[Closed] 1330 games , Should I be worried of my red line ?

Very interesting information guys.
So ... If I play in a lose table with many players who go all in and I don't call , it makes my EV line go down a bit ?

Also , If EV line was going up... and then for a long period of games its flat.. what does it mean about the play style ?! also , how many games are considered "sample size" in EV line direction ? I guess when I play 200-300 games in two days and I see EV goes down or BE it means NOTHING and I shouldnt be worried I need to fix my play untill I have like.... 3000k games ? 4k games ?

If you make good plays which do not involve getting all-in, such as stealing blinds from players who can't call very often, then both your red line and green line go up by the same amount. If you make bad plays which do not involve getting all-in, then both go down by the same amount.

When your EV line is going up, then it suggests that you are outplaying your opponents by more than the rake. When your EV graph is flat, this suggests that you are outplaying your opponents, but only by enough to make back the rake.

Adjusting for all-in luck is better than just looking at your results. It magnifies the size of your sample, perhaps by a factor of 3-4 for super turbos. If you play 300 tournaments, the luck-adjusted result is about as reliable as the unadjusted result after 1000 tournaments: Not perfectly accurate, but perhaps within 5%, and very likely within 10%.

huh, I'm also a bit worried about mine...

this graph are my last few months, since I had to restart my database in June or so. It includes 3249 $16s, 341 $27s and 180 $38s. I start with both lines going way up but now it's breakeven kingdom for both, with big oscillations considering my average buy in. Could this also be some kind of variance, or did I unlearn how to play

?

You know it really wouldn't be out of the question for those clever PT3 and/or HEM people to actually do the Nash ICM calculation for each hand and work out your EV that way. Then, if you push KK into AA blind versus blind, that would be a +EV play, even though it was -EV against his actual cards.

Originally posted by jbpatzer
You know it really wouldn't be out of the question for those clever PT3 and/or HEM people to actually do the Nash ICM calculation for each hand and work out your EV that way. Then, if you push KK into AA blind versus blind, that would be a +EV play, even though it was -EV against his actual cards.

Problem is that here is EV per tourney.. not per hand... no ?!

Originally posted by FWKanobi

Originally posted by jbpatzer
You know it really wouldn't be out of the question for those clever PT3 and/or HEM people to actually do the Nash ICM calculation for each hand and work out your EV that way. Then, if you push KK into AA blind versus blind, that would be a +EV play, even though it was -EV against his actual cards.

Problem is that here is EV per tourney.. not per hand... no ?!

well they could average it or something. I don't know. If they hired me as a consultant, I'd sort it out for them!

Originally posted by pzhon
If you make good plays which do not involve getting all-in, such as stealing blinds from players who can't call very often, then both your red line and green line go up by the same amount. If you make bad plays which do not involve getting all-in, then both go down by the same amount.

When your EV line is going up, then it suggests that you are outplaying your opponents by more than the rake. When your EV graph is flat, this suggests that you are outplaying your opponents, but only by enough to make back the rake.

Hey Pzhon could you elaborate on that?

What do you mean by good and bad plays? good in terms of succesful steal?

But what then are bad plays? Something like raise/fold?

It just sound strange that HM recognizes good and bad plays.

HEM doesn't plot the green line or red line from hand to hand, only at the end of the tournament, but you can imagine that it does.

When you gain equity, by winning a pot or by having some opponents collide, your equity increases, and you can imagine that your green line increases. If a player was all-in, particularly you, then there is an adjustment made to the difference between the red and green lines. If no player was all-in, then there is no adjustment made.

Good plays do not always work, but they are the plays which tend to make your equity go up on average.

To understand the red line its helpfull to just look at a complete tournament in hm (results -> dataview). As pzhon already mentioned, hm calculates the $ev for every all-in hand, adds these values up and finally adds these to your winnings in this tournament. The $ev-roi isn't your "real" roi, but it differs much less from it than your actual roi (shown with the green line) and so it reduces the needed samplesize to draw conclusions about your "real" roi.
You just have to be aware that theres still lots of variance in the red line (setups, villain wakes up with the topend of his range after a shove,....)

First 200 games since I started focussing only on SNGs again. All 4-tabling $6.50s. I'm looking to increase my volume by playing more tables, should this be safe enough to do with my red line going up? I don't want to make my downswing worse if I have major leaks, but if it is only a variance based swing I want to put more volume to overcome it.

That is a healthy-looking red line, showing about a 10% ROI adjusted. That is not yet strong statistical evidence that you are playing winning poker, but it is a good start.

Although the red line should be unbiased in one sense, and a more accurate indicator than the green line, the combination of the two may give you even more information. Usually, the red line and green line are correlated, so the red line is biased in the direction of the green line, due to uncorrected luck which affects both such as coolers and whether a draw hits on the turn when you are not all-in. Also, when you win all-in earlier in a tournament, this gives you more opportunities to use your skill advantage later in the tournament. Even if your play is exactly the same on winning days and on losing days, your red line will tend to be lower on losing days. So, seeing a high red line despite the losing green line should be quite reassuring.

Be careful that HEM has some bugs. One is a miscalculation of the all-in luck in multitable SNGs, which seems to lead to a red line which is unrealistically high. However, if these are single table tournaments then the red line should be reliable.

Ok pzhon, since u've mentioned MTT SnGs explicitly, here's a graph from one of my MTT SnG periods:

MTT SnG BIs Graph (sorry, still dunno how to upload the image directly ...)

How would these lines compare to single table SnGs? What kind of information can u gather out of this about my actual and adjusted ROI?

If the MTT all-in luck adjustment is still severely biased and not based on MTT equity at all, then I think you should ignore it.

I have seen many STT graphs where the green line is above the red line. I have not seen any such graphs for MTTs, and every time I have drilled into the MTT luck evaluation, it has been based on mistaken calculations which pretend that you are in a 50-30-20 STT. I think it is possible that there are biases introduced when tables are combined, or when players are added or removed, since these do not occur in STTs and many simple parsing tricks will break down out of context. I do not understand the bias completely, but it doesn't seem to be worth worrying about it until the MTT luck evaluations are based on MTT equity and not the bizarre assumption that you are in a STT.

What you might do is use the luck estimates for individual hands to get a better idea of whether you are winning or losing chips in particular phases of the tournament. Early in MTTs, gaining chips more closely corresponds to gaining equity than in STTs. If you are losing chips from the HJ on average, this is a sign that something is wrong, even though in theory you could be losing chips but gaining equity. The luck adjusted chip gains/losses should be slightly more accurate than the raw totals from each position, or at each blind level.

Thanks for the reply pzhon.

I've never given the red line much attention for MTT SnGs, in general they are considerably more complex than STTs which I assumed was also the reason why the red line of HEM is pointless for MTT SnGs.