*** April High Stakes Thread ***

[tripsONacid]

See Isildur, this is what happens when you play Omaha.

MODS- can we please move this to another thread. Ye can call it FlipsONacid as I was one of the original posters and I've had half a dot.

[poofinger]

Originally Posted by Mumbly

Yeah, I'm leveling. I guess "regression toward the mean" is a foreign concept here at NVG.

If I wanted to level you, I'd have gone the way of the guy who said he wouldn't play Aces because he had already won with them several times previously.

Not sure you fully understand the concept.

I think this is what it means-
flip 10 different coins 6 times each, on average you will get 3H's and 3T's. One coin might freakishly land H's 6 straight times, but if you flip that one freak 6 more times, it will regress towards the mean (or be closer to 3H's and 3T's). This is not the same as saying it will now favor T's to make up for it's previous freakish behavior.

[blocka]

Originally Posted by Mumbly

Yeah, I'm leveling. I guess "regression toward the mean" is a foreign concept here at NVG.

If I wanted to level you, I'd have gone the way of the guy who said he wouldn't play Aces because he had already won with them several times previously.

the law of large numbers and regression to the mean do not mean what i think you think they mean, i think.

[effuluck]

Well, how often will heads show up in that spot ike?

[AKingdom]

Originally Posted by poofinger

Not sure you fully understand the concept.

I think this is what it means-
flip 10 different coins 6 times each, on average you will get 3H's and 3T's. One coin might freakishly land H's 6 straight times, but if you flip that one freak 6 more times, it will regress towards the mean (or be closer to 3H's and 3T's). This is not the same as saying it will now favor T's to make up for it's previous freakish behavior.

That's right, the more times you flip the coin the closer heads will equal tails, not because tails will now be more likely (as each roll is an independent event from the previous one/s) but rather because the law of large numbers dictates that over a large sample, the proportion of heads and tails will approach it's EV.

[Mumbly]

So you're saying that the more times you flip the coin, the more it will even out. Gee, why didn't I think of that before?

[Folding Soup]

Originally Posted by Mumbly

So you're saying that the more times you flip the coin, the more it will even out. Gee, why didn't I think of that before?

I don't know. Maybe you're stupid or something.

[effuluck]

dont worry mumbly...highstakers and retards like to feel special and feel like they are above the rest of society because they play...gasp....poker!

[jack492505]

Theres either a lot of levels or a lot of really terrible mathematical skills ITT. We need less of both and more hand histories.

[blocka]

Originally Posted by AKingdom

That's right, the more times you flip the coin the closer heads will equal tails, not because tails will now be more likely (as each roll is an independent event from the previous one/s) but rather because the law of large numbers dictates that over a large sample, the proportion of heads and tails will approach it's EV.

Just to clear it up it should read the closer heads and tails will approach the mean.

For example you
60H/40T after 100 flips.

The next 1000 flips you still have 50/50 each time so expect 500H/500T

Adding together you 560H/540T, which is closer to the mean, but not closer in absolute difference between H and T. You could end up closer or further away in the absolute difference but the law of large numbers say you will finish closer to the theoretical mean.

[CalmDownSah]

April HEAD OR TAIL Thread.. yeaaahhhhh...

[blah_blah]

Originally Posted by Mumbly

Yeah, I'm leveling. I guess "regression toward the mean" is a foreign concept here at NVG.

regression to the mean must be the single most misunderstood statistical concept in existence.

and, just so it's clear, you don't understand it.

[effuluck]

Originally Posted by blah_blah

regression to the mean must be the single most misunderstood statistical concept in existence.

and, just so it's clear, you don't understand it.

Can you explain it without sounding like a pompous ****head?

much appreciated!
Thanks

[AKingdom]

Originally Posted by CalmDownSah

April HEAD OR TAIL Thread.. yeaaahhhhh...

This had me in stiches. Post moar plz.

[Fitzcat]

Originally Posted by effuluck

Can you explain it without sounding like a pompous ****head?

much appreciated!
Thanks

At the risk of being leveled, and also at the risk of being despised responding to the oft-hated effuluck:

An earlier poster had it right. If the first ten tosses are 9 - 1, then you expect the score after 1010 tosses to be 509 to 501. Which is a lot closer to 50%, but still an 8 flip lead for heads. The *ratio* regresses to 50/50, but the expected value of the difference (ie, the amount heads are ahead) stays the same.

Interestingly, the expected value of the *absolute value* of the difference actually grows. That is, you are expected to be more flips from even the more flips you do.