The Collective Win Rates of every role in the game(WIP)

Birdie Sànders·Jun 12, 2017·2 votes·515 views

WIP

Vanguards(UPDATED FOR PLAYRATES)

Zac: 53.75 Amumu: 52.62 Sion: 51.4 Skarner: 51.35 Rammus: 51.12 Leona: 51.03 Alistar: 50.37 Sejuani: 49.86 Malphite: 49.64 Maokai: 49.22 Gragas: 48.88 Galio: 48.74 Nautilus: 47.49 Rakan: 47.08 Poppy: 46.24

Collective Win Rate 46.73%(UPDATED FOR PLAYRATES)

Assassins:

Ahri: 53.04 Pantheon: 52.01 Shaco: 51.84 Nocturne: 51.23 Fizz: 50.01 Kha'zix: 49.56 Ekko: 49.25 Kassadin: 49.18 Quinn: 49.08 Katarina: 49.08 Talon: 48.95 Akali: 48.9 Diana: 48.35 Zed: 46.71 Leblanc: 45.13 Rengar: 44.46

Collective Win Rate: 45.9%

ADCs

Draven: 52.7 Twitch: 52.21 Miss Fortune: 51.97 Caitlyn: 51.13 Kog'maw: 50.87 Jinx: 50.33 Xayah: 49.81 Vayne: 49.61 Tristana: 49.09 Lucian: 49 Sivir: 48.71 Kalista: 46.95 Jhin: 46.51 Ezreal: 45.56

Collective Win Rate: 46.01%

15 Comments

Sp33d Zer0Jun 12, 20173 votes

Where are you getting collective win rate from?

Birdie SàndersJun 13, 20171 votes

Well I add them up, and divide them by the amount of champions in that role, for example if there are 10 champions in that role, than each champion is responsible for being in 1/10 if the games where that role is played. This algorithm is going to change, because I forgot to incorporate playrates.

Edit: the new system is, I must figure out the average play rate for a champ in that role. For example, there are 15 vanguards, therefore, if you pick a vanguard, there is a 6.7% chance of picking any single vanguard. If the play rate is higher that 6.7%, than I calculate the percent increase add additional win rate percentage because champions with higher play rates will have won/lost more games than others in that role

EndlessSorcererJun 13, 20171 votes

Which website are you using as your source for statistics?

Birdie SàndersJun 13, 20171 votes

Op.gg, since it appears to be widely accepted as accurate

Sp33d Zer0Jun 13, 20171 votes

I mean, if you looked at the collective winrate of all marksman, since they are only played bot lane for your selection, there winrate collective HAS to be 50%

LankPantsJun 13, 20171 votes

It would (approximately, Ziggs and Mord bot are still things) if you multiplied the winrates by the playrates and divided by the average playrate to normalise for playrate.

Sp33d Zer0Jun 13, 20171 votes

Both mord and ziggs are played so rarely botlane now that neither is classified as bot laners on stat trackers.

chipndip1Jun 13, 20172 votes

Ahri

Assassin

Quinn

Assassin

????????

DIANA

ASSASSIN

????????????????????????

RAKAN

VANGUARD

Delete this thread. Do it over.

Jo0oJun 12, 20171 votes

Don't you need to balance these numbers against play rate? Otherwise you give outliers too much weight...

Birdie SàndersJun 13, 20171 votes

Thanks, gonna change that

R7 FactorJun 12, 20171 votes

Shouldn't ADC's be A LOT closer to 50% since almost every game each team has 1 ADC?

Birdie SàndersJun 13, 20171 votes

Your right, it's because I didn't incorporate playrates, so the meta champions have more sway than the non meta champions

koshkyraJun 12, 20171 votes

Of course Draven is at the top of the list..

QuephaJun 13, 20171 votes

This statistic doesn't give any useful information. You are getting values other than 50% because you aren't weighting the champions by how often they are played.

ParoeJun 13, 20171 votes

@thread You realize by including the non-meta champions youre purposefully creating misleading statistics? Take the top 5 of each class and use THAT to determine average win rate. Which should almost always be around 50%.