|
|
 |
>3 aka < 0.13 aka wut aka unpossible aka mathz |
 |
0
 |
sawseech Canada. Sep 12 2010 21:27. Posts 3182 | | |
I played 2057 hands of MTT sng format today with buyins less than or equal to $ 4.4 with the $ 78 that I had left on the site, in order to attempt to make my nut, seeing no recourse, since there was no way that I'd be able to make my month-end target (covering the nut) continuing to play cash because there are not enough hours in the day because I have to do things like sleep and eat and write. I'd recorded an EV bb/100 of 44 over a tiny sample of penny NL but it doesn't mean shit when that winrate amounts to $ 3.50/hour cash in hand and, besides, this level of winrate is clearly unsustainable. I figure it's good for 2/hr, maybe barely 3. Throw up some error bars and it's not unreasonable to think that I would not have even managed to get to 25nl by the end of the month, nevermind make the nut, and even in that case I'm just making my nut and starting all over, and this further requires that I assume that I run not terribly at any point, which is really a fool's wager.
And so today I played micro 180s with a buyin of equal or less than $ 4.4 and I lost an average of 27.71 chips per hand.
Assume an average bb size of 100, mean median mode all pretty much agree with this.
My cEV for the 2057 hands was 5122.07. In turn, my expected result per hand can cautiously be interpreted to be 2.49 chips per hand. Sample size is an issue here, of course.
My actual EV bb/100 for 2057 hands was 12.36, clustered around the lower bb levels of course. This serves to imply that my true EV bb/100 is somewhere in between the two figures.
My realized bb/100 for 2057 hands was -0.48.
At an average bb size of 100, my linear count expectation of 2.49 chips per hand, a fairly strong though mildly sensitive figure, this constitutes an average expected gain gain of 249 chips per 100 hands, or 2.49 bb/100.
Start playing around with the numbers between 2.49 and 12.36 and we see an average maximal loss in bet count in the range of 430-520 , assuming that we are bounded by a Z-Score of 3 in trials. I have no reason to go below the expected figure of 2.49 bb/100 for this sample of 2057 hands because I have access to the remainder of my tournament data as it relates to tournament hand performance over a reasonably large sample.
And so at a true bb of 2.49-8 we see an expected maximal loss in the range of 500 with reasonable error bars over a sample size of 2057 hands.
Given that I am more present, more focused, and more skilled (or better able to make use of my skill) in my present condition than for the majority of my poker playing history, as evidenced by my ability to make this post, let's give me a true winrate in tourney chips of 3.8 bb/100, which I do not consider to be at all unreasonable given my historical per hand performance and ROI in all MTT formats over the course of my lifetime, which of course includes the past year. Even accounting for the past year, my lifetime ROI in MTT formats is comfortably in excess of 100. The SD of around 40 sounds pretty much spot on, I don't even need to check it. I plug it into the Z-Score of 3 bounded simulator which can be found here and instruct it to run it 1000x and I do it over and over and over.
At an average bb size of 100, I lost 57000 tournament chips for loss of 570 bets. Assume that there is no drop off in the quality of my play as the blind levels increase. In fact, I have traditionally excelled in late tournament situations, basically winning the tournament whenever I get ahold of any sizable number of chips. A performance far in excess of ICM interpretation, and in these games, I can take it as a given that I will outperform my opponents to a massive degree at every stage of play.
Give me a true winrate of 3 and I'm still in excess of a Z-Score of 3 by any reasonable statistical measure. OK, let's start to tinker with the average bb size. Still > 3, or at most spot on 3. Give me a true winrate in the range of 6 and the same applies. In order for me to exceed that Z-Score, my true bb/100 in MTT formats must be well in excess of 8.
I believe that as I played today and given the relative quality of my opponents' play, I exceeded the level of quality of play required to achieve that winrate.
Incidentally a loss of 570 adjusted bets, a very highly reasonable and robust figure in my view, amounts to a Z-Score comfortably in excess of 3 standard deviations given 1000 trials with a SD of 39.58. I ran the trials over and over and returned a similar result every time. In order for my loss to not be in excess of a Z-Score of 3 as evaluated in terms of basic statistics, I must be both the worst latestage MTT player in the world, and also the best early stage MTT player in the world. This fails the laugh test.
Now let's look at longest breakeven probability. Again, this tends to cap out at around 2000 hands, and given how latestage MTT goes (basically well or, uh, not) it's not really worth much over 2057 hands. Even so, again, I'm right around a Z-Score of 3 and even slightly in excess of it. Given that I'm going to be winning chips early, winning chips midstage, and winning chips late, and winning tournaments, I tend to believe the Z-Score on this too. Tell me how I'm wrong please if you are reading this and are of the inclination to disagree.
Additionaly consider that I didn't actually get to the endgame, which is where the real money is made. Adjust for that, and my cEV performance and my lack of opportunity to even have an opportunity to take in 30% of the prizepool and the actuality of my losing 570 adjusted bets becomes ... frankly startling to me.
All of this happened exactly today, which is also the day on which I have realized all of myself, and on the day that I consider myself to have played the best tournament poker that I have ever played in my life. Any way that you want to twist the data, it still comes out with the same Z-Score, or even larger, and this manifests itself in multiple terms.
Give me an ROI of >180%, for instance as accounted for by the impossible softness of the games and any reasonable cEV winrate and consistent form throughout, and then play 37 games, and return an ROI of -88%.
Give me an ITM of ~14 and then take me 25 games before my first cash, which was a mincash, and then give me another mincash at game 35.
-35 buyins for the day.
4.52 in the roll.
And so I'm going to see about getting my old job back tomorrow, a job that I must admit to find rather boring (but do really enjoy enough to not feel that it's a waste of my time), because when I work there I'm helping to keep people safe and teaching and helping others, and as I write this I must further admit that I am very seriously considering never playing on stars again. Except for the freerolls and the quarterly millions. I may have developed major negative feelings towards the site as a result of my experience today but I'm not stupid.
|
|
| lets go fucking mental la la la la lets go fucking mental lets go fucking mental lala la la | Last edit: 13/09/2010 00:21 |
|
| 1
 |
exalted United States. Sep 12 2010 22:18. Posts 2918 | | | |
|
| exalted from teamliquid :o | |
|
| 1
 |
bigredhoss Cook Islands. Sep 12 2010 23:10. Posts 8649 | | |
man....
i usually like your writing style but you could've made this WAY simpler to read lol
what's your old job if you don't mind sharing? |
| |
|
| 0
|
sawseech Canada. Sep 12 2010 23:11. Posts 3182 | | |
there is no way to make this simpler without losing information |
|
| lets go fucking mental la la la la lets go fucking mental lets go fucking mental lala la la | |
|
| 0
|
sawseech Canada. Sep 12 2010 23:16. Posts 3182 | | |
| | On September 12 2010 22:10 bigredhoss wrote:
man....
i usually like your writing style but you could've made this WAY simpler to read lol
what's your old job if you don't mind sharing? |
lifeguard, swim instructor, ain't done it in like 10 years, was very good at it tho. pretty boring job but helping kids learn to swim, to see the gain in confidence, to share that sense of pride with them, i found it to be rewarding at the time and i don't see why that would change. i pounded out ridiculously heavy hours and just rebuilt from the first time i went busto (no not poker, computer programming course that I finished but couldn't pay for and then found out that grinding code all day is fucking agony for me, nevermind being entry class and debugging the code of idiots as you try to figure out what this fucking retard is trying to do in the most ass-backwards and retarded way possible).
young girls are funny, i think a few times i think they pretended to be drowning on purpose, at least subconsciously planning it that way, just so that i could "save" them. i ain't mind, i know what i do to females.
young kids universally piss in the baby pool. it's disgusting. i just stopped smelling piss on myself after a while. |
|
| lets go fucking mental la la la la lets go fucking mental lets go fucking mental lala la la | |
|
| 1
|
bigredhoss Cook Islands. Sep 12 2010 23:18. Posts 8649 | | |
well, i read all of it and i still have no f clue what a z-score is
omfg just noticed your signature rofl |
| |
|
| 0
|
sawseech Canada. Sep 12 2010 23:24. Posts 3182 | | |
| | On September 12 2010 21:18 exalted wrote:
man what the FUCK |
the fuck u using that language in my blog for bitch. gtfo. i ain't make u open the page and i certainly ain't ask you to read it. did i hurt your socially retarded brain with my blog post? if so, good. |
|
| lets go fucking mental la la la la lets go fucking mental lets go fucking mental lala la la | |
|
| 4
 |
PoorUser  United States. Sep 12 2010 23:34. Posts 7472 | | |
| | On September 12 2010 22:18 bigredhoss wrote:
well, i read all of it and i still have no f clue what a z-score is
|
indicates where you fall on the sd graph |
| |
|
| 0
|
sawseech Canada. Sep 12 2010 23:46. Posts 3182 | | |
| | On September 12 2010 22:34 PoorUser wrote:
Show nested quote +
On September 12 2010 22:18 bigredhoss wrote:
well, i read all of it and i still have no f clue what a z-score is
|
indicates where you fall on the sd graph
|
oh pshaw we all know where you fall on the sd graff, stealth boast obv |
|
| lets go fucking mental la la la la lets go fucking mental lets go fucking mental lala la la | |
|
| 1
|
bigredhoss Cook Islands. Sep 12 2010 23:46. Posts 8649 | | | |
| |
|
| 0
|
sawseech Canada. Sep 12 2010 23:47. Posts 3182 | | |
muthafuckas should just rename the zscore > 3 pooruser land |
|
| lets go fucking mental la la la la lets go fucking mental lets go fucking mental lala la la | |
|
| 0
|
sawseech Canada. Sep 12 2010 23:53. Posts 3182 | | | |
|
| lets go fucking mental la la la la lets go fucking mental lets go fucking mental lala la la | |
|
| 1 | |
It really does seem like you have mental issues. |
|
| 1
|
Funktion Australia. Sep 13 2010 04:30. Posts 1638 | | |
How did your mean, median and mode all come to pretty much 27.71? I can see the mean being this but not the median and def not the mode? Also wouldn't you calculate this shit per level and not overall per hands? Deep run or bust early are going to skew your mean. Actually I really don't know which of the mean or median gives you good relevant information in this example. Mostly because I don't see the use of win/loss per hand using total hands. I skimmed the rest because I can't follow it basically (I have never done anything with error bars though) but I didn't see where you converted your own info to Z-scores?
I don't understand why you included that graph when none of your own calculations are on it? For anyone who hasn't seen the graph before it's about the most complicated way of displaying a bell curve with every possible way of writing what each number/figure represents ie/ numerous lines/95%/1.98std/+2std/etc.
|
|
| 1
|
Funktion Australia. Sep 13 2010 04:38. Posts 1638 | | |
| | On September 12 2010 22:18 bigredhoss wrote:
well, i read all of it and i still have no f clue what a z-score is
omfg just noticed your signature rofl |
It is just a fancy math way of standardising scores. Further to what pooruser said it's mostly used to give data relevance. It basically means that you can compare apples to oranges by converting each to a standard measure and then comparing the 2 scores. |
|
| |
|
|
 Poker Streams | |
|
Contact Users: 950 Active, 0 Logged in - Time: 21:58
,
,
,
,