Stats Discussion

[pooty]

So there a few typos in the knownplayer multipliers, and I tweaked them a bit more.

However, the balancer algorithim isn't ever going to work like this:
anon Vs Enyo
Stain1 Leon
Stain2 Avg Player
Stain3 Stain4

It works by looking at all the players and ranking.
(1,2) > Red,Blue
(3,4 ) > 3 goes to which ever team has lower over PPH, so that'd RED (1,3), Blue (2,4)
(5,6) > 5 goes to which team has lower overall PPH so 5 could end up on either team RED if (1+3 < 2+4)
And so on.
Last player goes to the lowest team, if odd number of players.

So all this if you get say 5 top tier players...it should split them 2/3 but it always means the really good players are always going to have other good players too.

And if you end up with all the tank whores and they have all the snipers, or all the good flyer users vs. only tankers.. its gonna be fubar. No way to balance teamwork and strategy. You can rack up lots of points still spamming with Mino.

[Judas]

I've finished the first part of the scrape of map by map player data. I still need to get the node stats and whatever else that looks relevant on those pages. The map names are clipped and there might be some other small problems I may have missed. After that's done, I can start doing some machine learning and data analysis nonsense.

You can look at it here: https://docs.google.com/spreadsheets/d/ ... sp=sharing

If you spot any problems with this dataset, let me know. If I incurred any meaningful web hosting costs, let me know and I'll throw a donation in.

[pooty]

If I incurred any meaningful web hosting costs, let me know and I'll throw a donation in.

Nope. They claim its unlimited, and we've never been throttled so all good!