June Stats

[peterbirks]

Before I start, I have details of Elkan's funeral. I will try to make it there for the whole thing.

A bus is being booked to take people from the Heath back to Highgate Spinney, so if you could ring Angie or Venice to let them know if you will want to avail yourself of this, it would help:

Friday 7th July 2006

11am Funeral Service at Burgh House, Hampstead

New End Square
Hampstead, London
NW3 1LT
Tel: 020 7431 0144
http://www.burghhouse.org.uk


12 noon Scattering of Ashes on Hampstead Heath

1pm Bus leaves Burgh House to take people to the wake at Elkan and Angie's house

4 Highgate Spinney, Crescent Road,
Crouch End, London N8 8AR
tel: 020 8341 1911



I began June with a simple target. I wanted to play 20,000 hands on Party at $2-$4 and to win $2,000. Given my performance over the previous couple of months, this would be my expected "earn". I tabulated a target of 400 hands a day (4-tabling) during the week, and 1,500 hands a day at weekends. It all seemed so simple.

As we know, it all went belly-up.

First, 4-tabling proved rather harder than I anticipated. I could just about keep up the pace, but a combination of beginning to slightly run bad and a lack of practice 4-tabling led to slight mistakes creeping in as my focus waned. The next thing I knew, I was stuck for a couple of hundred.

Undeterred, I resorted to plan B. THis entailed three-tabling on Party and sweeping up the easy money on Virgin.

Plan B, too, went tits up. I ran bad on Virgin. It was horrible, and it was at $5-$10, or E4-E8, or E3-E6. It didn't matter how dreadful the players were, I couldn't do a thing right.

It all culminated with two dreadful (ish) losses on Virgin.

There are two kinds of losses. One is where you know you can beat the game, but variance is kicking you in the teeth. The following morning, these losses don't bother you too much. But at the time, they can bother you a lot, as idiot after idiot seems to win money, and you can't win a carrot. I played one table yesterday where there were five donkeys. Four of them went broke, while the donkey on my right won about E200 (roughly what I lost on that table). I fel like kicking his head in.

The other kind of loss is the gradual defeat at the hands of players who you might have to admit are a bit better than you. There are about three or four of these on Virgin (not much of a bummer unless you are short-handed and these guys make up the majority of your opponents) and I think I must learn to steer clear of them, even if there are a couple of fish or potential fish at the table. Anyway, this kind of loss doesn't bother you so much at the time, but it should bother you in the long run, because you will lose to these guys constantly unless you improve your game.

Anyhoos, here's the gory details on the month:

		STAKE
SITE	Data	$2-$4	$3-$6	$5-$10	Bonus	$10-$20	$15-$30	Grand Total
Party	Win	-$242	$0	$0	$174	$0	$0	-$67
Ultimate	Win	$86	-$78	$0	$100	$0	$0	$104
Virgin	Win	$34	$79	-$860	$105	$57	$28	-$561
Stars	Win	-$179	$0	$0	$120	$0	$0	-$59
Paradise	Win	$215	$0	$0	$0	$0	$0	$215
Total Win		$-84	$1	-$860	$494	$57	$28	-$368
Total Hours		80.0	5.0	25.0	0	0.5	0.25	115
Avge per hour		-$1.05	$0.20	-$34.40	n/a	$104	$112	-$3.20

And here, in even gorier detail, is the session-by-session implosion:




So, let's cheer myself up with a six-monthly summary.


First there was Virgin, where I started off losing at $2-$4 (no fold'em hold'em), deposited another $500, and then started doing quite well. If you realize that I played only half as many hours on Virgin as I did on Party, you can see why I thought it was a money machine. Unfortunately, the last couple of days put a hole in this theory.

That's one of the problems with poker. I love those hypotheticals (such as with Caro) that begin something like: "suppose your true win rate at a table is $20 an hour... then the bad beats are irrelevant. All that matters is how many hours you put in. It might not feel that way, because of the variance, but that's the way it is".

As Sklansky might phrase it, "can you spot the flaw here?"

Here was I, happily assuming that my rate at Virgin was, say, $12 an hour, when I get two horrible days and the rate plummets to $5 an hour. So, which is the "true" rate? Is it $12 an hour and I've just hit a couple of bad sessions, or was I running really well beforehand (when I thought my real rate was $12 an hour) and all that has happened is that I have reverted to the mean, making my "real" rate $5 an hour?

I have no way of knowing apart from gut instinct.




++++++

Then comes Party. What a weird year it's been there. The site has definitely got tougher, but my profit distribution has been a real oddity, even allowing for the shifts up to $5-$10 and back. To be frank, I'm utterly clueless about my real rate on Party at the moment. It might be $10 a hundred bets; it might be $4 a hundred bets.

I played about twice as many hours on Party as I did on Virgin.




+++++++++++++++


The real star of the year has been the site on which I enjoy playing least (in the sense that the opponents are, ostensibly, the toughest). The cards are dealt too quickly and you can get through 110 hands per table per hour. The software pisses me off because every time you try to type a note, it steals your keyboard away from you.

And yet, here I am, playing half the hours that I play on Party, and yet winning not much less.

Part of the reason is the bonus, which kicks in at about 4¢ a hand at $2-$4, ever since they changed the rake structure at that level.





+++++++++++++

And finally the total for the year.

These also include the numbers from Stars, Paradise and Empire, where I have played less often. Empire and Paradise ran well (just over $1,200 from these) while Stars ran badly (loss of about $200).


I was targeting between $10,000 and $12,000 for this year. This chart probably shows the difference between playing as an amateur and playing as a pro. Over the past 8 weeks I have gone nowhere. However, I am $2,000 down on my "target". If I was struggling by at subsistence level on this kind of income, my bankroll would be $2,000 lighter now than it was eight weeks ago. And I would be starting to get worried.

But I'm not a pro, so although being below target is irritating, that upward-sloping target line is an artificial construct, something to desire rather than something I have to attain. I have as much money to play with now as I had eight weeks ago (not quite true, as I write $500 "to book" every month, so my "to play with" bankroll has sunk by a grand, but this is not a significant proportion of the bankroll, and most of it just sits in Neteller anyway).

All that said, I would imagine that most winning players, pros and amateurs alike, will know how frustrating this kind of sequence can be. The "bad" run probably started around May 23, with a false dawn recovery coming to an end early this month. In the grand scheme of things, even over 20,000 hands, it's statistically insignificant. We must not let ourselves get fooled by randomness. As they say at the end of AA meetings, the trick is to accept what you cannot change, have the courage to change the things that you can change, and have the wisdom to know the difference.





+++++++++++++++++++++++++

Comments

[(Anonymous)]

Pete,how many hands have you played in that 6 month period?

cheers

Dave D

How many hands?

[peterbirks.livejournal.com]

I played 87,800 hands of limit, Dave.

I could probably break it down into stakes as well. In hours it is 350 at $2-4, = 1000 table hours for $2,600 (excluding rakeback and bonuses). $5-$10 is 103 hours = about 260 table hours for about minus $200 (excl r&b) and $3-$6 is 33 hours = about 75 table hours for minus $200ish (excl r&b).

Other stake levels are either combined into the above (e.g. 4-8 euro, 3-6 euro, both as $5-$10, because that's the way they play) or are negligible. About 10 hours in total at various higher stakes (and some $1-$2 earlier in the year on UB) generated another couple of hundred bucks gain, excl r&b, in about 35 hours.

That probably means that, excl r&b $2-$4 is $7.80 an hour, $3-$6 is minus $2.75 an hour and $5-$10 is minus 80¢ an hour.

Including r&b my guess is that $2-$4 is coming in at about $13 an hour, $5-$10 at about $4 an hour and $3-$6 at about minus $1.50 an hour.

I just couldn't seem to do a thing right at $3-$6 on either Party or Ultimate this year, so I have (for the moment) given its unusual blind structure up as a bad job.

PJ

Your Month

[(Anonymous)]

Hi Peter,

The only way you can know whether your results for that sample size are just variance alone or not, is to review some of your play in your hand histories, especially for given types of hands like overpairs and draws, and see if there weren't some playing leaks contributing as well.

Also, on another topic, and especially given that you play in generally tighter games for the times you have available to play, there is an interesting theoretical limit HE article in this month's 2+2 magazine entitled "What If Everyone Is Playing By the Book?". I think you might find it of interest, although you might not think it has much application to the specific game conditions you play in.

BluffTHIS!

Re: Your Month

[peterbirks.livejournal.com]

Hi Bluff:

Unfortunately for me, at Virgin there is no hand history facility.

I suspect that a thorough-going review of Party at $2-$4 would increase my rate-of-return by an unknown amount (but, let's guess at 0.5BB per 100 bets per table), but herein lies a problem. If I intended to stay at this level in perpetuity, the research would be worthwhile. If I intend to move up, then the research might not be worth the opprtunity cost. It's a difficult one and I haven't come to a decision.

In my last serious review of $2-$4, I found that I was leaking by limping with stuff like 87s in late. A raise was profitable, but a limp with two or fewer callers previously was not. So I adopted a randomized raising strategy with these hands - sometimes raising, sometimes folding. This kept my "genuine" raising hands at the right game-theory level to make it unprofitable for players behind me (or the blinds) to play back at me.

My other conclusion was that AQ off (or less) in the small blind was not worth defending to a single raise unless I had evidence that the player might be a stealer.

A third conclusion, but this was tough, was that I might be calling too much on the turn and river. The book dictated a call, but it seemed that people in $2-$4 games were checking down if they had any doubt, and only a few were betting thinly for value. In other words, most of the players were less unpredictable than I had been giving them credit for.

I had one of these earlier today, actually. I picked up AKs in late and the betting went, call from UTG, raise from me, three-bet from the small blind, call from UTG, cap from me, call, call.

Flop came K74 two spades and small blind bets out without hesitation.

Now, I would happily bet at 1-to-10 here that this player (on whom I had no stats) had either AK or AA.

UTG calls. What does the book say? Raise "to find out where you are" (what a load of bollocks that line is) or call it down.

I chose the latter, and (with UTG presumably calling his flush and/or straight draw to the river and missing), when the bet went in on the river, I just said "that looks awfully like AA".

And it was.

So, why am I calling? I can't really hope to do more than tie, and if the flush comes I lose the lot, and yet I couldn't throw away TPTK because it was Saturday morning and there was a chance, just a chance, that this guy was talking total balls with his bets. The question is, what chance does that need to be? Here we get little help from the "names". Caro and Sklansky mutter homilies of "you only need to be right one time in x" (where x equals any number greater than six), without accepting that, these days, you will not be right one time in x (where x equals any number greater than six).

So, if you are talking about leaks, it's probably an inability to get away from TPTK when "the book" says you should call it down, but my gut tells me I should fold.

BTW, I did actually pitch AK on a board of A73, check-called the flop from in early against me in late and a very loose chaser in the blinds), followed by a king on the turn, check-raised without hesitation. I dumped the hand and saw the predicted 33 for a set appear two bets later.

I'll try to find the 2+2 piece.

PJ

The 2+2 Article

[peterbirks.livejournal.com]

As they say in the trade. Lol.

When I first played at $2-$4, back in early 2002, I noticed that most of the Paradise $2-$4 games during my afternoon were weak-tight. I didn't even have Pokertracker at the time.

Anyway, I kept a record of how often my raises first in stole the blinds, and I came the the conclusion that this guy came to, some four and a bit years later. In effect, I could raise with virtually anything and have a positive expectation.

I didn't take it to this extreme, but I started raising quite a lot when first in.

Too much, in fact. I won $800 (these were the days of single-tabling and no more than, ooh, 2,000 hands a month) over a couple of months and thought that I had cracked it.

At this point the other bastards cottoned on to me. The player base was smaller in those days, and all of a sudden I was getting three-bet.

At the time, I didn't realize what was happening. I just thought that they happened to be finding hands. After three-quarters of the $800 had vanished, I returned to 50c-$1 and rebuilt my bankroll for the rest of the year.

So, yes, the idea actually works against a complete set of by-the-book players. But even weak-tight people start to notice if you are raising more than the odds say that you should be. These days it wouldn't take very long for that to be spotted.

Therefore the best tactic would be to have accounts at a large number of sites, to play during the day (when you have most of the weak-tight players), to play "normally" for a round, maybe two, and then to do a hit-and-run set of raises.

Oh, and if you catch something, you ALWAYS bet for value at the end, because of the additional metagame plus of getting an opponent to fold and thereby concealing your raising criteria for as long as possible.

PJ

Re: The 2+2 Article

[(Anonymous)]

Well of course that is the rub. If you do too much of that light raising even in a tight game, they will adjust appropriately and start 3 betting you ligter as well, like with AQo which they would otherwise fold to a raise, especially one coming from early position. But still it has merit if you don't go overboard and also play postflop nearly perfectly, which should mean that you will be folding more to legit hands and not having your raising standards exposed except when you actually make a big hand.

Bluff