Life In 19x19

Anyway, back to this thread. I suspect Lee Sedol (I'm not on first name terms like Pippen) takes the jubango more seriously than this tournament so will put more effort in there, but still this result should give Gu Li a boost of confidence (Go game results are certainly not independent like coin tosses). Also I noticed Park Junghwan did well to beat Shi Yue and win the tournament on captain tiebreak as he has a poor record against him (as do many other Koreans).

I do not want to spam this thread, but the topic interests me, so if it's not wished, please delete or move the "probability stuff".

uPWarrior wrote:No. People have already told you that you are falling for the Gambler's fallacy, I'll leave you the link this time: http://en.wikipedia.org/wiki/Gambler's_fallacy

I do not accept this fallacy as a fallacy. I agrue that if we flip the coin 100 times and if we assume the first 20 times were all heads then within the last 80 tries there will be around 50 tails, while only about 30 heads. This is not what LLN says, this is my simple version of LLN! It says: Since the odds for tails or heads are 0,5, there will be averagly about 50 heads and 50 tails in 100 flips. So I see a tendency there and it matches with everyday experience, doesn't it? Therefore I could see a tendency for Gu Li to win a couple of games after Lee Sedol won already 4 in a row.

p.s. LLN itself seems weird, because either an experiment justifies LLN or LLN will say: do more attempts and I will be correct. Seems like an immune theory to me.

Pippen wrote:I do not want to spam this thread,...

Too late!

but the topic interests me, so if it's not wished, please delete or move the "probability stuff".

Please do it yourself. 'Off Topic' is holding an open slot for you. This does not belong under Professionals.

Just came to say that those caricatures are in a whole new stratosphere of awesome.

lemmata wrote:Just came to say that those caricatures are in a whole new stratosphere of awesome.

We were joking yesterday about how well they captured Choi Cheolhan's hairstyle

I suspect that moving the probabilities discussion to another thread is a good idea, but since I am not an admin I cannot.

Pippen, you are simply incorrect in your assertions. By the fallacious arguments you are putting forward, it is evident that you do not understand statistics and probability. There are a number of real issues here, but I am going to focus on only two.

1. In terms of raw probabilities, you are absolutely wrong. A coin toss is a 50/50 proposition no matter how many consecutive heads have come up. Period. This is demonstrable, provable, and a simple consequence of the most fundamental laws of probability. A sequence going 100T, followed by H and then T is just as likely as 100T followed by T then H. Count tosses are completely independent events. One toss does not in any way whatsoever affect the next. It is physically impossible - perhaps you wish to advance some sort of paranormal connection? I doubt you would argue that.

As such, then, given the fact that the probability of 100T + 1H + 1T = that of 100T + 1T + 1H, the chances of H or at on the 101st throw is 50/50. By your arguments a string of 10H after 100T would be more probable than a mix of 5T and 5H.

2. go games aren't coin tosses. Even if Gu Li and Lee Sedol are evenly matched, there are psychological impacts of a long run of wins or losses - and therefore real chances of any one game is probably not 50% for either party.