First thing first, it was 126 pages long when I downloaded the 62 pages of rank 1 people. Now it’s 145 pages long. It seems the database isn’t processed completely, a week after the season ended. Damn incompetence.
Anyway, there are 12500 14500 players participating in the ranked sprint and about half of them ranked out. I’m not surprised. Rank 7, 5 and 3 were irrevocable, ergo if you had sub-50% winrate with a lucky streak, you could advance and not lose it in an unlucky streak. I’m somewhat surprised by the low total player number (in EU). This event was as inclusive as can be and yet very few players participated – or WoWs has very low player count.
I put these ranked out players into 10 groups, ascending by number of games. Winrate – obviously – decreased heavily among these groups, but please note how frags, damage, XP or plane kills were barely affected. So it’s not like the best players do visibly much better. I also listed my own stats with below-average frags and damage, stellar plane kills (I’m actually #30 in plane kills in season 2) and XP.
You can see each individual’s performance on this graph:
The low R2 values show that what usually people consider “skill” – doing damage by learning how to penetrate which part – is barely affecting the outcome. Getting good division mates and scouting did. I played only with CV and you can see my very low damage. My bombers usually spent their time hovering over a destroyer or baiting fighters and sometimes even the torpedo bombers were on scouting duty. I believed – and the data confirms – that most damage happens in the mop-up phase when the battle is already decided. I mean if one team lose 2 destroyers and all fighter planes, their game is already over, despite losing less than 10% of team HP. This is why frags has the best correlation, that is the best proxy for destroyer kills.
I wouldn’t read much into the game stats, the season was too short and too chaotic for any serious conclusions.
I’m much more interested in the ship popularity stats.
The top one is Giulio Cesare at near 10%. It is a relatively new and expensive premium, deemed OP by most players.
Next is Oktyabrskaya Revolutsiya with 9%, she was a giveaway in an event a year or so ago.
Third Kamikaze R, 7%, not available for purchase for years, a reward in an event some 2 years ago (but it was extremely grindy, so that I didn’t make it even though I was playing a lot).
WG are being handsomely rewarded for selling OP ships. That goes without saying, but I’m surprised at the level. 10% of all participants is a massive number when most people don’t ever spend a dime on the game.
LikeLike
@Stawek: more importantly, the top 5 winrate ships:
Gremyashchy 56.15%
Kamikaze 56.1%
Kamikaze R 55.5%
Destroyer 54.98%
Giulio Cesare 54.31%
None of them are tech tree ships. So the P2W ships aren’t just popular, but are actually good.
LikeLike
The stats in win rates are misleading. All the Kamikazes and Fujins are the same ship and they are all old. There are exactly zero noobs playing them, same as Gremyashchy, as they’ve been out of the shop for 2 years.
They are good but they weren’t OP when they were sold. Kamikazes were pretty much copies of tech-tree Minekaze, only the latter got nerfed a few times since then. The difference is torpedo reload and speed.
I am actually surprised that Kamikazes had such a low win rate. At 7% popularity, they weren’t playing against each other all the time (which would flatten the win rate, obviously) and they have stats on par with Nicholas. When the mentioned noob-exclusion is taken into account, it seems that players with 2 years experience aren’t all that much better than the average OR all the ranked participants are very experienced themselves OR there is a skill ceiling that can be reached in reasonably short time OR there is a hard ceiling (like IQ) that can’t be improved with game experience.
LikeLike
@Stawek: newbies generally play bad, but they are pretty rare. Most “n00bs” have 5K games and no brain… and this is tomorrow’s post, because the statistics are too good for a comment.
LikeLike
“no brains” pretty much means that the game is beyond their cognitive capabilities.
Any reasonably complicated game is an IQ test.
LikeLike
@Stawek: not necessarily. They can just don’t care and approach the problem with “lol shoot redz XD” while listening to music and chatting with “friends” on discord
LikeLike