Observations on Excalibur LCD backgammon performance, including non-randomness in the dice and a few annoying bugs. Overall - pretty good game but needs improvement.
Excalibur LCD Backgammon
This is a shirt-pocket sized electronic backgammon game, claiming a tournament rating of 1536 (optimistic, in my view). Nice color display, but the game pieces can only be seen clearly from a very limited range of angles. In all, a fairly nice little game at $30 MSRP.
--
DaleBrayden - 02 Jan 2003 : Well, I've been playing it for about a week. I've had it set at its highest level the whole time, but I beat it about 3 games out of 4. It's trying to tell me my rating is 1750 - total nonsense, I'm not that strong, and there's no way that it could assign such a rating anyway, since it
thinks its own rating is only 1536. It's still kind of fun, but I'm thinking I should give it to a less experienced player.
--
DaleBrayden - 25 Feb 2003 : A quick update. I've continued playing this game. For some reason it's kind of fun beating the snot out of it. I've levelled off at winning 2 out of 3 games, with an accumulated point score of about 300 out of a total of 240 games. It is now rating me at about 2900 - which would be really awesome if it meant something. I'm amazed at the battery life: it's still going on the original 3 AAA batteries. I figure 240 games must have taken at least 30 hours to play.
--
DaleBrayden - 22 Apr 2003 : The original batteries are still going strong after 500+ games.
I have found a few bugs. Specifically, if it rolls doubles but is trapped and can't take all its moves, it will sometimes 'cheat' and move the piece anyway. It has only done this twice. Also, if the human player is trapped and rolls doubles it (one time) added a new piece to the player's side - then wouldn't believe that the human won when he removed all (16!) pieces. That was kind of a big grrrr. Still, it's not horrible.
--
DaleBrayden - 19 May 2003 : Batteries finally gave out after about 700 games.
I've detected a certain amount of non-randomness in the dice:
- The computer gets first move about 2/3 of the time.
- Computer's first roll (if it gets the first move) is 5-3 with higher than expected frequency. Seldom does it get 6-4 or 6-2.
- However, if player gets first roll, that roll is 6-4, 6-2, or 2-1 with higher than expected frequency. And in that case, computer gets 6-x with what I believe is higher than expected frequency.
- Computer likes to make its 3 point early in the game. Though this is really bad strategy and contributes to the overall weakness of the computer's game, it is able to do this pretty frequently because it gets 5-3 in either its first or second roll with high frequency.
- It gets worse: if player gets hit following his second roll, his first bear-on roll will be 3-6 or 6-6 with almost 50% probability, meaning he doesn't get on because computer has made his 6 and 3 points. This one is really aggravating. The actual odds of this should be 3/36.
- Computer seems to bear on after being hit with higher than expected probability. For example, facing a home with 4 points made we would expect computer to get on in 1 roll with probability of 20/36. But computer seems to get on at least 3/4 of the time in that case - this is subjective rather than measured judgment of course. Player seems to bear on in similar circumstances with far lower probability - again, my judgment on this might be biased. The thing that convinces me, though, is that player often cannot get on until the third or even fourth roll in these cases. The odds of not getting on after 2 rolls is 16/81 (about 1/5), and the odds of not getting on after 3 rolls is 64/729 (about 1/11).
- I don't want to sound paranoid, but it feels like computer gets favorable rolls generally. For example, often I will leave a blot that can only be hit if one particular roll happens (for example a 6-x) with probability 2/36. But I am convinced that computer gets that particular roll more often than 2/36 on an as-needed basis.
- Doubles seem to occur more frequently during bear-off than during other parts of the game. In this case the advantage seems to be for the player. It's not quite as consistent as some of the other non-randomness, but I've seen far too many games where one player gets 3 to 5 doubles during bear-off.
Picture of the Excalibur LCD Model 377 Backgammon game:
--
DaleBrayden - 25 Dec 2002
Books RSS
