PDA

View Full Version : Analysis of Streaks (I):Levels 1-3 (long)


Slim Pickens
03-07-2005, 01:56 AM
Background (http://forumserver.twoplustwo.com/showflat.php?Cat=&Number=1869590&page=1&view=colla psed&sb=5&o=14&fpart=1)

My general strategy in the early levels is to play few hands, just AA, KK, QQ, and AK always, AQ-AT, JJ, TT in position as appropriate, limp lower pocket pairs if I think I won’t be raised behind and fold if no set, occasionally limp a worse ace in position.

VPIP%(Overall/Phase One/Phase Two): 14.0, 15.8, 12.2
pre-flop raise %: 6.1, 6.2, 5.9
went to showdown when seeing flop: 28.6, 28.4, 28.8
won t$ in those showdowns: 51.9, 52.9, 50.8
Ok, the only thing I get out of this is that it’s basically 50/50 as to whether or not I’ll win the showdown if I see the flop. Since an average of well greater than two see each flop, I think this is fine. It indicates neither a lack of selectivity on my part or a disproportionate number of suckouts against me. I stopped limping bad aces altogether, which I think makes the difference in VPIP.

Levels 1-3. How often do I survive?
Phase One: 65/90=72%
Phase Two: 66/90=73%
Nothing to see here folks, unless anybody thinks those numbers are too low or high. Is anybody else significantly different than about 3 out of 4?

If I survive, how many chips do I have?
Phase One median chipstack: 1075
Phase Two median chipstack: 935
I hope the extra 140 chips don’t make a difference. I feel it shouldn’t matter as long as I play well later on.

How often am I seeing/winning showdowns where I’m within 60/40 either way?
Phase One: Three wins in four tries.
Phase Two : Three wins in five tries.
Conclusion: meh.

How often am I seeing showdowns where all the chips went in with someone a huge (>80%) favorite, and how often was I the dominating favorite?
sucked out on someone: 8, 2, 6
suckout against me: 14, 7, 7

How are the cards treating me?
The only notable difference I can identify is a dearth of pocket aces in Phase Two, only 5 of them where I expect 10 or 11, but I can’t say that ought to matter. My win rate when dealt AA, KK, and QQ is over 80%.

Overall my play looks fine, and I was not the victim of any vicious cold streak. This was a nice little exercise, but the difference lies elsewhere. Up next, Level 4 until it's 4-handed.

The Yugoslavian
03-07-2005, 02:06 AM
I think you can play tighter.

I also think that just playing back your last 200 HHs or whatever and stopping at crucial hands to further analysis would be more beneficial.

However, I applaud you for your analysis and willingness to self-reflect, /images/graemlins/smile.gif.

Yugoslav

SuitedSixes
03-07-2005, 02:14 AM
Let me be the first to post the obligatory, "Your sample size is too small."

I do think it's great that you are trying to figure this stuff out, but you're going to need WAY more than 200 tournaments to draw any conclusions that are meaningful. I am afraid you may end up correcting things that aren't problems and thinking things work that really don't.

This is the worst part about being a 2+2er, you can never take any comfort in short term results.

Slim Pickens
03-07-2005, 02:17 AM
[ QUOTE ]
I think you can play tighter.

[/ QUOTE ]
Agreed, and those numbers are a bit misleading. After my revelation about not trying to play A-crap and subtracting the times I think I need to start stealing on level 3, my baseline VPIP is about 13%.

[ QUOTE ]
I also think that just playing back your last 200 HHs or whatever and stopping at crucial hands to further analysis would be more beneficial.

[/ QUOTE ]
Oh, you better believe I've got those all logged and flagged. I thought I might discover trends that might not have been apparent to me right away. I was hoping maybe a specific hand like AJo or TT would jump out as a huge loser, but alas, I seem to play those for a profit...

Thanks for the input.

The Yugoslavian
03-07-2005, 02:21 AM
[ QUOTE ]
I was hoping maybe a specific hand like AJo or TT would jump out as a huge loser, but alas, I seem to play those for a profit...

[/ QUOTE ]

Yeah...I too hate it when I find those sorts of stats, /images/graemlins/tongue.gif.

Btw, SuitedSixes is onto something with this small sample size nonsense. Even the seemingly 'meaningful' stats aren't that meaningful, /images/graemlins/shocked.gif!

Yugoslav
PS to SuitedSixes....gl staying with my posts.....I'm heading for Pooh-Bah and ain't nobody stoppin' me!!

Slim Pickens
03-07-2005, 02:22 AM
[ QUOTE ]
Let me be the first to post the obligatory, "Your sample size is too small."

[/ QUOTE ]
Yeah. Don't worry. I'm not pulling out any numbers as statistically significant. I should be able to find the variance in a sample that small though. /images/graemlins/wink.gif

Barrett's Last Privateer
03-07-2005, 02:25 AM
[ QUOTE ]
If I survive, how many chips do I have?
Phase One median chipstack: 1075
Phase Two median chipstack: 935
I hope the extra 140 chips don’t make a difference. I feel it shouldn’t matter as long as I play well later on.


[/ QUOTE ]

Slim,

This may come out in future steps of your analysis, but I'm not sure I'd be so quick to dismiss the impact of this difference in stack sizes.

What if this 140 chips has led to you starting to push, on average, one orbit later at Level 4, and in that orbit one other player has been knocked out, or you've had one more orbit to find a better hand to push with, increasing the chances that push leads to a double rather than a bust?

Interersted to see the rest of your analysis.

BLP

Bigwig
03-07-2005, 02:31 AM
[ QUOTE ]
won t$ in those showdowns: 51.9, 52.9, 50.8

[/ QUOTE ]

Is this % of hands won at showdown? Or % of hands won that you saw a flop? I'm not familiar with PokerTracker.

Slim Pickens
03-07-2005, 02:42 AM
[ QUOTE ]
won t$ in those showdowns: 51.9, 52.9, 50.8

[/ QUOTE ]

Sorry. Correction: I believe that's TOTAL win percentage when seeing flop. It does not have to be a showdown.

Slim Pickens
03-07-2005, 03:19 AM
[ QUOTE ]
This may come out in future steps of your analysis, but I'm not sure I'd be so quick to dismiss the impact of this difference in stack sizes.

[/ QUOTE ]
I don't think this could be shown to be statistically significant, but...

Phase 1: median chip stack 1075
win rate with >1074 chips at the end of level 3: 26/33=79%
Phase 2: median chip stack 935
win rate with >934 chips at the end of level 3: 16/24=47%

Hmmm... I'll keep this in mind for later. I think what this really shows is that something is happening between level 3 and ITM. If it's nothing identifiable in that space than it may be this.

Barrett's Last Privateer
03-07-2005, 07:06 AM
[ QUOTE ]
Phase 1: median chip stack 1075
win rate with >1074 chips at the end of level 3: 26/33=79%
Phase 2: median chip stack 935
win rate with >934 chips at the end of level 3: 16/24=47%


[/ QUOTE ]

Slim, I assume you mean phase 2 win rate when above median = 67%, rather than 47%?

Statistcially, those two proportions can't be concluded to be significantly different, but out of curiosity, what was your ITM when >1074 in the second phase, and what was your ITM when >934 but <1074, and for the first phase, what was your ITM when >934 but <1074?

I guess I'm wondering if the probability of finishing in the money with stack 1074 at start of level 4 is actually significantly higher than with stack 934, despite there only being 140 difference. Is the correlation between stack size at this point in time and probability of cashing not linear, but it jumps at some figure, based on how long this implies before you start pushing etc, (though obviously there may not be statistical significance in the difference for your sample.)

Anyway, as you've said you'll look at L3-ITM, and if that doesn't show anyhting, this might then look more likley to be important.

FWIW, over 77 11s in Jan my median chip count when I mede it to level 4 was 845, and over 75 in Feb it was 770. ROI of 12 in both months.