A Slave to Variance

Wednesday: +4103, 1.25 hours

I’ve got a lot of mixed feelings about the last month. I started logging my results because I had been running flat over the last three or so months. The idea was to expose myself and my results so that I would not only play better, but also do a better job of self evaluation. From a purely results oriented standpoint, I did succeed in logging some excellent results. November was my second best ring game month ever, and more than twice as big as my third best month.

That feels good.

But on the downside, I did a horrible job of actual self evaluation. I did a poor job of tool development, and I can clearly point to three episodes where I knowingly played badly in a poor mental condition (AKA tilt).

Some of my failure here may be correlated to my positive results. “If it ain’t broke, don’t fix it.” The problem with that is that when it is broke, as it was yesterday, the first instinct is to play through it, and get back to even so you can quit. And when you quit, the last thing you want to think about is how poorly you played. You just want to turn off the poker bouncing around in your brain. It’s not an easy thing to do sometimes.

I’m not sure where to go from here. There is a lot of work yet to be done. On the other hand, I don’t want to come across as some kind of egomaniacal braggart — even if that may be closer to the truth than I might like it to be. I also dread the moment when I suddenly stop posting results because that $20K backslide was followed by another $20K backslide, and I lose all the nerve to post such misery. Some people like to whine about how bad they are running. I sometimes even boast about how bad I’ve run, wearing it as a badge of honor. But when you lose big, the last thing you want to do at the end of that day is tell the world.

In the end, this months results are an abberation. They are well out of my expected earn rate. But results this good just don’t last. Some might say I was drawing on all the +EV built up in my equity bank for the last three months. I figure I just got more than a little lucky.

I also don’t see myself being this open about my results over the long haul. On the other hand, it may be exactly this sort of external pressure which not only keeps me honest, but also in line.

I guess we’ll just have to see what happens next month. As for this month…

November: +52025, 91 hours

Tuesday: -$4657, 9.5 hours

Of course, the number doesn’t tell the whole story. It was more like:

2 - 20 + 4.5 - 4 + 3 + 10 = -4.5

Or maybe it was more like:

“I’m only thirty-one bets down now, I can quit!”

Yes, there was a $20K peak to trough yesterday. Needless to day, I found myself playing pretty poorly in the middle of that downswing. In other words, I was tilting. Tilting costs money. There’s a new TVB in town, and lord knows, not only do TVBs cause me troubles, they also tilt me. And often, it’s difficult to tell whether a new player is a fish, a TVB, or somewhere in between.

Needless to say, I try and be as much of a TVB as possible, but it’s a fine line.

Here’s a hand with me and the new TVB. Can you pick out the player on tilt?

omaha[L]-H/L $ 75/150   Poker Stars - Eumelos III    Tue Nov 29
         ante:0.00 blinds: 50.00 75.00  rake:1.00 pot: 1950 [Jd Kh 2h 8d 5c]
"new TVB"      1819  0 Bk   bc   rc   r            13    13 [Ks 4s Kc 4d]
*Andrew        1440  1 Sc   kr   bR   bc           13   -13 [3h Kd Jc 2c]

There’s at least $300 that I threw away in that hand due to bad play.

Live, and learn?

Monday: +$2459, 1 hour

Not a lot of playing. I sat with two big o/8 fishies early in the day and clean them both out in five minutes. That was nice. The rest of the play was mostly the usual fare; raise, bet, bet, bet kind of day nothing special.

I did get a chance to work on the code again though. In anticipation of actually reasoning about my play, I decided to finally fix some really icky that I’ve been working on/with since day one: hand history parsing.

Parsing is a pain, and parsing hand histories is a pain. It’s easy enough to get something that handles 95% of all the cases, but it’s that other 5% of weird hand logs that really takes tons of time. There are the hands where people sit in and post after the hand has started. There are those that post one blind, those that post two blinds. There are the interruped hand histories, the all-in histories, the timeouts, the fold to no bets. All kinds of little wingles to keep you hunting for hours. I think my Paradise parser has a reject rate of less than .01%, but my other parsers aren’t up to snuff.

So I’ve decided to move to a multi-level class hierarchy. Basically it’ll look like:

HandHistoryParser->HistoryTypeParser->SiteSpecificParser

Most of the hand histories that sites generate are in a straightforward log format, but there are other formats as well. I don’t play at Planet Poker now, but I remember that their histories were in a tabular format. The histories that I post here are also in a tabular format. [I actually prefer the tabular format because it makes it easy to use text processing tools on the hand histories. It’s also the format I got used to when learning poker on IRC.]

From there, the site specific parser includes all the data which is associated with the site, specific string matching rules, etc.

The idea is to push as much code up the hierarchy so that adding a new site to the parser requires minimal effort. So that’s what I spent yesterday on, and what I will be spending today on. In a lot of ways, this is about going back and dotting all of my “i”s and crossing all my “t”s. This code is code I’ve had to very activly maintain over the last five years as new sites are started and old formats are always in flux. The higher complexity of the code organziation is balance by the relativly small body of code which will have to be activly maintained. Hopefully my parsing maintenance will go down considerably after this is done.

Maybe I’ll squeeze in some poker too. :)

Overall, a fairly productive holiday. I’ve finally (re)built my Poker Stars bankroll to a size large enough so that when I tilt it all of again, I’ll be really fucking pissed:

Wednesday night: -$10, 5 hours

Don’t you just love making -$2/hr. It’s so great. It’s the new black. Or is it the The New Zoo Revue? As I said last time, it’s kind of a lame place to end a session.

Turkey-Day: +$519, 1.5 hours

Just a little squeezed in during the holiday fun. I even ate a slice of turkey.

Wine-Tasting-Day: +$776, 20 mins

Wine tastes really good, especially when your current earn rate is over $2300/hr.

Saturday of Cleaning: +$4034, 6 hours

I cleaned the bathroom, the kitchen floor, the living room floor, the comfortors. The nice thing about cleaning (for me) is that most the time is spent avoiding cleaning. That allows a good deal of time for playing poker.

Sunday: +$4867, 2 hours

Nuff said.

SongMonk asked about reading players, and here’s an excellent example of a player reading my hand and knocking me out for high.

omaha[L]-H/L $ 75/150   Poker Stars - Hecuba IV      Sat Nov 26
         ante:0.00 blinds: 50.00 75.00  rake:2.00 pot: 2175 [8h Ac 3s Td 6h]
 "Mr 42"       4752  1 Src  bc   kc   kr           11     3 [8c 3c 6d 2s]
*Andrew        2193  0 BR   r    b    kf            7    -7 [9c Ah 6s 4d]
 "nut low"     4028  3 cC   C    c    bc           11     3 [2d 3h Jc 4h]
 "uber nit"    1490  2 f                            0     0

It’s pretty clear that I don’t like my hand much on the river, and being the agressor I’ve likely got an ace, and thus am contesting the high half of the pot. The other passive caller is probably going low. So “Mr 42″ uses the nut-low bet to leverage me off of a medium two pair to steal half of a sizeable pot from me. In retrospect, I should have called. In fact, I almost did call but when I saw an unlikely straight on the board I pannicked and dumped my hand.

Oh well, I’ve made worse plays.

When I started playing poker, my roommate Jeff was one of my first poker mentors. He was a cab driver with a real outgoing personality. Tne regular game was the “cabbie game” which floated from one cab driver’s house to another. Sometimes it was even held in the break room at the cab company.

He had a philosophy when it came to results. After a night of poker, he’d rather be way up or way down. Finishing close to even was the worst fate of all.

Another mantra you frequently hear people say at the tables is: “I’m even I can quit now.” Usually, that sentiment is expressed after a long day of digging out of a big hole, struggling from deep in the red to just into the black.

Yesterday I finished about even, and upon achieving that magical place of evenness, I stopped playing. The twist here is that I had been up as much as $5000 over the course of the day, so instead of digging out, I was digging in.

Ah, the whimsy of poker.

Yesterday: -$85, 5.5 hours
Today: +$3519, 2 hours