The Roar
The Roar


Balancing games for fun and profit

League of Legends. How good is G2? (CHRISTOPHE SIMON/AFP/Getty Images)
11th September, 2018

Over the years, I’ve heard just about every reason you can imagine for wanting something nerfed in a game.

They usually boil down to two things – which are really just one, when you get under the surface.

“It’s too powerful” and “it’s lame” are basically synonymous with “I can’t beat this.”

Sometimes they are valid concerns, of course. Perfectly balancing an RTS or a MOBA with over 100 champions is basically impossible. Developers of a game like League of Legends are basically an under-staffed fire station, having to choose which fires are most pressing at any given moment (Zoe).

Something is always going to be best-in-class, or, at least, best-in-situation, but in professional play, that thing is usually self-policing. Pros know how to counter-pick and how to play around other parts of the map. If the thing gets too powerful they just ban it.

The other argument I have often heard over the years is that it’s good to maintain diversity. Pick, ban and play rates are often used to point out the dominance of a champion. Yasuo must be overpowered because he has a whopping 64 per cent ban rate in solo queue. The fact that he barely breaks even in win rate (51%) is moot to most people (more on that later).

Lee Sin was often pointed at in seasons gone-by for being ubiquitous in pro play. Not too long ago, pro junglers pretty much only played him, Nidalee and Elise, with the odd bit of Rek’sai thrown in. All of those champions have been nerfed right out of the meta-game for quite a while, in order to give others a chance to shine. Lee Sin has come back to prominence recently, but at nowhere near the same level as previously.

But in the last couple of years, a new argument – or a new spin on old arguments – was brought into the limelight. Brian Kibler, a member of the Magic: The Gathering Hall of Fame, and well-known Hearthstone celebrity brought it to my attention: “X is not fun to play against.”

Now, there’s an interesting concept: nerfing something for reasons of fun (or lack of). That is basically what’s at the root of all that stuff about something being lame or overpowered, after all.


Kibler, talking about Quest Rogue recently after the release of Hearthstone’s Journey to Un’Goro Crater expansion cut right to the chase. It wasn’t that Quest Rogue was overpowered (it had a sub-50% win-rate in ladder games), it was simply boring and repetitive. Every game played out in the same way, and you either had a deck that countered it, in which case you won, or you didn’t, and you lost. Yawn.

The Quest itself got nerfed twice. Kibler was far from the only critic of the deck, but he was the most visible, and his argument was better than most – more honest than most.

Going back to Yasuo, how does this type of argument apply to League of Legends? Well, Yasuo maintains such a high ban rate, despite his medicore win rate, because he is so un-fun to play against. Assassins, in general, are very frustrating to play against when fed. Yasuo is just the most hated of them all, but just look at a list of most-banned champions in solo queue and it reads like a who’s who of champions who snowball when ahead. The top ten alone features Yasuo, Zed, Darius, Akali, Master Yi, LeBlanc and Fizz. Of these champions, only one (Jax) has a win-rate over 52 per cent.

Fun is much more important in solo queue play than it is in pro play, but it is also similarly self-policing to some extent. Frustrating champions are just banned out, no matter how good or bad they are. The high ban-rates of these mediocre assassins shows how most players ban with their hearts, not their heads.

The highest win rate champion right now is Nami, and yet her ban rate is a minuscule 2 per cent. Nobody ever says “OMG I hate Nami, when will they nerf her,” so she is allowed to keep quietly winning games.

Does that mean that the “not fun” metric is a good one for RIOT to use when balancing the game? Well, yes and no. Part of the problem is that the word metric is so inappropriate – fun is highly subjective. And, even when there is a broad consensus, like with Yasuo, the community deals with the issue by banning him in two-thirds of games, so nerfs become unnecessary on that level.

Then there was Zoe. She has a lot in common with Yasuo. She is incredibly obnoxious to play against, snowballs like crazy, and destroys squishy champions in a heartbeat. She also attracts a lot of ire from the community. Yet her ban rate pales in comparison to Yasuo’s. After her recent nerfs she sits at just 11 per cent ban rate, but apart from a couple of weeks around her launch, she has never had a particularly high ban rate.

That’s because, as frustrating as her ‘miss every skill shot, kill you with empowered auto attacks and ignite’ pattern could be, the mechanics of the rest of her kit is actually quite difficult to utilize. Her Q and R gave her a high skill ceiling for the pros to make use of, but the floor granted by her passive and W was so forgiving that it felt like a monkey could pick up solo kills without having to do much. Her win rate doesn’t really reflect that, but the whole point of this article is that emotional reaction is just as important as statistical fact.


Don’t believe me? Take a look at Zoe’s win rate history. I hate her kit with a passion, I honestly do, but she has never had a win rate over 50 per cent in solo queue. Never. Only in high Elo, occasionally, did she peak over the midpoint, but across all Elos combined: never. RIOT still nerfed her anyway, but only due to community outcry. Oh, and her dominance in pro play, but that’s a whole other issue and this article is already getting long.

So, does RIOT balance around fun as well as power? You bet.