1st question - who are you?

2nd question - please explain your voodoo? :) If this is a ranking - I fear we're overvaluing some people if we look at straight points!

Assist points really don't say much, IMO.

Dave is valued too high, though tg seems just right. :)

outliers (imo)
negativezero might be an outlier since he only had 12 games over 5 seasons (the cutoff used was max 10 games) but otherwise this seems to be accurate especially on c.c. maps like elsdragon. Sure you cant measure really important stuff like blocking for meds and dodging or leadership and other intangibles but those will be hard to quantify. also long distance shooters might not get points out of range. dp who 'Does well up close and personal' is ranked preet-high. he also had only 11.5 games over 5 seasons (person with max games = 43 games). people with low games may have just had a few great games or were against bads (ahem) (or just might be that pro?)
probably should make the cutoff 12 games or something but it might exclude some others

a lot of good people may have not free agented up yet which might explain the voids at the top (idk)

the formula
final column=raw ppg/c*log(x) where x is (45,43.5,42 etc or 45-1.5*k)
45 is chosen since 45= number of people with 10+ games
so for example 1st row is 12680/12.9*log(45)=1625.01819183
1625.01819183 /6.5=250
2nd row= 28589/34.5*log(43.5)/6.5 =208.89= 209
3rd row= 22117/33.3*log(42)/6.5 =166

log(x) where x is (45,43.5,42 etc or 45-1.5*k)


find constants by dividing raw values to desireable e.g. 1625.01819183 / 250=6.5

edit. tg who is so good also ranks 8 in total assist points and has most total games (43.5) which means he likely played on low killing (ahem) squads against higher ranked ones.
actually idk. hmm... guess free agent stats really is assist points, why not total points (lol). that would explain the high med rankings, also werent stats from season 1 not recorded?
minus season 1 and 2 games stats he would be at 115 points

my numbers would be better with more games...i just shit wins. <mic drop>

The cat is outta the bag


I'm bad :(


Or one might say, The Gift is unwrapped

season 1 ppls stats r wierd

looking it over the math works and overall it seems to look good but like you said outliers, intagibles i assume factor in, like some very good players are lower then would seem and i dont think anyone would rate me #1 even though i am BY FAR the BEST to ever MSL/USL. i think that is a great metric but maybe that metric plus player opinion rated factored in although very impossible to do would be a good stabilizing factor in ratings.

Neg you suck!!


Miss you wanna vent later?

idk how to tell u this but with the adjusted season 1-.. stats the top 2 people are someone else. not revealing who they are yet so i guess negativezero yous be the top for now lol

my season 1 stats are on diff alias ;)

Which one of you is the Tom Cruise to rainman (the matrix)?

As much as I love advanced metrics - I'm not so sure points per game is a good barometer of effectiveness in USL. Meds and marines should be judged differently.

Take for instance if you're doing TS% in the NBA its a little hard to judge centers vs wings - because centers take most of their shots within 10 ft of the basket. Should be different formulas.

Perhaps for med you could add in ap% of total team ap or something, I think that would give a more accurate reading, and you'd have to add kills in for meds as well... because there's two different types of meds... meds like stav/jane/somewhat fax who mostly heal and don't shoot. And then you have meds like Sov/kippa/staley who shoot a lot and their pdws are basically another stream in team fights.

matrix i wasnt in the first seasons of FI i didnt come back to this til later or i would be farther in 1st!

cant post for some reason

-unfortunately dont have easy access to assist pts%, can see what u mean by med types though thnks
-idk why but people who played on season 1 (minimal stacking) and the KR map seasons (esp. meds) had generally worse stats since it is calculated per game

you do have easy access - if they play 4 games, just add all their ap% for those 4 games and divide by 4

this is the formula i had kicking around my head at work


avg ap%pg+kills+teamcomptotal*ownteamcomp


possibly add in competition rating so +1-7 or 8 depending on squad that season - higher = harder

find out a way to balance # of games - 6 games probably is going to be higher

mulitply score by ownteam comp 1-7 1 being highest

mecca tha god

^for the previous post column 1 - scaled value, column 2 - unscaled value

i see water you saying and i like the accuracy but i prolly think its easier to adjust for how well a persons team ranked in the post season but it would be impossible atm i dont have those stats

edit. mecca ranks 8th in this systems career stats