questions

Hey, I was just wondering if anyone else was experiencing certain fight patterns. I win 20 x in a row against a bot and then might lose the next 5 or 6. I would expect the losses to be interspersed amongst the wins but instead they often come in big blocks.

I was just wondering if anyone has noticed this or is it just me? Also, Is this just the way the game is and did this sort of thing happen in bots 2?

Thanks.

it's just coincidence.

I was just wondering if anyone has noticed this or is it just me? Also, Is this just the way the game is and did this sort of thing happen in bots 2?

Yeah, play for a couple more years and you might notice that some things that should be random appear to be patterns, they aren't though, it's just "bot luck".

^ what he said

in other words, this is how statistics work:)

^ Yep.

In a random cointoss (Heads/Tails), People think that HTTHTH is more likely to occur than HHHTTT while it the odds are exactly the same...

In a random cointoss (Heads/Tails), People think that HTTHTH is more likely to occur than HHHTTT while it the odds are exactly the same...

You are correct, although you should have used a more brutal example, like:

HTHTHTHT vs. HHHHHHHH

:)

You are correct, although you should have used a more brutal example, like:

HTHTHTHT vs. HHHHHHHH

:)

Um, those odds are not the same :)

You are correct, although you should have used a more brutal example, like:

HTHTHTHT vs. HHHHHHHH

:)

Um, those odds are not the same :)

I'll bite...why not?

Over time you will still expect 50/50 heads and tails, which was the point Mainor was making. The examples are over too short a time but it is the principle behind it that matters.

I have to disagree. The probability of either event is 1/2^8.

Since each coin toss is an independent event, prior tosses have no effect on later ones. It doesn't matter if you've flipped 7 heads in a row or if you've alternated heads/tails for 7 flips.

Maybe this is unrelated to what Mainor was trying to say, but P(HTHTHTHT) = P(HHHHHHHH).

Over time you will still expect 50/50 heads and tails, which was the point Mainor was making. The examples are over too short a time but it is the principle behind it that matters.

I don't think that was his point. Anyhow, the odds aren't affected by the amount of attempts. The odds of getting HH or TH are the same, and the odds of getting Hx2*10^15 or THx10^15 are also the same.

You guys are looking at specific events and not the overall picture. The probability of any specific event will be the same, I agree.

Looking at it overall though, say in 1000 tosses you get 1000 heads. Compare this to another 1000 tosses where you get 500 heads and 500 tails. Most importantly we expect the heads/tails to be 50/50. I would say something is very wrong in the first case where you have 1000 heads out of 1000 tosses.

What you'll are basically saying is that if you're expected to go 50/50 with a bot, and if you win 1000 out of 1000 fights then the probability of that happening is the same probability as winning 500 out of the 1000 fights. Not true.

Make sense?

Yeah, makes sense. I agree you would indeed expect a normal distribution of events if you're only counting how many heads/tails, whereas specific sequences of flips will have equal probabilities.

Statistics 101 thread..

^ Emanuel is right. That was not the point i was making. What i meant, and this fits in nicely into Enders post is what people EXPECT. Indeed we expect that out of 1000 trials like you said, 500 are heads and 500 tails. However, humans seem to think that this should also be the case in small (unrepresentative) samples, like the 6 coin tosses. We see the HHHTTTT as less likely, and therefore more noteworthy. Win, loss, loss, win, loss, win seems to be ok, and it thus not noticed. Winning streaks (In the case of BOTS4: losing)are therefore perceived as less random while this is not does not have to be the case. This might exactly be the reason why this thread was started ^^

You all seem awfully intelligent right now, so I may be making a wrong suggestion in that the difference in BOT stats would also play a very significant role in this matter. In comparison, making the 'heads' side a little heavier.