i don't think we will win the league

Originally Posted by Grimson

Put it this way: suppose you flipped a coin 9 times, and all 9 times it came up heads. What are the odds of flipping heads 10 consecutive times? Extremely low. So what are the odds the final toss will come up heads? 50-50. Dispute that and you shouldn't be allowed to vote, breed, or in fact live.

The probability of flipping a fair coin nine times and it coming heads each time is 0.5^9=0.1953125% or 1 in 512. So it would not be unreasonable to assume that the coin is biased. We can look at this quantitatively using Bayes' theorem if you like

Originally Posted by Serenity Now

The probability of flipping a fair coin nine times and it coming heads each time is 0.5^9=0.1953125% or 1 in 512. So it would not be unreasonable to assume that the coin is biased. We can look at this quantitatively using Bayes' theorem if you like

Let's leave out the first nine flips and progress to the tenth, shall we? What are the odds of it coming up heads? Thank you.

So let me get this straight. If winning last year means we'll have less desire to win this year, and losing means you would be more motivated to win in the future, why arent city massive massive favourites for the league?

I think that 'desire' aspect is overrated as well. United have new players who will want to win the league, and the returning players will want to show it wasn't all down to Ronaldo. And finally there's the old man in charge, who is unlikely to tolerate any satisfied attitudes.

the thing is though, having won them in the past doesn't negate the efect this has had and how hard it is tom just keep on doing it

its testament to them that they have managed to get it right three times on the bounce. it would be truly remarkable to mange it again

Originally Posted by Billy Baroo

the thing is though, having won them in the past doesn't negate the efect this has had and how hard it is tom just keep on doing it

That is true, and I agree. I just can't sit idly by while TT claims there is an actual statistical effect.

IMO the reason nobody has ever done 4 in a row is because nobody has ever had the best squad for 4 years in a row. If someone does, then they'll win it 4 years in a row, and no intangibles will prevent it.

Originally Posted by Serenity Now

The probability of flipping a fair coin nine times and it coming heads each time is 0.5^9=0.1953125% or 1 in 512. So it would not be unreasonable to assume that the coin is biased. We can look at this quantitatively using Bayes' theorem if you like

surely the probability of predicting any sequence of 9 coin tosses is 1 in 512?

Originally Posted by Grimson

That is true, and I agree. I just can't sit idly by while TT claims there is an actual statistical effect.

IMO the reason nobody has ever done 4 in a row is because nobody has ever had the best squad for 4 years in a row. If someone does, then they'll win it 4 years in a row, and no intangibles will prevent it.

its a fair point although i think 3 years ago chelsea arguably had a better squad than us, our first team just stayed fit, gelled together and players hit peak form

they had those management issues and player issues as well as a few bad injuries

All this probability stuff is so simple. Say we start now and say 'the probability of Chelsea winning all of the next four titles is very low'. That is true.

However, the odds of them winning any separate one of those four titles are all the same each time, regardless of if they win the first one, two or three. The probability of winning title no. 2 doesn't become lower if they win title no.1. The probability of winning title no. 3 doesn't become lower if they win titles no.1 and no.2. The probability remains the same as it would at any other time.

Players being affected by pressure and media talk, on the other hand, is a separate issue.

Originally Posted by Johnaldo

surely the probability of predicting any sequence of 9 coin tosses is 1 in 512?

Yes. When tossing a coin N times there there are N+1 possible outcomes and 2N possible sequences. However, for any number of heads n there are N!/n!/(N-n)! total sequences with that outcome. So for a nine toss series there is only one sequence of the 512 total possible sequences with the outcome of nine heads, but nine possible sequences that result in eight heads, and so forth.