You probably know this, but that line of reasoning is known to mathematicians as the Gambler's Fallacy.

In a larger sample (in your example) you are only nominally further away from expectation not relatively. After 100 tosses the Phat(heads)=.37 aftter 10,000 5194 tosses the Phat(heads)=.5194 which is much closer to what we expected it to be.

Say then that we are wagering with a sucker who lays us 1.1 to 1 (a relevant example for sng players since we all expect to have an edge in games). For arguments sake we'll also suppose we are on the losing side of EV both times.

1) We've chosen heads for 100 tosses and won 37*1.1 and lost 63*1 = net loss of 22.3

2) We've chosen heads for 10,000 tosses and won 4806*1.1 and lost 5194*1 = net gain 92.6

So in conclusion in games in which we have an edge even though we may have run below expectation our edge has allowed us to compensate for the added nominal bad luck in the long run. <---sorry this is the worst sentence structure of all time.

+1

DirDirDirty,

Your girl is super cute.... outkicked the coverage imo

I agree lol, but what does outkicked the coverage mean?

You're family is very good looking as well.

Edit: 1.outkicked his coverage - Having a date/mate that is significantly more attractive than you.

lmao

Is this a mathmatical fact? Not calling you out, and certainly not saying you're wrong. Just trying to wrap my head around this.
In your example, let's say that heads are wins & tails are losses. Aren't you saying that we've essentially "stolen" 50 wins? Then, over our next 4,950 tosses, we'll expect to run at 50%, with no "correction" of our freakish variance? For lack of a better term, shouldn't the numbers "work themselves out?"

It is a fact. The numbers work themselves out only in so far as 50 wins don't play a great role if you do 10 million trials. So although you booked the 50 wins the overall result will move ever closer to 50%.

"punching above his weight" is the common phrase for that from my part of the world

deudry, boss: ur both correct. that's why we move up after a bad run, to catch some of that good luck correction at higher limits. right?

The Dropkick Murphys. Have that Green Day/Ramones hard-driving sound, but also have multiple musical gears ala GD. Kind of Boston-Irish punk rock, with some traditional music blended in.

So over very large samples, the numbers do essentially work themselves out? There is a correction at work?

Is this a serious conversation?

Lol, yes I always do the same on the rare occasions when I take a shot in the casino and see 6x black in a row; me is betting red..

Lolled hard when I saw Dirdir's edit..
Well played Kev..


When you do this against a somewhat cocky rival you truly are zeee best:
http://www.youtube.com/watch?v=37qyvTRVus8

Also, prior lol rally:
http://www.youtube.com/watch?v=FLHfWGSkSIs

level... obv

Is this funny?

cool link. ty Josem

more than funny, its pretty much true

There is no correction, but the numbers work themselves out. I don't have time right now to get into this. The basic idea is whatever happened in the past is irrelevant in the long run. So if we hit tails 50 times in a row we expect to hit tails 50% of the time in the long run, whether we include those 50 or not. "The long run" is just a hand waving version of the mathematical concepts of infinity and limits. And when we look at infinite sequences, 50 wins simply don't matter.

Josem what happened to manly rugby heroic champion? (or was it aussie football)

Either way, its an intriguing switch to a guy who appears less manly - can you share the story?

you can tell it's aussie rules when they are wearing those ghey vests

the new guy is that athene/chiren80 fool

Yup, correction was not the right word. I'm familiar with independent trials, I guess it was just the way Josem worded it that confused me. Tks for putting up w/ my nonsense.

we just go with "that girl is too hot for that guy"

+1

we just go with "Girl, you can do better than this guy"