Quick LoL Thoughts: Oct 23
How we assess personalization, Preseason PBE home stretch, TFT things that didn't ship
These posts will often contain talk about future work we're doing, or planning to do, that isn't yet guaranteed to ship. The work or goals could change depending on what we discover, and projects mentioned may get delayed or even stopped. That said, I believe it’s important to share as much as we can even if plans may change.
Before getting into the details I wanted to outline our high level philosophy around position power. We look at positions in terms of game influence - which is their ability to impact the overall outcome of the game (win/loss) through good or bad performance. First off, we don’t believe that all positions should have the same level of game influence. Positions can vary in resources, map positioning, and skill ceiling which creates natural power positions in the game. We think rather than fully flatten the natural strategy and skill variety in LoL we would prefer to support the uneven role power but balance them into a tight spread so that all positions have the capacity to carry the game with good performance.
For example, we think a position like support, which has less intense skill demands throughout the game and lower resources, will have slightly less influence on win/loss than the other positions but they can make up for it by boosting their team’s performance through vision and utility. Worth noting that we aim for support influence being slightly low not extremely low influence as it was in the early days of League. Also we increased top lane influence this season, because we thought it should be on par with the other farming positions.
If you’re interested in our methods - the current metric we use to track position influence/power is how much an MMR difference (used as a proxy for good or bad performance) in that position affects the outcome of an otherwise even game. In a powerful position, a MMR difference can be highly impactful to the game outcome, whereas it has less of a difference in a weaker one.
Current state of the game, we see top dramatically improved from the last two seasons, and jungle remaining powerful. Our next focus is on Mid and Bot Carry positions, which aren’t far off but could use some slight adjustments.
Bot carry power has decreased slightly this season, so we wanted to put a small amount of power back into the role. We are exploring a few options, and we may not put power directly back into XP.
We think mid should be among the strongest positions in the game, but currently it is carrying a bit too much power. Slight nerfs should bring it in line, and we still intend for mid to be a highly influential role. We will likely have to change some mechanics (not just tune numbers) to effectively nerf mid without indirectly hitting top.
We’re currently working on a few more improvements to matchmaking - an improved algorithm and ranked seeding.
We're improving the matchmaking algorithm to close in on a player's skill much more quickly as they play games. This'll reduce the amount of time every new player goes through the initial calibration period where the system doesn't yet know how to accurately matchmake them. It also reduces the amount of time smurfs are in matches well below their actual skill level.
Ranked Seeding will additionally improve the speed that we find a player's skill by using previous matchmaking data from normal queues to “seed” a player with a better rating when they first start ranked. Before this change, we always used a fresh rating (around high silver) for new ranked players, which led to problems with both smurfs and players that are truly new.
We’re planning to iterate on and test these changes slowly to make sure that they don’t have any unknown negative effects, but our goal is to have these fully rolled out some time this summer.
Thanks again everyone for playing. Stay safe out there.