devtrackers.gg
ye
ye
Excited to have you!
How does this team measure success?
I think of Veigar and Senna from just a patch or so ago that got random buffs they didnt need, just to follow swift nerfs.
Is that a piece that your team owns? or are you just purely providing data and someone else is making decisions based on the analytics you provide?
Predictive accuracy is the relevant KPI for the examples you mentioned. We do make predictions on changes going out in a patch (typically for how much winrate change they'll result in, sometimes broken down by role or skill bracket). The Summoner's Rift Team makes the final decisions based on the data we provide, but it's a very collaborative relationship - we usually agree on things. Veigar and Senna were examples of missed results here. In Senna's case it's not shocking because the more creative higher scope changes are more likely to result in needing some followup, but Veigar is a sadge.
Honest question, were these action items always a thing? Or did they come about recently because of the more varied abilities and skills league has gained over the years?
We've had champion release plans for a while, I think Pyke was when we started doing them. Skimming through it, we had plans for:
please communicate more with the community because it is so weird when we see amumu get gutted because of "low elo" when he is objectively bad everywhere else. Why should amumu, the most telegraphed of all junglers be nerfed because low elo players can't dodge a bandage toss?
People that don't play all the classes or certain champions don't understand how difficult it is to play certain things or how it feels when certain things are nerfed.
In my case, I don't think there is a single Qiyana player are riot games which I think is concerning because some of the changes were so weird. We should know that all the voices of our champions are equally represented.
The most recent damage changes were good but nerfing base hp and hp regen is so crazy because if you have played the ahri matchup or the zoe matchup you would know this makes the champ unplayable in those matchups
We try to balance the game for every player as best we can. For Champions, they can't be allowed to be OP in any bracket of play, but also have to be viable in at least one bracket of play. Amumu is a lower elo skewed champion, so whenever he's too strong that's where he's too strong, even if he's already weaker everywhere else. The changes we did made him balanced in lower elo, but we definitely overshot the targeted nerfs.
For Qiyana, we actually do have a big Qiyana player on our team who consulted on all of the changes that have been shipped for her recently. I can't say there's a main for every champ on League (we don't have 158 League game designers or QA) but I can't think of any champion that no one who works on the game balance in some way plays. I know it can feel bad to see changes not agree with the context, but we don't do any changes without hearing multiple perspectives first.
Has former pros or even world champions applied for the job?
Former pros - yes. World Champions, I don't believe so. Maybe later this year when NA wins Worlds.
What happens if you drop out of Diamond? Do you just get fired or do you release an OP champ that you can abuse for climbing. Actually that would explain a lot of things... :P
monitor my op.gg - when I hit platinum that is when yuumi 2 will launch
How do you measure "Diamond 4 or higher"? I might have a D4 rank, but perform high gold level when not playing my main champion(s). It seems like versatility would be a requirement for playtesting.
Mostly just table stakes, other context will matter like versatility & experience but it's unlikely that someone who hasn't demonstrated they can reach Diamond 4 in whatever the most recent season is will be able to keep up in the role. Not impossible but we'd probably look at other candidates first.
He implied that they usually have these action plans for potentially disruptive abilities in general, so they probably do do this for similar things (like viego passive, akshan w, sylas r, etc)
yep
I mean then you always have the different definitions of high elo and low elo.
It's really hard to define at what point you are 'high elo' since the difference in skill grows larger the higher you go up the ladder. So many challenger players say that the Gap between Masters and Challenger is huge. So you could say that by this standard diamond is not really high elo. At the same time Dia4+ is already the 3-4% top of the whole server which you could define as being high elo. Idk.
The four brackets we typically use are:
"High MMR" for me would be in that Skilled+ range, aligning with your latter take (the top % of the ladder is meaningful despite what popular sentiment often says). With these brackets you see meaningful differences in data, which makes them useful. In Elite specifically the data is less reliable due to sample size, particularly on more niche champions or strategies.
I have a question. One of the main reasons Riot was founded was to be that players company and go against what the data says. Over the years has the SR team become more data driven rather than data informed? Being data informed was (to my knowledge) a major ethos of Riot.
My personal take as someone not on the SR team: over time, yes, League has become more data driven when it comes specifically to patch-to-patch balance. There are pros/cons to the approach, like how it removes bias in the process, but also how it removes room for creativity and doing things that would just be cool to do. On the whole though, I'd still confidently say that League and Riot aren't blindly driven by data.
What team requires that? Is it balance team?
The Game Analysis Team. To apply we look for a rank of Diamond 4 or higher so that we're able to keep the playtests reasonably competitive. We playtest and analyze most content that goes into Summoner's Rift to predict and prevent risks, and also empower designers to take risks by providing evidence that they'll work out (example: Akshan's revive won't break the game).
just last patch the 3rd one happened to riot galaxysmash on twitter for nerfing pyke mid and they pulled the nerfs bc of it sadge
pyke mid is op btw
Whew, my job is safe
ironically, you're on a team where being good at the game is a job requirement :)
I've seen many people flame Rioters for being low elo which makes no sense to me. You can be good at game design without being skilled at the game. And they don't make changes based on their own games but look at stats from all levels of play and consider opinions from different people.
A lot of Rioters who design Legaue aren't even low elo either. I still remember when Meddler made a quick gameplay thoughts and linked the op.ggs of the Summoner's Rift Team lol
You are basically saying that the best balance team would be a super-AI. Which I agree with but its unrealistic, so the point is kinda pointless.
Yes if we had ALL the data, we knew exactly what to do with it then yes we would not need to play the game. But we are humans and we cant do that well (riot is a prime example lol)
One thing I'd add to being entirely data driven is that at the end of the day, it's humans playing your game and what they think and perceive matters a lot, arguably moreso than what data says. This is one reason why we look at banrate when considering changes for a champion - something can be balanced "objectively", but banrate is an indication of if players are rejecting something for subjective reasons, which we should also take into account. And we do act on it when something is banned beyond a certain threshold, regardless of what the other data says.
Read moreIf any of you have came from the Destiny 1 & 2 crowd you will know how it feels when there is that disconnect from the developers to the playerbase. The beginning of the game for both D1 and D2 felt unplayable and even during the lifecycle they add in things like stasis subtrees where any brain dead player could easily see how overturned they are, the same kind of foresight the chemdrake and map implemented into LoL was what stasis felt like (unplayable unless you had it).
I remember watching a dev stream on upcoming content and Deej the community manager was trying to equip two pieces of exotic armor and couldn’t figure out why it won’t let him. Something a person who plays like 5 hours of the game would understand.
To Mort’s point about work life balance for working on a game and also being good at it, I’ve always wanted gaming companies to have at least one person on the decision making team where their job is to be good at the game. Just playing the game every...
Riot in general is starting to invest more in that space, and League has been pretty good about it for a few years now. Our Game Analysis Team (Diamond+ QA) gives feedback on all of the gameplay content that goes into Summoner's Rift and gets to pitch changes for every patch, sourced both from data and from experience playing the live game. They can't all be worked on, but there's always an earnest discussion about priorities and what's painful or going really well in the live game, and that helps keep the game designers in touch with the high rank perspective. Many of them play a lot on their own too in anywhere from Silver to Diamond, which gives us a diversity of perspectives on the game that lead to better changes.
if I ask nicely can you make remove the restrictions on building stacking items or building items with overlapping passives. i dont see why i shouldn't be able to build 6 cosmic drives if i wanted to
Not in the near future with the current items, sorry :(. They'd have to be tuned around the possibility of getting multiple of them, so individually they probably couldn't be as strong, so you wouldn't get to experience the 6 cosmic drive fantasy you want anyway. Having those restrictions lets them be cooler.
THORNMAIL IS RED YOU ARE NOT ALLOWED TO MAKE IT GREEN>:((((((((((((((((
you have successfully identified one reason we kept it red :)
what is so bad about 50% winrate? Why can't they wait until it actually goes to like 54%? I don't see how is she broken
she has literally no damage and is next to useless without that ult
She's already there in higher ranks. Average winrate is a misleading metric in a champion's first patch because the first ~ week of people not knowing what to build or how to play the champion weigh down that average. We usually look at winrate by day by MMR bracket (Normal, Skilled, and Elite when there's enough sample) in the first patch.