| |
BGonline.org Forums
Whither Luck Ratings?
Posted By: Michael In Response To: Whither Luck Ratings? (Albert Steg)
Date: Sunday, 1 October 2017, at 8:56 p.m.
It's not a prerequisite for the winner of a match to be more lucky in terms of MWC%. Both Players could be equally lucky or unlucky in which case the player who made the less errors would win.
Notice PR is just the average error per unforced move (and close or actual cube decisions) measured in terms of normalized equity. It's not a measure of the MWC%. A better measure of skill imo could be to measure the player's error in terms of MWC% at a fixed match length. So 2 identical PR plays are very unlikely to yield the same MWC% and sometimes the difference is substantial.
Looking at the Luck figures could help you identify a certain kind of cheaters in online games i.e those who hack the dice in various ways.
Doing extensive statistics on the relation between luck and skill could give us more accurate possibilities of winning compared to what we have today.
We all know backgammon is a game of both skill and luck. But is the effect 50-50%? It's not! Between players of equal skill, luck decides the winner 100% of the times. This percentage gets lower the more the skill difference but not linearly. A total amateur can win the bot at 1pt match 35% of the times!
So doing your own limited statistics you will learn that by improving your skill so that instead of making say 15% MWC errors in a 7pt match you make 10% you will not increase your winning chances by the same percentage. Simply because the effect of luck is usually multiple times more than the effect of skill.
Other than that looking at the Luck figures in the bot's analysis only clarifies how lucky or unlucky you were.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.