An experiment with Babylon and LMNO

[Cramulus]

1 L
2 L
3 L, H
4 L, H
5 H
6 H

If BH guesses H, the odds of him being right are 4/6
If BH guesses L, the odds of him being right are 4/6

I suck at stats, but I'm pretty sure we should see a 66% success rate

[AFK]

Yes, if he made 100 guesses, it is very likely he would get 50 to 60 of them correct. 

[LMNO]

Yeah, like I said, the probability he'll be right is 66%.

But the probability his predictions will have any effect on his final score seems to be 50%.

The only time a prediction can be considered "wrong" is if he says rolls of 1&2 are "High", and rolls of 5&6 are "low".

That really seems to me to be a more elaborate way of calling heads and tails.

[AFK]

Yeah, like I said, the probability he'll be right is 66%.

But the probability his predictions will have any effect on his final score seems to be 50%.

The only time a prediction can be considered "wrong" is if he says rolls of 1&2 are "High", and rolls of 5&6 are "low".

That really seems to me to be a more elaborate way of calling heads and tails.

It essentially is, except it is a really wide coin that lands on its side roughly every third flip. 

[LMNO]

For example, I could roll the die, and ask, "is it 1&2, or is it 5&6?", and simply re-roll if it's 3&4, because either answer would be correct.

So, with a randomized rolling process, it should come up 1&2 equally as often as 5&6.  That's 50%, right?

[Cainad (dec.)]

If you want to PM them to me, I can add it to the spreadsheet.

Sent.


And it does seem like an awfully complicated version of a coin flip. Why couldn't we do that, again?

[LMNO]

Babylon claims his magic works best if he starts with a 2/3 chance of being right.  Which, technically, this does.

[AFK]

For example, I could roll the die, and ask, "is it 1&2, or is it 5&6?", and simply re-roll if it's 3&4, because either answer would be correct.

So, with a randomized rolling process, it should come up 1&2 equally as often as 5&6.  That's 50%, right?

Sounds right.  Though, if you want to get technical, I think it would be 50% over a series of 100 rolls.  Cause in any one set of 100 rolls, you might, by chance, get an inordinate amount of one of the sets of numbers.  

[LMNO]

True.  Probability and actuality are fairly different.

[Jenne]

If you want to PM them to me, I can add it to the spreadsheet.

I'll forward Thorney's shall I?

[LMNO]

Go for it.

[Jenne]

K, on it.

[LMNO]

Cheers.

[LMNO]

So, uh...



Anyone hear from Babylon today?

[Thurnez Isa]

no why? spreadsheet done?