Tournament Poker Poker Forums

So true. I feel like either Holdem manager calculates it wrong and I'm living in illusion I'm doing everything right, I'm just running like megadoo-doo or the poker site f*cks with me (...).

Right now according to HEM I have played 612 6max superturbo sitngos with EV (ev roi 19%) of +118 buy-ins and results being -21 buy-ins. The EV graph is pretty nice looking, quite straight for that sample size. I can post the graph here if it is possible. I have one friend who might be able to help me with checking out the tournament histories to find out if it's calculating everything correctly, but otherwise I'm not sure how to check it myself. He actually reviewed quickly the EV of one sitngo which at first glance seemed to be correct, but that's too little to prove anything.

If I had the certainty that the EV is truly correct, I would start putting in tons of more games knowing it'll turn good at some point. But right now I don't have the courage. Mostly the fact that I only had a tiny upswing in the beginning (green line following strictly the red line) and then everything started going "wrong" makes me feel self-doubted.

I am pretty sure that HEM is calculating it right and that you are just running megadoo-doo with the allins. Seeing as you are playing Super Turbos the 20% allin ev is very unlikely to be sustainable, I would say that even 10% is difficult at these but if currently your 'true' roi is 10% you are havng a very, very poor start.
I'll have a go a trying to calculate the chance of hitting -21BIs if you have a10% roi after 600 games later - I suspect it is a very large amount under 1% (but I might be wrong)
If you do want to publish a graph you have to upload it to an image hosting site and post a link on this forum.

Thanks BaseMetal for your replies! Yeah, I know 20% is not realistic in long run. I will coninue with the games and see what happens.

Assuming a true roi of 10%, so say a true (in the long run) finish distribution of (20%, 21% and loss 59%), it pays 65% 35%, let’s say the cost per game of $1.10 with 0.10 as the rake. This actually produces a roi of 11%, and the average result of one game is $1.22 (ie, mean of one game in the long run)
You can calculate the variance of one game and this comes out as 2.0066.

Clearly the finish distribution of one result is very spikey (3 spikes one each for 1st 2nd and 3rd) and not a Normal shape (ie, not the bell curve graph) but there is a stats law that comes to the rescue The Central Limit Theorem, and from this you can say that if many games, say N, are done the overall result will be a Normal graph with a mean of N x (mean of one game), and a variance of N x (variance of one game)
The more games the closer to Normal but with 600+ games it is still reasonable to expect this to be a decent model.

Packmax has played 612 at $1.10 and from this has –21 x $1.10 which is an actual result of (612 x 1.10) – (21 x 1.10) = $650.01
The expected (11% roi) mean for these was, 612 x 1.22 = $747
So currently (747 -650)/1.1 = 88BI's down

N x variance = 612 x 2.0066 = 1228
So Normal (mean, var) => Normal(747, 1228 )
You can use this to calculate the spread of expected results and how often certain scores will occur and using a maths package the chance of getting $650.01 or less is 0.00282025 or about one in 350.
edit: I corrected the result

This is pretty bad but playing tournaments is reasonably high variance and you have to survive bad times without tilting and spiralling down, you have had an early bit of bad luck but these are like rites of passage – if you play long enough you will hit a spell where this one seems like a walk in the park – the joys of tournaments . Also remember that there is likely to be someone out there who is 1 in 10000 over the last 612 games and also that we really don't yet have too much of a clue as to your true roi, after about 5000 games you start to get an idea.

I hope I did the maths fine for this but I may have made a mistake or two.

yes basemetal has a point. 1/10k odds happen to people cause millions play poker