Disturbing Roy Cooke Article

[Kevin J]

I apologize if this has been covered already. CardPlayer arrives late to my card room and I just read the article "The Good and the Bad".

I also want to make clear that in no way do I mean any disrespect to Mr. Cooke. On the contrary, I greatly admire someone who is secure enough in themselves and their poker ability to publicly discuss their own adversity at the game. Also, I go out of my way to read anything this guy writes. I've never played with him, but I suspect he's gotta be one of the top limit players. This is why I was so disturbed by his article.

I'm wondering if I even understood it correctly. It seemed to say that Roy went on a 365 big bet slide in just under a month in his 30-60 game. Roy states that he doesn't play full-time anymore. So you do the math. Something's not right. It seems almost statistically impossible for a great player with a decent edge to go on this kind of slide. In what, 80-100 hours? Did I misunderstand something in the article? Has Roy played enough poker where this can be expected (once in 80-100,000 hours?).

This is the scariest poker article I ever read.

[34TheTruth34]

The first time I read the article, I got the same impression, however I think it makes more sense now that I read it again. I might be wrong, but what I got from the article is that he was -365 BB compared to where he was in January. He also stated that January was one of the most profitable months of his poker life. This huge 365 BB swing probably doesn't leave him stuck that much for the month of February.

For example, he could have won, say, 345 BB in the month of January and be down "only" 20 BB in February. Theoretically, he could still be ahead. Let's say he won a monstrous 465 BB in January and is ahead 100 BB this month. I can only speculate as to what a huge win is for a player of his caliber at those limits.

Either way, what he is probably trying to say is not that he's down 365 BB, but that he is 365 BB behind where he was in January.

[Kevin J]

Right, but that's still a 365 big bet downward spiral. And if he won +365 big bets the prior month... I don't know. Those are some huge swings for someone who plays as good as Roy Cooke. Even for 30-60.

What I'm saying is that from any point (regardless of what you've done previously), a 365 big bet slide figures to be outside the bell curve of 3 standard deviations, meaning it's very unlikely.

[34TheTruth34]

I guess I don't understand your question then. It's not unreasonable to assume that during his long and successful poker career that he's won 365 BB in one month before. Heck, if awful players can go on runs like that, then he can. It's also reasonable to assume that some point during his poker career he's broke even or even lost for the month. That's a difference of over 365 BBs. What's the difference if they come in back-to-back months or not??

[mikelow]

Let's say that he plays 100 hours during this deep losing streak. If one standard deviation is 10 big bets per hour, and his expectation is one BB per hour, his one SD range is -9BB to +11BB. His SD for 100 hours of play is 100 big bets. The
one SD range is 0 to +200 big bets. The two SD range -100 to +300 big bets.

The three SD range is -200 to +400 big bets, which has a less than 0.2% of happening.

What can we conclude from this? That the standard deviation is more than 10 big bets per hour. Even if it's 15 BB per hour, his result would only occur about 2% of the time.

So, he must not have played as well as he thought he did during his losing streak. Kind of like going on tilt and not realizing it.

[Kevin J]

Maybe it's just me, but I think sliding 365 big bets from any one point is friggin HUGE!! This is more than a reasonable bankroll. And to think this happened to Roy Cooke playing part-time in under a month?

Yes, it's mathematically possible. I suppose if you played over 100,000 hours of poker, you could expect it to happen about once. Put another way, if you start from scratch with a 300 big bet bankroll 100 seperate times, you can expect to go broke about once. This assume I understand standard deviations and bankroll correctly, which I might not. If I don't, I'd appreciate someone clueing me in. Thanks.

[Kevin J]

It's worth noting that a player of Roy's caliber should have an hourly standard deviation that is less than 10 big bets. This makes such a swing even less likely.

[J.R.]

I read it as meaning Roy Cooke has not undergone a 365 big bet slide, but as meaning that Roy Cooke won x (lets say 182 bets) in January and in February Roy Cooke was down y (lets say 183), meaning Roy Cooke is 365 bets below his January pace (which would be winning 182 bets by now),and has only experienced a 183 big bet down turn from an arbitrary point in time, the beginning of February. While still hefty, a 183 big bet slide is more within the realm of possibillity than a 365 big bet slide over the same time frame.

[PokerPrince]

It IS possible for a very good player to go on a very bad losing streak for a very large amount of BB's in a very short period of time. It's unlikely, but possible. This is the reality of poker.

PokerPrince

[Kevin J]

Right, but +182 to -183 is still a 365 big bet slide which I would think is running well outside 3 standard deviations of a very good player.

[Kevin J]

Unfortunately, you're very right PokerPrince. It's just disturbing (depressing?) to read of it happening to a player as good as Roy C. I also wonder how much of a role poorer play (which he admits to) had in such a catastrophic streak.

[brad]

ok thats about say 14k in a 20/40 game.

lets look at that.

say you play fri/sat nights.

you play 7 nights total that month.

say you play you know about 6-18 hours or whatever, (avg about 8 i guess).

if youre hot you can easily win 2k every time.

thats 14k.

now losing 14k, thats harder, but you dont hit any hands (cooke said he hit no hands) thats a rack or two right off the bat no problem. so thats 7k right there.

add to that 30/60 has higher blinds and plays more aggressive than 20/40.

also if i remember right, cooke said it was his biggest losing month ever i think, so you know you have to think about that.

[34TheTruth34]

[pudley4]

He's not saying he lost 365BB during Feb. He's looking at each month independently and saying that his performance from Feb 1 - 20 is 365BB behind his performance in January.

Say he wins 275 BB in January. Now from February 1 - February 20 he loses 90 BB. His result for February only is 365 BB behind his result for January only. This is what he's trying to say. Notice that +275 for one month and -90 for one month both fall within 2 SD (according to another poster above).

He obviously didn't lose 365BB during February because later in the article he states "On Feb 21...I got myself even for the month in one session. " He's not winning 365BB in that one session.

[Piers]

A loss of 365 big bets in a period of 100 hours is pretty big. At a rough order of magnitude guess I reckon it will only happen to about a third of break-even players in a lifetime, and. maybe in the lifetime of one in every hundred players beating the game by around a big bet an hour.

The above assumes of course that the players continue to play at their best while going through the loosing streak.

[Kevin J]

You must be right. My brain just wasn't working that way. If I were to win 10k in Jan, and lose 2k in Feb, I might say I'm up 8k for the year, or I'm averaging 4k per month, but I would never think of it as I'm 12k below my Jan performance.

I also saw his statement that he got even in one session and thought it was odd. I should've known then that I must not have understood what he was saying. Sorry for wasting bandwith. Thanks for clarifying it for me pudley.

[pudley4]

np.

He's not the most clear writer in the world. Clarkmeister pointed this out in an earlier thread regarding Roy's use of the term "EV"