4Game

BLUF: Keep stats so you know what's happening.

"My last play session was 20 games of 5-15/15-5 and I hate it." We've all felt that way. Losing sucks, and winning a blowout isn't much fun. But how often does it really happen? Player stat sites like tomato.gg and wotlabs.net can tell you a lot of things about your results, but they aren't very detailed.

Enter WOT Replay Analyzer. Its creator, Aim_Drol, isn't supporting it anymore but the really important functions still work. You can export a large selection of battle information to a .csv and open that with your favorite spreadsheet program. Once there, you can filter and sort and contextualize all your data.

Some key points: only complete replays can be analyzed, so if you hop out of a battle as soon as you're dead it won't contribute to this data; it looks at replays from your replay folder so the more you have the better your sample size; it gets a little slow the more games you have so be patient while it's opening. Exporting data is fast once you make your selections.

Once your battles are listed, click on one and then press Ctrl-A to select all of them. Press Ctrl-E to open the export dialog. Click the "Export to Excel" tab to choose what data is sent out. The program remembers your choices so you won't have to change this next time you run it. Make your choices and export. Almost to the fun part.

Open that .csv and start sorting! Let's look at an example. I chose to include Game Type, Score, Finish Reason, and Duration among other things. Game Type is used to filter out skirmishes and Steel Hunter. Looking at only the games where Finish Reason is "all vehicles destroyed" I have a sample size of 1,936. Of those, 28% were 15-5/5-15 or worse. Only 12.8% were five minutes or less (surprising).

So what's the point? I know that when I'm keeping track of things losses don't feel quite so disheartening. Data helps give some perspective that your feelings might need. It also helps you identify vehicles you're doing well in and places you can improve, and is a good complement to the stat sites.