中文

How PR is calculated

PR (Personal Rating) compares a player's damage, kills and wins in each ship with that ship's expected values, weighted by battles, and turns it into one number. Higher means the same ships played better.

Formula

PR = 700·nDmg + 300·nFrags + 150·nWins
  1. For every ship the player has played, take the expected values — average damage expDmg, average kills expFrags, win rate expWins — multiply by the ship's battles, and sum.
  2. Ratios: rDmg = actual damage / expected damage, rFrags = actual kills / expected kills, rWins = actual wins / expected wins.
  3. Normalise: nDmg = max(0, (rDmg − 0.4) / 0.6), nFrags = max(0, (rFrags − 0.1) / 0.9), nWins = max(0, (rWins − 0.7) / 0.3).

The formula is wows-numbers.com's public PR algorithm (wows-numbers.com/personal/rating), credited as it requests.

Colour scale

0+
Bad
750+
Below average
1,100+
Average
1,350+
Good
1,550+
Very good
1,750+
Great
2,100+
Unicum
2,450+
Super unicum

Win rate uses the same colours at 47 / 49 / 52 / 54 / 56 / 60 / 65%.

Expected values

Dockside computes its own expected values every day: ships with enough battles in the current version use that version's server average, others use the lifetime server average, and ships without enough lifetime battles are left out of PR. Dockside PR therefore differs from wows-numbers PR; don't compare the two directly.

PR coverage: the share of battles played in ships that have expected values. With low coverage, PR describes only part of the record.

Current expected-value set

Loading expected values…