Loads of doubles

[Romain672]

By "last 10 games" i meant the moment I wrote that text.
It was way weirder then, and it repeats frequently.
10 games are surely not enough for analysing doubles...

I took that into account. I took game dated as 'one day ago' when I wrote my post, and the last game were one of the 98%, so it seems it was the games you wanted.

Thanks for letting me know...
I really needed that value of yours, on my comments.
Good luck in your camps.

You ignored all the other part of my post.
You got something ordinally, which you though was unlikely. I don't know what to add to that.



edit: I even forget to remove half of the occurences of the value. So the probabilities became:
{6>54, 7>48, 7>51, 10>63, 5>44, 10>33, 3>22, 6>37, 5>34, 11>38}
14%, 37%, 30%, 45%, 18%, 97.3%, 38%, 49%, 40%, 97.0%
So from two probabilities of 98% we goes down to 97%, which is a big difference.
These numbers represent how far from the average were the number of doubles. If it's close to 0%, that mean there were really few double. If it's close to 50%, we are exactly to what you expect. If it's close to 100%, that mean there was a lot of doubles. And more it's close to which value, more the generation is weird.
Two 97% isn't weird.
Per example a '8>37' (=8doubles for 37 rolls) has value of 79%. 9 is 89%. 10 is 94.4%. 11 is 97.6%. 12 is 99.04%. 13 is 99.66%. 14 is 99.89%. 15 is 99.96%.
That means accross all generations of 37 rolls, there will be about 0.04% of generation which will have 15 or more doubles.

I'm starting to be more aggressive but meh, I'm much more confident I'm right so meh

[Ze Monstah]

[Ze Monstah]

[Jest Phulin]

So i ask you: What is the probability to get a sum of 10 doubles in a game with a total of 27 rolls? Thanks in advance.

Well, according to https://stattrek.com/online-calculator/binomial.aspx, the probability of an event with a 0.1666 single probability happening 10 times out of 27 is about 0.00629. Or about 0.63% of the time. Or about once out of every 158 games. Since you've played over 3000 games of backgammon, it should have happened about 20 times.

[Ze Monstah]

[Jest Phulin]

i do not consider your arguments worthy, despite probably being mathematically correct,

OK, if we are going to disregard mathematics in a discussion about probability, .... Wow. I don't even know where to start with that. Can we discuss climate change without talking about rainfall and temperature also? How about we discuss soccer/football, but ignore goals?

[Ze Monstah]

[Benoit314]

And question: what would you say if a game was with 24 doubles out of 25 rolls?
You would consider it... rare, but acceptable, right?

On thousands of games played on BGA it could happen. Maybe it already happened.
An unlikely event doesn't mean it is impossible. It could even be 25/25.

[Ze Monstah]

[Raal]

Any perfect random generater have biased results in sampled short part.
This means that any results are not being affected by historied results.
In big samples, it means broken generater that they don't have such rare part.
For example, pi has "999999" at first only 762 digits (0.0001% possibility)
In fact, 0.63% is higher than three continuous 6s(0.46%).
60 games have only about 3037 dice rolls.