A Character's Average Winrate Has NO BEARING ON THEIR BALANCE

Battlecast Sona·11/10/2016, 9:18:22 PM·2 votes·1,090 views

I am so tired of seeing "x champ is at a 48% winrate, they could probably use buffs."

I am going to explain to you mathematically why a statement like that is ridiculous, in a really easy way.

First we have to discuss the most important point: a champion's winrate is dictated by ALL GAMES played, not only experienced players. Even people with 0 previous games are included in winrate calculations. This means that if you try to use the average winrate you're including as a balance variable the fact that people are just bad with the champ. A champ being hard doesn't make them bad, so this is a useless variable that needs to be discarded. But how?

It's simple: Performance and Loyalty. Performance and Loyalty refers to how much your winrate will theoretically go up if you practice. Most sites do this in 15 games increments, showing the winrate fluctuation as you have more games played. This is the important statistic in determining a champions balance: If a champion has a 47% average winrate and playing 100 games on them only increases the winrate by 1% to 48%, they're probably too weak (mastery is not improving their ability to carry a game). However, a not insignificant number of champions have some interesting fluctuations...

For example, one of the most spikey examples of this metric: Azir.

Azir's average winrate is a baffling 42.5%. How is this even possible rito?! How could you ever let a character be this weak! how could you... oh wait, at 30 games played his winrate spikes a full 8.25% to 50.75%. It's almost like players who have never played Azir before or who have little practice are ill-equipped to play him appropriately, or something. (not saying Azir couldn't use QoL/buffs, just pointing out how innacurate average winrates can be in general).

This happens ALLLLL over League of Legends. Ivern, our newest addition, has a 47% average winrate that spikes to 55 to 57% at 30~50+ games. That is almost a 10% increase in winrate!

I hope you now understand a little better why so many people roll their eyes and shake their heads when you say "x champ needs a buff, their average winrate is only 47%!"

tl;dr: consider how much a champion's winrate improves with experience before spouting that they're weak based purely on their average winrate.

Source: all winrates from lolking

13 Comments

Astôlfo11/10/2016, 9:22:08 PM2 votes

Average winrate over all factors averaged is sort of meaningless, as it is too vague. But looking at specific factors and how they influence winrate (experience, game length, matchups, etcetera) help in determining balance.

CerealBoxOfDoom11/10/2016, 9:19:28 PM1 votes

A winrate does in fact have meaning for riot's balance

What people don't understand though is that they have no fucking clue about most champs winrates or the mechanisms that drive said rate

TehNACHO11/10/2016, 9:24:43 PM1 votes

OKAY MR. MATHEMATICS I GOT YOU.

Sites like champion.gg in fact isolates win rate by games played. You can see the distinct differences between the common beginner to the average win rate to the win rate of true mains.

This is besides the fact that champion.gg only uses Plat+ players, meaning that the general skill floor is in fact there. Hell, if I looked hard enough, I can probably find a League of Legends stat site that can isolate for every rank.

When used correctly, Winrates can say a LOT about champion balance. Do not throw win rates as a whole under the bus because you want to play smartass and point out how incorrect usage of data can lead to incorrect conclusions. This is common sense.

Sariat111/10/2016, 9:26:22 PM1 votes

You're totally right! Kalista is Obviously OP and needs nerfs!

Primaquarius11/10/2016, 9:29:11 PM1 votes

Champion winrate should only be relevant per person, not as an overview