Tracking the number of players and the winrates of players using the Anonymizer over 50 battles

The Anonymizer

Don't ask me why, but I decided to log the winrate of players using the Anonymizer over 50 battles to see just who is using the Anonymizer

All games were played between T8 and T10 and no player names were logged. I did not log the winrates of players not using the Anonymizer during these games. All details were taken after battle and entered manually. There may be errors in the data, however I've screened for obvious mistakes (There's no extreme data)

I did not use the Anonymizer during these games to not screw the results, however this does mean in theory the maximum number of players using the Anonymizer per game is 29.

Once I had the data I broke down the winrates of players in to brackets of <45%, 45-49.99%, 50-54.99%, 55-59.99% and >60%

So, what happened?

Over the 50 battles I encountered 258 players using the Anonymizer. This works out at an average of 5.16 which is 17.20% of players

Over the 50 games, the most players I saw using the Anonymizer was 12 and least was 1

The average winrate of a player using the Anonymizer was 50.85%, with the highest being 65.57% and lowest being 43.53%

I also logged the highest winrate and lowest winrate of players using the Anonymizer in each game, which gave me an average winrate of 55.18% for the highest and 46.99% for the lowest winrate players using the Anonymizer

For each winrate bracket the counts and percentages using the Anonymizer were:

<45%9 players3.49%
45 – 49.99%102 players39.53%
50 – 54.99%107 players41.47%
55 – 59.99%35 players13.57%
>60%5 players1.94%

So based on this, I decided to compare what I saw with what I *should* have seen based on the EU server population's winrates, assuming 17.20% of all players are using the Anonymizer

The percent of players by winrate % on the EU server are:

<45%10.90%
45 – 49.99%61.30%
50 – 54.99%24.10%
55 – 59.99%3.17%
>60%0.57%

*Note – This is based on wot-news' EU server statistics

The expected counts for 258 players based on this works out as:

<45%28 players
45 – 49.99%158 players
50 – 54.99%62 players
55 – 59.99%8 players
>60%0.26 players

*Note – I rounded these to the nearest whole number. With >60% the count was below 0.5% so I rounded this to 1

If you're still reading well done, I'm not good at presenting data in an interesting way

Now, if you compare the actual counts to the expected counts there's 2 significant differences: Less than expected players below 49.99% winrate were counted, and more than expected players above 50% winrate were counted

The differences between the actual count and expected count worked out as:

<45%68.00% decrease
45 – 49.99%35.51% decrease
50 – 54.99%72.09% increase
55 – 59.99%327.95% increase
>60%400.00% increase

*Note – if the actual expected count of 0.26 was used for the >60% players, the actual increase was 1791.32%!

So what does this mean?

For me, it shows that the lower winrate players are either a) Not interested in using the Anonymizer, or b) Not informed about the Anonymizer, and that higher winrate players are more likely to use the Anonymizer

It also shows that although higher winrate players are more likely to use the Anonymizer, due to player winrate distribution on the EU server players above 55% winrate are essentially invisible amongst the masses, and those above 60% winrate are pretty much impossible to spot when anonymised

Although I have no evidence, I suspect that lower winrate players are less interested in the game and therefore much less likely to read any news articles, follow any blogs, watch the WG YouTube channels, read patch notes, or come on to Reddit. My guess is these players probably hit x on every popup in game, select a tank and hit battle, as they want to spend more time playing than reading/learning

I also suspect the inverse is true for higher winrate players; that they're more likely to be up to date with any news and changes to WoT and more likely to want to test any new features

Again, although I have no evidence, my exerience was that players with lower winrate using the Anonymizer were more prone to raging, insulting and ping spamming during games. This might link in to not being as well informed about how it works, believing that they're actually anonymous in game. Or maybe they just act like that anyway

Things to note

Obviously it is early days on the Anonymizer. I would expect the percent of players using it to increase in time, but probably not drastically.

My sample is only 50 games. There were times during the run when the percent of players was around 20%, however it did settle to 17-17.5% around game 35 and stayed there until the end

For anyone who wants to run their own analysis on the data, I've added it to a shared Google Sheets doc here: https://docs.google.com/spreadsheets/d/1Mjf11mj6aQPwf-QWTs0dnO2xyonOGMTpFKq2z8hSs9U/edit?usp=sharing

The data also includes tier played and tier spread

So what do you guys think?

Source: https://www.reddit.com/r/WorldofTanks/comments/earxtl/tracking_the_number_of_players_and_the_winrates/

leave a comment

Your email address will not be published. Required fields are marked *