales" (big, rich donating fishes), but I haven't been able to bust any of them. Either someone else busts them or they leave the table before I get the cards to do it. I've also made a handful of evil mistakes that have cost me. It's funny, one of the things that bugs me about being break-even this week is that it hurts my win rate. I think of my win rate as like my skill rating, almost like my stats in an RPG, and by going break-even it just drags down the win rate. I need to post a good weekend and get it back up. (The dollar level you play at is your "level" and the # of hands are "experience points").

02-17-06 [poker] - 4

02-17-06 [poker]

So, Party Poker did this big software upgrade yesterday, and they seem to have really fucked it up. There are a number of really major bugs, including crash bugs, and one bug where it can delete all the images in your "C:\Program Files" !! Fortunately I haven't been hit by these, but it motivated me to do a full backup of my hard drive. I wonder what's gonna happen with the problems. The users have no real legal recourse, but people are having problems where it crashes during a hand and they lose a bunch of money. It's a bit worse than your typical MMO updates. Usually Party will compensate people a bit. The image deleting thing is pretty horrific.

02-17-06 [poker] - 3

02-17-06 [poker]

Obviously poker is made of a lot of different skills and decisions, but I like to categorize your moves into two broad categories that make up the flow of play : 1) Making the most possible when the cards let you. This involves value betting, it also involves having an aggressive image so you get paid off, it involves bluffing when you can steal the pot, etc.. and 2) Losing the least when the cards go against you. This involves not bluffing when they won't fold, not value betting when only better hands can call, being patient and not playing junk preflop, folding when you're beat, etc.

02-17-06 [poker] - 2

02-17-06 [poker]

Against semi-good players just pushing your good hands works great. They think "if he had a monster, why would he just push?" and they call with a mediocre hand trying to catch you pushing a bluff or a draw. I know this well because I do it all the time myself, and I'm usually wrong.

02-17-06 [poker] - 1

02-17-06 [poker]

Pretty interesting hand at Dustin's live tournament last night. It's the very first hand of the game, I just got there. We all have 150 in chips and the blinds are 1/2.

I limp in UTG+1 with 4d 6d . Several other limps and the blinds check. Dustin is just to my left and limped behind me.

Pot $12 , 6 players
Flop is 4h Td Jd

Bingo! I have a pair + flush draw, a monster hand, I want to get chips in the middle.

SB checks
BB (Jessica) bets $6
I raise to $24
Dustin calls (!!)
all fold

That's really funny, Dustin just cold-called a reraise. That means he has something really big. He limped in and now cold-called a reraise. He must have a flush draw, str8draw, maybe str8+flush draw, or possibly a set, something like 44. TT and JJ are pretty unlikely because he would've raised them preflop, but it's possible he got tricky with TT. Also TJ is possible but less likely.

Pot $66, 2 players
Turn is 3d
(Bingo???)
I bet $40
Dustin calls

Again very suspicious, Dustin just calls a big bet on the turn. If he just pushed all in here I'd definitely call, because it looks like he's protecting against the diamonds and maybe has the 44 or TJ.

Pot $146
River is 9h
Hero ????

I have only about $84 left, but I can't put him on anything but a higher flush! What do I do ? If I check, he'll check behind with any worse hand, so I'll miss some profit there. If he has the flush, he'll surely bet, probably all in. If I bet, he'll call with a wide range. He might be able to fold if he doesn't have the flush, but he might call with a set. I really really felt like he had the flush at the time, but I just couldn't imagine check-folding a flush, so I went ahead and pushed it and he called with the Ad5d.

It's easy to say "flush over flush is just bad luck", but I really put him on the flush on the river, and I think I could've check-folded. Of course if I check-folded and he showed me 44 I'd feel like a real moron. Anyhoo, I'm okay with the mistake of running a flush into a higher flush. At least I didn't show up with top pair or some nonsense.

You can read Dustin's version of some of the hands at The Chapel Perilous .

2/16/2006

02-16-06 [poker] - 5

02-16-06 [poker]

I had to remind myself of this today - against guys who are super loose preflop but play well postflop, you have to punish them with raises preflop.

02-16-06 [poker] - 4

02-16-06 [poker]

Well, I think the new Party Poker GUI is a complete POS. It's an improvement like the Windows XP GUI - slower, bulkier, balloony, with lots of unnecessary effects that detract from functionality.

In positive news, I very quickly got GoldBullion working on the new GUI. They broke a few things - I can no longer read hand histories while the game is running. They used to log each line of the hand history as it happened, now they way for the hand to be over and then log the whole thing. Also, their buttons are all fucked up which is currently breaking the autoplay, hopefully they'll fix the buttons in general. Also, the GoldBullion data mining DOES still work. So, despite Party's efforts to break data mining, I still can. Nya Nya.

02-16-06 [poker] - 3

02-16-06 [poker]

How to : get ride of sidebet button in Party Poker.

02-16-06 [poker] - 2

02-16-06 [poker]

Stuff Party Poker needs to fix :


1) fix the rebuy exploits (being able to rebuy when all-in)

2) fix the way the buttons/blinds move when people get knocked out (the BB should move, not the button)

3) fix the way the blinds are when heads up at cash tables (eg. SB should be the button)

4) fix alert sounds - you don't get any action warning sounds at inactive tables
(chip sounds and stuff should remain only at the active window)

5) add a time bank for the cash games, give me 40 seconds a day or something like that

6) fix the window activation & click bugs where sometimes you have to click many times on a button to get 
	the action to go through

7) get rid of "hide me from search" ; I want to find my buddy listers!!

8) give me an option to auto-rebuy at cash tables when I bust out so I don't
miss hands, better yet give me an option to stay at a full buy in any time I
go under full.

9) let me move seats at a table without sitting out and rejoining the table

10) let me get up from my seat without leaving the table, eg. leave the table window open just with not me at it
	(this would make #9 unnecessary)

11) don't auto-minimize the lobby when I open a table; also don't auto-open the lobby when I close a table
	(let me control the lobby myself)

12) better buddy list, like "select all for search"

13) heads up sit-n-gos

02-16-06 [poker] - 1

02-16-06 [poker]

Well, Party Poker has dropped the bomb today. They've upgraded the software and broken PartyMine. The upgrade's not done yet, so I haven't seen the new software, so I don't know how badly it will break GoldBullion or how hard it will be to fix it. The new PP is resizable, so all their graphics must be scalable. Hopefully that doesn't mean they've stopped using standard Windows widgets, which I rely on to scrape the windows. The standard edit boxes and list boxes and such are great for scraping in Windows because you can just send them a message and ask for the contents. If they're now using a raster drawing library to do their scalable graphics, that would suck balls.

02-16-06 - 1

02-16-06

Since I can't play, I'll write up an idea Dan & I had a while ago. It's a TV commercial, it goes like this :

Hot chick stumbles out of a club in an alley in some big grimy city. It's late at night, she yells goodbyes to her friends in club and stumbles down the alley obviously drunk. Lots of scary dark shadows, water dripping in puddles, a rusty gate swings in the wind, it should sort of feel like a horror movie where she's obviously unaware but the audience is thinking "oh no, don't walk down the alley!". Suddenly someone jumps out of the shadows, we can't really see them they're still in the dark. They press her up against the wall and say "take off your pants!", camera pans down to her pants, tight jeans that make her ass look great. She protests and the assailant puts a knife to her "take off your pants!" , she, sobbing, complies. The assailant pulls the knife away and grabs the pants, the girl collapses to the ground sobbing and covering her face, her legs akimbo hiding her panties. The assailant throws the jeans over her shoulder and we see the camera pans out as she walks away and we see assailant is another club-girl. Zoom in on the jeans pockets slung over her shoulder so we can see the logo.

2/15/2006

02-15-06 [poker] - 3

02-15-06 [poker]

I wonder if my poker app is sellable. It basically does everything that Poker Tracker + Poker Ace HUD + PartyMine does, and those cost about $70 total. On the other hand, those all have much nicer GUIs (I write very functional GUIs and command line stuff, which is how I prefer it), plus I think they've pretty well saturated the market already. Even if I sold for $20, they've got the mind-share, market-share, advertising, etc. etc. and much of the customer base already owns those things. My app has a lot of nice features that those don't have, but that's not enough to get someone who already bought those to get mine too, My app also takes about a minute to load up with a million hand database, while Poker Tracker takes 30 minutes or so (!!). But, PT has a full database so you can do custom queries and I don't (though I guess I easily could with some freeware DB code). I think the potential market is pretty big. I'd guess that Poker Tracker has sold around 10,000 units, maybe more, at around $50, which = $500k in sales, which is almost all profit.

If I were to commercialize there would be lots of annoying code things I'd have to do. Aside from the GUI issue, I'd have to make it work with all the other poker sites. Right now it only works with Party since that's all I play. I've never sold my own shareware apps so I'd have to learn how to do installers and lock apps with registration codes, etc. etc.

02-15-06 [poker] - 2

02-15-06 [poker]

Some strategy thoughts :

Flush draws really suck as calling drawing hands because they're so obvious and people are very scared of the flush card. Because of this I prefer to use flush draws as semi-bluffing hands. If possible I'd like to get all-in on the flop with a flush draw when it's reasonable. Of course this is moronic at low stakes unless you have a combo draw or have some good fold equity.

Straight draws are great draws to take off a card and try to spike it because the implied odds are good. Also, in Limit hold'em it's bad to draw to straights when the board has flush draws, but in No Limit the opposite is true. You can represent the flush draw, so you actually have outs like a str8 flush, since the flush outs are bluff outs for you.

With big combo draws on the flop I really want to get all in. Out of position I usually lead and hope they raise, rarely I'll check-raise. In position I'm thinking that I might start min-raising more. If they reraise, I can push. If they just call they probably check the turn to me and I bought myself a very cheap free card, I can check behind. I'll also min-raise with monsters and with air when I put them on a whiffed c-bet.

02-15-06 [poker] - 1

02-15-06 [poker]

If someone folds to cbets about 45% of the time, it's still +EV to cbet them. You're not getting enough value just from their folds, but even when they don't fold you have some equity in the hand. eg. if you're on whiffed overs, you may still have 6 outs, so 45% of the time you pick up the pot, and the rest of the time you still have some pot equity.

Now if their fold to cbet is down around 25% or less, the equation changes. You're getting much less value now from taking the pot, and need to have a hand that can win. You have to just bet for value, bet when you think you have the best hand (even if it's AK no pair).

2/14/2006

02-14-06 [poker] - 4

02-14-06 [poker]

Funny coincidence, with all my recent thinking about Prahlad, the WSOP Circuit event he was in just reran on TV. He gets heads up with Chris Ferguson at the end and their stacks are nearly identical. Unfortunately Jesus makes the nuts and Prahlad makes the second nuts, so we don't really get to see what a match between them would be like. I actually think Prahlad has a big edge in that matchup. Chris is awesome in tournament play against a field (and under-rated, IMHO, I would take him in my Fantasy Poker league), but heads up against a wild aggressive player like Prahlad, I dunno. Would've been great to see without that horrible hand. Anyhoo, it struck me that Prahlad really really reminds me of Justin Hall who is a wacky goofy white guy freestyle rapper Scandinavian-cum-bay-area vegetarian hippy just like Prahlad is.

02-14-06 [poker] - 3

02-14-06 [poker]

It's very good to have an image of pushing hard at pots. You want everyone to think that if they play a pot with you, you might bomb it all the way. This makes them scared to play marginal hands. If you can get them to fold all their weak top pair hands (and worse), you've won the battle. You don't want them thinking that you'll play timidly and they can get to a cheap showdown and win with their mediocre hand.

02-14-06 [poker] - 2

02-14-06 [poker]

I'd like to get to the $1000 NL game this year, by steadily moving up in levels and demonstrating a good win rate at each level. So far I'm kicking ass at the $100 NL level, but I keep making really amateurish errors. My win rate rocks despite that, but I'm not letting myself move up to $200 until I can get a hold on these errors. I wonder if any of you have good training ideas? Whenever I make a mistake I review it and try to analyze what I did wrong, and why, but I'm not sure that helps very much. I wonder if there's some dumb psychiatrist trick that could be useful, like teaching myself that ringing a bell is really horrible, and then ring a bell whenever I make an error. I could condition myself by hiring a dominatrix to ring a bell and stomp on my nut sack ...

On a related note, the one really bad thing about the poker lifestyle is that when I have a losing session or make a big mistake, it makes me really depressed for a short while, and I usually have to drink booze and eat chocolate, and it makes me pretty worthless for doing anything at that point. When you win your brain injects endorphins and you get this short term high, but when you lose or fuck up your brain punishes itself and injects some sort of massive downer chemicals and you can't get off the couch.

02-14-06 [poker] - 1

02-14-06 [poker]

I have this phobia of playing mediocre hands out of steal spots. I have no problem playing pure junk, or of course monsters, but I find mediocre hands extra hard to play. For example, something like AT on the button. It folds to you and you make a normal raise. You flop an ace, and then the big blind check-raises you !? He probably just puts you on a steal, so you have a much better hand than he thinks you do, but it's still a crap hand. In order to play correctly you have to read how he's adjusting to the fact that you're on a button raise, which is so hard to do.

02-14-06 - 3

02-14-06

Patri's crazy old story about losing a fortune at the WSOP in a side game. Also have fun browsing his site for all sorts of wackiness. He's a true bay area weirdo, much like myself.

02-14-06 - 2

02-14-06

I hate Valentine's day. Fucking Hallmark.

02-14-06 - 1

02-14-06

Luge requires all the skills of being able to lie on your back and let gravity pull you down hill. Those guys are amazing. How do they go so fast? Oh yeah, they just lie there and slide down the ice.

2/13/2006

02-13-06 [poker] - 2

02-13-06 [poker]

This is related to the previous thoughts on Prahlad - what if you only push sets and flush draws on the flop? In that case, it's an easy easy fold for top pair, since top pair is way behind the sets and only 65% to win against the flush draws. What that means, and what Prahlad does, is that he can actually push with a lot more hands, such as straight draws, gutshots with overs, etc. Even with those extra hands, his range is still favored over top pair. By adding more hands he increases "Shania" - he gets more value for his gutshots because he gets a lot of folds with them, and he also gets more value for his sets, because people know he pushes crazy draws so they're more likely to call his sets.

Anyhoo, in Low Limit this is all way too fancy. A practical thing that I've noticed for myself, is that when you play against these "solid" low limit TAG players, they will only raise big with their sets and very good draws. In that case, you can just fold top pair to them. Yes, you fold to the flush draw some times, but over all it's a good fold and you don't need to sweat it. They feel clever by sometimes semibluffing, but really they're not bluffing very much and you can correctly fold there without giving up value. Also, since they love to semibluff flushes but almost never semibluff gutshots, it's pretty easy to tell when they might be semibluffing.

02-13-06 [poker] - 1

02-13-06 [poker]

I was telling Dustin about this the other day - it seems like all the online poker sites screw up the blinds when cash tables get down to 2 players. I know Party does it, and I hear Stars does it too. Heads up (2 players) the button should be the SB preflop, but they make the button the BB, and the out of position guy the SB. This is very weird and should change your preflop strategy a LOT. Heads up, position is even more important than it is in full games, so being on the button is a big advantage. That means when you're the SB out of position preflop, you want to fold almost any two cards!! Obviously play super-premium hands, but even something like ATo you could fold in a deep game. Conversely, on the button/BB you have the advantage of position, plus you are defending your blind. You want to call most raises preflop, and raise any time he limps. You want to play almost any two cards, down to things like Axs and 64s in a deep game. (those are my wild guesses in the dark about preflop strategy, anyway).

2/12/2006

02-12-06 [poker] - 3

02-12-06 [poker]

Let's say your opponent folds their BB 70% of the time, and folds to continuation 50% of the time. (these are pretty standard stats). So if you raise to $4 on the button, 70% of the time you get +1.5 in the hand. The other 30% they call and the pot is now 9.5 BB and you're -4 into it. On the flop they check and you c-bet 7 BB's, and they fold 50% of the time. In that case you're +5.5 BB. The other 50% of the time they raise or call your bet and you have to play from there. How often do you have to win in this final case (they didn't fold to your cbet) ?
EV = 0 = 0.7 * 1.5 + 0.3 * ( 0.5 * 5.5 + 0.5 * ( -11 + P * 22.5 ) )
=
0.225000 + 0.3 * 0.5 * P * 22.5
=
0.225000 + 3.375 * P
You don't have to win at all !? If you just check/fold every time they play back at you, you're +EV !!? This is because they're folding their BB too often, but more importantly they're folding the flop too often when you c-bet. That actually really surprises me, I thought P would have to be at least 25% or so to break even. If P is 25%, this is right around EV = +1 BB. Note that P will of course be < 50% here since they have chosen to only play their better hands, while you're on a big range, but I suspect it's probably actually > 25%.

One thing I find is that even when I feel like someone is running over me with c-bets, and I start playing back at them, I still fold about 50% of flops. Playing back more than that just feels like going maniac, which of course is what you have to do!! Similarly, when you're running over someone c-betting and they start attacking back, notice that it's actually still pretty rare, and you can just fold to their raises. They just won't have the balls to do it very often because they'll be afraid that you pick up on it and come back over the top. Also, you're getting them to commit a lot of chips with marginal hands, so when you do show up with the hand it's a big bonus.

Clearly giving up the pot to the preflop raiser 50% of the time is too much. Roughly 2/3 of the time you should not give up the pot so easily. Either lead into the preflop raiser, or check-call or check-raise. With big hands, mostly lead, and occasionally check-raise. With junk (whiffs), occasionally lead and occasionally check-raise. With mediocre hands, mostly lead and sometimes check-call (lead turn). With big draws mix all three choices.

Of course one way to deal with blind steals is not to try to defend too much, but simply to make sure you steal enough when it's your turn. There's an inherent disadvantage to being in the BB. There's an advantage to being on the Button. The Button is worth money, and you can't do anything about that. If you're in the BB, the Button is going to make money from you on average. That's fine, you just have to make the same amount when the situation is reversed.

02-12-06 [poker] - 2

02-12-06 [poker]

Over the weekend I've been thinking about the importance of the blinds in 6-max hold'em. You pay 1.5 BB's per 6, which is 25 BB's per 100. That's right around the win rate of a good player. So, at an aggressive table, if you didn't defend your blinds, you'd be giving up a fortune. On the other hand, against passive/tight players, stealing their blinds can win you a fortune. Unfortunately, defending you blind means playing mediocre hands out of position against the raiser, which is nasty.

02-12-06 [poker] - 1

02-12-06 [poker]

In defense of the minraise. On later streets, when it's unlikely he's drawing, the min raise can be very useful. It's a good way to extract value when you have a monster, because it's such a small raise they'll call with a lot of hands they might dumb for a real raise. Also when you have a mediocre hand that you want to show down, it's a cheap way to slow them down. Say they bet $16 on the turn. If you just call, they'll bet $30 on the river. However, if you min-raise, they'll check the river to you and you can just check behind. You make the same minraise with a monster, but you then bet the river when they check. This is a powerful move in position. Out of position, it's a horrible move. You always have to play much more straightforward ("solid") out of position.

02-12-06 [life] - 1

02-12-06 [life]

We went down to Santa Barbara for the weekend. On the way back we went to the Red Rock swimming hole off the 154. It was a nice warm day out, but the water is still icy cold, swimming in it for a few minutes gave me a head ache.

Didn't play any poker this weekend, and I feel like I missed out. The weekend games are so juicy. The whole setup with Danielle working a regular job really sucks. I can't play poker at night or on the weekend, which are the best times to play. I play during the day which is much, much tougher. We can't just take off in the middle of the week and go skiing or camping. I'm spoiled and I really don't like doing those things on the weekend when there are bloody human beings around. Camping/hiking/skiing/the beach are all so nice during the week when it's quiet and deserted.

We ate at "Roy" in Santa Barbara which was disappointing; I've always thought it was a cool place, but the food is very ordinary. Came home today and I made much better food (poached salmon with dill and capers). A funny thing happened at Roy. I was sitting at the bar with Danielle while we waited for our table. Dan got up to go pee, and this Urban Cougar sat down next to me. We were chatting a bit and she was practically purring (I think she tried to rub her cougar musk glands on me). Anyway, Dan came back and I got to do the "oh, this is my hot young girlfriend. Get lost, cougar".

2/11/2006

02-11-06 [poker] - 2

02-11-06 [poker]

An old 2+2 thread I started : When folding a set might be right . Here's the conclusion :

To wrap up again, I think there are a whole bunch of factors that come together to make this hand very special :

1. Villain has obvious "set miner" stats
2. The board is very separated, two pairs are unlikely
3. There's no raise preflop, so high pairs are unlikely
4. Villain is passive, so pushing a draw is unlikely
5. I have a respectable TAG image so he doesn't put me on junk
6. No one's on tilt, we have no history of playing back at each other
7. I have the bottom set
8. There's no high card on the flop so TPTK is impossible for his flop action (and he wouldn't push that on the turn anyway)
9. The Q on the turn also makes hands like JJ very unlikely to push there
10. I haven't made any big laydowns recently so I'm not perceived as weak

So, if I'm ever in a hand again where all those factors come together, I think I just might fold the bottom set and be confident about it.

I thought about this because of the Prahlad post that Dustin mentioned. The other case where you can fold a set is when your opponent is very smart and can read you well and knows from your action that you must have a set, and yet is raising in a situation where he's unlikely to be bluffing or semi-bluffing.

02-11-06 [poker] - 1

02-11-06 [poker]

I played a hand recently which is really bugging me. It seems like a totally trivial hand, but something just feels wrong.

Party Poker $100 No-Limit Hold'em, $1 BB (6 max, 6 handed)

UTG (Maniac) ($89.46)
Hero ($98.50)

Preflop: Hero is MP with As, Js. SB posts a blind of $0.50.
UTG (Maniac) raises to $5, Hero calls $5, 4 folds.

Flop: ($11.50) Jc, 7h, Th (2 players)
Maniac bets $11, Hero calls $11.

Turn: ($33.50) 8s (2 players)
Maniac bets $31, Hero calls $31.

River: ($95.50) Kd (2 players)
Maniac pushes all-in ($42.46)
Hero calls $42.46.

Final Pot: $137.96

Maniac was super-aggressive and bluffed a lot. I make top pair top kicker (TPTK) and call him down to trap him if he's bluffing. Sometimes he'll have a better hand, sometimes he'll have a worse hand. (actually this board is really bad, since KJ got there on the river, AQ got there, any 9 makes a str8, etc.). So, what's wrong with this hand? The thing I don't like is I'm allowing him to play perfectly. I'm not sure if he was just a crazy maniac or if he was a clever thinking maniac. I've made a mediocre hand and let him double up with his good hand. I didn't get any information to find out if he had a good hand or a bluff, which means he's getting to decide when the stack goes in. If he's clever he's only putting the stacks in when he's got it.

I don't like the idea of letting him decide when to play a big pot. He could stab early and then back off without a big hand. Certainly I know he's open-raising a lot of junk and often c-betting with it. He will usually fold to a raise on the flop if he was just c-betting with junk.

Let's look at some extreme cases. What if he open-raised a big range, and then would push the whole way with anything? In that case I should call behind preflop with any good cards, and call down with any pair. It would also be +EV to just call behind with pockets, and only call him down if I flop something super-good. That would reduce my variance but also reduce EV. What if he open-raised a big range, but then would only bet the whole way with a decent one pair or better? In that case I should call behind with lots of hands, and only call down if I flop two pair or better. If I don't flop two pair or better, I should raise some of those hands as a bluff, because he'll fold with no pair.

2/10/2006

02-10-06 [poker] - 7

02-10-06 [poker]

How do you keep your big pairs from being cracked by sets? Reraise them *big* preflop, big enough to make setting wrong. Now, good players will just fold if you do this, but bad players will look you up or reraise. You'll often get action from AK. Your AA can get action from KK and QQ. If you were in a tougher game where people were watching you carefully, expand your reraise range to some more hands. In a live game you can just pull this big reraise once with pure junk and show it, then the rest of the time you do it, have the goods (QQ+).

02-10-06 [poker] - 6

02-10-06 [poker]

I found a very cool old thread at 2+2 about a hand with Mahatma/SpiritRock (aka Prahlad Friedman). There's a bunch of junk posts but also some very good thoughts on his style in general. You also see just how much fear and confusion it creates. The big problem is the idea that Prahlad is good at reading hands, so when he pushes and you've defined your hand as good, he can't be bluffing right? Well, then of course he'll bluff in exactly that spot since he knows you know that. More generally, Prahlad often severely overbets. This might seem like a donk play, but what it does is greatly increase his value. The reason is that he pushes with a range which is +EV vs. any reasonable range you can have for calling. He pushes with some bluffs, and that gets him action, but he mainly pushes with premium draws and very good hands. He also can read you well, so he doesn't push when you're likely to have something very good, he pushes when you're in a tough spot. If his equity is 55%, by overbetting he wins 55% of the huge overbet instead of 55% of the smaller pot size bet. Prahlad also counts on you calling with hands like top two when he pushes, because he's a big bluffer, you have to call with top two right? He makes his money because you fold hands like top pair weak-kicker. The correct way to decide on calling him is how your actual hand strength compares to the range that he puts you on. I'm still not sure what the right answer is to that hand in the thread; it's pretty easy to talk yourself into either decision.

02-10-06 [poker] - 5

02-10-06 [poker]

The biggest leaks of very good players are 1) not reraising enough preflop, and 2) picking play frequency based on hands, not based on situations. Lolo maximizes against #1. SpiritRock maximizes against #2. I'm not talking about individual leaks that people might have, such as being tilt-prone or whatever, I'm talking about game-theoretical leaks in their entire style of play. If you look at the way most people play "solid" NL hold'em, these are very common errors.

Lolo's strategy is to always open raise in late position, and always continuation bet. Then take it from there. In early position or behind a raise, lolo is actually very tight, but opening the pot late lolo can have literally any two cards. It would seem like this is easy to counter. Just call with decent hands, try to hit something. Even if you don't hit, lolo will usually have junk, so you can just check-raise and take his c-bet. That works great against bad LAG's, but not against lolo. Lolo will sometimes have a hand and reraise you all in. Lolo will sometimes reraise you all in on a bluff. Lolo will sometimes just call with a mediocre hand or draw. Lolo will sometimes just call with junk to see if you keep it up on the turn, then bomb the turn if you check.

Lolo's strategy is easy to counter, but no one wants to do it. The counter is to reraise big preflop with anything decent. It's not good enough to just reraise big with super-premium hands, you need to reraise with a lot of hands, down to things like K9 that most people just can't stand to reraise and possibly play for their whole stack. This is a lot like David Sklansky's "All in Hold'em" Casino game. The optimal way to play is to push huge bets with lots of hands, but no one is willing to risk that much variance in order to get an EV bonus. Against lolo, in a $1000 buy-in game you should be reraising preflop to $150 with a very wide range, which effectively puts you both on short stacks for post-flop play.

Prahlad's strategy is to bomb the pot when he has monsters and also when he has junk. The correct way to counter this is to call when Prahlad thinks you have junk. Let's say he can read you well and puts you on a correct hand range. If your hand range is all so good that he can expect you to call with anything in that range, then you should fold!!! Say for example he puts you on a range like { two pair, a set } and he then bombs it. You should fold!!! On the other hand, if he puts on a range that is mostly so weak that he expects a fold, then you should call with the upper end of that range. Say your range is { ace high, one pair, two pair }, and the board is scary with possible straights and flushes, then you should call with the two pair, because he expects you to fold most of your range but he's pushing anyway.

02-10-06 [poker] - 4

02-10-06 [poker]

You want to play big pots when you're in the upper end of the range your opponent can put you on. Let me elaborate. Your opponent can figure out from your action that you're on a certain range of hands. He knows that range, so if he is still calling or raising against you, he must think he can do well against that range. That means you can put him on a hand which beats the average of your range. So, if your actual hand is in the lower end of your range, you must fold. If your actual hand is in the upper end of your range, you can play. Note that this may mean you are folding some very very good hands, but playing some very weak ones. Also note that the bigger your range is, the more hands you can play!! For example, if you completely define your hand to a small range, and your opponent shows strength, you cannot play. Say for example that you only ever play AA. Now if your opponent shows strength against you, he knows you have AA, so he must beat that, so you have to fold. On the other hand, what if you play AA and also 23o. Now if he shows strength against you it just means he can beat that range, which means your AA is way ahead of him. In simpler terms, when you define your hand and your opponent then shows strength, he's saying he can beat your hand, and you must fold. The ranges are just a better way of saying this.

02-10-06 [poker] - 3

02-10-06 [poker]

I started writing this post about how I hate "one pair". I lose a ton of money with one pair, even good one pairs like TPTK and overpairs. I was thinking I might just fold any one pair hand on the flop. But, of course I'd still continuation bet with one pair. And I might try to catch other people continuation betting. If my c-bet is called, I'll probably check the turn, but if they bet on the river, I'll call a small bet to catch a bluff. Okay, so I won't really bet hard with one pair. Hmm.. that's exactly how you're supposed to play one pair.

02-10-06 [poker] - 2

02-10-06 [poker]

The human perception of frequency is very fucked up. It feels like any time you fold some junk, it seems to always hit the board. Everyone you meet named "Barry" seems to act just like your friend Barry you had when you were a kid. It seems like astrologers have some knack of saying things that perfectly fit your life. This is a well known fallacy - of course these things really don't happen very often, but you don't notice it when they don't happen, you only remember the positive events and thus incorrectly think they happen far more often than they do. This is the human flaw which makes us think "coincidences" are some magic thing that can't just be random events, because we latch onto the coincidences, and ignore the millions and millions of events which fail to be coincidences.

This comes up all the time in bad player's poker play. It seems like some donk is always hitting his draws on you. No, of course he's not, he's just folding when he misses so you only see it and remember it when he does spike it. It seems like this aggressive guy keeps taking pots from you. Maybe so, but not as often as you think, because you're not remembering all the hands he just checked down. If you're the fishy donk at the table, it feels like the pros are always picking on you. No, not really, you just didn't notice the vast majority of hands where they just folded preflop or folded on the flop, you're only remembering the few where they raised you big. Because of that you think they're raising all the time with junk, which of course they're not.

02-10-06 [poker] - 1

02-10-06 [poker]

I'm beginning to realize that I could realistically make about $1k / day playing poker, even at low limits ($100 NL). That's actually more than I made at Oddworld, though not as much as I could theoretically make in software as a tech director or some sort of management position. I make about $30 per table hour at poker now, and if I could 4-table that (not unrealistic at all, some people 10-table !!), that would be $120 per real hour, 8 hours a day = about $1k average. If I could get my table count up I could earn the same with fewer hours. And of course moving up limits I could play fewer tables and fewer hours. Currently I'm mostly just 2-tabling at 6-max because I want to have the full focus and make sure I'm not making mistakes. Of course with variance there would be days of $0 and days of $2k, I'm just talking about long term averages here.

2/08/2006

02-08-06 [poker] - 3

02-08-06 [poker]

Holy crap, I'm now folding AQ in the SB to a raise from good tight players. What's wrong with me?

On a related note, I hate it when people use "good" as a synonym for "weak/tight". As in "when the board pairs, you can put in a big bluff and a good player might even fold a flush!". That's not good, that's easily bluffed and too tight.

02-08-06 [poker] - 2

02-08-06 [poker]

A guy on 2+2 said something the other day which has clicked a light on in my head. Part of the reason why being out of position (OOP) is less profitable is because you sometimes fold the best hand.

Consider the case where you have a good but not great hand against a very aggressive, decent opponent. For example, you might have something like top pair top kicker (TPTK). Normally you want to play a small pot with this hand, but he won't let you. You have to decide whether to just fold, or go ahead and play a big pot with it. The reason is he will attack weakness. So you can either A) bet hard the whole way, leading at him and playing a big pot, or B) bet weak (or check), he'll probably attack, and you call trying to pick off his attack with a weaker hand, or C) try to make some probe bets and if he attacks big, you fold. Against this opponent option (A) is probably the worst. If you play a big pot, you'll win some and lose some, but the problem there is you're letting him decide whether to play a big pot with his hand. If you just take option (C) and fold to any big aggression from him, you will often be folding the best hand.

Now, I used to think that folding the best hand in this spot is an EV disaster. That's not true!!! Folding the best hand OOP is part of a normal winning game. Of course you also want to balance that by making other people fold the best hand when you have position.

02-08-06 [poker] - 1

02-08-06 [poker]

Poker players assume that unknown players play the same as they do. That is, a tight/weak guy will assume that an unknown is also tight/weak. So, if you're unknown to him and you put in a big check-raise, he's going to fold almost any hand. Similarly, a loose/aggressive guy will assume that you're often bluffing or calling with marginal hands, so you should bet,bet,bet hard for value against him, and not bluff. Once they get to know you they will adjust to your actual style, but this is a sure tendency to be aware of.

I made a big score today, more than making up for my disaster yesterday. I was reminded again of how stupid I get when I'm losing. Big wins come from big hands (or other forms of good luck - such as the right scare card hitting for you to bluff, him missing his draw, etc.). When you're winning, poker is easy, and you realize that all those little hands like QJ with top pair J's, you don't have to fight over so hard, you can let those go and still take his stack when you have a big hand. When you're card dead and not making any big hands, you start thinking you have to fight over all those little pots, and you wind up pushing too much to win pots that you can't win and lose a lot more. I've written almost this exact same thing here before and it still doesn't sink in when I'm losing. When the cards let you win - win the maximum. When the cards make you lose - try to lose the minimum. Don't try to win when the cards won't let you, it just makes things worse.

2/07/2006

02-07-06 [poker] - 3

02-07-06 [poker]

Live at the Bike is a stream of a NL cash game at the bicycle casino. I've just watched a tiny bit but it looks pretty cool.

The flop is a great place to push your good hands, because nut jobs will call with all kinds of draws and ideas of things they could do on future streets. The river is a great place to bluff, since no one will call with a draw, but of course that doesn't work unless your line the whole way is believable.

The turn is sort of a middle ground. People are much less likely to chase on the turn so it's a better spot to bluff. You also have your hand much more defined with only one card to come.

02-07-06 [poker] - 2

02-07-06 [poker]

You should be c-betting about 70% of the time you open raise. Part of the time you open raised you had a pair already. The rest of the time you have unpaired cards. 33% of the time you hit a pair, and most of that you go ahead and c-bet. The 66% of the time you had a pair and miss, you should bet about half those. 5% pairs + 33% - a few + 33% = about 70%. This means that of the time you c-bet, you have a hand slightly more than 50% of the time. That's enough to make it wrong for people to attack your c-bets.

02-07-06 [poker] - 1

02-07-06 [poker]

Well, I just had my first day of horrible donkey mistakes that I've had in quite a while. It was embarassing, sickening. I blew 4 days worth of work. Poker is uniquely frustrating in that your accomplishments are your ammunition, and it only takes a few minutes of screwups to wipe away days and days worth of toil. Worse than losing the cash is that I've allowed that horrible donkey back into my game, which I thought I was making progress on. Oh well, back to the toil. Playing $100 NL, I make about $33/hour which is a pretty solid win rate for that limit. You blow off one stack, that's 3 hours of work.

I like to try to figure out what led me to make a mistake. Just like in programming, when you code up a bug, it's not just a "mistake" that was sort of random, it happened because of certain conditions. Can we look at those conditions are perhaps change something such that we're less likely to make that mistake in the future? We're not looking at the particular bug, that in itself is sort of the random manifestation of some underlying factors which led to the bug. This is like the mushroom body underground, when the toadstools pop out they are not the real beast we want to address, they're randomly there because of the body underground.

So, what led to my poker mistakes today. I think the biggest factor was forcing things trying to stack terrible players. I saw them make just attrocious plays, and it threw off my game. I got into that "frustrated weak/tight player" mode, where you cling to hands that are good starting hands too much. This happens when the loose/bad player keeps sucking out on you, and you just get sick of folding postflop and starting calling or pushing with bad hands. This is a very very bad place to be. I need to look at the donks not as stacks that are going to come to me. I need to just play a good game. The presense of donks just makes the good game more profitable. I do not "go after them" or "take their stacks". I just play my good game and their stack flows to me. I don't have to bust every player. One flaw is when I sit down and identify a major donk, I want to bust them. I might bust them, but I might not. I shouldn't actively try to bust them, because it makes me play too "fast". I should just play, and whether they go bust or not is of no major concern to me. Just play good poker.

The other bad thing I was doing was playing from the stats too much. I have all these great stats on people and whenever I get to a decision that's not obvious, I look at the stats. Is this guy loose? a bluffer? I let that make my decision about whether to call or not. That's wrong. I need to just play the good game, and only when it's sort of a marginal tough decision that could go multiple ways, then I look at the stats, and they can nudge the decision one way or the other.

In terms of micro actions, almost all of my big mistakes are not folding enough. In particular, not folding *early*. I raise preflop, I c-bet. He raises me. Okay, now just fold. I need to fold on the flop when I'm probably not good. Even if I probably am ahead but am out of position and don't want to play a big pot, that's okay, just fold while the pot is still small.

2/06/2006

02-06-06 [poker] - 3

02-06-06 [poker]

In deep stack NL, play on the turn and river is crucial. If the bets are roughly pot size on each street, then bets on the river are typically around 20-40 BB. That means one river mistake is equal to 20-40 bad calls preflop !!! This also means that you can take the worst of it preflop for 1 BB if your opponents will make mistakes on the river. A lot of people take this fact the wrong way. They think it just means you should chase draws. That's not really true, as I tried to show in previous post about implied odds. Rather it's more crucial with hands like AJ where you hit your ace. If you're going to pay off a big bet with that hand on the river, that could be a huge mistake, even though playing it preflop was only very slightly -EV or even slightly +EV. Another example is when people will attack hard or call big on the river if you show weakness on earlier streets. In this case you can give them a free card on the flop or turn (which is a small -EV play), which will cause them to make a large error on the river when the pot is bigger. Note of course that all this is only true if you're building the pot up to the river.

I'm playing around 500 hands a day right now. I need to get that up to 1000 at least, and then go a lot higher. A lot of online pros play 5000 a day regularly.

02-06-06 [poker] - 2

02-06-06 [poker]

I think I've grasped one of my big leaks in the heads up game. When we were deep stacked I think I was playing at least as well as Dustin, but when we were short stacked I often blew off my stack, particularly when I was behind. One of my leaks was this situation :

I have some sort of decent hand on the button (like QJ or KT or A8). I open raise. Dustin calls in the BB with any decent hand. Now, the pot is very big compared to my short stack. I don't really have any moves. If I make any bet I'm pot stuck. I whiff the flop. Dustin checks to me. I bet whether I hit something or not. Most of the times I'll have missed. Dustin calls with any pair. This is great for him, because he's just risking that preflop call, and if he hits a pair, he gets my stack. It's terrible for me because I'm basically pushing my whole stack all the time, and if he doesn't hit a pair he can just fold the flop and I only win a small pot. Obviously on a short stack I'm way too predictable about pushing hard from the button no matter what comes on the flop. I need to check behind on the flop more often, with whiffs and also with good hands. I might also try some limping preflop when I'm on a medium stack, and perhaps also some more folding preflop.

This happens to me in that horrible region of M between 5 and 10. M below 5, you just push, okay. M above 10, you can still play poker. Between 5 and 10 I have a really hard time, and that's where I lost most of the matches with Dustin.

The numbers : if I have a 13 BB stack (M = 8.7), I open raise on the button to 3 BB. Dustin calls, the pot is now 6 BB and I have 10 BB behind. If I make a decent bet on the flop, like 4 BB, I'll only have 6 BB left and the pot would be 14 BB and I'm totally pot stuck. On the other hand, Dustin in the big blind only has to call 2 BB to see the flop, and if he flops something he can get my whole stack, +13 BB, which happens about 1/3 of the time, which is just hugely +EV.

He's risking 2 BB to get +13 from me, and I'm risking my 13 to get +3 from him. That's way out of whack and very bad for me.

02-06-06 [poker] - 1

02-06-06 [poker]

There's a lot of weird things that happen in 3-way pots. What hand do you think is best here : { A7o, Q9o, 55 } ?? The answer is the Q9 !! just because it makes the most straights (JT would even be better). Actually TJs vs. A7o and 55 is a huge favorite, 41% to win the pot!! Very strange, since Q9o is a dog to 55 one on one, and a bigger dog to A7.

02-06-06 - 1

02-06-06

King Yao has a pretty cool blog on sports betting.

2/05/2006

02-05-06 [poker] - 1

02-05-06 [poker]

I'd like to work more on my use of image in poker games. I think generally I have a pretty good image vs. people who have seen me a while. I'm sort of intimidating, and they know I can make some big bluffs, but that I'm also tight and smart, and the combination just makes them scared to play with me and makes them mess up. However, against super-fishes and people I haven't played much, I don't have much image, and I can't really control how they play against me. Ideally against really bad players, you set up an image and get them to play one specific way so that you can manipulate them and read them. There are two interesting choices here.

One is the "asshole" who actually wants you to attack him. I don't know if it's intentional, but I think Sean Sheikan and Phil Helmuth have this, and sometimes Dustin gets this in the local game. Basically this guy acts like just a real ass at the table, mocking people, berating them, and really pisses people off so that they really want to bust the asshole just to get him off the table. This makes them go after the asshole. To counter, the asshole just plays super-tight. He needs to keep making little stabs, but if you come at him big, he can make huge laydowns, and you get frustrated and keep attacking with worse and worse hands, and then he calls you and you're busted. This really works best in live games where you can really get under someone's skin, though Drew used it successfully online.

Another is the "maniac" or gambler. Basically you create the image that you're a fish who wants to gamble it up, you want people to think you're bluffing all the time. This is the classic LAG strategy used by many top pros. You play a lot of small pots and make lots of little stabs. You also intentionally take gambles where you might be a slight dog (but not a big dog), you turn over some cards and show bluffs with junk. Now everyone thinks you're a nut and they want to call your bets with mediocre hands and play back at you with junk, so when you get cards you get a ton of action.

One that I use sometimes is to act like the "mouse". People who know my game aren't fooled by this, but I can use it in games I'm new to. The "mouse" is super tight/weak, and will fold to your bluffs and give up big pots. I create this image by just not bothering with small pots that I think you're bluffing, I let you have them. Then comes a pot where I build up the pot, then I check the river, and you bluff it big, and I call with a mediocre hand and pick off a huge bluff. This is a great way to exploit bullies who just can't resist attacking perceived weakness. It's sort of like a Judo approach, you sucker them into over-committing their weight, and suddenly you're on top of them and you've got the kimura.

02-05-06 - 1

02-05-06

The fucking numpad is ruining my elbow. The MS natural keyboard would be almost perfect if its right-most key way the right Ctrl key. Having to reach out way past the numpad is disastrous ergonomically.

this mod rocks

2/04/2006

02-04-06 [poker] - 1

02-04-06 [poker]

If your opponent pushes with a typical semi-good push hand, { TT+, AQ+ }, would you rather have JJ or AK ? The answer is AK. AKs is 51% vs. that range, AKo is 49%, and JJ is 47%. A funny thing to me is that if you drop the AQ out of there, it doesn't hurt AK that much (relative to JJ). If the range is { TT, AK }, then JJ is 43% to win and AK is 41%. I thought maybe the AK was beating JJ just because it's so strong against AQ, but that's not it. Part of it surely is that by having AK you greatly reduce the chance of facing AA or KK.

02-04-06 - 2

02-04-06

The big game of the week is not in Detroit. It's at the Wynn in Las Vegas, where Andy Beal is playing $50k/$100k limit hold'em against the syndicate of pros named "The Corporation". The best coverage is at Bluff Magazine though you have to pick through some garbage to find the updates from the moderators.

addnedum : the match only lasted a few days and was rather a disappointment. They didn't play very long, and the little swing could've just been variance. Also it seems like Andy may have gotten tilted, which is lame. It would've been nice if they played a good long match and played solidly.

02-04-06 - 1

02-04-06

If you're a big free-market guy, you must see there are quite a few huge economic advantages to national health care. By disassociating the health care from the job, it gives workers more freedom to leave jobs and change jobs. This increased fluidity is great for economic engines; any "binding" or friction factors in capitalism are major impediments to functioning markets. Another is the massive amount of administration burden both to the companies and the workers of managing the health insurance and benefits packages that employees get. Some of that same work would have to be done by the national program, so that would zero-sum, but a lot of it goes away, because you're never changing programs, you don't have to figure out new policies and benefits and get new doctors, etc. etc. you're always on the same plan.

Health care costs have been rising way faster than inflation. Some of the rise in Medicare costs are because there are more and more old people. The solution to that is obvious - kill anyone over 70. Okay, I'm joking (mostly), but even if you normalize out that factor, the cost of health care for a middle aged person has gone up way faster than inflation. That's very strange. Naive people say it's because care has gotten so much better, so it must cost more, right? Nonsense. In almost every other discipline, gains in quality are roughly matched by gains in efficiency, so that costs stay similar.

Part of the problem is surely the insurance model, in which a very few people who receive super-costly care drives up costs for everyone. I have no idea to what extent this is actually happening, but it makes some sense. The same is true for things like car insurance - the majority of the costs come from the rare massive payouts. Perhaps less than 1% of the population incurs super-high health care costs.

2/03/2006

02-03-06 [poker] - 2

02-03-06 [poker]

The net winnings of everyone on Party Poker that GoldBullion has watched is -$200,000 over roughly 1.5M hands. That's -$13 big blinds per 100 hands (-6.5 PTBB/100) (per person). That doesn't mean I'm watching losing players - that's just the rake!! Every hand I see the money goes from one guy to another, so the net balance should just be $0.

02-03-06 [poker] - 1

02-03-06 [poker]

People often overestimate implied odds. The problem is that the amount you make up when you hit is reduced by the chance of hitting. This makes implied odds particularly bad when your chance of hitting is low. In Limit Hold'em, implied odds are usually pretty thin. In No Limit they can be very good, but you have to be sensible, especially on the turn.

Say you're on the the turn with a gutshot. You have 4 outs, so your chance of hitting is 8.7%. Say the pot is like $20 and your opponent bets $5 into it. That's a puny little bet and you're thinking that you might be able to call with implied odds. On immediate odds, you have to call $5 to win $30, so if you win no more the EV of immediate odds is : -$5 + $30 * 0.087 = -$2.39 , obviously a bad call. Can you make that up with implied odds? You have to win at least 2.39 over all, which means you have to win 2.39/0.087 when you hit!! That's $27.47 extra you have to win when you hit. That might be possible, but if you imagine that he'll fold to a big bet about 50% of the time on the river, it means you have to bet $55 into a $30 pot, which he's pretty unlikely to call even that 50% of the time.

02-03-06 - 5

02-03-06

That Troy Polamalu (sp?) is one hell of a damn good safety, BUT he does that horrible lazy thing where he dives at people's legs to tackle them, and he turns his back. That works often, but sometimes the runner can sort of hop back and lead into it, and Troy just rolls off his legs and the back keeps on going. He needs to dive head first and get his arms out to wrap. Pussy.

02-03-06 - 4

02-03-06

Printf is the "leak" in my coding game. Those of you who have worked with me or seen my Game-Tech talk on using C++ to make self-checking code know that my coding style is very rigorous, and I try to engineer things such that the compiler catches the errors, rather than the run time. That is, as much as possible, I want any errors to be compile time errors which I can easily fix, rather than run-time errors which I have to test and debug. While working on GoldBullion I've hardly ever had a bug which wasn't some sort of complicated logic situation. However, there is one glaring exception : PRINTF !! I still make heavy use of printf because IMHO it's the only decent way to format text in C, but it's lack of type safety gives me a stupid bug several times a week. Obviously they're easy to fix, but they cut into my dev cycle a lot, because they're bugs where I have to run the app, see there's a problem, and go fix it. If it was compile-time type-checked the iteration time would be so much better.

addendum : Wes pointed out Boost format as an option. Just reading about it, it looks pretty darn cool. I think I'll try it out. It is 5X slower than printf (!!) and I imagine the compile time hit is not zero because it is a big boost template beast, but those are probably minor trade offs.

old rants