all ins are killing my results

this is the last two days play but generally i'm finding that when i'm getting my money in I'm getting sucked out on. Does the red line indicate i'm generally making money and the all ins are letting me down?.


http://img216.imageshack.us/img216/1...00hands.th.jpg

the following is my "luck" on my all ins since the start of the year when I've really started playing. Most of my hands have come this month as I've got more comfortable playing cash games and have played longer sessions.
http://img3.imageshack.us/img3/6202/luckl.jpg

top line is the cumulative expected value at all in stage, middle line is the cumulative actual value attained, and the bottom line is the net difference between the expected and actual which is currently -$17.

You are running below expected value. However, this sample set is WAY too small to draw any conclusions from.

If you're a winner overall, and you felt you played well, assume that it's just variance. When you get a much larger sample you can start looking at what you can change. The first graph shows that you're making more money when there is no showdown in the hand. It could mean that you're playing a little on the nitty side, but you can't draw any real conclusions from 700 hands.

The reason we yell and scream about 10k hand sample sets is that over this sample Phil Ivey could be up a few hundred grand or down a few hundred grand. It doesn't mean anything. Hell 10k is small for most uses.

Over the short term, you'll see a pretty dramatic swing, as one hand will have a dramatic influence. For example, here's my last three days. It's pretty easy to spot the suckout:


http://img25.imageshack.us/img25/7074/3dayev.jpg

When I go back and look at my cummulative data, it shove's I'm only running slightly below expectation. It all evens out.

I fully appreciate that its still a small hand sample but it gets pretty disheartening to grind lots of small pots and then the money is going in favourite and buyin and table profits disappear due to variance .
this is my graph of all hands so far and I think that from about the 1400 hand mark ive sort of broken even with non showdown play, its just those dam big losses that make the chart look ugly. I think I'm running a cumulative $-12.45 on cash play.
http://img3.imageshack.us/img3/5020/allcashhands.jpg

Going all in during multi-table tournaments, any time before the turn( within the first 2 hours of the tournament ) works out poorly for me about 7 out of 10 times... so I have to be really cautious and only go all in when I've been able to play a few hands for the table to see, but even then, there is always a risk of some crazy person calling on a gut feeling that their 6 8 suited will hit, and sometimes it does; :?
but calling all ins when I have the position and / or the chips to risk usually works out for me when I roughly calculate the pot odds.

What stakes do you play? How many BIs are you running below EV? This is such a small sample.

you're still negative in non-showdown pots, and it's a very small sample, so this means nothing

Erpel very kindly sweated me last night which had a dramatic effect on the graph with a 6BI gain in roughly 400 hands. This was helped by flopping nut flush and drawing to a flush for stacks.

I talk a lot, but I don't really pretend at poker wisdom. I think more than anything else having someone to talk to staves off boredom and tilt. And of course you ran extremely well. I still think we managed to fold some hands on early streets that could have cost big pots if they were played out of boredom.

I think the main thing in regards to this post is that you have a small sample size because you mostly play a single table. So you play for hours and end up with a very unimpressive number of hands that is hit hard by variance. It's hard not to be affected by variance when you see the results in relation to hours and hours of effort. This is something you just need to get better at.

In my experience analyzing ev-grafs accomplishes not much besides fucking with ur mind. And as said before, such a small sample means nada. I´m a decent number of buy ins below ev for like 150k hands, so don´t expect to run within ev or you´ll set yourself on tilt.

Those graphs intrigue me immensely.

I want to start playing poker more seriously in the future and start using poker tools. For now I am happy keeping it all "manual".

A couple of questions:

1. Do the HEM (or what it is) actually calculate your EV for every decision?

If so that is an incredible powerful tool. Modifying your game based on results is very difficult, because of variance. But EV is not affected by variance so we can make much more honest analysis of our game.

2. Is that EV calculated using villains actual hand?

If so, EV calculation is incorrect. When me make decisions we can only make estimations and must base our decisions on limited information. If HEM calculates EV based on villains actual hand the EV calculations becomes unrealistic.

3. If 1 and 2 is true we must be talking about Sklansky bucks here- EV IF we can see villains pocket hand.

It seems that this EV calculation also is victim of variance. If we assign ranges and villain keeps showing his top of his range at showdown, what was really a EV+ decision, according to our limited knowledge, becomes a EV- decision according to HEM.

At the bright side of life: If EV graph is similar to profit graph we have become masters at estimating E and FE.

Maybe I misunderstood the graphs, but you seem like a solid winner according to EV graph. Negative variance is just the poker Gods testing you out.

The only purpose of those graphs is to put you on tilt. And they don't say much. They don't tell you how many times you got a set vs. overpair instead of the other way around, or if you made a -EV call before the money went in and sucked out.

[insert generic poker wisdom here]

The EV graphs in HEM are for AI EV only, and only based on when you went AI. For example, on mine I went AI on the turn with a set of Jacks against villains QQ overpair. I had tremendous equity in the hand, but he hit one of his two outs on the river and I lost.

It's not there to tilt you, IMO. But it does help you keep your heaters/coolers in perspective. Generally, when I'm running below EV, it let's me know I generally played the session well. If I'm running over EV, it probably means my luckbox was working and I need to seriously review some of my decisions.

Quote:

---

Originally Posted by oskar

The only purpose of those graphs is to put you on tilt. And they don't say much. They don't tell you how many times you got a set vs. overpair instead of the other way around, or if you made a -EV call before the money went in and sucked out.

[insert generic poker wisdom here]

---

Well, I need to do some more research on these graphs- but my first thought is that I disagree with you.

1. The number of times you get set under set, AA lose versus KK etc, evens out in the long run. That is why these graphs probably do not mean as much with small sample sizes. But with 50K hands it should even out pretty much.

If u lose with AA versus KK when all money went into pot preflop:

EDIT: OOPS

Pot: 200 USD, E: 80%, EV: 80 USD

If we have this situation occurring thousands of times our profit will get close to 80 USD on average.

2. The same concept is true for EV- decisions. Over a large sample size they will prove unprofitable.

My point is that short term variance does not negate the meaning of these graphs and that the more hands we play the more they mean to us.

"The EV graphs in HEM are for AI EV only, and only based on when you went AI."

Ok.

Then these graphs will be very helpful in analyzing all-in situations, but not your overall game. The profit graph over a large sample size will be the most helpful in analyzing whether our game is improving or not.

As you said KBcoolman, the EV graph in relation to profit graph tells us whether we run good or bad.

I tried PT3 for a free trial, even though it was not compatible with my current sites, and found out it was incredible sophisticated. It is a very powerful tool to analyze your own game.

In the future these poker models will become more advanced. When the models start to incorporate many more factors into the calculations the edge we gain from using poker tools will be tremendous. Maybe the models will even assign ranges on people based on history, stats, tendencies etc.., and calculate your EV based on that.

OP is obviously running bad and needs to work on his luck.

I think I need to clarify my reasoning on these graphs. I'm not raging about the results but more interested in whther I was ahead when I got my money in.It's reasssuring for me that although my net profits were negative atthe point of posting , in general I had put my money in ahead and then lost.
As I think KB said the 'luck' should even itself over time and greater sample of hands, my non showdown winnings have turned upwards and are climbing back towards profitability , and with the session that erpel sweated me im now in net profit, despite my 'luck' figure now being $-21.
I'm taking it as reassurance that my cash game play is improving, if however my luck was positive and I was still down then I would know that I had some serious issues to deal with.
heres the latest graph.
http://img3.imageshack.us/img3/953/latestu.th.jpg