If you turn that sideways it is a christmas tree with red lights.

It is important to keep in mind that 55 and 80 hour samples are both lol meaningless sample sizes.

Having said that...

You've just went on a 900bb downswing in 1/2 NL. As you can see from my giraffes for laffs above, it's pretty easy to plummet within a short amount of hours. Those giraffes above show my two instances of 955bb downswings and they both happened in less than 80 hours. However, those are the only two downswings I've had above 800bbs in 3766 hours of playing 1/3 NL (where I've posted a 7 bb/hr winrate overall). Game conditions (i.e. stack sizes, aggroness of game, etc.) as well as your style (LAG vs TAG, etc.) will have a lot to do with how often you should expect this to happen, but it also depends on your overall winrate as well as the amount of hours you've put in. If you've played a decent amount of hours over 7 years at an ok winrate, you're probably simply very lucky it took you so long to encounter a 900bb downswing.

Good luck digging out. Just play your game that has always been working for you and you will, eventually.

GgoodluckG

All right, here is my latest update. Finally I hit a bit of a downswing after a crazy hot December.

Total Stats: (since starting playing again in Sept.)

All at $1/3 NL $300 Max buy-in.

Hours: 157
Won/Lost: $9436
$/Hr: $60.10
BB/HR: ~20
Std Dev: $283/hr

After running mega hot in December with over $100/hr over 55hrs or so and picking up $6k, I hit a $1700 downswing in January. I definitely experienced some negative variance, but also was pretty aware that I was still making plenty of mistakes which no doubt exaggerate any downswing.

Downswing lasted 5 sessions and about 560bb. Last session though I picked up another 230bbs, so hopefully downswing is over, though I still have a little hole to dig my way out of before I reach a new all-time high.

I was telling myself that I would start taking shots at $2/5 when I hit $10k, but I am re-tasking that to after 250hrs of $1/3. So probably two more months before I take some shots. I am wanting to be very conservative with my roll, as it will not be easily replaced. (Well, the wife would strongly resist it's replacement).

Thanks for the thoughts. Yes instead of viewing myself as "unlucky", I do think it would be more fair to say I've just been lucky it hadn't happened before. I think I posted a while back (probably when I was on a heater) that I didn't have much variance in my results, but this has definitely humbled me.

When I went on my first 955bb downswing at my 1,774 hour mark of 1/3 NL, I had never even been on a 500bb ($1500) downswing before.

Geverysamplesizewe'llputinduringourlifetimewilllik elybetoosmall,imoG

80 hour downswing is going to be happening more often than you thought.
If you kept a record of your hands that involved a lot of money & you were a favorite when you got it in, then it's all variance. However, how many times did you read your V wrong & got it in bad?

Read an article in Card Player by Brian Devonshire over a year ago. He had an 18 month downswing where his bankroll was less at the end of the 18 months. He did not say how by how much, only that he didn't take any money out for expenses.

So, his bankroll could have been X & he ended up with 95% of X after 18 months. I can't wait to try & maintain my mental discipline vs. that kind of downswing.

That's like a 2/5NL player, with a $40K bankroll, having $38K left after playing poker for a year and a half.

Now Devonshire plays in a totally different kind of game & most likely doesn't play 100+ hours of cash a month, but still, I think it's possible.

At the 2|5 level a professional will never have such a downswing, or they should probably not be a professional. Mathematically this is beyond the point where one could suspect either rigging or sucking at poker (or their game just sucks). In any case one should quit playing poker for money.

It may be totally reasonable for certain other types of players to experience such a downswing, such as the big game players who don't get much volume and have lower edges. It's not reasonable for a 2|5 pro.

Edit: It's probably close to 1 in a million for a very good player depending on the parameters you choose.

I agree 100%. 18 months of losing poker for a 2/5 professional? That would be somewhere in the 2500-2750 hour range. There's just no way anyone who is good enough to legitimately be a pro poker player runs that bad for that long. No way, no how.

Ah the naivety of youth

Why is it so hard to believe math?

I'll borrow a quote from YTF in the Breakroom thread: "You can’t reason someone out of a position they didn’t reason themselves into."

Being 1200 hours into a BE stretch, I can tell u. It doesn't matter if you believe in the Boogie Man, he believes in u.

0% chance a winning pro loses money after 3,000 hours at 2/5. Zero percent.

Nobody is saying its not possible to break even for 1200 hours. I know lots of guys who break even or lose a **** ton of money over 1200 hours. But that is just not going to happen to a player who is a 6-10+BB winner and if you're not a player who wins at that rate you shouldnt be playing for a living.

Mines PLO, bit apples and oranges I suppose

Are you trying to make a living off of poker?

This is bad argument. Maybe you have little interest in this discussion which has probably come up many times already in the past, which is fine. However, I don't think it is fair to discredit someone without challenging their claim.

Of course it is possible for a professional to break even for a very long period of time, but there is a point where such a professional can no longer have an expectation of sufficient profitability in their market. I do not know exactly where this point is, but it comes before 18 months.

The possible length of a breakeven stretch can be modeled mathematically as a function of SD/winrate. We can start with the function that models our bankroll over time. u is winrate, sigma is standard deviation, t is time in hours and z is a constant.



Set B(t) = 0 and solve for t:



If we set SD/u = 10 and z = 5 standard deviations, we get t = 2500 hours, which is reasonable to accomplish in 18 months. This means that according to our model there is only about a 1/3.5 million chance of this occurring given the assumptions.

My point is not that it is impossible for this to happen, but that if it does happen to you need to seriously consider that you may not be capable of playing poker professionally in your market. Ask yourself what is more likely, that I ran worse than millions of other people or that my assumptions about my capabilities are way off?

[QUOTE=browni3141;53398121]Are you trying to make a living off of poker? [QUOTE]


Yup

Those saying what is impossible for winning players should probably run some distribution simulations. bip! had some great ones ITT, but the images are now dead.

With all respect of math, statistics and simulations: i dont think its neccesary to run ultra complexe simulations to gauge pretty precisely that its close to impossible for a proven long term 6-10 BB winner to go on a 2500 hour breakeven run. Its not a constructive exercise to do either- because i have the impression that those who starts these train of thoughts with these absurd variance issues is looking for excuses when they realize that their own results isnt up to par.

Yeah, sure we can lower that mark to maybe 1000 hours to get into a closer situation about what is really possible where we would need simulations to get a non biased accurate answer.

But i have said it before and i will repeat it once more: mostly what its the case with players that is talking about all this absolutely insane variance stuff like go breakeven over 18 months,2000 hours or 2500 hours of play or whatever- is overestimating their own skills and abilities as a player. The likelyhood of you having big leaks that you arent aware of is pretty big in my opinion. You likely have blindspots, that you need outside help from another person to pinpoint for you. Variance isnt the real problem these players are dealing with, the real problem is that they arent as good as they may think.

I don't know about the egyptian hieroglyphics posted itt but I do know the assumptions were pulled out of thin air and are likely incorrect.

His hours were mentioned nowhere (and this is a bad metric anyway, it should be #hands. He plays alot of mixed games and therefore this volume is going to be very low)

His win rate was mentioned nowhere (it aint 10bb/hr in nosebleed mix games)

His stakes were mentioned nowhere (nosebleed mix games)

His stdv was mentioned nowhere (stats was 10 years ago and **** me if I'm going to try to remember that nonsense to win an internet argument)

Main point:

The stakes played are most important. I played about 300 hands last year that impacted 50% of my 2017 earnings.

The other 50% was about 12,000 hands.

No where anywhere did anyone say that someone winning 10bb/hr at 2/5 who played 2500 hours could lose. That was a strawman.

The statememt made was that a poker player could break even over an 18 month period. That is a very true statement.

Simulations are pointless because they standardize something that isn’t standardized. No two sessions are ever the same and “win rates” don’t exist the way people think they exist. You have observed results and confidence intervals. There is also selection bias at work and the fact that everyone you see “crushing” is running way above expectation which goes back to observed results vs. win rate.

The math can be very helpful. As browni3141 points out, the ratio between WR and SD is the key. GatorXP can use math to evaluate how poorly he is running IF his believed true WR were true. This information can be used to help him with the reality check on whether his beliefs about future expectation may or may not be realistic.

At one point when I played in the same game all the time I thought I had a pretty decent handle on my win rate and max possible expectation.
Now that my game type and opponent type very so greatly, haven't the slightest.

ZuneIt implied it when he said:

"That's like a 2/5NL player, with a $40K bankroll, having $38K left after playing poker for a year and a half.

Now Devonshire plays in a totally different kind of game & most likely doesn't play 100+ hours of cash a month, but still, I think it's possible."

Maybe he didnt mean a 10BB 2/5 winner but the way he made that statement in relation to the rest of his post about a guy losing/breaking even for 18 months it sure sounded like he thinks a 2/5 pro could lose or break even over 18 months.

If you are learning new games, growing, improving, etc, then worrying about how poor you might be running could be relatively unnecessary. But, if you were to determine that your current expectation is likely < X/hr, and you would then decide to pursue other avenues for income, then doing some math could be worthwhile, especially if it's nagging at you or causing self-doubt.

Obviously, multiple games makes the exercise more challenging and the results less precise, but could still be worthwhile.

I don't track such things, but my guess is that my sucky results last year could have very easily been acceptable (and perhaps even normal) had I simply ran near expectation in the ~12 biggest pots I played last year. ETA: Just realized your point was more with regards to how you run at particular stakes, although I think the same applies to how you run in big pots.

Gpokerisalotofsmokeandmirrors,imoG


Yeah, I'm kinda more on board with this. I guess the bigger sample size you have within a smaller time frame within a game that is remaining the exact same, the more confidence you'll have in the results. But it's unlikely we play in that environment, especially rec players getting in lol small hours over many years in an every changing game.

Git'spossiblewe'realllongtermlosersovera100Khoursa mpleanddon'tevenrealizeityet,maybe?G