I have two files: one taken 15 minutes ago and one 36 hours ago. The sample size I got was just over 400 games, so too few for any conclusions (though the fact that overall corp win percentage across whole OCTGN is 62,6% is a bit worrying), but I also noticed one thing that makes me suspect something is wrong with the page itself. I put all the data into an excel file to calculate the differences and in some cases they were negative (for example, Weyland BaBW has 4 wins and 16 losses less now than it had Saturday evening). Does anyone know explanation for this?
Yeah, db0 changed it to do the rolling past 6 months thing, and as a result we cant actually get useful things like ‘what are the win rates during set X’ anymore.
Why is this so? I can’t think of any advantage this has over running total from some constant point in the past, unless someone is interested exactly in stats from past half year but I see no reason why they would be.
It was changed because apparently there were too many matches to crunch and the system was running out of memory attempting to generate a cached page. That’s why the stats were down often before a few weeks.
Out of curiosity, how does Slaghund get rolling 6-month data? Is there an API interface to OCTGN? If the data isn’t more than an order of magnitude or so larger than the last data dump, I might investigate writing a polling script and mirroring the entire dataset on my local server so that I can run my analysis code whenever I want.
