>>18543962>>18544238To continue on this thoughts, what it would take to really measure the luck vs. skill dependence of a game would the following:
1. You need to be able to measure the skill of a player.
Now here's already a big problem. As long as a game does not have a perfect strategy "skill" is something that can't be measured objectively, because you dont' have an objective reference (the perfect strategy) that you can compare with the playsers decisions. You can only deduce it from the amount of times he wins, but this is dependent on the skill of his opponent as well as his luck, two factors you don't want to have involved in this measure. Now the luck factor is not a big problem because you can simply let the players play many games, which leads to the influence of luck getting lower. But the opponents skill is tricky, because you can't control it. If you could, you wouldn't even need to measure skill in the first place.
This means that at this step you already need an approximation for the measure of skill.The best possibility is probably to let players play against many other opponents with the hope that the opponents skills will be balanced out. You could now use an ELO-system, or simply the number of victories as the final measure.
OK, so now you have approximated the skills of several playsers by measuring their win ratio against several other players.
