Backgammon Online

29-Jun-2010
00:58 PDT

http://zooescape.com/backgammon.pl?v=200&gid=201143

Start whereever you like but at roll 13 you see a perfect board. a 6-1 creates odd numbers and then a 6-5 forces a open man(odds of the 6-5 are only 5%). He of course hits first chance (odds of only 30% but he hits).he is now behind a 5 pt board and he only has a 3 pt board but he doesn't stay off my 5 pt board hits again a second time. I get in and try and keep him cornered and he hits again for the third time and fortunatley for him gets a 6 to get past the 5 point board from off the bar. he gets by the 5 pt board with ease and I stay off his much weaker 3 point board. that happens unfortunately but that's the fun of backgammon. Now he has many rolls to roll a 4 or 5 or 6 or 7 or 8 or 9 with two rolls to leave a shot but he rolls a 2-1 to be safe when many numbers leave a triple shot. He then he rolls a 6-5 to get by and with all those numbers to leave shots I roll 1-1 and then he rolls 6-5 to get by.

It took three straight hits off the bar plus two 6s to get past a 5 pt board and then me stay off a roll giving him time and then a pair of high rolls to get both men by a triple shot. He now he can gammon me as it is not done. I had a perfect board and he hit every shot first try.

The odds of this scenario can be computed but it is incredible. Just to get one man past the triple shot without leaving a shot is only 7 rolls out of 36 or 19%. So for two men it is 19% for each but twice in a row should happen only 4% of the time. This does not count the two rolls to get a 6 past my 5 point board nor the 3 straight hits off the bar while I stay off his 3 point board once..

Incredible luck from behind a 5 pt board.

10-Jul-2010
10:49 PDT

You seem to know about odds for rolling di. Where can I find a table of odds for rolling the di?
I have one that shows odds for rolling 2 di but it's not very comprehensive.

10-Jul-2010
16:52 PDT

http://www.paulspages.co.uk/bgvaults/tips/dicerolls.php

11-Jul-2010
05:04 PDT

That's some incredible analysis right there.

I tend to go on the analysis of shit happens - if you're on the receiving end of bad luck in one game, then maybe the next you'll get the benefit of the dice rolls.

12-Jul-2010
17:07 PDT

For fun I started keeping track of two things once they started happening regularly. (I must confess that I have never played so many games at once and so it is really emphasized since these are all 1 pt matches and therefore every situation is magnified and recorded as a win or loss. When you play 5 pt or longer matches these games get hidden in the longer matches and do not record as a win or loss immediately. In one pt matches there is no recovery from a loss--it is done.)

The two things are #1) doubles to change the winner and #2) bare off hits to change the winner. Although it is no ones fault the dice frequently come up doubles at the end --many times two sets or 3 sets to bring someone from far behind to win or doubles last roll --it happens to everyone. However, in my games I have been fortunate to win 40 games to date and extremely unfortunate to lose 104 right at this time. I fear the end race every time I am in it due to the frequency of doubles at the end.

The other criteria is the bare off hit when a miss likely means a gammon or at least a victory. That number is also part of these totals. Is it just luck and I will catch up to even things out or is it any other explanation? I don't know. I have been told that it will even out over the long haul --if this is so --don't play me as I have 64 more wins than losses in these situations to come--lol--Unless I am in these position more frequently than others so it happens more often to me. I have to learn to make light of this instead of being frustrated by it so bare with me. I will keep track and if anyone wants to know how the numbers keep going --just ask. I will have no problem mentioning it if I catch up and things even out. I also must point out that I have received a number of these 40(fortunate) wins against two or three players and I apologize to them as the dice won them for me and also have a number of players who have killed me with these miraculous come from behind wins that I want to keep playing them hoping it turns.

That game posted earlier was one of 4 or 5 I lost to that player --all of which it took a sequence of rolls to turn my win into his win. None of it was his fault of course.

Jokingly, had I won half of these magical game(which many were gammons) I would have had a 2000 rating. Then some of those higher rated players would play me when they won't now--hehe. Now I am barely hanging on to the 1800 level.

I am going to buy new dice --

Have fun everyone!

12-Jul-2010
23:04 PDT

One hypothesis that comes from this discussion is that too many doubles are being rolled by the person who is behind once men start being borne off. To investigate this, I looked at about 70 of my most recent games to see if I could find a pattern. In each game, I started looking at rolls once bearing off began. I counted only the rolls of the opponent of the person who started bearing off. (For example, if I was the first to bear off, I looked only at my opponent's rolls from then on.)

Everything looks exactly as you'd expect: about 1/6 of the rolls were doubles (17.2 percent to be exact). Were there an unusual number of double 6's or double 5's? No -- only 2 percent of the former and 3 percent of the latter. In fact, for each of the 21 different rolls that could be made, the distribution was well within what you'd expect from random variation.

Now these games only included about 500 rolls of the opponent of the person who first began bearing off. It's therefore possible that something statistically significant might pop up with 5,000 or 50,000 rolls, but there's nothing glaring about the outcome from this sample of games.

I should also note that these were a boring set of end games. Several times someone hit from a holding or back game and came back to win. I counted only once when someone rolled a couple of doubles to pull out a victory, and even in that case the person was not very far behind when bearing off began.

Maybe next is doing the same type of analysis using CharmedOnes's games since she has sensed the pattern in her games. I'd also be interested in any other testable hypotheses anyone cares to throw out there that suggest the dice are not random.

17-Jul-2010
02:03 PDT

Charmed one I can feel your pain, but imagine this... I recall a game played in a head to head match 30 years ago for cash, when my opponent rolled... six sets of doubles to beat me in a race by one roll!!! As statics go get enough data and you will see everything possible or impossible happen!