You're So Unlucky!

amen to that! I'd like to make my brother read this :)

From: PhilSimborg   

In answer to your questions Djangtchoub, Snowie has a button you can push called "dice" which tells you how much your position has improved or decreased for each possible roll. You can see, for example, how a 6-6 might cause you to become an 80% favorite or a 1-1 might make you a 52% favorite, etc. etc. Also, after you evaluate a match you can click on "luck" and see just how lucky you or your opponent was. The way it defines luck is by how well you roll over and above an average roll each time. But again, the biggest luck you can have is finding an opponent who thinks backgammon is all luck!

Phil

From: PhilSimborg   

Roderick, in effect we are saying the same thing in different ways. When you say you evaluate your opponent and try to, in effect, remove the luck factor by realizing what was lucky, you are actually focusing on your opponent's skill or lack thereof and making sure you don't think of him as good just because he got lucky, or bad just because he got unlucky. In every match it is important to concentrate on your opponent's decisions, as identifying weaknesses in his game can help you explit him in the future.

And you also say you don't let luck bother you....here again, we agree...but I find it easier to make sure it doesn't bother me by not considering it at all. Bottom line is we both focus on the skill, and that's what is important.

Phil

From: CKT   

BG is all about minimizing risks and maximizing chances. It´s a calculated risk and safety management. The rest is in the cups unless one uses a special kind of Georgian dice

From: RichMunitz   

When analyzing your play, yes, it is useful to look at the luck evaluation to help you confirm your suspicion that you or your opponent got lucky. However, ultimately it is your own error rate that you should truly be looking at. The dice are out of your control. But as long as your error rate is low (particularly if it is lower than your opponent's), it is ok to dismiss the result as being due to bad dice. However if you made lots of errors, you need to take some responsibility for the result.

Ultimately, Phil's point was that any factor that contributes to a player winning the game is indistinguishable from luck. So in effect, the difference in error rates between the two players can be thought of as additional perceived luck in favor of the player that played better.

Finally, a factor that Phil did not mention but that is important is that a strong player not only makes plays to maximize his own good rolls and his opponent's bad rolls, but a good player will steer the game towards complicated positions that increases the likelihood that his weaker opponent will make more and larger mistakes. Since errors are indistinguishable from luck, this serves to further magnify the apparent luck factor.

From: PhilSimborg   

Richard Munitz brings up some great points, as usual. One cannot measure skill simply by the Snowie error rate...all that really matters is the difference in rating between the players. If you give a cube that is not a cube and your opponent drops, your rating will go down, but for the situation and opponent, you were clearly right to double.

There is one other aspect of luck and skill that was not mentioned, and it is a key one to keep in mind. One thing that greatly contributes to how well you played (according to Snowie) is how difficult the rolls are in a particular match, and there is currently no way to measure that. Take two players of identical ability and give one of them many very tough cube and checker play decisions, and the other is lucky enough to have fairly simple plays to make, the one with the tougher decisions will almost always have a poorer rating. You simply cannot draw conclusions about luck or skill from a short sampling of games or matches, but only after many, many matches.

Going back to the main point of the article, however, and I believe most responders agree: the more skill you apply, the "luckier" you will be.

Phil

From: PhilSimborg   

Dear Turk01: Nobody, including me, will say that there is not an element of luck. That is not the point of my article at all...my point was it's not all luck. If you lose the match because of a single "lucky" roll, it is quite possible that it was just luck, but it is also possible that it was your lack of skill, or your opponent's superior skill, that allowed him to get to the position where he could win the game or match with that roll.

And I am saying if you seek to find what skill elements were involved you will a) learn something; b) play better in the future; c) win more, and d) enjoy the game more.

I know that most of the time, when I win, it was largely because of luck....I was lucky to find worse players to play me!

Phil

Sir,

It's such a pleasure every time to read an article from you.But THIS subject is fundamental,core definition of BG values.Whoever did not follow this path or disagreed even partially with the concept skill vs luck would better reed more on BG (or reexamine his/her RELIGIOUS values!!). Now I have a different question, related with the LUCK issue. Would you modify your play (SAFEST vs best rated GNU/SNOWIE play) at match point, or do you think the bots take into consideration the fact it's a match point?Hope I don't ask a question with an obvious answer. REGARDS:Eu_Bg_Giudjen1 (V.Gudgenov)

TYVM, that's all!I'll do my rollouts. Regards:V.G.