What are people's thoughts on how long a tournament should last? Just curious.

I agree with all of yinz on this topic!

My guess would be that the random number generating function is fine. After all, think how hard it would be try and write a random number function but actually write something that produces the results we are seeing. The bad code would be obvious. Instead, I suspect that the code is not being called properly. Consider this outline for the code:

Whose turn? - Player 1
What do we need to do (accept double, accept draw, roll dice, etc.)
Roll dice
Show dice
Player 1 moves
Player 2's turn
What do we need to do (accept double, accept draw, roll dice, etc.)
Show dice
Player 2 moves

It would be pretty easy to bury a bug that could give us the results we are seeing with something like this.

Nothingness: I sure hope you're using autopass! ;-)

Doesn't doubling when you're ahead and a move or two away from the end of a close game help to keep your opponent from getting a lucky win with doubles, which increases the skill factor in the game and reduce the luck factor?

gogul: Shouldn't it be 'terahertz'?

Is there an easy way to pull all the dice rolls and analyze them? I *NOT* accusing Brainking of doing anything wrong, I just think some of us math geeks might have fun seeing the results.

grenv: The solution is actually simpler than that. In a way, a game that encounters an auto-pass move would live in two states. The two states would be the last states viewed by each player. One of these would be the current state of the game. The other would be an old state that the other player simply hasn't viewed yet. Also, add a 'skip ahead to the current state of the game' for players who don't want to scroll thru every auto-passed move. That solves the problem of players who like to view all their moves vs. players who do not. It would be better, in fact, because I could look at one or two forced moves, then skip to the end. ;-)

No offense to everybody, but this will never happen. Fencer does what he wants and he'd never take the time to study the math to do all that's involved here. I'm not speaking for him, of course, I've just been here long enough to know how he operates.

A better strategy would be to ask Fencer to give us access to player statistics (but no personal information) and then someone could program up any stats we want. You could have any or all of the ideas mentioned below. Since all the info relating to past moves and current games and such is already public (except for private games), Fencer would not be giving anyone access to information that isn't already available, so it shouldn't be a problem.

That's my $.02.

Ah, yes, I read about that.

Novice question: Wouldn't having both players use the same dice fix that? The game would be fair even if the dice weren't. Or would that create new problems (other than breaking tradition)?

And as for my original question, the reason I mentioned it is because I was wondering if moving three checkers instead of four would make for an interesting variant. It seems like having to move three checkers would force a player to leave oneself open to a hit more often than when moving four checkers. That might balance out the higher pips per roll of doubles. I thought it would be an interesting variant.

Ok, you sorta lost me there. Why are we changing the dice?

My conjecture is that players who roll more doubles win more games but does anyone have a way to test that?

alanback: Right, I forgot the combination factors. Oops! Hurrying = Bad Math!! ;-)

Kili: The probability of getting exactly ten doubles in 45 throws is (1/6)^10*(5/6)^35 = 2.79990879 × 10^-11.

The probability of getting exactly five double-6s in exactly ten doubles in 45 throws is (1/36)^6*(5/36)^4*(5/6)^35 = 2.89408577 × 10^-16.

Someone correct me if I'm wrong. ;-)

It's reported: http://brainking.com/en/ReadBug?bgi=1499

grenv: Pawns can't delete posts. I had the same trouble when I downgraded my membership before.

Czuch: It's a 7-point match. The cube is now at 8, so the advice is moot.

I'll report it as a bug. Priority 1, right? j/k. ;-)

alanback: Yeah, I'm not too worried about it, I just wanted to ask here before I reported it on the bug tracker page.

I have a prime block in my home row and my opponent just doubled while on the bar. I accepted. Now it's his turn. Shouldn't autopass fire and make it my turn? Here's the game, although things might not be evident once he takes his turn:

Backgammon (Thad vs. sascham)

Czuch: I'm pretty sure it does.

wetware: Put the match score in (parentheses) or [brackets] and/or in a different color. IT could also be in a different size font.

Why doesn't a tournament show the score of the match instead of a useless dash?

grenv: Ok, but how hard can it be? And what other changes are being made that keeps Fencer from doing this one?

rod03801: I've read that too. In fact, I used that method in some programs back in my coding days.

Another vote for revealing the rolled dice. Do we really have any for not revealing them (other than folks who think Fencer shouldn't implement it because he should use that time to work on something else instead)?

Czuch: I highly doubt that the dice are unfair. To make them so would require a lot of code. Plus Fencer probably has better ways to repercuss those he does not favor!

However, I would surmise that the dice are not random! After all, it's pretty hard to get a computer, which is designed to do the same thing exactly the same way every single time, to do something repeatedly and get a different result.

But I would guess that they are pretty close. ;-)

I'm sure they are close enough that it would take some serious statistical analysis to prove otherwise.

And even if there is an advantage or disadvantage to the dice, either players is probably just as likely to receive whatever benefit there is, so I believe the game here is fair.

alanback: So you actually think BK's current implementation is better than it would be if the dice were revealed as soon as possible?

alanback: Why is this such a big deal. We are here for one thing... fun. The game would be better if the dice were shown straight away when they can be because that would be... more fun! So do it already. What more is there to say?

Czuch: I agree. If the dice are rolled, show them. It doesn't matter whether either player or someone else is the first to look at the board (as long as it's not private). If anyone looks at the board, BK should show the current status of the game (which includes the rolled dice). Anything else is just dumb.

fakar10: Good idea. I'd join if I wasn't a pawn. I hope you get lots and lots of members and that Fencer sees that (1) autopass should be a feature, not an option and (2) that he should expand the power of autopass to other situations, like endgame moves.

Vikings: According to grenv's calculations,
Backgammon (grenv, 2007-10-04 14:05:59)
it takes 17 rolls minimum to get all your checkers home without any doubles. If you went first, your opponent would get 16 rolls. That's 33 total. The odds of not rolling doubles on any roll are 5/6, so the maximum odds of rolling a doubles free game are,

(5/6)^33 = .002438 or 1/4 of one percent, about once every 400 games.

I suspect the typical game is close to 25 or 30 moves/side, so the odds are probably around,

(5/6)^50 = .00011 or one in 10,000.
(5/6)^60 = .000018 or one in 50,000.

Geez, I didn't mean to stir up such controversy.

A player once asked me why I resigned the previous game of a match with only two or three moves to go. I said I did it because I was mathematically unable to win. She said, "Oh, ok." Maybe she didn't notice or didn't want to (or couldn't *gasp!!*) figure that out. Point is that there are all kinds of players with all kinds of skills.

Sometimes it's just easier for me to click a few times than think if the game is really over. If my opponent is online and playing quickly too, does it matter? I suppose it would be different if my opponent was taking like 9 days to make every move, but I try to avoid that kind of game.

What's the proper etiquette concerning resigning at the end of a backgammon game? In chess, it's proper etiquette to resign when you're clearly defeated by laying your king on its side. Of course you can't do that online, but it's still considered inconsiderate to drag out a game that's clearly lost. In pente (the game I'm most familiar with) one puts his stone far away from the other stones when the only other option is meaningless captures or blocks that would only delay the inevitable. But what's the proper etiquette in backgammon? Or is it ok to finish any game no matter how far behind you are?

Here's one I lost in 16 rolls,
Backgammon (trambu4ka vs. Thad)

It this rare or can someone top it?

AlliumCepa: Except the checkers on the eight-point and 13-point will necessitate an additional rolls, I believe.

With doubles:
To bear off the two checkers starting on 24 will require two rolls of double-six and for the opponent to clear the 12-point. Double-fives would be satisfactory here too.

Two more double-sixes, double-fives, or double-fours will get four of the five checkers from the 13-point home. One more double-six, -five, or -four will get the fifth checker home and also two checkers from the 8-point.

Now another high double roll will get the final checker home plus three checkers borne off, leaving 12. Three more high enough doubles will clear the board. So you could do it with nine rolls of sufficiently large doubles and possibly some assistance from your opponent clearing a point or two.

Good luck.

Vikings: What is the minimum number of rolls needed to clear one side? Without doubles? I'll think about it later, but someone will probably come up with the solution sooner...

And her first move in the next game was with 6-6!

Someone asked about doubles on several consecutive rolls about a week ago. My opponent rolled 5-5, 6-6, 6-6, 2-2, 5-5, 6-6... right after I doubled the cube. She probably would have rolled more doubles, but that was the end of the game.

Backgammon (trambu4ka vs. Thad)

grenv: My opponent opened the game with 6-1 and blocked his seven-point. I moved 24/21(2) and 13/10(2). It seemed a little more important to advance my back checkers with that seven-point blocked.

Czuch Czuckers: Ok, well, I suppose you're right. My opponent won the opening roll, then I rolled the double 3s. I suppose what I am really wondering is is it better strategy to pull your stones out of your opponent's home with a good double roll early in the game, or use the double to establish blocks of your own in/near your own home?

When I roll doubles on my first roll of the game is it better to advance my stones farthest from home? Or should I have constructed some blockades? How do players feel about advancing stones out of the opponent's home vs keeping one or two in there for hitting purposes, especially early in the game?

Does Fencer keep a record of the dice rolls? If not, perhaps he could. If so, does he publish them? It would be interesting to run some statistical analysis on them to see how random they are.

If you think about how many games have been played here, the odds might actually be pretty good for someone to have thrown 1-1, 1-1, 1-1 by now like Andersp did. It probably should have (and probably has) occurred several times.

It seems to me like more doubles are rolled near the ends of many of my games. I'm assuming that it just looks like that to me, though.

alanback: I don't care to argue the point with you. How about a match instead. ;-)

alanback: Don't you ever want to 'look ahead'? ;-)

Say you just made a nice blockade with a hole in it. Maybe your opponent can get thru it, but only with a roll of 5-3. If it's a slow-playing opponent, you might not see the result for several days. I'd like to be able to see the dice roll right away (when possible).

It would enhance the gaming experience! ;-)

The problem is that the way Fencer wrote the original code, the dice weren't rolled until a player looked at his turn. It needs to be changed to implement many of the changes we'd all like to see, but that's a lot of programming.

alanback: I made a request for something to fix this a while back. I thought my idea was good, but I guess not enough others did.
Feature requests (Thad, 2007-07-15 06:55:25)

alanback: Ok.

tippyc: Should it?

pgt: No, the biggest weakness is that tournaments don't progress once all the meaningful games are complete, but having to roll the dice when you have no legal move does come in second. ;-)

alanback: The thing that is so stooooooopid is that you have to roll the dice. The choices should be double and autoplay the turn. Then you wouldn't have to click twice. I know clicking twice is no big deal, but it's the dumb implementation that irritates so many players. Fixing that needs to be done.