PDA

View Full Version : All in EV sinks like a stone



NZfisherman
03-20-2011, 11:26 PM
my winnings on pokerstars is about -60k . My all in EV is atleast 80k out < So it reads +20k > . So basically just counting all in ev alone its making the difference between a winning player and a heavily losing player. Ive played nearly 200k hands now and I also beleive my cooler ratio has been too high and the times u get outdrawn but arnt all in by fish is amazingly off balance also .So in every sense of the word "luck" im coming up far short.

I used to mostly play 50bb max games and altho my EV graph was ever widening out of balance ( slow but pretty consistent getting diddled ) it wasnt going up so wasnt sure whether I should be winning but after including other non all in EV factors I assumed I should be making atleast a small consistent profit instead of small consistent loss I now play almost exclusively cap games and now all in EV is the predominant factor with my consistency of misfortune. Sometimes losing more than than 10k EV a day playing 10/20 and 25/50 cap and have had -23k EV loss in one day. With nowhere near the same result when I occasionally run ahead like 1/5 sessions

Does anyone else run this bad? . And does anyone run ahead or well ahead of their EV ( winning more than they expect to mathematically ). If you have 100k hands of absolute shaftage is it possible to turn a corner?. I beleive I have a winning edge in the games I choose to play in but the bottomline is eating my available funds and my confidence quickly.

Any thoughts, comments, or personal accounts are welcome as Im scratching my head and need some advice

Patvs
03-21-2011, 08:21 AM
We had a big EV bug affecting 1.11.05 -up to- 1.11.05d. If you ever imported hands into those versions of HoldemManager, please update to the latest version - http://www.holdemmanager.com/Downloads/HmUpdate.exe

Create a NEW empty database--> and reimport all your hands (by running a import from folder on your archive folder).

---


To see how bad you're running, and if it's mathematically possible, compare your winrate (in BB/100) to your EV BB/100 winrate. And read: http://forums.holdemmanager.com/manager-general/58261-interesting-find-hem-reports-proof-foul-play.html

As a cashgame player, you actually can use the Variance Simulator from evplusplus.com.

---


The "swings" you experience in 100k-200k hands are normal. Get used to them!
The trick is to adjust (moving down stakes, or experimenting playing with a different stack size) before you've lost more than half your bankroll, affecting your confidence.

ktx49
03-21-2011, 08:24 PM
hey PATVS-

if my findings are somehow invalid, then you should have no problem explaining why EV in terms of BB/100 is not relevant at SNG or Tourney tables when looking for anomalies.

the ONLY reason given before the thread was mysteriously closed was that your all in EV somehow becomes irrelevant since it doesnt matter how much or how often you lose in tourneys, it matters when you lose.

this gaffe almost went uncorrected since you locked the thread for no reason at all, but here i can finally show how laughable your response was:

here this is probably the best way to show you how absolutely retarded this statement was.....if i had shown you that very same data and told you it happened at cash game tables, what would you think then?? the data would be alarming because it would show that i am losing an unacceptable amount of all ins over a large sample, period. what we care about is EV, and it doesnt change simply because we are in a tournament setting.

....i am not worried about the timing of when i lose or win here since that would not really prove anything as far the shuffle/cheating/security/whatever....but what i CAN show is that i am losing all ins at a rate far outside the acceptable distributions shown on a simulator designed to show the next-to-impossible.

can u please explain how WHEN i am losing would be more relevant than the all in EV of these hands over large samples as far as determining something like rigged software or cheating?? i am very interested to hear this one....

and NZFISHERMAN-
let me guess, this is happening on POKERSTARS?? rofl

ktx49
03-21-2011, 08:50 PM
let me ask you this as well.......

which set of data would be more alarming and relevant in terms of determining whether a coin is fair:

A.) I flip a coin all day long tons and tons of flips all the while recording each one...at the end of the day the tally shows that the coin landed on heads 90% of the time.

B.) I flip a coin all day long tons and tons of flips all the while recording each one....at the end of the day the data shows that the coin landed on heads without fail at the top of each hour(1am, 2am, 3am, etc).

Patvs
03-21-2011, 10:47 PM
@ktx49: I have no idea why your thread was locked.
I unlocked it. So you can reply @ http://forums.holdemmanager.com/manager-general/58261-interesting-find-hem-reports-proof-foul-play.html

Patvs
03-21-2011, 11:24 PM
In SNGs, especially turbo/super turbo with fast structures, other factors screw up All-in EV:
-stacksize, when playing a push/fold strategy will result in more coinflips. If you have for example three 60-40 allins in every SNG, where you push twice with the best hand (and are called by an inferior hand).... and you call an allin once where you call with the best hand.
It's completely irrelevant you have the best hand in all three situations. You're mostly not going to survive all three all-ins.
And I'm talking about STTs, it's even worse playing (45 man) MTTs.
Though your BB/100 (and EV BB/100) will show you're playing profitable. Which is why focusing on BB/100 with a 'holecards" filter has very little meaning in SNGs.
-Winning an all-in against a shortstack doesn't make up for when you loose an all-in against a bigger stack.


The only thing that matters in STT/MTTs in a positive ROI, that can only be accomplished by finishing in 1th/2th place often enough to compensate for the many more games you will lose outside of the money. The worst that could happen to you in a 45 man MTT is winning every all-in hand, and then still finishing in 7th place.

NZfisherman
03-22-2011, 04:19 AM
thanks. I have already been through the process of downloading the latest versions and reimported all my hands a few weeks ago so there is no bug problem with my graph.
I have a good understanding of the potential swings in poker . But swings generally mean you have large lucky upswings also.
As I said I probably run ahead of my EV around 20% of the time and during those times I never gain anywhere near some of the worst days when the variance is on the bad side, especially now that I play cap games quite often my ev line rises sharply while my winnings shoot almost a vertical straight line down!!
There just seems to be no end , Ive always thought pple who think sites are sumhow biased or rigged were just fish and I still dont actually think they are (rigged) although the consistency of my so called downswing begs the question.

If I continue to try my best, learn improve, and get it in as good as I do on a daily basis I tell myself surely its time to start winning again. But sure enough it may take several thousand hands but the broll gets crippled and most of the time its not under my direct control ( the moneys in and im flipping or winning much more often than behind ).

I also know that I make more +EV plays than my average competition ie betting large enough to deny a fish the proper true and implied odds to chase his draw etc. Which in the short run doesnt mean much but im talking a pretty big sample size. Its been like this for about 150k hands now. All in ev was only one factor in non cap for me being a slight long run loser as opposed to a slight or moderate long run winner. those who understand that in non cap games there are other big factors that cause bad variance which im certain ive sustained over 100k sample ( bad cooler raito and outdraw ratio while not allin )

Does anyone have a history of running terrible for sample sizes of over 100 - 200k hands and then bouncing back with no major change in playing style or table selection??

ps - Yes I do mean stars but Ive also had the same run on tilt although its been a much smaller sample size ( about 20-25k hands ) Im not even gunna go there. I dont beleive each site makes a difference. Ive seen it all on both

Patvs
03-22-2011, 12:26 PM
Fixed Limit, Pot Limit Omaha,... and CAP games can also screw up All-in EV:
Often you will deny a fish proper odds (because you 3bet big with the best hand preflop, or make a substantial bet on the flop basically committing yourself with the best hand).

If the fish then hits his hand. The remainder of your stack will go in when you don't have the best hand. So according to All-in EV you suck and always get it in bad. This is one of the pitfalls of All-in EV. (as supposed to other EV methods like "EV by street")

(while you will fold preflop or on the flop, if you don't have the best hand, so there are less opportunities to suck out on your opponent)
Therefore, similar large upswings (like the large downswings) are more rare. It just takes playing an endless amount of hands to make up for it.

Another downside of CAP is: if you lose many buyins (because your opponents keep hitting the board), it's more difficult to make up for those losses. Since you will never have implied odds to chase a draw yourself. And even when you hit your draw, there's hardly any extra money you can win since the pot is capped.

---

100k hand downswings are normal. (so are 1000 game swings with STT/MTTs) Including bouncing back with no major change in playing style. But that can only be done if you had a super-conservative bankroll management that can handle the swings.

You should consider to experiment with stacksize, and stakes.
If you play 4x NL200. Try playing 8x NL100.
If you play CAP NL200. Try playing with a 60 BB stack at regular tables NL100.
Or play with a smaller stack, but play more tables.

You can beat variance, by actually sacrificing part of your high BB/100 winrate. By playing more tables or playing a more conversative style at lower stakes.