if i toss a coin and it comes heads 15 times in a row .......

I'd have a good look at the coin with a view to taking odds on the heads.

depends what odds you give ... ?

^ a live one at last

1000s would take 1.7 just for the sake of a bet

it is still 50-50

a coin has no memory chaps.......................

bit like a lot of the punters on hear :)

bound to be tails, chances of 16 in a row are so small that it will never happen

remitrom 26 Dec 11:52
it is still 50-50



TheProphet 26 Dec 11:53
a coin has no memory chaps.......................



You're not willing to give any consideration to the fact that the coin may be biased then?

remitrom 26 Dec 11:52
it is still 50-50


We have no evidence that it was 50-50 in the first place.

Took me five minutes to google hypothesis testing before deciding to just hit submit and finish behind the investor in the race to point out the obvious :(

Okay so assuming the coin may be biased and we see 15 heads in a row, what odds tails on the 16th throw?

Well the trouble is that you might be betting against someone who knows if it is/isn't biased, so you wouldn't want to lay much above 2.00 unless you yourself had additional info, or were happy that the nature of the coin wasn't known by anyone else.

Would happily take a chance laying 2.00 though.

I am making a note of all this for the next time I bet on the sport of coin tossing.

Assuming nobody knows and assuming tails is at least possible?

I'm laying at 2.02

is it a 2 headed coin????

If this is purely an exam question the the answer is evens as the 16th toss is an independant event.

If it happen in reality and the first 15 tosses of that coin produced heads, the chances of that happening is about 32,00 to 1. I would call in the stewards.

i would happily take 1.99 on heads in this scenario

Tails 19 times well done ...... It dont matter the next toss heads or tails is still even money you cant name it .....

Phill if you took 1.99 on a even money shot you would go bust ......

The 1% would kill you ........ may take a while .. but it would wipe you out

We know a coin has no memory, but shouldn't we take into account the way the coin is being spun?

Hence the odds needing to change?

"Phill if you took 1.99 on a even money shot you would go bust ......"

Presuming it's an even money shot and that you can't arb that 1.99 anywhere.

If this individual just happens Russian official whose next coin toss with the same coin at the same location decided the fate of a Russian national hero against a known fixer at a Russian event you wouldn't be as quick to dissmiss heads at 1.99...

if you take this to the extreme and say there had been 500,000 heads in a row, i think you would be wise to take 1.99 on heads.

sadblokes go 10/11 heads 4/5 tails

Some would bet heavy odds on heads again. Some would lay close to even money. Who is right its all down to opinion. Some at the very mention of a coin automatically go into stats mode. I find the same with football and fi xed games. Certainly if its too good to be true i would worry. As for the market price it will be fairly close to true probability if there is enough liquidity involved. Maybe the best thing to say is take whatever few u like on the true odds but as soon as a decent market is out one needs to respect its opinion also.

view = few .... u get the gist lol

On info 100% heads.

I'll bet £10k heads at 1.99 as long as you promise not to lie and say it's tails if heads comes up

boba how can who is right be down to opinion, its even money each of 2, anyone who bets lays differentley is insane imo.

so , your giving evens... no bet from me.
value is everything.

15 times in a row I would have to commend you for that, yes you sound like a right**.

T OSSER

touche

Using a Beta(1,1) prior I believe you would come up with a Beta(16,1) posterior distribution for example. A Beta (16,1) has a mean of 16/17 so it would not be entirely unreasonable to give an estimate for heads on the next toss of 94%.

Tails - ffs are people still betting on this

It is easy to remove bias in coin flipping by using von Neumanns trick, and use two flips.

^^^^^^^^^^^^^^ I bet you lot dont half pull the ladies

How many times do you have to flip a coin to show that it is not biased ?

Depends on your confidence range.