Since it is impossible to have a ROI of less than 100% the red line shouldn't drop more than 1BI. I hate when I see my red line droping something like 22$ in a 6.5$ SNG. How is it possible for me to lose that much?

The same way that it is possible for you to win more than there is for 1st place... ducy?

But the red line goes up more than 3.5BIs?

Also in HU you cant lose money so every play where even if your EV- you win some small amount of $. Like you are freerolling for 1st so if you have 1% of winning where the prize is 100$ you should win 1$, don't know if it is already like this.

It is possible because you can suck out multiple times.

Say you are playing a $100 SNG

Assume somehow magically you move all in and get called and you only have 1% equity
So after luck adjusted you would only have about $2 equity in a $100 SNG.

You hit that 1% and double up and now have $180 equity. So ev difference is $2-$180 = -$178

This is more than a buyin, no?
And you can also suck out again in later time to have that ev line drop further.

It is impossible to have a ROI of less than 100%

.

You cant lose 7$ in a 5$ SNG

In real money it isn't possible.
However it is possible to lose $7 all in ev in a $5 SNG.

The redline only tracks your allins. If youre getting really lucky in allins and then losing a bunch of chips without being allin, your redline will and should drop lower than a bi. I think the main problem a lot of people seem to have with the redline is they think it isnt "fair". Its not fair that when your AA holds up vs 72 your AIEV is less than your actual gained equity. Its not fair that you can lose more than a bis worth of equity in a game. This is a really silly attitude to have. AIEV has been analyzed by a lot of great statistical minds, and discounting for the limitations of ICM and mtsngs, its gonna be accurate. Trying to poke holes in the redline because things seem counterintuitive to you is a waste of your time.

I asked a question about the red line a while ago and this was part of HEM's response:

"Also, what's special about Luck Adjusted Winnings in HEM is the $Won can be higher than what you can actually win, and lower than the actual buyin. In a $10 DoN, it's possible your $Won is $30 (although you can ONLY win $20) and it could be -$20 when you can ONLY lose $10. This may seem strange especially in DoNs. But it will result in a more accurate EV line in the long-run."

They're aware the red line can drop (or go up) more than 1 buy-in/tourney, and they seem to like it that way.

Given the fact that pretty much everyone here posts graphs where their red line is way above their green line, could you imagine how much higher those red lines would be if the potential -$EV was limited to just 1 buy-in per tournament? The general consensus is that the red line does a good job of reflecting true $EV over a large sample (and also that it coincides with the green line over a truly significant sample) despite the potential to lose well beyond the buy-in amount in $EV for a single tournament.

Now, if you're citing the fact that you lost $22 in $EV for a $6 sng as a reason to show that the red line could not possibly be accurate in the long run, you are wrong. Here's why:

It is a zero-sum game. If you lose 3 buy-ins worth of red line equity in a single hand, your luck cost the remaining players at the table a cumulative 3 buy-ins in equity (their red lines, added together, will be 3 buy-ins higher than their green lines for that hand). If you want to complain about having your red line go down 3 buy-ins when you can actually only lose 1, then you also have to complain about the red line equity you GAIN when an opponent sucks out and his red line goes down 3 buy-ins (because it means that your red line goes up a disproportionate amount by your logic).

If you are a winning player, you will be in a position to have your red line plummet far less often than the combined situations of your opponents. You might not see their red line spew massive $EV to you because it is divided amongst you and the other players at the table, but it happens far more often than you losing a massive amount of $EV. If you aren't good, you will have a short stack on the bubble quite often and both your red and your green lines will ultimately reflect that lack of skill.


p.s. For those who responded to his question by saying that it is theoretically possible for the red line to go above the 1st place payout, that would never ever ever balance out the times when one's red line drops 3 buy-ins for a single 50/30/20 tournament (and I still doubt it even for DoNs). While it does depend on the structure, I have fewer than 10 instances in 75,000 tournaments (50/30/20 payout) where my red line went above 1st place $.

thanks for your response, my graph is way above the red line anyway, but that tilts me

I been playing the new Fifty50 format and up untill the other day HM has been doing the ICM cals as if the games were a standard 9 man (50/30/20). They have updated this to the correct structure in the lastest build (60/10/10/10/10).

Ive updated to this newest version, created a new database and reimported all hands (ive actually done this twice cos i thought there was a problem the 1st time).

After 5k games the EV difference between the old 50/30/20 and new 60/10/10/10/10 is like $100. I play $50's and $100 mostly. Plus they both look pretty much exactly the same shape wise.

Does anyone think this is normal? Theres gotta be something going wrong right?

wayneking7,

That does sound suspicious that they're that close. But 60/10/10/10/10 isn't really correct either is it? Maybe there's no better way to model it since the payouts always vary though.

60/10/10/10/10 is correct.

Two prize pools, the first 10/10/10/10/10 and the second effectively 50/0/0/0/0. Since both are linear they can be added. See the Fifty50 strategy thread if u want a better explanation.

Ok I'll check it out.

Ok I see it's using chip EV for half the prize pool. The 60-10... thing threw me off, but yes that makes sense.

I don't have an answer and am curious what others think as well.

I am willing to prop bet over my next 10k games at 6-9 non turbos that I will run above my EV line on HEM, any takers?

what kind of odds will you give me?

Thank you guys in advance for your help.

I just want to know how do you interpret my redline compare to green in terms of :

1) What am I doing wrong for it to be breakeven ( i know it is a small sample ), what mistakes/leaks can cause that from your experience.
2) What spots should I check to improve it.
3) Should I be worried in general.

Obviously graph says I am running pretty good but red line really worries me.

I am reviewing my games everyday and I can`t find a lot of bad allins. They all seems absolutely fine to me and to sng wiz. I found out that I am getting called a lot when I am in desperation mode and shoving very wide range from SB. Maybe thats the case and i should adjust it/ tighten up a bit ? Basically what would you do I your graph would look like mine ?

Any advice is appreciated.

just ignore it imo.

Even money up to $500 if you want anything $1k-2k 3:2, 3k+ 2:1

@ Riven: You should ignore it until you have a not-insignificant sample. Assuming you're playing turbo/non-turbo tournament structures, you're still going to need at least 2-3k games for your red line to give you a good indication. Variance comes in a lot of forms, and the ones that are not directly quantified by the red line still effect the red line.

Guys, can someone explain me, how the red line would be accurate?

The red line doesn't track my winnings if the player doesn't call my shove, they are neutral. 0 $ev. So shoving on a 600 blind and get a fold is 0 $ev?

But the red line does track my all in results, which obviosly most of the time are at a disavantage. So am i always losing money?

Does anyone know how you would go about figuring out your luck neutral ROI using chipev? Would you weight 3 and 2 handed play more heavily since thats where the most equity is gained and lost?

I have a 1.5% ROI after ~1250 games. Heres what my stats look like.

Chips won / Luck adjusted chips won

Overall 118k / 159k
6 players 2k / 22k
5 42k / 24k
4 2k / 17k
3 82k / 40k
2 -10k / 56k


So just looking for any insights you math guys can give. I realize not all 3 handed situations are gonna have similar bubble factors, and the sample is pretty small to start with. Just trying to figure out what my roi will be in these and dont want to wait 10ks for that. thanks

If you get a fold, your red line isn't 0; it goes up. It goes up by the same amount that your green line goes up. The number that you are looking at is the DIFFERENCE between the red and green line.

Anyone having problems with F50's and the red line? No matter how bad I run at times, my green line is almost always above my red line, and it's been like that for over 5k games.