OT: Analysis of my recent results (long)

[pineapple888]

In the tradition established by good2cu, I am going to post to STTF one day after saying I was done with the group.

There are several reasons for this, three of which are:
1.) I'm in a good mood for the first time in several months. Partly because I had a nice rebound after my latest horrible downswing. But mostly because of a hottie at Vons supermarket who was trying to sell me gift cards. And was nice to me. Which is pretty rare in my life right now.
2.) JacKnight21, over my protestations, insisted on reading highlights of several STTF threads to me while I was eating lunch today.
3.) I don't want to have to explain all this shizzle at STTF-USP.

So, what follows is a rather lengthy analysis of my recent results, including the conclusions I've drawn, and responses to some of the posts that JacKnight21 read to me.

Note that I still won't be reading any other threads, won't respond to any replies to this thread, and won't read any PMs. In a few weeks, maybe. But have at it anyway, of course, if you feel like discussing this information further.

I'll be discussing the results of the 1261 Turbo $60 STTs I have played since moving up from the 27s on November 5. This sample size is not enough to establish an ROI, but I think it is enough to be interesting, and to draw some conclusions.

First, my Finish Distribution:




Note the $26 profit, for an ROI of 0.0% before rakeback.

The first interesting feature of this graph is that, until we get ITM, it is not very far out of line from what we've come to expect from a winning player. The bubble percentage/ITM is a little low, but nothing too horrible.

Then, once you look at ITM, it's clear that the 3rd vs. 1st is out of whack. Typically, the two percentages would be close to reversed. So, if a player I respected showed me this graph, I would say "Dude, you're doing fine, just having some bad luck once you are ITM, stay the course."

It's interesting to look at what the results would be if we swapped the percentages for 1st and 3rd. In that case, I would have a profit of $5175, for an ROI of 7%. Not outstanding, but solid, given that I was running a little cold on the bubble.

Now, what would have had to happen for this swapping of percentages to occur? As an approximation, let's say that there were some key showdowns that led to me getting 3rd instead of 1st. This is not really accurate, of course, but it is a reasonably proxy I think, given that the actual key showdowns may have given me 4th or 5th instead of 1st, i.e., no money at all.

It turns out that if 34 of these key showdowns had gone my way instead of Villain's way, the percentages would be swapped. I was ITM 472 times. So, if I could have reached in and changed just 7% of these key showdowns, I would be running just fine and dandy.

In other words, it doesn't take much to turn a crappy run into a decent one, and vice versa.

Now, on to the Winnings By Tourneys:

Results graph:


The overwhelming question that this graph should raise in your mind is: WTF IS WITH THE VOLATILITY?!?!?!?

I mean, I've seen some volatile results graphs, but this one is just ridiculous.

A key concept to realize here is the following: assuming independence of tournaments, Finish Distribution Encapsulates Results.

What I mean by this is that the details of my actual playing style are completely irrelevant to the volatility of this graph, which is solely determined by the Finish Distribution above, plus random variance.

I'm not interested in proving this result again, there was a huge thread where it took me a week to convince Slim (a Ph. D. candidate in Nuclear Physics) of the truth of this concept, somebody can dig it up again if necessary.

But wait a minute, you might say, what about the assumption above about independence of tournaments? Surely I must have been on Monkey Tilt during my downswings, right?

No. There are several pieces of evidence I can cite here:
1.) I wasn't. You simply have to believe me on this one.
2.) If I was on Monkey Tilt during my downswings, that means that outside of my downswings, I pretty much must be the best poker player in the history of the universe, by a very large margin, if I am going to manage to break even, and this is just absurd.
3.) JacKnight21 has sweated some of my sets over the past couple of weeks, and he sees the unbelievably horrible runs I get on, where I get rivered 7 or 8 times a set by 2- or 3- outers, and sure this happens to everyone 1 or 2 times a set, but not 7 or 8 times, and not 5 or 6 or 20 sets in a row.
4.) This crap happens to me so often that if I had been tempted to go on Monkey Tilt during a downswing, I would have long ago learned how to deal with it.

So, WTF?!?!? Really, I have no idea. Extreme volatility just seems to be my own personal curse, and IMHO, there is no reason for it, and it should end Real Soon Now. But when you are in the middle of it, and it lasts for months, well, it sure does suck balls.

Now, on to some other issues that I have seen raised at various times, or that JacKnight21 raised today:

1.) My play is too aggressive.
First, as I noted above, my playing style should have no effect on my volatility.

Second, the regulars here should realize that I play very, very close attention during my tournaments and take detailed notes on every player, which means I know how good you are, and I know what your leaks are if any. If you are good, you are going to be calling tight in many situations, so I am going to push on you, and you are going to be pushing wide in many situations, so I am going to look you up frequently, and if I see a spot where I can exploit one of your leaks, you can be damn sure I'm going to put pressure on you.

If you are really curious, sweat my play when there are a bunch of nobody donkeys left at the table. You'll see me just sitting there, not doing much of anything, folding into 2nd, and then opening up.

Summary: Selection Bias.

2.) I should continue to work on my game.
I do.

3.) I'll review your hand histories for you.
Thank you. I have had several people do this. They have found differences in playing style, and maybe a few small leaks, but nothing remotely resembling big leaks.

As I said, if any of you were sweating me over the last few months as I played, you would have seen the huge number of absoultely stupid, horrendous beats I take, and all this talk of analyzing my play for leaks would become laughable. I get my money in as a clear favorite nearly all of the time, and I lose more showdowns than I should. End of story.

4.) Are the 60s beatable currently?
IMHO, the 60s are currently beatable for approximately 10% ROI by a top-notch player. I have stated this before in other posts. But there are not very many top-notch players at the 60s from what I have seen. AtlBrvs4Life (who isn't around much) and maybe a few others. Most of the regulars have quite significant gaps in their game IMHO that will prevent them from making any kind of decent money at the 60s long-term. You might think it is unspeakably arrogant of me to make this judgement, and you are right, but I'm making it anyway.

5.) Will the 60s be beatable by a less than top-notch player in the future?
Who knows? I'm waiting until after Jan. 1 personally, to see what develops.

6.) I should use table selection.
See 4. above. I see little point to avoiding most of the regulars, as long as there are still a few fish at the table, and it would take me 3 hours to fire up a set otherwise, so it's impractical in any case.

7.) It's stupid to play STTs.
Agreed. Cash ring games are the place to be. I'll get there eventually. I have my reasons for playing STTs at the moment. Mostly bankroll-related. But unless you are one of the few top-notch STTFers AND are having bankroll troubles, really I see no reason to continue to play STTs. You might think this statement is self-serving, but it is my honest opinion. Really, I can't wait until I leave these damn things behind.

[blackize]

[ QUOTE ]
It turns out that if 34 of these key showdowns had gone my way instead of Villain's way, the percentages would be swapped. I was ITM 472 times. So, if I could have reached in and changed just 7% of these key showdowns, I would be running just fine and dandy.


[/ QUOTE ]

It's a shame you won't be reading this, but I raised a very similar point with Ryan in #STT a few months ago.

Essentially I was beginning to think that given the nature of SNGs your finish distribution will be determined by just a few hands. I interpretted this to mean that the necessary sample size to get close to a true ROI has to be almost astronomical to get around the enormous swings associated with so few hands entirely determining your winrate. As you have shown, just a few of those key hand going the other way can swing things to a very favorable outcome.

[Semtex]

Personally I've been having the same ITM problem. My luck just seems to go south when I get there. I'll win every coinflip and race during th body of the tournament, make the money, then instantly lose an 8-8 v A-2 race, or push A-9 on the button only to have BB wake up with queens. Going to keep grinding away though...

[ManChild]

[ QUOTE ]
[ QUOTE ]
It turns out that if 34 of these key showdowns had gone my way instead of Villain's way, the percentages would be swapped. I was ITM 472 times. So, if I could have reached in and changed just 7% of these key showdowns, I would be running just fine and dandy.


[/ QUOTE ]

It's a shame you won't be reading this, but I raised a very similar point with Ryan in #STT a few months ago.

Essentially I was beginning to think that given the nature of SNGs your finish distribution will be determined by just a few hands. I interpretted this to mean that the necessary sample size to get close to a true ROI has to be almost astronomical to get around the enormous swings associated with so few hands entirely determining your winrate. As you have shown, just a few of those key hand going the other way can swing things to a very favorable outcome.

[/ QUOTE ]

AKA none of us have enough of a sample and we are going to go on gross runs and monkey tilt?

[ManChild]

It turns out that if 34 of these key showdowns had gone my way instead of Villain's way, the percentages would be swapped. I was ITM 472 times. So, if I could have reached in and changed just 7% of these key showdowns, I would be running just fine and dandy.

okay, thats nice, and if i won 60% of my coinflips with the PP then i would make more money than winning 55% of the time, but mathematically that doesnt work out. so what is the point of this statement

[AMT]

[ QUOTE ]
It turns out that if 34 of these key showdowns had gone my way instead of Villain's way, the percentages would be swapped. I was ITM 472 times. So, if I could have reached in and changed just 7% of these key showdowns, I would be running just fine and dandy.

okay, thats nice, and if i won 60% of my coinflips with the PP then i would make more money than winning 55% of the time, but mathematically that doesnt work out. so what is the point of this statement

[/ QUOTE ]


i think hes saying if the numbers came out the way they should have based on the odds, not "wow i wish i was a luckbox"

[Bastian]

[ QUOTE ]
It turns out that if 34 of these key showdowns had gone my way instead of Villain's way, the percentages would be swapped. I was ITM 472 times. So, if I could have reached in and changed just 7% of these key showdowns, I would be running just fine and dandy.

okay, thats nice, and if i won 60% of my coinflips with the PP then i would make more money than winning 55% of the time, but mathematically that doesnt work out. so what is the point of this statement

[/ QUOTE ]

He's not talking about coinflips. What are you talking about?

[ManChild]

[ QUOTE ]
[ QUOTE ]
It turns out that if 34 of these key showdowns had gone my way instead of Villain's way, the percentages would be swapped. I was ITM 472 times. So, if I could have reached in and changed just 7% of these key showdowns, I would be running just fine and dandy.

okay, thats nice, and if i won 60% of my coinflips with the PP then i would make more money than winning 55% of the time, but mathematically that doesnt work out. so what is the point of this statement

[/ QUOTE ]


i think hes saying if the numbers came out the way they should have based on the odds, not "wow i wish i was a luckbox"

[/ QUOTE ]

nowhere does he say that statistically he SHOULD have won 34 more of them, he just says IF he won 34 more.....
( and your high)

[ManChild]

[ QUOTE ]
[ QUOTE ]
It turns out that if 34 of these key showdowns had gone my way instead of Villain's way, the percentages would be swapped. I was ITM 472 times. So, if I could have reached in and changed just 7% of these key showdowns, I would be running just fine and dandy.

okay, thats nice, and if i won 60% of my coinflips with the PP then i would make more money than winning 55% of the time, but mathematically that doesnt work out. so what is the point of this statement

[/ QUOTE ]

He's not talking about coinflips. What are you talking about?

[/ QUOTE ]

it doesnt matter if hes talking about coinflips
the point is IF you won more showdowns than you actually did youd have more money - not a very hard good point

[blackize]

[ QUOTE ]

okay, thats nice, and if i won 60% of my coinflips with the PP then i would make more money than winning 55% of the time, but mathematically that doesnt work out. so what is the point of this statement

[/ QUOTE ]

How long do you think it is possible to win coinflips at a rate of 48%? It's entirely likely that if you play enough poker this winrate of coinflips will continue for hundreds or thousands of hands. Since most of the times we are in coinflips are times when it has opportunity to drastically affect our finish distribution one way or the other this is a rather large deal and as Pineapple has shown could drastically reduce one's ROI.

This all goes back to what I had to say about sample size. With so much of your ROI riding on so few hands an extended downswing can really change your finish distribution and make your TRUE ROI impossible to determine.