This month wertrew even topped the alternative deadliness** rankings (Fluttershy and jurgen were excluded):

4.514 wertrew
3.385 bcmatteagles
3.219 the die
3.079 Soromon
2.995 Noobie
2.925 Im Defeated
2.732 isaaktas
2.405 mrspeed
2.298 Dengar2
1.899 Jacking
1.716 cadpilot
1.663 Impact187
1.550 charlatan1983
1.467 OneShot7
1.454 tubzmcgee
1.449 tstj
1.371 rap1d
1.284 bivo
1.244 K8Dice
1.117 Rafkrik
1.114 EvAngelion27
1.092 vino_en_carton
1.050 Enzo
1.019 Marco_TNT
1.003 xanax
0.881 diceptr
0.807 spanky6
0.802 Carl101
0.778 Lil Johnson213
0.773 Dude111
0.643 2K13
0.517 Carloos
0.477 morrinson
0.473 special k dice
0.448 Jack Barrows
0.404 derivative
0.402 Rebel9
0.385 Honyo
0.317 twinky
0.274 SHO-GUN
0.208 63 belly
0.200 PRESIDENT N° 1
0.188 stakaboo
0.153 Seljacina
0.142 Krazy_koala
0.127 franklyghost
0.094 tehpoop
0.059 Nippelking
-0.020 negramarta
-0.038 Zard0z
-0.084 M3r0w1ng3r
-0.117 alex_sfb
-0.180 Monsanto
-0.227 BackstabberNoob
-0.233 hothercule
-0.238 Nokia3310
-0.251 omalg
-0.292 delikadir
-0.329 happytoscrap
-0.418 Ninjamonkey
-0.481 gogi
-0.587 serrik
-0.588 Albeer
-0.613 fuzzymcfuzz
-0.659 Fonias
-0.717 Who'sNutz?
-0.731 JiNn
-0.748 moneymaker$$
-0.749 michele
-0.765 Whitehawk
-0.791 cant be
-0.830 cyob
-0.864 ehervey
-0.879 1a2b
-0.995 seeb
-1.001 MadHat_Sam
-1.018 ScimitarNZ
-1.048 dragonfly500
-1.064 Qaziyev
-1.079 @MikeTamburini
-1.331 {A}Monkey SLayer
-1.357 Juhokki
-1.437 moulue
-1.472 CEO1979
-1.550 Clempson
-1.552 Voo
-1.617 cemt2307
-1.800 procutter
-1.810 chclau
-1.872 IFIGENIUS
-1.992 mike29
-2.018 spman
-2.074 burcin
-2.154 parsifal
-2.172 peacefrog
-2.627 apignarb
-2.705 moshecohen
-4.842 pHaNToM-R

**kills per root game played minus [games^0.5]*(7611/21552), where 21552 was the sum of the number of games played by the top100 players and 7611 was the number of kills achieved by the top100 players.

E(k/g) = 7611/21552=0.35314 is the global top100 k/g average a player is expected to beat for a positive score; (k/g - E(k/g)) is the margin by which they beat it; and sqrt(games) is a multiplier that reflects the confidence of our judgement. (The same k/g average calculated over a higher number of games is more likely to indicate differences in skill and less likely to indicate differences in "luck".)

I am being bullied by the stupiest statistic in KDice history. Nice.

If it's any consolation, it's not the stupidest.

That isn't consolation. But thanks.

Alright, I will also include your score if that cheers you up a little.

It should be #DIV/0… oh wait!

that's right I killed you and I will kill you again!

Query: Would not E(k/g) have a bias because of multiple counting of games played in by the top 100? Suppose two top 100 players are in one game and they each score one kill. In the sum over the top 100, the number of kills would be increased by 2 and the number of games would be increased by 2

The E(k/g) should be multiplied by the expected number of top 100 players in a game.

No, it should be multiplied with the square root of the number of participants in the first Freeroll of the month and then added to my personal points highscore.
Only that way you get a true indicator of skill.

skrum: Yes, a game where two top 100 players play would count towards the #k/#g of both players and would increase the sum of games played by the top 100 players by two. (E(k/g) is a weighted average of the k/g scores of the top 100 players.)

However, I'm not sure if "z := the expected number of top 100 players in a game (| provided at least one of them participates)" could be calculated from aggregate data or if multiplying E(k/g) by z would even be helpful. E.g. Calculating (k/g - z*E(k/g))*(sqrt(g)) would just put the break-even k/g higher, leaving fewer players with a positive score. The "intuitive interpretation" (do "better" than "the average" to get a + score) could no longer be used and what would it yield in return?
__________

About the silliness of these rankings: I think it's pretty obvious for everyone that kill rankings/scores (kill counts, k/g, kprg or whatever other indicator you can think of) are not really indicative of "KDice talent" in themselves. Deadliness in itself is more like a trait/characteristic than a skill.

Being more aggressive and deadly only helps (on higher tables) if someone doesn't have to risk (or sacrifice) precious 1st places (or dom/reputation etc.) to get the kill, and even then these +50s are only nuances on high tables.

IMHO true skill could be best measured by combining K/G or KPRG with PPG or PPRG (points per root game) and tourney efficiency scores, and ranking the players according to a composite indicator. Of course, one would expect KPRG to have a relatively small weight in such an indicator since deadliness only determines a tiny fraction of the total score of a typical top 100 player.

The expected number of top 100 players in a game would not matter if the idea is that we are merely looking for a mean of kills per game per player (or kills per appearance) and the top 100 is the largest sample available for a month. It would matter only if we were looking for a measure of top 100 players pitted against each other.

Deadliness would be a combination of inherent dangerousness and likelihood of encounter. For example, an encounter with a tiger is more dangerous than an encounter with a black widow spider but black widow spiders are more deadly because the likelihood of encounter is much higher.

I suspect that deer on the highways at night are more deadly than both tigers and black widow spiders.