Re: Method to prove no dice manipulation on on-line BG sites

monty1945@xxxxxxxxx wrote:
Max is making my point, which is that if you are going to trust a
company in a foreign country (and with no oversight) with your money,
then you want to be able to verify fairness, since it is a game of
"chance" to a large degree (in money games and short matches).

It is *always* a game of chances, even in longer matches.
I think you're just unwilling to accept that notion.
The stronger player does not win 100% of the time. Period.

Suggestion to Max: read my posts a bit more carefully.

Ufff, that's requiring a bit too much ...

Some have
claimed that certain sites are "rigged," but I am specific. For
example, on CG it seemed to me that there were just too many double
rolls, and that it could be a software issue (for all I know about the
software being used).

Checking for unbalanced dice is rather easy if you have some notions of statistics. Here you can find an example about testing if a coin is fair:
http://en.wikipedia.org/wiki/Checking_if_a_coin_is_fair

Your "too-many-doubles" case is almost identical (for a coin you have two possible results, head/tail, with probability 1/2 each, while for bg rolls you have 2 possible results, double/no-double, with probability 6/36 and 30/36), hence you can apply the same method. Take the time to record a few matches ... In fact your case is a binomial test, described here (note that the article describes almost exactly your case):

http://en.wikipedia.org/wiki/Binomial_test

Just as an example, in the case of the coin, if you want to estimate the probability of head : count the number of heads H and the number of tails T in N experiments (N = H+T), then the probability of head is estimated by p = H/(H+T). Now, if you want an error of 0.01 (1%) and a confidence level of 99.90% (i.e. you're 99.90% sure that the probability of head lies in the range [p-0.01 p+0.01]), then you need to record 27255 samples.

It may be that 3 point matches are just too
"luck oriented," because my winning percentage in 7 or 9 point matches
is much, much better, but I play these longer matches on sites other
than CG or XG (I did play a 9, a 7, and a 5 point match on CG for
small stakes and lost them all, though I outplayed my opponent each
time - obviously too small a sample, but why should I play there when
I win on other sites?).

3 short matches is nothing. Last year in MonteCarlo, the final was a 25pt match, and to get there you must pass at least 8 rounds ... even here, most of the players agree that the winner is not necessarily the best player.

MaX.
.

Relevant Pages

  • Re: British Money is so strange and German too...
    ... I precis this personally as the "Teston" being the coin, ... Henry VII, a coin which got its name from Italian, testa, ... head, because it was stamped on one side with the head ... may be an attempt to give money a more human face, ...
    (rec.collecting.stamps.discuss)
  • Re: probability concept
    ... that it is a fair coin. ... the probability of a head is known to be exactly 1/2. ... No Jack, the assuming that the coin is fair doesn´t ...
    (sci.stat.math)
  • Re: OT: Probability Problem
    ... pre-reveal chance of a head vs a tail is no longer relevant. ... They are of uneven probability IF and ONLY ... Coins HH, reveals H ... person decides which coin to reveal. ...
    (rec.gambling.poker)
  • Re: probability concept
    ... That assumption implies that the probability of a head is ... Jack *** My ... the assuming that the coin is fair doesn´t has much ...
    (sci.stat.math)
  • Re: Who help me to decipher what Bob really to say?
    ... that it is a fair coin. ... the probability of a head is known to be exactly 1/2. ... What Bob and I are saying is that since the OP knows that the coin is fair as a given, then he knows that the probability of a head is 1/2. ...
    (sci.stat.math)
Loading