RANDOM thoughts

[Wacken]

science is not sure yet about randomness.

If you flip a coin, it is influenced by many factors that will decide what it ends up like. All those factors are also influenced by other factors. Etc.
All those things not random at all [img]/images/graemlins/smile.gif[/img]

However, if we have one atom of a substance with a half value time of 10 minutes, what will have happened to that atom after 10 minutes? We don't know if this is random or not.

Not that this all is very important, but for this very reason i do not at all discard the idea that everything might be predetermined and nothing random at all. And i am not religious or otherwise supersticious at all.

[jgunnip]

BUMP

[ QUOTE ]
Bonus question for reasoning:

Consider two players Alice and Bob. They are going to play a fair coin flipping game. But instead of one winning on heads and the other on tails they decide to add drama and win only if they get their triplet of result first. Bob chooses to win if he gets THH. Alice chooses to win if she gets HTH. They keep flipping until one of them wins.

I.e.,

If they flipped TTTHH then Bob would win [because TTT is no one's, TTH is no one's, but THH is Bob's]. If they flipped HHTH then Alice would win. Are Alice and Bob playing a fair game?


[/ QUOTE ]

Alice and Bob would be playing a fair game IF each separate game constituted three flips of the coin. However, in this particular game where they play until one person wins, and then start over and play again, Bob is at a disadvantage because unless the first flop is T his needed sequence of coin flips to win is actually TTHH because when the flop is HTHH, Alice of course wins with HTHH.

[bearly]

a couple points: para 1) who is 'science'? para 2. i imagine it's shape will be called 'pretty worn out' at the end. para 3) you could say this about anything, ie. why is the football player still sitting on the bench, and where will he be 10 mins from now? para 4) little work on the spelling here. now to the meat, i believe this to be one of the finest of irie's or anyone else's posts. wish it could be stickied for a week......h

[tjh]

I think that the hidden gem of insight that I get from this post is that.

People believe in strange unscientific things. Given the effects of "random" events in poker, other players can do and will continue to do stupid things.

Using the power of science and reason we can play better than these folks. In the long run we will make money. In the short term we can also be fooled by random events.

I disagree with the opinion that results are meaningless.
One SNG is significant, and means more than 0 sng's. The more you play the more significant your results become.

I also disagree that a "good" player can watch a game and decide who is playing well and who isn't. We are playing a game of incomplete information against highly unpredictable opponents. There is no "one correct way" to beat SNG's. Aggression is good but can be dangerous, tighness is good but can be overdone. TAG is supposedly the best but within that there are various playing styles.

As for Daniel Negreaneu (sp?) being lucky and not talented... I don't know. Surely his style of play might not work at our level but at other levels it just might. Also the highly publicized plays of his may or may not being indicitave of his playing style overall. Each game has different weaknesses that we can take advantage of.

In my opinion results do matter. It may not be a perfect measure but it is a close as we can get.



--
tjh

[bearly]

hi. i think these will prove to be very prescient observations, both by irie and eastbay. i wish they would take a moment and expound further. i know i could pm them but i feel this material is very important for all those who are studying sngs in particular and on-line poker in general; especially as it relates to the profit possibilities of both. tia guys...........h

[AliasMrJones]

[ QUOTE ]
The reason people trust ICM is because of it's base in a solid mathematical foundation. It starts with the assumption that if two players of equal skill are heads up, each person's chance of winning is in direct correlation with the number of chips he has (an assumption that has recently been challenged but is probably relatively accurate). From there, we are able to calculate the approximate percentage of times that each players will finish in each position and hence, determine his expect percentage of the prize pool. Its limitations seem to lie in the fact that it doesn't account for differences in skill level or strategical advantages gained or lost by having a chip stack of different sizes. For something like SNGs however, this model seems to be relatively accurate and a good general indicator of whether or not you should call or raise in a given situation. Mason Malmuth gives an example of how to do a simple version of these calculations in Gambling Theory and Other Topics.

[/ QUOTE ]

This isn't entirely true. A chip equity=prize money equity model also starts with the above assumption. We don't use a chip equity model, however, we use ICM which is a chip-as-lottery-ticket model.

Back to the original Irie post. It seems to me to be a restatement of a previous post in which he talked about reviewing the results of many players, max long-term attainable ROI, etc. and basically said that most players on the forum aren't as good as they think they are because it takes so long to play enough SnG's to get any kind of precision/confidence interval. Plus self-reporting/selective memory... If you think about how few major tournaments there are every year and then think about how much stock people put into who won the WSOPME this year or who won 3 WPT events you're starting to get the picture.

[AliasMrJones]

[ QUOTE ]
I think that the hidden gem of insight that I get from this post is that.

People believe in strange unscientific things. Given the effects of "random" events in poker, other players can do and will continue to do stupid things.

I disagree with the opinion that results are meaningless.
One SNG is significant, and means more than 0 sng's. The more you play the more significant your results become.


[/ QUOTE ]

I think you're missing the real point of his post. The fact that poker often rewards mistakes and thus will encourage many players who don't understand this to make bad plays that good players can take advantage of is fairly well understood.

What isn't is to what extreme randomness-caused variance can play a role in what many people think is somewhat long-term. Unless you're a pro multi-tabling full time it maight take 10 years to get enough games under your belt to be able to say that you are a very good player. Yet virtually noone has played that long because internet SnG's haven't been around (or at least played by very many people) for long enough. So -- if you're not a full-time multi-tabler, you might never really know if you're any good. And, those pros you hear about all the time, have a fair chance of not being nearly as good as is generally accepted. Yet, people on the forum often take as gospel things put forward people because viewed as "long term winners."

[rydazzle]

This thread has left me with a nagging question:

1) Can a poker hand be characterized by random variables?

Wait, take a second to think about that. Consider bluffs, varying styles, human error and emotion, number of players, chance of seeing all cards on the board, etc, etc, etc.

Suppose yes: there should be some "normal" variance based on the combination of these random variables. How would our variance relate to this "normal" variance? What would it mean to have a variance above or below "normal"; would that reflect skill levels? Does this suggest the smaller the variance the better the player?

[GrekeHaus]

[ QUOTE ]
This thread has left me with a nagging question:

1) Can a poker hand be characterized by random variables?

Wait, take a second to think about that. Consider bluffs, varying styles, human error and emotion, number of players, chance of seeing all cards on the board, etc, etc, etc.

[/ QUOTE ]

The answer to this for me is a clear "yes". Suppose you're playing against a player who will always call to the river no matter what his hand, then bluffing is always wrong. On the other hand, suppose you're playing a player who never calls without the nuts, then bluffing is always correct.

Most players will fall somewhere in the middle. If you can accurately assess the range of hands that a player would have played in the manner that he played them, calculate the probabilities that he has each of those hands, and calculate the chances that he will call a bluff with each of those hands, you can mathematically predict if a bluff is the correct play. It's true that a player might call or fold on a whim, but you can still use percentages to predict each of these events.

[ QUOTE ]
Suppose yes: there should be some "normal" variance based on the combination of these random variables. How would our variance relate to this "normal" variance? What would it mean to have a variance above or below "normal"; would that reflect skill levels? Does this suggest the smaller the variance the better the player?

[/ QUOTE ]

Generally the ammount of variance a player has is determined by the game he is playing. In MTTs for example, a good player will have a much higher variance because he will be winning the tournaments much more frequently. STTs are the same, but less extreme. If you know that a place will place 14%/13%/12%/61% (1st/2nd/3rd/OOTM), then you can easily calculate that persons variance, and in general, it will be higher because of the fact that he's getting more firsts than average.

Many people confuse variance with the number of winning/losing sessions a player has, but this is incorrect. A player with a 20% ROI will have greater variance than a player with a 10% ROI. However, the player with a 10% ROI will have more losing sessions (assuming the same number of tournaments per session). Variance only refers to the deviation from the mean, so even though a player with a higher ROI will have fewer losing sessions, he will have a lot of sessions which are below average, but still winning.

[Misfire]

[ QUOTE ]
How many hands would you have to see to convince yourself that the 10% ROI player is better than the 20% ROI player. 1? 2? 10? 100?

What if they played 1000000 SnGs? Would you still need to see how they played to know who is best? Because if your argument is a sample size argument, I'll grant you that 5000 SnGs might not be enough. But somewhere along the line, there would be enough SnGs to tell you who is the better player, SOLELY on the basis of their ROI.

[/ QUOTE ]

Certainly as # of SNG's approaches infinity, the confidence in saying how good the player is based solely on his results approaches 100%.

I would suspect that as we increase the number of SNG's, the probability of a bad player maintaining a good ROI would drop exponentially. Like coin flipping, although the odds of winning and losing are the same per trial, the odds of maintaining the same lucky result time after time decline sharply. The more times you flip the coin, the less likely your overall results are to deviate from your true ROI (0% in this case)...that is, of course, unless you have some kind of coin-flipping skill.

In poker, after a certain point, the odds that the player's ROI can be attributed to "luck" have got to be so small that they're not even worth arguing over.