***Official HEM Red Line Discussion Thread***

[jukofyork]

Originally Posted by Glitlr

That's a good point. HOW THE HELL DOES HE NEVER FALL BELOW THE RED LINE??!?!?!?!

A lot of people seem to have the misconception that the $EV_luck line should converge towards zero - it will not and in actual fact as n gets larger so does the chance that it will be further away from zero. What does converge though is your adjusted $EV_luck/n, or in our case our adjusted ROI.

That is a sick graph though... I thought my own lack of variance was doing good til I a saw that. The poker gods certainly are nice to some people.

Juk

[bagclip2007]

juk thanks for posting. when holdem manager started implementing the luck thing i still ran ur program to make sure because i believe it is pretty damn close.

honestly, juks program has been the only reason ive kept my sanity while playing stt's i mean i dont mind losing if i was supposed to be winning. only reason i dont use it anymore cause hem has pretty grafsss

[bagclip2007]

Originally Posted by NJD77

Simplicity, that graph makes me feel sick. You have basically never once run below expectation in 10k games. I on the other hand have never once run over expectation in 3500 games.

This thread just serves to highlight the horrible effects of variance on stts.

i think this says alot about his consistant play as well man. i mean there are times when running bad i also began to play bad which also hurt me quite a bit. i mean if u look at alot of graphs from players like him and sippin they pretty much just start consistant and go straight up.

[Glitlr]

Originally Posted by bagclip2007

i think this says alot about his consistant play as well man. i mean there are times when running bad i also began to play bad which also hurt me quite a bit. i mean if u look at alot of graphs from players like him and sippin they pretty much just start consistant and go straight up.

Players like simplicity, sippin, venetian, loon, gravy, neil_rick and others motivate me so much because they prove that if you set out on a mission it can be done.

[SlackerMcFly]

Thanks guys for all of your help. Finally got the Beta version working.

BTW, ~100k hands (small sample size) took 1.5 hours to export/import. Kinda slow, but worth it.

My graph isn't cool like Simplicity's and I'm still unsure if any of it is correct, but will keep working on it. I'm ITM over 40% of the time in STT/MTT play, but have a very small profit. That profit is nearly 12% below EV fwiw.....

Gotta put in some fresh hours and run a report based on just a few games I guess to get a feel for the reality of the program.

Thanks again. Slacka

[NJD77]

Originally Posted by bagclip2007

i think this says alot about his consistant play as well man. i mean there are times when running bad i also began to play bad which also hurt me quite a bit. i mean if u look at alot of graphs from players like him and sippin they pretty much just start consistant and go straight up.

I wouldn't for one minute suggest simplicity is not a very solid consistent player, but this exercise is about luck. How good and consistent your decisions are should have nothing to do with how lucky or unlucky you are. For him to have never once dipped below expectation in 10k games is remarkably lucky imo.

Or is there some truth in the old addage...."you make your own luck"?

[Glitlr]

Originally Posted by NJD77

I wouldn't for one minute suggest simplicity is not a very solid consistent player, but this exercise is about luck. How good and consistent your decisions are should have nothing to do with how lucky or unlucky you are. For him to have never once dipped below expectation in 10k games is remarkably lucky imo.

Or is there some truth in the old addage...."you make your own luck"?

I just think I chose the wrong SN. Should been Difficulty8. I also think his skill level makes these results more likely but even the most skilled players would drop below imo. At least, I THOUGHT SO.

[NJD77]

Originally Posted by Glitlr

I just think I chose the wrong SN. Should been Difficulty8. I also think his skill level makes these results more likely but even the most skilled players would drop below imo. At least, I THOUGHT SO.

I just don't get why his skill level would affect how lucky or unlucky he gets, surely that is just in the hands of the same poker gods who deal to us all.

[Glitlr]

I mean, obv the better a player is the less variance they will get (to some slight degree) b/c they have less bad plays mixed into their game. Everytime I make a bad 3To push (haha) it adds to my variance. But I still agree that somewhere along the line things should probably fall below expectation. coach me for free simplicity? nah, probably not

[th1986]

Obv not trying to take anything from Simplicity here , as i respect that he is a better player than me , but i do believe luck is gonna have a ton to do with a lot of the sample sizes us high volume grinders put in. Like obv 20k games at one level you'll have an idea of your true ROI but I'm sure you can be quite a few points off depending on luck. I mean , take for example this player i play quite a bit with. 14% ROI iirc in 18 mans from $16-$27 , he plays like 30 vpip and 9 pfr , which is pretty lol. And to add to that he will limp utg with like 7bb and fold to a shove , or limp/call raises bb50 from utg on like a 900 stack with 910ss. Or limp the sb with 5bb and fold to a shove. All things that point to a bad player , but he has managed to sustain a 14% roi over like 8000 games. There has to be a huge luck factor that is making this happen. Either way this inspires me to re load my database , but first I wanna make can anyone tell me if this will for sure properly work given a large portion of my database will be 18 man SNG's. Since I've start using it (about 2000 games) it's pretty accurate on the profit mark . and while i feel like I've ran like complete garbage for my last 4k games , it just doesn't seem possible for me red line to be so far off my green line. I'm gonna filter this month and figure out how to post a picture because while i have 2k games that i can graph it's pretty off profit wise for a lot of them , i might have missed importing some of making some kind of mistake of maybe it doesn't work right with 18s. BRB gonna post graph of this month.

[th1986]

Here's this month graphed in buyins

[th1986]

yeah and i guess im deciding not to reimport all my hands. i did it once this year already and than prosteg or w/e messed up and like a big dummy i freaked out and deleted it off my computer , while deleting my recently re imported database that took freaking 20 hours. so now i have like 2 months worth of hud stats but oh well. i should have a 10k game sample again within 4 months so i can report back here and see my green line ever caught up.

[angry_man]

Originally Posted by NJD77

For him to have never once dipped below expectation in 10k games is remarkably lucky imo.

Not particularly. If an unusually good (or bad) spell of, say, 1K games coincidentally occurs very near the start of the data sample, then you could almost *expect* the lines not to cross for quite a long time. I think if you saw a graph of the *difference* between the two lines, it would tell a different story.

[Hood]

thc have you even read this thread or do you just want to complain about how bad you run? There have been at least two other posts in this thread showing signficantly worse results than you. you are 80 buyins which is nasty or something below which sucks but it's not that bad. this thread is not a BBV thread, it's about how we can read this information. however if oyu do feel the data may be incomplete, if you play on stars or full tilt (i think) you can also run your data juks program and compare. Perhaps there is an issue with 18 mans.

I mean, obv the better a player is the less variance they will get (to some slight degree) b/c they have less bad plays mixed into their game. Everytime I make a bad 3To push (haha) it adds to my variance. But I still agree that somewhere along the line things should probably fall below expectation. coach me for free simplicity? nah, probably not

I think you're confusing 2 issues. If you are making mistakes and misclicks, your true ROI is lower.

Basically:

Someone with an 12% true ROI is just as likely to run at, say, Y% below expectation of X tourneys (let's say 4% below, so results of 8%) as someone with a true ROI of 2% is likely to run 4% below (so -2%).

Yes we can say - the best players don't start tilting and thus play worse and having a lower true ROI etc. But we're not discussing that - we're looking at all-in adjusted luck.

[th1986]

Originally Posted by Hood

thc have you even read this thread

for the most part , i have.

Originally Posted by Hood

or do you just want to complain about how bad you run

and i do.

thanks for posting instructions though , I'll PM you and get approval before i post next time!