Tweet
Statistics of TWL 18's regular season. Last year I included terrs and sharks, but I wasn't happy with the correlation between performance and the stats in those ships, so I've excluded them this year.
Obviously, stats don't tell the whole story - there's no metric for shooting through the holes in coffin while cramming or chasing down an enemy spider in the mid while breaking - so please don't spam my inbox, thanks!
The stats are pulled from the TWL website. The "Rating" category is an excel formula that is heavily weighted around the "Kills Per 30 Minutes" column. I think it's the strongest metric available, as it doesn't punish you for blowing out other teams (whereas Kills divided by Games Played would), gives a pretty good idea of how efficient the spider is at killing, and doesn't punish hard pushers for low KDA.
Additional weights to the "Rating" include Teks, % of kills that are Spiders, Games Played (for confidence in your KP30M), and KDA.
These are all lightly weighted. I don't think KDA is a particularly valuable stat, as it punishes extremely pushy spids like Hellkite or Ease. KP30M is a much better stat for evaluating their contributions. Teks is somewhat finicky. For example, Kate Upton has 9 Teks, but 6 of them are from one game vs. Warhable. Moreover, you can put 2-3 shots into a terr and someone else gets the final hit and takes credit.
I've also given some weight to the percentage of spiders killed, but not too much. Spezza, for example, has a ridiculous 68% of his kills are spiders, which I think should be more valued than Minotar's 44%, which is pretty low.
That said, Mcvicar's 50% is especially low for an underspider, but I know that's because he dominated under all year which led to a ton of cram time which in turn led to him killing a lot of sharks. He also has the 3rd most Teks in the league. So I don't want to weight %Spider heavily and punish success and cram time, but I do want to reward high %, so it is weighted.
The heavy weighting of KPM is a bit of a detriment to underspids, even with some weighting for KDA, Teks, and %Spider which all benefit them. You can see Spezza going from under in 2013 to over in 2014 boosted his rating by 7 points and his KP30M by about 14.
That said, Acidbomber was a monster under last year and would have ranked 5th this year in Rating, despite being under all year. His spidering has really fallen off this year, and you can see the huge drop from 86 Rating last year to 68 this year. His KP30M went from 91 to 75, meaning he's way less effective and the light weighting of KDA means his 2.02 can't mask the drop in output, while losing even more Rating going from 11 Teks to 2.
Due to that handicap I've added a "U" tag in blue to everyone I'm aware of that was an Underspid this year so you can compare them separately from the Overspids if you want.
These could all be weighted differently to produce different ratings, I tried to find the best balance for one list. Nevertheless, everything except the "Rating" category are raw statistics and not the result of a subjective formula.
I also lightened the weight on Games Played and lowered the threshold to make the list from 1 hour to 45 minutes to include as many people as possible. You could make an argument that this is too low, but I found KPM to be more or less consistent from 2013 to 2014 whether the spider had played 2 games or 7.
Really the only thing you need to know is that JammyJames is a beast.
gj
Comment