Wednesday, September 28, 2005

How to measure progress?

I've created a lot of goals for myself over the last 9 months. A few I've met (mostly just bankroll); a fair number I've abandoned after further consideration (mostly live or online MTT related ones).

In terms of actually quantifying progress, I've really only been keeping my eye on a few metrics:
- online LHE, BB/100h: the sheer number of hands I play (not absolute compare to internet pros, but relative to other forms of poker I play) allows me to use this simple metric. For given playing limits, I am somewhat comfortable with gauging results after 25k hands. I know that many 2+2er's advocate a much larger sample set, and I probably would too if I played full time, but given my time bank I settle for 25k.
- live LHE, I make my best effort to determine and track the number of mistakes I make (measured in BB) per hour of table play. Things like missed value bets, incorrectly protecting my hand (e.g. if I incorrectly protected my hand on the turn in a 3 way pot giving another player with a 5:1 draw better odds to see the river, I would penalize myself 20% of the pot size on the turn even if I win the hand), paying off when certainly beaten, folding reasonable winning hand on river, etc... (The pace of live play is so slow that I can do substantial analysis of all interesting hands, whereas online I to put less work into reviewing individual hands. Typically, I only mentally review some online hands that occur towards the end of a session when I end up doing something very mindless, like driving somewhere, immediately after playing. I should invest a little more time in reviewing my online play.)
- online and live MTTs: ROI and avg hourly rate, my sample size is probably way too small

My current short term goal is related to online bankroll so this is easy to measure.

I guess my high level goal is something like this:
- To develop the capability and resources to have an expectation to earn $200/hour online and $100/hour live from poker

I am envious of those who have quickly and successfully made the climb to mid-high stakes level. My emphasis is on quickly. I do wonder about the duration and peak of this poker boom.

Some people, like Dr. Schonnmaker, are reminding players to be make the necessary plans to be successful regardless of how the environment changes.


eric said...

As you may have found from your play, winrate can vary w-i-d-e-l-y over 25K hands.

I've had 25K+ hand stretches at 4bb/100, and 25K+ hand stretches at ~0bb/100.

I'm sure your well aware of confidence intervals. Just plug your WR/SD/#hands into one of those spreadsheets and see the numbers it gives you. They don't lie. If you're at 3/100 after 25K hands and it gives you a CI of 0.5-5.5...that's really the range. Although, realistically, the range is probably 0.5-3.5.

Don't focus so much on those who have risen quickly. That's related to the above: win rate can vary widely over a given seemingly large sample of hands. Bad players can get very lucky for a vrey long time and do surprisingly well. Many bad payers have a lower win rate (usually negative) but a much higher standard deviation. A guy with -1/100 WR but a 20SD will have more "big" winning stretches than the guy who wins at 2/100 with a SD of 15. But he'll also have many more "big" losing stretches.

If he gets a nice long extended winning streak, he can easily think himself the best player in the world and rise rapidly through the limits. I mean why not? He won $2000 at 5/10, moved up to 10/20, won $4000, moved up to 15/30 and just won another $6000!

In live play, this phenomenon is even more pronounced, since 25K hand stretches can take years for a casual player.

I myself have moved up fairly quickly time-wise, but not so quickly hands-wise. I had 80K hands at 2/4 and 3/6 nearly 100K hands at 5/10, and about 80K at 10/20 and 15/30. And now I'm at 20/40, with still some 15/30 thrown in. Plus I've played a bunch live. All that's happened in a year or so, but that's still 300K hands...

d said...

Eric, Thanks for the reality check.

You are, as usual, correct. Numbers don't lie.

I should probably use 2.5BB/100 as the lower bound of my CI as a gating requirement for my BB/100 goals.

I suppose that SD is going to be much higher in 6 max games, but I still prefer these since they require more decision making.

d said...

I previously marched up to the $15/$30 online level (and I'll be honest to admit, in some cases without even the aforementioned 25k set size), and then subsequently cashed out the lion share of my online bankroll to use as a seperate live bankroll for $20/$40.

As I reclimb the online LHE levels using statistically accurate confidence intervals, it will be a kick in the gut if it turns out I can't even claim to be a winning $3/$6 player!

eric said...

A couple more things:

1) You say you want to be able to earn $200/hr online and $100 live. $200/hr online is probably easier than $100/hr live. To make $200/hr online, "all" you have to do is beat the 10/20 six-max games for a decent rate and get rakeback. Four tables means 400 hands/hour, so if you can make $42/100 hands, that is 2.1bb/100, then you should be able to pull the other $32/hour from rakeback.

An equivalent live level would probably be 30/60 or 40/80. It's probably a little tougher to make $100/hour there, because then you're talking about 1.25 or 1.67BB/hr, which is quite good, long term.

2) Re: poker environments, in my limited experience, there are a lot and I mean a lot of donkeys playing NL right now. That might be something you could look into. Seriously, the 2/5NL games here in Vegas are full of home-game heroes willing to call $50 preflop raises and $300 flop bets with whatever random hand/draw that guys even at 4/8 limit are folding. Heck, from what I've seen, you might be able to hit your magical $100/hr live goal there...