I created a metric that would kinda, sorta, maybe measure skill quantitatively based on a lot of data gathered from Lockdown games. Since nobody has the data, you can just make up values that someone you're competing against agrees with. Then use the metric to figure out your relative skill index.
Hypothetical application:
"I'm better than you."
"NUH UH."
"Fine let's use this metric. You agree that the average Lockdown player gets __(list of assumptions)___?"
"Yeah, sure."
"Okay, let's plug in our own data to calculate our skills then."
(does calculations)
"HA, I got a higher Skill Index!"
"WTF, this metric is full of crap. You're still a nub. Deleted-Knight is also a nub for making this crappy metric."

____________________________________________________________________________
For the purpose of this metric:
"Skill" = The ability to maximize one's contribution to winning, averaged over many different games (played with different people).

Significant Assumptions
1. Everyone is trying to play their best every match. (not accurate, but hard to quantify "effort", so yeah)
2. There is a wide range in skill levels within the LD population.
3. Luck factors disappear when we take the average of many games.

Some constants that are assumed to be known:
DGP = mean damage score for the general population.
CPP = mean caps per game for the general population.
DFP = mean defends per game for the general population.

Lockdown shields must be ignored when calculating the following constants:
DBP = mean weapon damage buff used by the general population.
--> Take the damage buff for guns, swords, bombs, and average them. Then average that over everybody.
SBP = mean weapon speed buff used by the general population.
--> ASI for guns and swords, CTR for bombs. Calculate in similar way as DBP.
MSP = mean movement speed buff used by the general population.
--> Again, calculated in the similar way as above.
ATP = mean AT usage by general population.
--> For every person, assign a value 0-1 where 1 is always use AT, and 0 is never ever use AT. Average over everybody.
HPP = mean base health of the general population.

IMPORTANT NOTES:
1. Average buff values should be expressed as percentage (where 100 is no buff, <100 is debuff, >100 is buff). Ideally the average should be weighted across different loadouts based on usage.
2. The "population" should be chosen so the "significant assumptions" are as accurate as possible without making the pool of players very small.

Player-Specific Values:
LT = mean latency (ping time) of player's connection in seconds.
JT = mean jitter (ping uncertainty) of player's connection.
DG = mean damage score for that player.
CP = mean caps per game for that player.
DF = mean defends per game for that player.

DB = mean weapon damage increase. (expressed as percentage as before)
SB = mean weapon speed increase.
MS = mean movement speed increase.
AT = value from 0-1 defined same way as the description for ATP.
HP = mean base health.

WR = win rate = (# games won)/(# games participated)

________________________________________________________________________________
The Formula

CNX = MIN[(sqrt(JT/LT) + LT), 1] --> Assume nobody with 1 second latency or 100% jitter will play LD.
SCR = (1/3)(DG/DGP + CP/CPP + DF/DFP)
BNR = (1/3)(DB/DBP + SB/SBP + MS/MSP)
TKR = (1/2)(AT/ATP + HP/HPP)

Skill Index = (SCR + 2WR) - (BNR + TKR)(1 - sqrt(CNX))
________________________________________________________________________________

Benchmarks and Comments

Connection Strength effects on Skill Index*
400 ms latency and 10% jitter ==> ~1.69
200 ms latency and 10% jitter ==> ~1.44
100 ms latency and 10% jitter ==> ~1.29
50 ms latency and 10% jitter ==> ~1.21
25 ms latency and 10% jitter ==> ~1.17

400 ms latency and 20% jitter ==> ~1.84
200 ms latency and 20% jitter ==> ~1.61
100 ms latency and 20% jitter ==> ~1.48
50 ms latency and 20% jitter ==> ~1.41
25 ms latency and 20% jitter ==> ~1.37

*Values are for hypothetical player which has equipment and average scores equal to that of population mean.

Jitter really screws up timing in fast-twitch games, so extra jitter affects performance more than extra latency. The metric tries to compensate for this.

Caps/Defs/Dmg effects on Skill Index
1. Doubling average on any of the three (while matching average in the others) will yield +0.33 to Skill Index.
2. Getting double the average value on all three will yield +1, much more than any penalty resulting from good latency.
3. Skill index doesn't change if you have 1.5 times the average dmg and lose half the average caps or defs. (NOTE: average = population avg, not personal avg). Likewise you can have 1.5 the average caps or defs and half the average dmg and still be equally skilled.
4. Each category is an equal reflection of skill due to symmetry in the formula.

UVs/Trinks/AT effects on Skill Index
1. Health bonus and AT are lumped together because both directly make room for more mistakes. One in defense, the other in offense.
2. Speed and damage buffs are lumped together because they all are secondary buffs with benefits that depend on playstyle. I assumed any such buffs are on a loadout because they are useful. (eg. A pure swordsman will not wear full gunner buffs, all BKC/chaos users use guns, swords and bombs)
3. Buffs can only help so much, so they are allowed to reduce the Skill Index only down to a certain range.