With Warships.today having been down for sometime (though the site is at least up at the time of writing this article) many players and tools (Matchmaking Monitor) have moved to using the WoWS Stats & Numbers Rating – PR, or Personal Rating to measure a players skill. Very similar in concept and construction to WTR and WN8 for WoT, PR gives you a score of 0-3000 based upon Damage, Winrate, and Kills (Frags). 1
What is PR?
As mentioned earlier, PR, or Personal Rating is a metric that takes a Players average damage, frags and wins, compares them to expected averages, normalizes them for deviations across the population, weights them, and finally combines them. While I don’t have full statistical data, I can tell you that the True average based on the supplied formula (below), is 1,150. This assumes that you are getting the same damage, kills, and winrate as the average. 2
PR = 700*nDMG + 300*nFrags + 150*nWins
In PR, damage is weighted very heavily, even more so than in WTR (700 vs 500), with frags weighted at 300, compared to WTR’s 200. However, as I am a bit of a ‘purist,’ I do wonder why he adds a weight of 150 to the winrate. I question this as the total of all weights, when correct for 100, equates to 115 (hence the PR average of 1150 instead of 1000). Perhaps Wiochi likes that number more, or there were statistical reasons at play that I just cannot see that led him to use that combination of weights, I just don’t know, but I felt it was worth mentioning. To continue on the the point of weights though, and adjust for the 1150 mean, damage accounts for ~60.87%, frags for ~26.09% and winrate for ~13.04%. This means that if you can control your damage and kills, and boost them to or above the average of the population or your ship, you can control ~86.96% of your PR. That’s pretty significant.
The Good and Bad
- Well known scale
- Easily recognized color scheme/scale
- Easily reproducible results
- Heavy weight on Damage can lead to damage farming
- Heavy weight on Frags can lead to kill securing/stealing
- Does not account for potential damage (likely can’t)3
- Does not account for spotting damage (likely can’t)
- Does not account for Capping/Defending Caps
Though the creator or PR does mention the “number of destroyed planes should be ignored – in WoWS AA is automated so no skill is needed to destroy plane” as a reason for not considering plane kills in his calculation (something that WTR actually does [either an oversite from my original article or something that has since changed]), I only partially agree with this assessment. First off, yes it is automated, so very little skill is required to shoot down planes directly. However, a players choice in captain skills for a particular ship, and their ability to manually focus a flight squadron are certainly indicators of skill. Furthermore, any damage done to enemy attack aircraft directly results in less damage done to teammates. So my personal opinion is, plane kills should not be overlooked, but just not weighted heavily. Another negative to this metric that I’ve noticed is in the normalization. Though I can only make observations on what is published, and not on the authors original reasoning behind the values chosen, the normalization actually pushes the weight more in favor of frags. This may be due in large part to the single expected values for frags. Therefore, if you can double the expected frags in a ship, lets say average 2 kills compared to the 1 of the Worcester, or even average 3 kills, that can boost your PR substantially. This is just a pure numbers game as it’s far easier to achieve (in theory) 3 kills in a game than roughly 246,000 damage required to triple the average damage.
Overall, PR is not a bad metric. Like WTR, it lacks some data considerations that I would like to see in a more inclusive metric of a players overall skill. However, I still prefer the WTR over PR in terms of it being a fairer metric for skill, though to be honest, I much prefer the Stats & Numbers site over Warships.today, even when Warships.today is working, due to the amount of data that is present on both individual players, but quite significantly, clans.
So if you want data for a specific player, I might recommend WTR over PR. However, if you want information about a clan as a whole, the use of the site WoWS Stats & Numbers is the better choice.