PDA

View Full Version : Luck factor



VurtAddicted
08-11-2011, 11:53 AM
Hi there,

imho EV winnings don't work very well when I want to consider if I am in a lucky or unlucky moment.
This is true mostly in deep stack cash games for two reasons:
1- "Luck" doesn't care of money, it's just a question of cards. How much money I put into a pot has nothing to do with luck, it's my own decision.
2- EV winnings doesn't consider river game, which is very important in deep cash game.

Let me be more clear with 2 examples:

1) If I play 12 allin hands like the following:

Pot 1000$, EV 95.5%, won = EV +45.5
Pot 1000$, EV 95.5%, won = EV +45.5
Pot 1000$, EV 95.5%, won = EV +45.5
Pot 1000$, EV 95.5%, won = EV +45.5
Pot 1000$, EV 95.5%, won = EV +45.5
Pot 1000$, EV 95.5%, won = EV +45.5
Pot 1000$, EV 95.5%, won = EV +45.5
Pot 1000$, EV 95.5%, won = EV +45.5
Pot 1000$, EV 95.5%, won = EV +45.5
Pot 100$, EV 60%, lost = EV -60
Pot 100$, EV 60%, lost = EV -60
Pot 100$, EV 60%, lost = EV -60

SUM = EV + 229,5$
Have I been lucky? From the EV it seems like I've been VERY lucky, but I actually wasn't. I just put a lot of money when I was far ahead and just a few when I was in a coin flip.
Actually I won 9 times a 95.5/4.5 while I lost 3 times a 60/40, so I was VERY UNLUCKY.

2) let's say I have a hand which is 95.5% ahead on the turn, we don't go allin, but then the opponent hits his/her 4.5% on the river, we go allin and I lose 1000$. I was VERY unlucky, but this isn't considered in EV winnings/losses.

So here comes my proposal: the luck factor.
The luck factor should be:
- if you win the hand: 100 – equity%
- if you lose the hand: -equity%
Where equity% is either the allin equity or the turn equity if the hand went to the river.

So, in my first example the result would be:

Pot 1000$, EV 95.5%, won = LF +4.5
Pot 1000$, EV 95.5%, won = LF +4.5
Pot 1000$, EV 95.5%, won = LF +4.5
Pot 1000$, EV 95.5%, won = LF +4.5
Pot 1000$, EV 95.5%, won = LF +4.5
Pot 1000$, EV 95.5%, won = LF +4.5
Pot 1000$, EV 95.5%, won = LF +4.5
Pot 1000$, EV 95.5%, won = LF +4.5
Pot 1000$, EV 95.5%, won = LF +4.5
Pot 100$, EV 60%, lost = LF -60
Pot 100$, EV 60%, lost = LF -60
Pot 100$, EV 60%, lost = LF -60

Global Luck Factor = -139.5

In our second example we had a -95.5 luck factor.

It’s a pure number, doesn’t care about money, so I think it’s much better to understand if I am in a lucky or unlucky moment.

GB223
08-11-2011, 04:24 PM
It’s a pure number, doesn’t care about money, so I think it’s much better to understand if I am in a lucky or unlucky moment.
+1
sounds interesting and makes sense....

melles
08-11-2011, 05:18 PM
sounds interesting and makes sense....
+1

udbrky
08-11-2011, 08:08 PM
I'm a little confused here.

Also your logic in example 2 seems flawed because the first 9 pots in which you won the 95/5 were $1k pots, while the 3 you lost were $100 pots, they need to be weighted to show truly.

VurtAddicted
08-12-2011, 05:43 AM
I'm a little confused here.

Also your logic in example 2 seems flawed because the first 9 pots in which you won the 95/5 were $1k pots, while the 3 you lost were $100 pots, they need to be weighted to show truly.

This is the point: EV winnings/losses DO take care of the money I put in the pot. Because of this, it's not a good indicator when we just want to evaluate pure luck.
Simply because luck has nothing to do with the amount of money in the pot, it's my own decision.

Take it in another way: the luck factor is the EV winning/losses considering my bet always 100.

Pure Presence
08-12-2011, 06:59 AM
There is nothing pure about this as it suffers from severe selection bias. If you are a huge calling station on the river, you will appear extremely less lucky than if you can make good folds.

To give you an example: if you bet small on the turn with TPTK and pay off a huge river bet when the flush got there, you supposedly got unlucky because you had ~80% equity on the turn, while if you had made the right play and folded the hand would not have been included in the luck calculation because there was no showdown.

The converse also holds, if you make good folds and your opponents don't, you will appear way more lucky than you actually are.

Patvs
08-12-2011, 07:16 AM
You make two statements:
-1 EV shouldn't care about money
-2 EV is flawed (especially in deep stack poker, but also in CAP games, or FIXED LIMIT) when you commit 92.5% of your stack on the turn. And commit the last %% on the river after your opponent has hit his 2 outer.

Statement #2 is absolutely true. There are several ways to calculate EV. HoldemManager uses "All-in EV". There are many disadvantages of All-in EV. Read: http://forums.holdemmanager.com/manager-general/24059-ev-not-all-situations.html

#1: to not care about money is interesting.
Let's simplify your example. You only play two hands:
hand 1: AKs vs 22 (preflop allin) = 50%-50% coinflip = you win the hand.
hand 2: AKs vs 22 (preflop allin) = 50%-50% coinflip = you lose the hand.

With your method. The "luck" would always even out and end up as zero.
With HoldemManager's All-in EV if both hands were $100 pots, the end result is also zero.
But if hand 1 was a $1000 pot, and hand 2 was a $100 pot, it will show you're lucky.

And you have been lucky. Since in poker we do care about money! If you win a coinflip when you just happen to take a shot at a much higher stake... or the pot is extremely large because both you are your opponent are extremely deep, that coinflip contributes more to your overall winnings. (than a coinflip where your opponent is shortstacked, or you're playing at a lower stake)
The fact that both are coinflips, and two coinflips--> have an expected value of zero (when you're playing heads or tails with a coin) do not apply.

melles
08-12-2011, 07:49 AM
Check out the program Cooler Hunter

VurtAddicted
08-12-2011, 08:57 AM
Hi there and thank you for answering.

I'm not complaining about EV winnings/loss should be as it is, I'm just saying that we should have more information on how the cards play, regardless of our bets.

The point is: I'm trying not to put myself in a big coinflip while I'm going to play a lot of allins where I'm far ahead. If I reach my goal, my EV winnings will be very positive, but if I look strictly at the cards, I haven't been lucky at all, I just played as I wanted to play, putting a lot of money far ahead, just a few in coin flips.

This "luck factor" should be something completely separated and considered beside EV winnings, not instead of.

For instance, if I have a poisitive EV winnings, but a "luck factor" close to zero, it just means that I choosed the right moment to put my all my stack in the pot. Putting a lot of money in the coin flip I win and just a few in those ones I lose, like a real coin toss, doesn't work in the long run, so why should I care about it? I know it's not going to happen.

Using turn equity when the hand goes to the river can also be a separated stat, having a "standard luck factor" and a "river on turn luck factor".

I think it's very cheap to do (maybe someone out there can write a report?) and it can be usefull.

VurtAddicted
08-12-2011, 09:02 AM
There is nothing pure about this as it suffers from severe selection bias. If you are a huge calling station on the river, you will appear extremely less lucky than if you can make good folds.

To give you an example: if you bet small on the turn with TPTK and pay off a huge river bet when the flush got there, you supposedly got unlucky because you had ~80% equity on the turn, while if you had made the right play and folded the hand would not have been included in the luck calculation because there was no showdown.

The converse also holds, if you make good folds and your opponents don't, you will appear way more lucky than you actually are.

Are these situations somehow considered in EV winnings as it is now? Is there any number that does, in HM, take these things into any kind of consideration?
I'm not saying the factor works for every "bad luck" situation, but it's another information which is, IMHO (and others), useful.

VurtAddicted
08-12-2011, 09:08 AM
Check out the program Cooler Hunter

very cool! :D

That's what I really needed, but still I guess the luck factor could help somehow.

Tyvm

udbrky
08-12-2011, 07:10 PM
My point about the pot size is,

Poker's not about being lucky or unlucky, it's about making $. You should make the examples using bb rather than $, because I assumed both were at the same stakes.

The real way to examine is to put yourself up against a likely range and see how much your winning % against that range compares to your winning % actually - Using Phil Galfond's G-bucks vs Sklansky bucks.

Sometimes, that fish just happens to get AA when you get KK, and if you suck out, it looks like you ran super hot, when really, you did run pretty bad to get KK vs a 60/40 w/ 35% 3bet in 150 hands when he had AA.

EV/luck/running good/bad is so dependent on many factors that it's really hard to put into one neat little package.

johncc
08-16-2011, 05:11 AM
i know there are lots of differing situations in poker but how about a simple report that shows wether u won or lost with the best hand when the chips went in pre flop

udbrky
08-16-2011, 05:47 AM
21221