Table of Contents
/Dev: Ranked and Matchmaking in 2020
"Hey everyone! Cody “Riot Codebear” Germain here, Product Lead for Competitive Gameplay.
Back in January, we started a conversation around improving ranked for 2020. Today we’d like to run through our big goals, how we’re planning to meet those goals, and share where we’re going with our next steps.
GOALS
Over the past year we’ve been collecting your feedback to help us build clear goals for a more competitive, transparent and rewarding ranked system in League of Legends.
Specifically, here are the areas you called out:
- Improve queue matchmaking quality without compromising queue time and availability.
- Improve transparency around ranked and matchmaking systems.
- Improve progression satisfaction and skill expression in our systems.
- Make rewards more recognizable and relevant for time spent in League.
- Players can play with, and find, others they want to play with.
ALMOST HERE: MATCHMAKING
We're starting with the first goal, which covers matchmaking. We’ve all been in games where we feel like we’re at a disadvantage from the start. We want to take a look under the hood of our matchmaking system and make sure that the current system is balancing player matches out the gate to the best of its ability.
Our first two ranked matchmaking improvements are in testing and nearly ready for launch. Expect the following features within the next few patches:
Given the sensitivity of matchmaking quality and the impact it has on players, we’ll be constantly tuning these features behind the scenes as necessary. Once they've been live for a while, we'll circle back and let you know how they performed—potentially alongside a status update on...
- Autofill Balance - As announced in Season Start, these updates aim to balance the number of autofilled players on each team.
- Duo Balance - This update aims to balance the number of premades on each team.
IN DEVELOPMENT: NEW AND AUTOFILLED PLAYER ADJUSTMENTS
There are a couple standout areas where our matchmaking service can be improved to more accurately assess skill level. We've begun exploring system improvements to better handle them.
- New Account Seeding - For players entering ranked for the first time, we think we can better identify true skill level by looking more holistically at their playstyle.
- Position Informed Autofill - Though our current system does not specifically measure you against all potential positions, we do understand that position proficiency has a significant impact on the game. We’re balancing the number of autofills on each team, but we would also like to balance the positions autofilled if we can. We think there are small adjustments that we can make without putting risk on areas of the system that are already working well.
THE SECOND HALF OF 2020 AND PRESEASON
Past the four specific improvements above, we’re also deeply investigating opportunities to fix some more complex pain points. The products of those investigations will be coming later in the year (anything big will wait for preseason), but expect updates as we go. Here's where we’re working:
WHAT SHOULD YOU EXPECT NEXT?
- Game Ruining Behavior - We’re attempting to develop better ways of dealing with disruptive behavior. Examples include better detection and punishment for behaviors like intentionally feeding, AFKing, win trading, and griefing, in addition to giving more frequent feedback when we’ve taken action on unsportsmanlike behavior you reported.
- Rank and Matchmaking Rating (MMR) Transparency - The hidden relation between player’s visible ranks and matchmaking rating has been causing confusion and frustration ever since the tier system launched years ago. We’d like to provide more clarity to players around their rank and updates to where we can fix the disconnect in meaningful ways.
- Promotion Series - We know that too many promos have become a source of frustration. We're seriously examining the current promotion system and will be making changes to shift the role it plays in our competitive ecosystem.
- Flex Queue and Organization - Hovering somewhere between a social competitive queue for groups and a quasi-competitive queue for those looking to try out some new things, it's unclear exactly what Flex is supposed to be best at. We’re taking an immediate look at loosening up restrictions for group formation to make it easier to play with your friends, and how we can tie Flex to the greater competitive ecosystem.
The Competitive Gameplay team is dedicated to making League the most rewarding and competitive experience out there and we’ve got a lot of work ahead of us. As we do, please keep the feedback and criticisms coming. Many of the changes we're working on are frankly overdue, and we want to make up for lost time by being as attentive and responsive as possible as our updates roll out through the year.
You’ll be hearing from us again in a few weeks once we've got results to report for autofill and premade balance. See you on the Rift!"
/Dev: TFT: Rise of the Elements Learnings
"As we approach the end of Rise of the Elements, it’s time to reflect on what we promised for the set, how we did, and what new lessons we’ll take forward as we move into the next chapter for Teamfight Tactics. (In case you missed it, the next set will be TFT: Galaxies. Head here for an intro to the new set mechanic or here for a preview of a few of the upcoming origin traits.)
This is a long one, so here's the TLDR:
Looking Back…
Looking Forward…
- TFT: Galaxies will have a LOT less random targeting for individual spells and will be more focused on specific things you can play around.
- Balance was better in Rise of the Elements than in the launch set (more viable strategies, fewer drastic shifts), but there's still room for improvement.
- Traits with strong disables (like Glacial) were less frustrating in Rise of the Elements, but we're still keeping an eye on them moving forward.
- Soft counters are good, hard counters are not so good. Expect us to continue making more soft counters and even fewer hard counters.
- Galaxies will feature one big mid-set expansion instead of the smaller one-off trait and champion additions.
- Bugs are still bad, and while there were fewer of them in Rise of the Elements, we need to be even better.
- We want spells to have higher impact, have greater clarity, and be more exciting visually.
- Expect more viable carries across the cost spectrum.
- Major traits (the big ones) should be a viable option, but should always be a little bit weaker than a truly creative combination of traits.
- The Spatula is coming back to the carousel but in more controlled and… unique ways.
- The new ranked season will start immediately with Galaxies, and the reset will work differently.
Looking Back
Let's talk about how we did applying our major learnings from the first set to RotE, including ways we hope to continue improving going forward.
We want to keep random effects in positive and controlled spaces.
We made progress here, but not as much as we could have. While it’s true that some of the more insane things like Phantom and Hextech were gone, there were still some effects that were random and fight-deciding. At the beginning of RotE, Mountain was often considered “Reverse Phantom,” though it didn’t end up being as impactful as was originally feared.
The bigger issue was our spell targetting.
Having so many champions with randomly-targeting spells led to a lot of extreme variance in fights. Taliyah and Azir were especially rough because they started with very high mana pools, which led to fights being decided by where their first cast went. Some random spells worked out quite well (Malphite and Lux) and added a healthy amount of fight variance. AoE Spells with variance make positioning a rich decision-making challenge, while single target CC spells that start early end up feeling like dice rolls.
For the next set, you’ll see a LOT less random targeting on spells and more focused on specific things you can play around. For example, one new spell is always aimed at the enemy with the most attack speed. This will allow you to use your game knowledge to outplay opponents more often.
Hold us to a higher standard for balancing the game so that a wider variety of comps can win in a given patch.
This is another one where we made significant progress compared to the launch set but didn’t nail it. The balance of Rise of the Elements was overall much better than the launch set, with way less outliers. But it also had its share of misses. Singed at launch, Brand in 9.23, Amumu in 9.24, “blender” Nocturne and Azir, and Yorick & Zed all jump out as clear balance issues that were a bit dominant in their given patch.
That being said, two things greatly helped balance in RotE. One is that the set’s design allowed counter options to come up in a lot of cases, so even if something was dominant, there were traits like Mystic and Cloud that could still give you a leg up. The other was that we were quicker to adapt when things became dominant. Strategic B-patches to target big outliers were helpful in bringing things to a more balanced state quicker. We also took a lighter-touch approach to balance, which led to less dramatic meta shifts, so people weren’t as thrown off patch to patch.
So while we made progress, we can still get better with balance. Continue to hold us responsible in making more comps viable patch to patch.
We’re aiming for less frustrating and fewer disables overall.
This is one we pretty clearly met. Compared to the launch set, there were just FAR fewer disables. The only really frustrating one was probably Olaf being Glacial Berserker. Someone with a 1.5 attack speed (or more) who AoE cleaves and freezes multiple champs was probably a bit over the line. Even when he was balanced or even sometimes underpowered, it just FELT frustrating to have your whole team frozen. Disables are healthy for keeping the pace of the game in check and creating tension in the battles, but when they get overwhelming (such as Launch 6 Glacial) that balance goes away.
Moving forward we will still have disables, as they are a key part of TFT champ design, but we’ll be keeping an eye on them and providing strategic and tactical ways to play around them.
Rise of the Elements will have more traits in the “soft counter” space.
This was a success, but also taught us a lot. Gone from the launch set were things like Dragon (100% magic damage reduction), and in its place were things like Mystic and Cloud. These traits allowed you to adapt and make smart decisions based on your opponents, but weren’t foolproof win conditions against certain comps. This was successful.
But we learned that even better-designed traits like Mystic still leave people feeling like they were hard countered. In addition, there were still a few traits that had that “hard counter” feel. (4 Desert vs 6 Warden being the biggest example). With this in mind, expect us to continue using soft counter traits even more while avoiding hard counters altogether.
Rise of the Elements will ship less total content patches, but each one will be larger.
While it’s true that we didn’t ship any content patches with only one champ like we did in the launch set, we’re still not happy with the way our content patches are coming out.
There were basically two major patches in Rise of the Elements that added content:
- 9.24: Added Lucian, Senna, Amumu, Soulbound, and UI improvements.
- 10.1: Added Leona, Karma, Lunar, and Item Reworks.
Other than that, patches were typically balance changes and small systemic adjustments (such as removing Spatula from the carousel). We feel like having two content patches that aren’t that big isn’t the best we can do, so for Galaxies we’re going to take a different approach. We’re going to launch one mid-set expansion, focusing our efforts on one specific cool moment to come back and see all the new stuff.
B-Patches will be extremely small and only for nerfing strong outliers.
We did this one really well. 9.22 and 9.23 didn’t have B-patches. 9.24 had a large one due to the holiday, but otherwise would've been fine. 10.1 had a small B patch that did exactly as we described (only nerfed 3 outliers), and 10.2 didn’t have a B-patch.
Our goal is to keep changes at a smaller and healthier amount than what we did in the launch set, and we seemed to hit that with RotE. The next step is to try to lessen the number of actual balance changes needed each patch so that things aren’t shifting around as much. That said, the B-Patch rules will continue into the next set, with us only using them to nerf outliers in extreme cases.
Rise of the Elements will have less bugs and issues.
http://www.surrenderat20.net/2020/02/red-post-collection-dev-ranked-and.html
No comments:
Post a Comment