Determine if you are a fish or if you run bad

VPIP YES :

Normal Graph :

NL20 SH .

Stats 21/17/ 3BET-8% .

Originally posted by Desultory

Originally posted by ipeaceonu
Here's an stt graph to muse over, I have no idea if the filtered vpip can help gain any info.

Fitlered for vpip

Spoiler

Unfiltered

Spoiler

Figured you cash guys were having all the fun discussing all of this ♦

Interestingly, your red line is super super straight. Blue line is generally going in an upward trend - no obvious blueline potential downswing trend, just up and down. Its not a lot different from a theoretical long term cash graph, but I would have expected blue line to be way more variant in SNG's. I'm sorry to say that I suspect you haven't experienced the full brunt of variance just yet.

I am thinking the red line is so straight as people don't call anywhere near as wide as they should therefore I pick up a tonne of blinds/antes pre flop. I think the unfiltered line being so far below zero is due to being so nitty when stacks are deep. I am not convinced my red line has anything to do with decent post flop play

The blue line is pretty weird as when I am called I would say I am crushed pretty much 65% of the time on average as I am shoving crap and being called by strong hands.

This sample is only 1500 games anyways as I have just moved back to stt and haven't imported old games into PT4, I may import the 180 man games I played last year and see what that one looks like, interesting even if not that much use to us sng guys, thanks for the response.

Originally posted by booomm
here is my filtered graph, am I the only one with red line above blue ?

Spoiler

was wondering about the same

Originally posted by ipeaceonu
I am thinking the red line is so straight as people don't call anywhere near as wide as they should therefore I pick up a tonne of blinds/antes pre flop. I think the unfiltered line being so far below zero is due to being so nitty when stacks are deep. I am not convinced my red line has anything to do with decent post flop play

The blue line is pretty weird as when I am called I would say I am crushed pretty much 65% of the time on average as I am shoving crap and being called by strong hands.

This sample is only 1500 games anyways as I have just moved back to stt and haven't imported old games into PT4, I may import the 180 man games I played last year and see what that one looks like, interesting even if not that much use to us sng guys, thanks for the response.

All those all in: fold situations I imagine. Less non showdown variance than cash, would have thought be countered with more showdown variance.

Interesting about blue line given you think you are usually crushed when called. Surely that means blueline should have a negative gradient in SNG's?

I'm glad you posted. If any more SNGer's, MTTer's, PLO'ers want to post please do

Hmmm.. Finally a place where I can get some feedback.
I started thinking I must play really poor to have these results, but getting the VPIP filter on gave me some hope.

Filtered

Non-Filtered

NL20 SH Ipoker
Winrate: -3.37bb/100
Vpip/Pfr: 23/21
3bet: 8.5
WTSD: 24.5
W$SD%: 51.7

I would really appreciate some opinions, thank you.

NL25 Not filtered:

NL25 Filtered:

VPIP 19 PFR 17 3bet 6.5 WTSD 23 W$WSF 45 W$SD 58 AFq overall 53.5

I should read the thread again because i could not quite understand what is good and what is bad .

Originally posted by M1rCea
Hmmm.. Finally a place where I can get some feedback.
I started thinking I must play really poor to have these results, but getting the VPIP filter on gave me some hope.

Filtered

Spoiler

Non-Filtered

Spoiler

NL20 SH Ipoker
Winrate: -3.37bb/100
Vpip/Pfr: 23/21
3bet: 8.5
WTSD: 24.5
W$SD%: 51.7

I would really appreciate some opinions, thank you.

I can't really tell you much other than your gradients look pretty smooth over the course of all of your hands. Doesn't mean you haven't run bad over that period of hands, just means there is no evidence of an obvious long term sudden change in 'luck' as I was theorising.

Using my models, Your blue line and red line should both be higher on average given your W$SD and WTSD compared to other players. This may reflect invisible luck variance which I go on to talk in the next post. But equally it could absolutely mean nothing as my model is not based off enough players samples to be anywhere near confident.

I did some more regression last night and the conclusion is I don't have enough players with big samples, and likely will never have enough players with big samples to mathematically do anything useful.

The best opinion I can give you all is speculation, which quite frankly isn't very helpful.

Looking at all of my hands divided into 100k samples and 50k samples, I suspect that variance has a much bigger part to play in poker than is predicted by standard deviation. If true, what this means is that poker is way way way more variant than anyone realises. Using standard deviation models show incomprehensible variance and I'm saying that variance is even bigger than this. To what extent, I have no idea.

I'm sure some of you will have big leaks in your game and this information may be of the detriment to you because you may go away thinking your losses are purely luck when they probably aren't and will continue to play on, making the same mistakes (gamblers fallacy). But I am believing that there will be some of you who are playing similarly to the big winners out there, yet breaking even or even losing at whatever limit you may be playing at.

If you take the guys on PTR making 14BB/100 over 100k samples, using standard deviation confidence intervals states you can be 99% confident win rate is >21bb/100+ over that sample. It leads you to the belief that your true win rate is >21bb/100 but I am 99% confident through logic/experience that this will never be the case so I think StDev fails big time. On the other side of the coin, if you are -5bb/100 over 100khands, using confidence intervals you can be 99% confident that you are a true win rate losing player over that sample. I've had 100k samples of -8bb/100 playing no different strategy from the times I've had +11bb/100 so I am confident there is something involved that Stdev does not factor. I call it 'invisble luck', the amount of times you get dealt the nuts/TP vs the times Recreational player gets dealt a hand he is going to stack off with for instance.

In search of this 'invisble luck', I noticed on my negative samples that my W$SD was much lower than in my positive samples. I also noticed my blue line flattening or decreasing in my negative samples and peaking in my positive samples. My redline is generally straight. So I was thinking these specific stats might reflect invisible luck somehow. After running the stats though, I am thinking that redline also has invisble luck in it albeit less so than blueline.

Allin EV like St dev fails imo. So we as professionals are playing on without ever finding out how truly successful our current strategy is or how truly successful changes in our strategy are. Thats a worrying thought for me.

Originally posted by Desultory
I did some more regression last night and the conclusion is I don't have enough players with big samples, and likely will never have enough players with big samples to mathematically do anything useful.

The best opinion I can give you all is speculation, which quite frankly isn't very helpful.

Looking at all of my hands divided into 100k samples and 50k samples, I suspect that variance has a much bigger part to play in poker than is predicted by standard deviation. If true, what this means is that poker is way way way more variant than anyone realises. Using standard deviation models show incomprehensible variance and I'm saying that variance is even bigger than this. To what extent, I have no idea.

I'm sure some of you will have big leaks in your game and this information may be of the detriment to you because you may go away thinking your losses are purely luck when they probably aren't and will continue to play on, making the same mistakes (gamblers fallacy). But I am believing that there will be some of you who are playing similarly to the big winners out there, yet breaking even or even losing at whatever limit you may be playing at.

If you take the guys on PTR making 14BB/100 over 100k samples, using standard deviation confidence intervals states you can be 99% confident win rate is >21bb/100+ over that sample. It leads you to the belief that your true win rate is >21bb/100 but I am 99% confident through logic/experience that this will never be the case so I think StDev fails big time. On the other side of the coin, if you are -5bb/100 over 100khands, using confidence intervals you can be 99% confident that you are a true win rate losing player over that sample. I've had 100k samples of -8bb/100 playing no different strategy from the times I've had +11bb/100 so I am confident there is something involved that Stdev does not factor. I call it 'invisble luck', the amount of times you get dealt the nuts/TP vs the times Recreational player gets dealt a hand he is going to stack off with for instance.

In search of this 'invisble luck', I noticed on my negative samples that my W$SD was much lower than in my positive samples. I also noticed my blue line flattening or decreasing in my negative samples and peaking in my positive samples. My redline is generally straight. So I was thinking these specific stats might reflect invisible luck somehow. After running the stats though, I am thinking that redline also has invisble luck in it albeit less so than blueline.

Allin EV like St dev fails imo. So we as professionals are playing on without ever finding out how truly successful our current strategy is or how truly successful changes in our strategy are. Thats a worrying thought for me.

Wow... This is some heavy stuff, very interesting insights sir.. Yeah I also think variance is muuuch more invisible than currently perceived, I also see it as postflop spots u get into, setups etc. Sometimes to ur advantage, sometimes to others'.

So I guess the conclusion is focus on what you can control: work hard on ur game to minimize loses maximize winnings and put in big volume to compensate for variance.

PS. The example u gave of running -8bb/100 in 100k hands inspires me to keep pushing forward.

I was thinking about something, aren't more aggressive players (that make most of their winnings through NON-SD) less exposed to variance?
I mean they win by applying pressure and not seeing as many SD's, making opponents fold stronger hands.

Originally posted by M1rCea
I was thinking about something, aren't more aggressive players (that make most of their winnings through NON-SD) less exposed to variance?
I mean they win by applying pressure and not seeing as many SD's, making opponents fold stronger hands.

I see why you think that given what I've said in this thread, but logically if you think about it, they are exposed to more variance not less. Just because their proportion of winnings mainly comes from non showdown doesn't mean their proportion of variance decreases. Their blue line will be lower in general, but it will still reflect variance same as a higher line would.

It could result in that their redline contains more variance because of the way they play.

Not filtered:

Filtered:

NL10 & NL20

Have been trying quite aggressive laggish style lately (20k hands)

Hi,

Here are my stats for:

NL5 Speed Poker (6-Max) - 93,391 Hands:

WTSD - 31.1%
W$SD - 53.0%
WWSF - 45.6%
cbet flop% - 58.3
cbet turn% - 49.2
BB/100 - 4.88

NL10 (6-Max) - 23,886 Hands:

WTSD - 27.0%
W$SD - 53.0%
WWSF - 49.0%
cbet flop% - 58.8
cbet turn% - 61.3
BB/100 - 21.57

NL20 (6-Max) - 51,924 Hands:

WTSD - 27.5%
W$SD - 52.1%
WWSF - 45.7%
cbet flop% - 55.3
cbet turn% - 48.3
BB/100 - 5.70

I would be interested in your thoughts.

Thanks

Originally posted by Ghartman
Hi,

Here are my stats for:

NL5 Speed Poker (6-Max) - 93,391 Hands:

WTSD - 31.1%
W$SD - 53.0%
WWSF - 45.6%
cbet flop% - 58.3
cbet turn% - 49.2
BB/100 - 4.88

Spoiler

NL10 (6-Max) - 23,886 Hands:

WTSD - 27.0%
W$SD - 53.0%
WWSF - 49.0%
cbet flop% - 58.8
cbet turn% - 61.3
BB/100 - 21.57

Spoiler

NL20 (6-Max) - 51,924 Hands:

WTSD - 27.5%
W$SD - 52.1%
WWSF - 45.7%
cbet flop% - 55.3
cbet turn% - 48.3
BB/100 - 5.70

Spoiler

I would be interested in your thoughts.

Thanks

Your sample is interesting because given the stats, I have you as an outlier from the current average population. My model is predicting you to have much lower win rates closer to 0. Negatve in the first sample. Does that mean you are running good; maybe. All it actually means is that you share more common statistics with the losers in my sample than the winners in my sample. Which basically means nothing because my sample is too small.

I suspect that your 2nd and 3rd sample are simular in play style. Your 2nd sample is definitely run good imo. This may be obvious for someone just looking at the win rates/graphs but I haven't even looked at the graphs in detail. Just my models.

My model says you have improved from 1 to 3.
It says you have drastically improved from 1 to 2(76%+improved), and then very subtly improved from 2 to 3. (4%)
None of this is accurate though. Just massive estimations.

Very interesting. So basically there has been a steady improvement from samples 1- 3. The biggest improvement being from samples 1-2. Also extremely interesting that you say I share characteristics with losing players. Can you elaborate a little more on this if possible please? I've worked a lot on hand-reading and my red line and seem to have good sense for picking off bluffs. It may also interest you to know that I don't use a HUD at all, just using Hem2 for hand/stat reviews and logging my progress.

So what could/should I be working on based on the stats I have provided, do you think? I can provide more stats if necessary. Thanks a lot for your analysis.

Cheers

Does the non SD line (filtered VPIP) have meaning to your theory?

I propose: for agressive players (non SD higher than SD line) your theory should apply better for their non SD line instead of SD line

Originally posted by Ghartman
Very interesting. So basically there has been a steady improvement from samples 1- 3. The biggest improvement being from samples 1-2.

Only improvement according to my model. I want to reiterate that my model is based off almost nothing. Nothing concrete anyway. Its basically saying that your statistics are moving towards the profitable players in the sample, which if an accurate model, would mean you are improving as a player. It would kind of make sense that you improve over time - probably don't need a model telling you that but I think its the massiveness of the change between 1-2 that is interesting. If you were playing well to begin with, I suspect logically there wouldn't be such a massive change. 2-3 samples are almost the same according to my model - the 4% is negligible.

Also extremely interesting that you say I share characteristics with losing players. Can you elaborate a little more on this if possible please? I've worked a lot on hand-reading and my red line and seem to have good sense for picking off bluffs. It may also interest you to know that I don't use a HUD at all, just using Hem2 for hand/stat reviews and logging my progress. So what could/should I be working on based on the stats I have provided, do you think? I can provide more stats if necessary. Thanks a lot for your analysis. Cheers

I'm not saying you share statistics with all losing players. Just the losing players in my population sample, which again doesn't mean a great deal as the sample size for each of the players in my population is likely not big enough to determine anything.

I wouldn't want to say that I think you should change specific stats based off this, because I am not confident in it. I just thought this summation may be a good end to the thread. A kind of explanation why all this research likely means nothing. I just do it because I get a kick out of stats really.

Originally posted by patszerdonk
Does the non SD line (filtered VPIP) have meaning to your theory?

I propose: for agressive players (non SD higher than SD line) your theory should apply better for their non SD line instead of SD line

The non SD line is factored into my model yes. So it has 'meaning'. As I've said before I can only speculate that meaning and speculation means nothing really.

I did also say in an ealier post that maybe redline has more of an indication of variance in aggro players than non aggro players, but I have no statistics to support this. It seems logical given you also came to the same conclusion.

Hi Desultory,

Do you mind telling me your thoughts on my Zoom graph so far this year?

I'm a long way under EV and feels like I've run badly for a long time. I play a very loose aggressive game though so would imagine my variance would be very high?

I put a lot of study in around 60k hands, stopped raising a ton of hands in UTG + MP as well as a few from the CO. I also switched around some of the hands I was flatting to 3betting and vice versa and upped my 3bet from around 5 to 7%.

Stats were originally something like 36/30/3.25 now playing around 31/25/2.65, can give you more info if you want. Sorry to drag this up if you're not still using this thread.

Thanks

Millsy

Originally posted by Millsyuk
Do you mind telling me your thoughts on my Zoom graph so far this year?

If you look through this thread at my past answers, you will be in a position to self diagnose your own graph.

Originally posted by Millsyuk
I'm a long way under EV and feels like I've run badly for a long time. I play a very loose aggressive game though so would imagine my variance would be very high?

You cannot gain any reasonable feedback from EV. For what we are trying to figure out it means nothing. You would be better off removing it from your graphs because its just confusing for anyone trying to work out which is your actual redline, so not only is it worthless, but its a pain in the ass line.

Poker variance is incomprehensibly high. Standard deviations are no where near factoring for bigger variations, 'invisible luck' variations, as I mentioned earlier in the thread.

No matter whether you play laggro or taggro or passive or whatever. You will experience ridiculous variance swings. The common poker statement is that if you play laggro you have more variance, which makes logical sense if looking at it from a really small minded perspective. But I think, this 'relative play style' variance is completely over weighed by 'invisble luck'. The differences between a 20/17 player and a 36/30 player is only like 16%/13%. Its not a particularly different style of playing imo. Playing 70/50 would be markedly different and likely accompany much more variance. Therefore I think differences in variance of arbitrary lag and tag styles are massively overplayed.

TBH, pretty much all poker theory is overplayed and reiterated parrot fashion without enough re-thought or in depth research. Not much of it is logically questioned and this is all because of poker media convincing you that you should play a specific way to make the most money. Funnily enough the poker media is only there to try and take your money.

In your poker career, if you don't experience mass swings in both directions you are likely to continue thinking you are either better than you actually are, or worse than you actually are. All of you posting in this thread are on average likely to be on the side of thinking you are worse than you actually are. If you are even reading the thread, you can take that as positive news that you are probably not running good. Because if you were running good, you can have some pretty serious leaks and still be making making a shit load of money. I'm not talking about the 40vpip 5pfr type bleeders - their skill level seems to be so different that they lose their money pretty quickly. I am talking about the 32vpip 24pfr type players - these players with good luck can think that they are profitable and they may quit work and call themselves professional players. I should note that 32/24 could be the most profitable way of playing and I am only being dismissive of it because my logic and research has dictated that I think playing this way is bad in the current totality/mentality of today's games.

If you experience both sides of the swing, like I have, then you will truly understand what I am writing about. Until then, you will probably not believe or comprehend.

hey guys. This is my first post in English part of community. So hey all:)

I play nl10 fast on Ipoker, and I´m a little bit down. So i made filter like you say on first page.

Do you can tell me from this graph something?..ty very much