News:

Play our  New Fibsboard Position of the Week --> perhaps give your comments/reasons thx..here's the link  http://www.fibsboard.com/position-of-the-week/

Main Menu

Backgammon variant with reduced luck

Started by takeonme79, June 11, 2015, 09:57:37 AM

Previous topic - Next topic

takeonme79

I've read that backgammon is about 75% skill and 25% luck. Of course it's hard to give exact figures to the skill/luck ratio. However, assuming these figures are correct, I would like there to be a variant that pushes the skill/luck ratio to 85/15 or 90/10.

I suggest the following.

(1) Doubles count as 2 of moves instead of 4. This eliminates some of the drastic turnaround in games when someone is fortunately enough to throw a double at the right time.
(2) For each game, each player has one opportunity to roll his dice again instead of playing the dice he just rolled. Alternatively, he may once per game ask his opponent to roll his dice again if he feels that his opponent was particularly lucky.

I feel the second suggestion is particularly interesting in terms of how to program software to make the decision to roll again or whether to save the opportunity for later. I think it's something that humans will naturally have a good feel for whereas software will struggle to assess.

I'm not aiming to get rid of luck from backgammon, just to reduce it's influence.

dorbel

Unfortunately you have failed to demonstrate in what way either of these new rules would reduce the amount of luck in the game. In any event, even if you could do so, the game is popular precisely because there is a lot of luck! We are all fans of the big turnaround when it happens for us, but in any case most of the luck is of the small drip variety rather than a deluge. Good play catches more of the small drips than poor play and over time, that's what makes a successful player.

ah_clem

The essence of backgammon is evaluating probabilities and if you reduce the randomness (or luck) the game loses a lot of what makes it so interesting.

That said, if you're looking for results that are more skill-based than luck based, play Nackgammon with long matches - I think you'll find that the skill factor outweighs the luck factor.


takeonme79

I think the opposite. By reducing the amount of luck you need to rely better on correct evaluation of position and probabilities. As I see the game at the moment, your opponent has a good chance of beating you just through 1 or 2 lucky throws at the right time and your calculation means nothing if that happens. This happens too often for my liking. The proposals I made don't remove any randomness from the game since you're still dealing with 36 outcomes each roll, but they do reduce the effect of a couple of lucky throws. You still have to analyse and calculate as much as before.

dorbel

You still haven't given any reasons why your proposals will mean that luck will be less important. You also haven't shown why you think bg is 75/25 skill/luck.
Also
QuoteBy reducing the amount of luck you need to rely better on correct evaluation of position and probabilities.
It's an interesting assertion, but do you have any reason for thinking this?

takeonme79

Quote from: dorbel on June 11, 2015, 09:31:12 PM
You still haven't given any reasons why your proposals will mean that luck will be less important. You also haven't shown why you think bg is 75/25 skill/luck.
Also It's an interesting assertion, but do you have any reason for thinking this?

The 75/25 ratio is something I've read and I feel that's about right although of course there is no objective way of calculating a precise ratio. However, the specific numbers do not matter. I suggested that *if* 75/25 was correct then I'd like to push the ratio to 85/15 or further. If 75/25 was inaccurate and the ratio was something like 65/35, then likewise I would like to push the ratio somewhere towards 80/20. The precise numbers don't matter, just that I would like to push the game towards more skill.

Quote from: dorbel on June 11, 2015, 09:31:12 PM
By reducing the amount of luck you need to rely better on correct evaluation of position and probabilities.
Also It's an interesting assertion, but do you have any reason for thinking this?

Let's say a game was 5% skill and 95% luck. Then it's hardly worth your time thinking as the results of your strategy are not going to be a major factor in the outcome of a game based mainly on luck. I've just extrapolated that to conclude that the more skill a game has the more important the strategy becomes to the outcome of the game.

I agree that the argument is far from watertight. However, let me rephrase what I want my variant to achieve:

'I would like there to be a variant of backgammon where consistently playing better moves than the opponent and evaluating positions better than them is more likely to win you the game than in standard backgammon'.

Now that I have changed my requirement to the above, I believe that my variant would achieve that. Doubles often turn the losing side into the winning one and is often the culprit behind why a strong player might sometimes lose to a weaker one. My second proposal will also eliminate one particularly lucky throw and introduce a new element to the game which I think computers may find hard to handle well.

Whether my proposals achieve what I want is testable! Find a person you know is weaker than you by a reasonable margin. Play 20 games of standard backgammon and 20 games of my variant and see whether you win a significantly higher number of games under my variant. I haven't tested it yet but I would love to if I manage to find a willing lab rat.


takeonme79

#6
While pondering my previous reply I've thought of a new variant that is totally deterministic.

Before the game, a table of many successive dice rolls is made, using a random number generator or by rolling the dice yourself and recording the results.
The table is observable to both players.
Instead of rolling the dice during the game, the players use the table to see what their roll is. They may also look ahead and see their opponent's next throw, their own throw after that and so on.

The players calculate as far as possible, but just as in chess, you find there are limits to your calculation and therefore you need to maintain good structure on the board to help in the long term.

Interestingly, if both sides have perfect calculation until the end of the game, then this variant is a 100% luck game as one side is destined to win because of more favourable dice. In practice you can play the same dice throws again but with reverse colour moving first so that each player has a go with each of the sets of dice throws.

I call this variant Chessgammon. I wouldn't really want to play it myself, since I used to play chess and I've moved away from it because it involves too much move-after-move calculation. However I just want to put it on the net and claim it in case no one else has suggested this before (someone probably has).