A post to put all of your results in to perspective...

[45suited]

This thread got me thinking...

I have an idea for casting of a remake of an old comedy classic... Eastbay could play Mary Swanson and Mr J could play Lloyd Christmas:


Lloyd Christmas: What are the chances of a guy like you and a girl like me... ending up together?

Mary Swanson: Not good.

Lloyd Christmas: Not good like one in a hundred?

Mary Swanson: I'd say more like one in a million.

Lloyd Christmas: So you're telling me there's a chance?

[img]/images/graemlins/smile.gif[/img] [img]/images/graemlins/grin.gif[/img] [img]/images/graemlins/smirk.gif[/img] [img]/images/graemlins/wink.gif[/img]

[WarDekar]

Not entirely sure if this has been discussed or not before, but don't you think it's possible SNG results don't follow a normal distribution? Truly you don't believe it's as simple as you've stated above.

[eastbay]

[ QUOTE ]
Not entirely sure if this has been discussed or not before, but don't you think it's possible SNG results don't follow a normal distribution? Truly you don't believe it's as simple as you've stated above.

[/ QUOTE ]

My calculation assumes nothing about distributions, normal or otherwise.

What it does assume is stationarity of the underlying process on a relevant timescale. Sue me.

eastbay

[yeau2]

[ QUOTE ]
I'm definately 15%+ longterm at the $22s.


[/ QUOTE ]

You can't say that.

[awr000]

Look at the big brain on brad....


been waiting to use that line [img]/images/graemlins/smile.gif[/img]

[Irieguy]

Well, Mr J's game is at least twice as good as other $22ers whom i've seen post 20% ROIs over whatever number of SNGs.

But it doesn't really matter... he was just trying to illustrate that a good player can run really poorly. But that's a futile illustration on this forum for several reasons, not the least of which is the fact that everybody will assume that poor results are due to poor play when somebody posts them. It's human nature; nobody wamts to believe that a good player can do poorly for very long.

Eastbay had the most poignant post in this thread: there's a 3% chance that Mr J is simply unlucky over that sample size if he is "really" a 20% ROI player.

The problem is that 3% is perceived as a small number. Mr J makes his exclusive living off of wagering on games that have large components of chance. He will lose to a 3% event hundreds of times a year because he is making tens of thousands of wagers.

People tend to underestimate the size of the pool from which they may pluck events that can be considered unlucky.

If I see a patient who is pregnant, has diabetes, and also has sickle cell anemia I don't spend the rest of the day contemplating with awe how unbelievably unlucky she is... or how unlucky I was to have come across something like this. If you look at the incidence of these conditions and then calculate how likely it is that a single person will have all 3 at the same time... the odds are more than a million to 1 against.

It's a very challenging task to evaluate a fabricated collection of data and then assign meaning to fabricated metrics that can be statistically extracted from that collection of data. It becomes even more difficult when you attach arbitrary characterizations to the metrics:

Result A for metric A means you are "good"
Result B for metric A means you are "bad"

Mr J is a winning player, and over a large number of SNGs he won money. That is an expected outcome. The details and magnitudes should not be mindblowing within that basic framework.

This is why Eastbay's comment is so great. He says "there is a 3% chance that results X can be due luck..." and the dialogue immediately reverts to whether that means that somebody sucks or not.

Irieguy

[Mr_J]

[ QUOTE ]
Did you just explain to me what 3% means? Just checking.


[/ QUOTE ]

I'm sure you know what 3% means [img]/images/graemlins/smile.gif[/img]

I'd like to know the chance that an 17-23% player will go 12%ish over 1300 sngs.

Don't mean to come across as someone who thinks they are gods gift etc, just confident that I am a better player than my results reflect.

[eastbay]

[ QUOTE ]
[ QUOTE ]
Did you just explain to me what 3% means? Just checking.


[/ QUOTE ]

I'm sure you know what 3% means [img]/images/graemlins/smile.gif[/img]

I'd like to know the chance that an 17-23% player will go 12%ish over 1300 sngs.

Don't mean to come across as someone who thinks they are gods gift etc, just confident that I am a better player than my results reflect.

[/ QUOTE ]

X%'er getting 12.5% or worse over 1350:
16.4% -> 19.3%
19.1% -> 7.2%
23.6% -> 0.7%

eastbay

PS My 3% post was a tad mischievous. I sort of knew that a fight would break out about what the 3% "meant." It doesn't really mean anything other than what it says.

[WarDekar]

My point was, whether I stated it correctly or not, is that even what it says can't necessarily be taken at face value. We don't always play the same way from SNG to SNG, or from 500 SNG block to 500 SNG block.

[eastbay]

[ QUOTE ]
My point was, whether I stated it correctly or not, is that even what it says can't necessarily be taken at face value. We don't always play the same way from SNG to SNG, or from 500 SNG block to 500 SNG block.

[/ QUOTE ]

You don't have to, as long as the variation is fast enough to wash out.

The OP is talking about his confidence in "his skill level" which seems to indicate he has a pretty well formulated game. He's not talking about playing bad, or tilting, or what happened when he tried to adjust or play a new style. So it's not crazy to think his game is reasonably stationary.

The only other thing which can move here is the player pool, and I just don't buy that the player pool moves significantly over the times of interest here.

So I stand by my estimate, and challenge you to produce a better one.

eastbay