[Closed] NL4 Party (almost) has me defeated - advice? [Stats&Graphs inside]

Some people enjoy the humiliation. But there are clubs in Amsterdam for that sort of thing, and I'm sure if that's what OP wanted he wouldn't be asking for it here.

I am new to these forums (1 week) so if what I am saying is teaching my grandmother to suck eggs, please forgive me.

I can relate to what you're saying. It feels like shit when you keep losing like that.

First off, I usually try to find ways to make myself feel better.
You asked at one point is it variance?
Have you run the numbers on that to find out to what extent it could be?
I usually find that helps me feel better.

Look at your standard deviation, and at what is your win rate + or - 3 standard deviations. That gives you the bounds of 99.8% probability of your win rate. ( PT gives you the SD but not the confidence level by %, I don't know about HM etc. But anyway its easily done on excel with just the wr and teh SD). From all that you can work out the probability that your win rate is actually positive and that this is a downswing.
However, even that model is flawed and that fools a lot of people - as Economics Nobel Memorial Prizewinners Scholes and Merton foundout to the cost of billions ( LTCM) and as the flood defence planners for New Orleans found out six years ago today ( The Dutch protect Amsterdam much better and are not fooled by the normal distribution).
What fooled them is that it only holds as long as the base assumptions/conditions don't change. And in poker they usually do - but often for the better. Over the course of a few thousand hands your game may improve so your "true" winrate may improve but unless you can freeze all the conditions and play zillions of hands without learning anything new you can't accurately measure it.
Another change factor is something like a change in the opponent demographics - eg a black Friday changing the type of opponents. Or perhaps some marketing campaign for the site bringing a sudden influx of fish.. none of which the normal distribution takes into account. Some of these might improve it, some decrease it.

But in general look at your probability envelope, see what the probability is of it being positive, and maybe adjust it for known factors. Take that as a rough estimate, but don't treat it as gospel. Then you have a rough idea as to your probability of a downswing.

If you like reading, read Fooled by Randomness by Nassim Taleb. His other book The Black Swan is good too.

Apologies again if you know and think all this stuff already - this exercise helps me when I get these kind of situations and I find I beat myself up less as a result of rereminding myself of this. The intensity of the situation can lead me to forget.

Keep Calm and Carry on. Good luck!

[ ] decent sample size

You really shouldn't be adjusting a strategy you are sure works just because you have lost over less than 3k hands......

You should understand that bigger swings than this are an integral part of the game, no matter your skill level.

gl sir

yeah that's true but it's not just losing, it's so consistent that something's bound to be wrong. I did experience some ridiculous suckouts, but was certainly not playing the optimal.

My strategy on there had some good elements (isolating fish with high cards IP, valuebetting thinly, torturing regs) and some that were really bad (3betting OOP with no FE, levelling myself vs. fish, playing to loose), I have to adjust to reduced Fold Equity, among some other things. I'm on it.

Cheers.

Originally posted by CountBrass
Some people enjoy the humiliation. But there are clubs in Amsterdam for that sort of thing, and I'm sure if that's what OP wanted he wouldn't be asking for it here.

I am new to these forums (1 week) so if what I am saying is teaching my grandmother to suck eggs, please forgive me.

I can relate to what you're saying. It feels like shit when you keep losing like that.

First off, I usually try to find ways to make myself feel better.
You asked at one point is it variance?
Have you run the numbers on that to find out to what extent it could be?
I usually find that helps me feel better.

Look at your standard deviation, and at what is your win rate + or - 3 standard deviations. That gives you the bounds of 99.8% probability of your win rate. ( PT gives you the SD but not the confidence level by %, I don't know about HM etc. But anyway its easily done on excel with just the wr and teh SD). From all that you can work out the probability that your win rate is actually positive and that this is a downswing.
However, even that model is flawed and that fools a lot of people - as Economics Nobel Memorial Prizewinners Scholes and Merton foundout to the cost of billions ( LTCM) and as the flood defence planners for New Orleans found out six years ago today ( The Dutch protect Amsterdam much better and are not fooled by the normal distribution).
What fooled them is that it only holds as long as the base assumptions/conditions don't change. And in poker they usually do - but often for the better. Over the course of a few thousand hands your game may improve so your "true" winrate may improve but unless you can freeze all the conditions and play zillions of hands without learning anything new you can't accurately measure it.
Another change factor is something like a change in the opponent demographics - eg a black Friday changing the type of opponents. Or perhaps some marketing campaign for the site bringing a sudden influx of fish.. none of which the normal distribution takes into account. Some of these might improve it, some decrease it.

But in general look at your probability envelope, see what the probability is of it being positive, and maybe adjust it for known factors. Take that as a rough estimate, but don't treat it as gospel. Then you have a rough idea as to your probability of a downswing.

If you like reading, read Fooled by Randomness by Nassim Taleb. His other book The Black Swan is good too.

Apologies again if you know and think all this stuff already - this exercise helps me when I get these kind of situations and I find I beat myself up less as a result of rereminding myself of this. The intensity of the situation can lead me to forget.

Keep Calm and Carry on. Good luck!

Wow that's great mate, you've defo put the effort in to make a fish feel better. ♦ I'll invite you to friends so we can talk about the books you mentioned.

Thanks a lot!

EDIT:
PS. I do stats at Uni so I'm familiar with those figures, although I've never applied them to poker. I might go ahead and do it with some of my graphs. I use HM so it doesn't give me SD (or so I think), but it ain't hard to calculate it.
Like I said i did experience some harsh variance in that sample, but the losing graph is undoubtedly caused by other factors as well - such as bad play in a form I described in the previous post.
Anyways cheers mate.

Yeah actually I made an error in that post. I did stats at uni too but am a bit rusty, lol.

Look at your standard deviation, and at what is your win rate + or - 3 standard deviations. That gives you the bounds of 99.8% probability of your win rate.

Its the variance in the true mean from the measured one that you need, not a simple addition of the sd to the mean.
which is in excel:

upper limit = wr - normsinv[(1-c)/2]*sd*[1/sqrt(n/100)]

lower limit = wr + normsinv[(1-c)/2]*sd*[1/sqrt(n/100)]

Where wr is your BB/100
c is your confidence level ( eg 0.99 or 99% )
n is the number of hands.
sd is your sd/100.
evplusplus have some variance tools, I also saw something there that said how you can find your sd in HM. Though like in PT it seems well hidden.

Yes, I prefer to think that I have a 99% probability of a win rate between 18 and -28 BB/100 than to just think of a win rate of -5.1 - assuming that I am no better than when I started, and my opponents are exactly the same. Meaning I take it with a large pinch of salt.

cheers

that's great mate, thanks.

I've read up on variance-related issues in Mathematics Of Poker by Bill Chen and Jerrod Ankenman and they give you probabilities of some winrates over a given amount of hands, and I think that to calculate that they use some assumed variance per hand (or sd? ♦ seems I need to dust off my Andy Field Statistics book) and some assumed winrate that hero is able to achieve on average. It's on limit poker (hence the calculations revolve around BigBets/100) but the message is clear nonetheless - poker is a high-variance game.

By the way, Mathematics Of Poker is a G-R-E-A-T read.

EDIT: PS. I'll defo have a go on those equations you've given me. And the website! I'm already looking it up. Great stuff.

You are hemorrhaging money from the blinds. I'm not sure if this is due to the small sample, but I suspect that you are losing a lot of money on unprofitable blind defenses. If you can't recover at least half of the blinds you post, there is a problem, but much of the recovery occurs when you are dealt a premium hand in the blinds.

By the way, the standard deviation for 6-max NLHE depends a lot on your playing style, the effective stack depth, and how your opponents play, but some players have a standard deviation of about 90 big blinds/100 hands.

This is odd how theres tons of thesekind of topics in here from Party players. I have problems beating party micros too. Maybe pokerstrategy has overfloated the party micros with the free 50$ and theres tons of starting TAG´s and we are just about breakeven there, getting just enough money from fish to pay the rake ♦

I've been playing ~28/22 on NL5 SH and winning at 16bb/100 over 12K or so.

I definatly think the more LAGy side of TAG is optimal, the tables at stars are filled with so many nits from what I can see and the TAGs nit up once 3bet and no one adjusts all day.

Bet/3Bet/Cbet/Barrel All day vs nits/tags (=non-showdown winnings)

Bet/3bet/Cbet/Barrel Widely for value all day vs maniacs/fish etc. (=showdown winnings)

100% unbalanced lines and betsizing. (pre and post flop)

Abuse the BTN like crazy - your btn PFR could be doubled , opening 60%+ is good.

Go out of you way to play as many hands against the fish as possible. (eg. hero is UTG , tightish players behind and fish in the blinds go ahead and open K9s etc)

Isolate limpers likes its going out of fashion.

Play the opponent and you can do what you want really imo.

yea i got extremely confused recently. i moved over from stars to party after building up enough of a br to be rolled for 4nl. so i moved over to get the 100% deposit bonus on party and i started taking shots at 4nl and i just got taken to the cleaners. it was so different to 2nl which i had gotten used to and whenever i thought i was getting my money in good i get folds. and the rare times it's not folded around it is because the opponent has me crushed. it's extremely nitty and i've actually had to move back down to 2nl where i seem to run well - not fun.

on a plus side it has made me study a lot more. not exactly sure what to do about this tho - i'm thinking about going back over to stars.

Play 50k more hands theb come back for help.

As a "solid" player you should know that your sample = nothing.

great advice, @CheckBop.

Cheers @DrDunne, indeed every limit is unique and has different spots where we make money.

@Rainbow Ponny - I do realize that what I've given there is not big enough of a sample, but losing so consistently seemed a bit odd to me, as even in my break-even-bad-TAG days I haven't experienced anything like it before (i played about 350k hands as the bad TAG).
And I was right - it was abnormal - the consistency of those loses can be attributed to more than just variance (mainly bad play on my part), me and others have pointed them out repeatedly in our posts (if you'd read them you'd know).

Cheers for all the responses
I'm on a break from cash now, grinding SNGs on Stars and enjoying an improving mindset!

Matt

Originally posted by METALzx
You play too loose for this level. Nowadays something between 10/12 - 16/14 might be the best for nl4-5sh I guess. Fold more

I can't disagree more with you. I played 24/21 and had much more better results than 16/13, however I play on Ipoker.

Of course it always depends on table, on one 10/10 is perfect and on another 27/22.

cheers

I had the same problem as you like about a year ago. First of all what Checkbop said is pretty much how i ended up playing myself and managed to win at a consistent rate of 10bb/100 (i remember some crazy nights where i would make 12bi+ in a session). But the thing is that you dont want to start of like that immediately taking in consideration where you are coming from atm. You want to get there by following a "process".
To start off i'd say tighten up the dif between your VPIP and PFR, and have something like 20vpip18pfr, and play solid poker (if you ever come to a point where you feel questionable about plays you are doing you should stop and clear your doubts and then keep playing). Once you feel comfortable with your play and really feel by your own account that you are playing very solid THEN you should loosen up and start seeking for extra edges over your exploitable NL4 opponents.
Many are saying that 3k is a small sample, it is, but if you loose more money than you'd ever expect in only a "small" sample then obviously something is wrong.