going pro: how often can one have a losing week? (MATH)

I just quit my job to play poker full time (again).

I thought about how often I can take a big loss and decided to do some calculations. I figured the ideal situation would be to only take a losing week 1 in 10 weeks. Knowing this would be unlikely, but possible, I did the math.

I took all the data from a recent PT DB i made for a smaller site (41k hands).

so the question is, "How many hours do I have to play per week to only lose 1 in 10 weeks?"
What I'm looking for is a distribution with a lower bound(LB) of 0 at mean-40%, so
z=1.28
WR = 2 (2ptbb/100)
stdev = 26.3679 (standard deviation for bb/100 hands)

LB = WR - z * stdev / sqrt(N)
N will be the number of 100-hand increments I'll have to play per week.

0 = 2 - 1.28 * 26.3679 / sqrt(N)
N = 284.7808

at 500 hands/hour I'd have to play 56.9562 hours per week. For me that would be burnout city.

I posted this for 2 reasons, 1 is to get feedback and make sure I did the math right. 2 is to contribute something, as messy as it is, it might give some perspective to players who think about going pro.

wat

Just have a lot of buyins and forget about all the Maths imo

if I earn 2K per month, I live like a Prince.
At 4k, I live like a King.
At 8K, I live like an Emperor.

Taking Emperor level, for instance, means netting almost 100K/ year.
To be able to withdraw 100K per year, taking into account the inevitable downswings that occur, I'd need to have won 200K+ that year (Without counting my beginning roll). I'd then withdraw the full amount at the end of the year, and store it in a bank account somewhere, making sure that there is a clearcut division between life-roll and degen-roll. Remember, money makes more money.

Also of note will be that it will be damn near impossible to lend money from the banks etc. You'd need something huge as escrow, plus demonstrate you need some sort of steady monthly income. Forget about mortages, et al. Just win tourneys instead.


That's all the math that matters to me. The bankroll to be constantly sufficiently padded, etc, are also of importance.

yay low living costs ftw


Bonus: there is also a "set for life" level, which you'll reach once you make 50M total profits, just deposit that shit in the bank and live off interest for the rest of your life in the Caribbean country of your choice.


This is obviously how i'd approach things, you can obviously do whatever you want

Originally Posted by gametight

Just have a lot of buyins and forget about all the Maths imo

also it's not normal so redo your maths or read some stats

Assuming you are a winning player:

*some weeks you will lose.
*some weeks you will win lots.

make sure your withdrawal per week is less than you win per week on average.

also, 500 hands/hr sounds kinda light for playing pro unless you're playing mid-high stakes...

I like to have a little wiggle room. I get about 750/hr on most sites, but I find some sites are harder to multitable than others. (I can do 12-15 on PS, 10-12 on cake, 6 on UB )

Originally Posted by dev

Checking the math for you:
What I'm looking for is a distribution with a lower bound(LB) of 0 at mean-40%, so
z=1.28
WR = 2 (2ptbb/100)
stdev = 26.3679 (standard deviation for bb/100 hands)

LB = WR - z * stdev / sqrt(N)
N will be the number of 100-hand increments I'll have to play per week.

Correct to here.

0 = 2 - 1.28 * 26.3679 / sqrt(N)
N = 284.7808

Yeah, correct all the way.

at 500 hands/hour I'd have to play 56.9562 hours per week. For me that would be burnout city.

I posted this for 2 reasons, 1 is to get feedback and make sure I did the math right. 2 is to contribute something, as messy as it is, it might give some perspective to players who think about going pro.

Someone said "it's not normal," and that's true, but it's "normal enough" with these sample sizes. The means samples are normally distributed regardless of the shape or characteristics of the underlying distribution. Your assumption here is that every sample of size 100 comes from the same distribution, which is correct enough - it's all samples from you playing poker at the same sites at the same levels.

As to whether or not its doable, I don't know. But the math is correct, and the assumptions are valid. Could the win rate be higher? That's the thing that will "improve" these statistics the most. Your standard deviation is already pretty low (mine is 40+).

I have mentioned this before in other threads about going pro. Daniel Negreanu used to play live cash games before he became known. He used to play 5 days a week and make $80/hr. He never looked at wins and losses. He would add up his winnings at the end of the month, 3 months and 6 months and it was always $80 an hour. He liked the winrate, he knew he had an edge and he just wanted to keep playing and winning. So he played at $80/hr for over a year and built his roll and lived. You need to think about your edge and winrate. When playing for aliving you may be better suited to drop a level, maintain winrate and have plenty more buyins for comfort. There is no sense in grinding $200NL for a 3 PTBB winrate when you can hit the $100NL tables for a 8PTBB winrate.

I'd just like to add that there are a lot of multi-tabling bot types that find what jyms just mentioned in the last sentence of his last reply true when going from 200nl to 400nl in FR. That is, their hourly at 200nl is similar or even better than 400nl. I'd assume that's because of the relative skill difference as you start pushing into middle stakes.

Thanks for the replies, guys.

Robb:
Sometime soon I'll probably fix the math and do it over for each limit. I'm playing a smaller site at multiple limits at the same time now, so I'm waiting for a big enough sample set. I never actually intended to play 50+ hours per week. I don't imagine I could do it. Right now I'm aiming for 20, and that's been difficult. You're input is appreciated, most of my knowledge of statistics is from poker, so it's shaky at best.

Jyms:
I read about negs' early years. It was kind of a turning point in my thinking about bankroll management. BM should be inate, and although it took me years to really get that, I'm there now. I don't even want to play games outside of my roll. Something about stories that trumps the math in the right brain.

Jyms/spoon:
I play on smaller sites right now for rakeback (which I didn't add into my calculations. It's almost 1/3 of my earn, so I'll have to redo the math for myself). The player pool is not vastly different from 200-1000nl.

I want to get back to stars, but the VIP->RB sweet spot is 100nl and that will hurt my overall earn. Also, I understand the players are generally tighter hence, not going back to stars any time soon.

I'd like to actually get enough data to figure out where I make the most, then concentrate my other poker time on higher limits from there.

Ive been a pro for about 6months and i have roughly 2-3 sessions as winning sessions. If you have a session stoploss that helps also. Math really isnt important in the equation. Whats important is the ability to put in long sessions when your running good and play your best for prolonged periods of time.