I’m familiar with another kind of statistical normalization from my World of Tanks day, win rate curves. Basically, for any given vehicle, you plot players overall winrate versus their winrate in that specific vehicle. Here’s an example I found on Google:
The resulting curve tells you a decent amount about the vehicle. If it’s a perfectly diagonal line (The dotted reference line above), the vehicle is perfectly balanced. If lower win rate players do better than normal, it’s got a low skill floor. If they do worse than normal, high skill floor. Same goes for high win rate players. Most vehicles end up with one similar to the one above.
A vehicle only needs addressing if either A) The entire curve is above player’s average win rates, or B) one section is exceedingly higher than normal. Otherwise, deviation is fine, especially on the lower end of winrates. Not every vehicle is going to be equally good in all hands.
I bring this up because the Type 81 is the ideal example of a low skill floor vehicle. It’s very easy to use, simply lock the target and fire the missile. If it doesn’t hit a wall, that aircraft is almost certainly dead. Doesn’t require any skill to lead the missile to it’s target like other SAMs.
However, it’s also a low skill ceiling vehicle. There’s nothing good players can do to improve the aspects of it. At best, a more aware player is marginally harder to sneak up on. Otherwise, they’re as limited by the seeker head as everyone else. This also means it’s trivial to counter if you know one’s around, especially for helicopters.
Accordingly, as lesser skill players are a greater proportion of any group of players, you’re going to see overall win rates increase for this (And countless others, Zeros, Vampires, etc). Thus, Gaijin (For whom the concept of statistical normalization is a foreign concept, apparently), just moves it up.