Question For People With A Good Grasp Of Statistical Signifigance
Started by Errinovar, Jan 26 2015 03:29 AM
83 replies to this topic
#81
Posted 28 January 2015 - 06:12 AM
I am surprised to see a reasoned explaination on the Forum. I'll be watching you Miz!
#82
Posted 28 January 2015 - 07:56 AM
Mizeur, on 28 January 2015 - 06:08 AM, said:
Nope, that's all factored into the outcome. You are either able to choose mechs and play them in a way that is correlated with winning or you aren't.
ok i'll use that definition. I like to out of habit quantify each factors contribution to the whole. But yes mech selection definitely has an affect on W/L record. I like to parse pilot skill separately then mech selection. "skill" in the mech lab is part of ELO.
It also explains why you start seeing the same mech types over and over again.
Edited by Tombstoner, 28 January 2015 - 08:06 AM.
#83
Posted 28 January 2015 - 08:20 AM
The danger of parsing things separately is that you miss a critical variable. Or you don't weight them properly against each other. Or you don't weight them properly for how good the opponents were.
Using the macro stats for your rating, wins and losses, means that every conceivable variable that is correlated with winning and losing is accounted for. Also, your Elo only changes if the outcome is different than what the matchmaker predicted. And the change is more significant the less likely that outcome was.
Using the macro stats for your rating, wins and losses, means that every conceivable variable that is correlated with winning and losing is accounted for. Also, your Elo only changes if the outcome is different than what the matchmaker predicted. And the change is more significant the less likely that outcome was.
#84
Posted 28 January 2015 - 08:28 AM
If you are looking for a mathematical answer rather than a seat-of-the-pants guess, good luck. The number of major factors in whether you win or lose probably exceeds 10, and the smaller ones are easily in the 100s. That's if you consider all other players to be equal and non-unique. I think even trying to get a computer to do an ANOVA and then determine the impact of you as a pilot is going to be nearly impossible.
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users



















