I think people are more prone to being bias that way.

What exactly is taken into account? Because I think reputation of that specific individual would play one of the biggest factors in someones rating. Without squadding with each specific player, you cannot really know what everyone offers to each individual squad.

Simplify ratings and cap, roster cap being 50-60 each player rated on a 1-10 scale, easy day :D

edit: as for raters please overhaul whatever committee is doing the last few ratings and awards for that matter.

Also - watch people regularly? do you mean in scrims too? If someone drops 20 kills every scrim but performs sub par in matches should their rating go up?

If people are taking reputation (if you mean personality wise) into account they shouldn't be rating players.

Some things to consider outside of stats: play maker, battle opener, leader, support player, etc.

And no, I do not mean SBs. Just matches.

Lets face it, aside from most of the newer players all of us have played or been around long enough to see what everyone offers to a team, however finding sub categories such as Activity, Shot calling/leading, clutch, awareness etc... grade all those on a scale of 1-10 and then average them out for a rating.

I wouldnt necessarily say scrim kills mean their rating goes up but I like the whole activity factor dave had implemented(just needs to be tweaked)

Apollo rate these players without looking it up.

Kills Deaths KDR Games KPG
Player 1 79 45 1.8 10 7.9
Player 2 71 41 1.7 9 7.9

w/l ratio?

7/10

edit - stats alone.. better than average.. but not top tier.

Player 1 is Shasta
Player 2 is Uzo

Heres how it should be done, given those categories chris mentioned there should be around 5 raters each person gets their own spread sheet in which they send in once theirs is done to be averaged with the others, thisll help avoid people just copying and pasting ratings.

I dont agree with the blind rating solely on the fact that people may have similar stats but one may be leading etc. This does however kind of let people be bias. Therefore the raters should be chosen carefully.

See Kevin, I was accurate with my blind assessment

"I dont agree with the blind rating solely on the fact that people may have similar stats but one may be leading etc. This does however kind of let people be bias. Therefore the raters should be chosen carefully."

Thats why you get a blind rating and then after the fact add a leadership bonus to it or something.

Not really. I'd give shasta (if out of 10) an 8 and uzo a 5. From a skill-level/activity/awareness point of view. Knowing the name plays a role.

Uzo is better than shasta skill wise
Shasta is smarter than uzo

you cannot factor in activity from the numbers you provided. Shasta instantly seems like the better fit because of reputation. One can argue that Shas performed better because of the squads he was on compared to Uzo.

I can factor activity, if i can see the name. Those numbers are from when they were on the same squad.

One more reason I take issue with blind stats is if we were to do VCso rating. You have no stats on site because he doesn't join in game or care for them, but he's been active the past few seasons and played well

And it's not just VC, FnI played a few games for bh this season but has no stats as well. So which season would you use for them? Who were they playing with at that time against what type of competition? Etc etc

Yeah Kevin, I was on that squad too that season Uzo was active. Luckily I think it's judging the last 3 seasons, if that was the case in your scenario.. yes I would rate them differently. But that season - they were equal imo.

And Eric that's not really a good reasoning.. to cater to people who dont bother taking the time to accept an invite.. lol If it were up to me I wouldn't let people play if they were not squad joined in game, but were on site (like they used to in every other league)

If salary does stay, plz don't try to make an entirely new rating system.

You have a better chance of getting accurate ratings if you make small adjustments to the existing system after each season rather than making a whole new one every year.

Also the more raters the better, gives a more accurate of the public consensus of skill