Red Post Collection: Client Cleanup: 2020 Recap & What's Ahead, Ask Riot & More

Posted on at 1:02 PM by Aznbeat

Today's red post collection includes Riot Scruffy's quick gameplay thoughts for February 12, a dev blog on the Client cleanup, this week's Ask Riot, a handful of 11.3 mid-patch updates, new Lunar Revel merch, and more!
Continue reading for more information!

Table of Contents


Quick Gameplay Thoughts: February 12

Here's Riot Scruffy's quick gameplay thoughts for February 12 - "Smaller Update Pipeline."
"Disclaimer 
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. 
Smaller Updates Pipeline 
In 2021 we’re interested in increasing the amount of smaller scale design work on Champions, Items, and Runes. We feel like our bigger projects (New champs and VGUs) and patch-by-patch balance changes are on a good track, but there's a middle zone that can give us a little more flexibility to improve the game. 
  • The types of projects this may include are:
    • Overly skewed towards Pro Play
    • Champions that have pattern-level (as opposed to power level) issues that create difficulties for long-term balance
    • They're picked frequently in professional games but have very low solo queue win rates, making it painful when we nerf them
    • These champs often display a pattern where “the more skill you have, the more you can remove the counterplay” as opposed to having difficult-to-execute combos that pros are just better at. As you might expect, at the pro level we’re seeing the weaknesses mitigated too far
    • Examples here - Upcoming Tahm Kench changes, Sylas changes in season 10
  • Game health or counterplay issues at all levels of play
    • These champs are often balanced (as opposed to the Pro Play skewed champs who end up underpowered for regular play) and even sometimes popular. However, they degrade the overall game quality because they have insufficient counterplay or promote strategies that lower the depth and interaction of other players
    • Examples here - Aphelios (s10 range and aoe burst reductions), Akali (s9 invis under tower removal)
  • Older champs that need love
    • Often these champs have reasonable power/win rates but players just feel they aren’t keeping up with the current game
    • These updates will often look to strengthen the identity of the champ and find ways the make their current power more satisfying, clear, or skill expressive
    • Examples here - Upcoming Rammus changes, 11.3 Jinx changes
  • Items not hitting their goals
    • Not appealing - Sometimes an item can be balanced but doesn't deliver on its fantasy in a satisfying way, so players don't buy it even in cases where it'd be good.
    • Unhealthy - Removing counterplay or skewing the shape of the game too far (like burst, sustain, range) are problems we want to fix. (Hextech Gunblade)
    • Unclear purpose - The item either doesn't have a clear set of circumstances where it's strong, or that circumstance is overly narrow. That makes it either generically weak, or a must-buy if tuned to be generically strong.
  • Runes not hitting their goals
    • Warping pattern - Some runes can be under picked mostly because they demand too much from the player and end up feeling like “playing the rune not the champion”. (ex. Omnistone)
    • Dominating choice - Some runes ruin the choice structure by being too generically powerful for too many different types of champion. (Conqueror has run into this a few times in the past.)
    • Game Health issues - like with items, runes can undermine important gameplay. Nimbus Cloak is a recent example where it was breaking the intended weaknesses of many fighters and tanks by giving them too much target access.
We’re hoping to establish a small pipeline for these projects, which should lead to a bunch of them shipping over the course of the season. 
Thanks again for playing League with us. Stay safe out there."

Client Cleanup: 2020 Recap & What's Ahead

Here's the most recent dev blog on the client cleanup initiative - "We’re recapping 2020's progress and going over plans for 2021." 

"We've made a commitment to share recurring updates on the work we're doing to improve the performance and reliability of the League Client. 
Today's update is the sixth blog in the series. (See the first, the second, the third, fourth, and fifth here.) 
TL;DR: We spent 2020 improving client start up time, architecture, champ select, end of game, memory leaks / client crashes and bug-fixing. We're in the middle of planning for Q1 2021, with a core focus on improving the social panel and continuing to improve the overall stability of the client. 
2020 Recap 
Our team formed late in 2019, and a lot of our early work was going through the process of understanding the state of the League Client. As a part of that process, we decided to focus on which parts of the client players globally (all of you!) were most vocal about to determine what we work on to make the client better. We still have work to do, but looking back at 2020, we were able to make meaningful improvements to the Client. 
  • Launched our global client player survey, which we use to determine what to work on. Most accomplishments you see below are a result of what players wanted to see addressed.
  • We got client bootstrap (startup) time down to 16 seconds for 90 percent of players globally (from 29.5 seconds at the start of 2020).
    • This was accomplished by architecture work, including a 63% reduction in bootstrap ember apps (ones that load on client start-up), and a 57% reduction in plugins within the client. In other words, we cleaned up a bunch of code.
  • Improved the responsiveness and reliability of champ select when picking and banning champions
    • Fixed performance issues while filtering through champions, and made improvements to the champ select grid itself
    • Investigated the usage of certain features in Champ Select. We'll talk more about this below.
    • Improved time to lock-in & ban champions
    • Fixed memory leaks in and around Champ Select (highlighted below)
  • Addressed the client completely freezing and black-screening when players return to it via the End of Game (EOG) flow after a game ends
  • Addressed the client hanging on the ‘Reconnect’ button during EOG, when it shouldn't have showed up at all
  • 175 bugs fixed in 2020, ranging from small stuff to issues that've been around forever.
    • Player names showing as “...” in chat
    • Taskbar not lighting up when receiving a message in chat
    • New players missing the Social Panel
    • Settings resetting when a player logs on from a different computer
    • Gift notifications repeating on login
    • Players returning to a black home screen after various actions in the client
  • Improved responsiveness of the Collections tab and each of its sub-sections
  • Reduced the Runes tab loading time by 40%, made the Runes interface smoother and more reliable to use
  • Addressed 30 memory leaks across the client. This is a particularly difficult area we discovered later in the year so we're especially happy we were able to ship improvements before the holidays.
    • Memory leaks are caused when a part of the program fails to release it’s memory after completing a task. This leaves less overall memory for other programs running on the system. The compounding results are always bad, but the causes are notoriously difficult to pin down, especially in large applications.
    • Prior to our first memory leak fixes in 10.23, players in Riot regions experienced 193,000 out-of-memory crashes each patch. By the last patch of 2020, patch 10.25, we saw a 55% drop in these crashes!
As we mentioned, we used your responses to the global client survey to guide decisions on where to focus our efforts. The above improvements represent progress in the areas you were most vocal about, but ultimately this only matters if you all feel things are moving in the right direction. That's why we're happy to report that, in addition to game-based metrics, we've seen positive survey trends across the board, meaning more and more players are feeling a less frustrating client experience! 
...Less frustrating. We still have a lot of things to accomplish this year, and we'll continue to provide updates. 
Low-Use Champ Select features 
If you had a chance to read our previous post, we mentioned we were investigating removing some low-use Champ Select features to improve overall performance. After evaluating favorites, missions, and emotes, we decided against removing any of them and will instead work on optimizing them. We've already fixed an issue with emotes, and are currently working on making sure favorites and missions are working as expected. 
What’s on deck 
We're currently finishing up planning for the first quarter of the year, but here's what we currently aim to address next for the League Client: 
Social Panel 
We've seen an uptick in issues regarding the social panel over the course of last year. In December, we fixed an issue where hovercards for friends in-game weren't displaying the game type or champion. Here are the next issues we plan to address:
  1. Social panel not connecting
  2. Friends list not loading, incorrect status being displayed, friends not populating
  3. Player won't be sent a friend request if their friend list is full
  4. Ranked information disappears from hovercard when a player is in a normal game
  5. Chat stutters when typing messages in quick succession
  6. Folders reorder themselves between sessions
  7. LoR and Valorant folders go missing when sorting by "Group Games & Servers"
  8. Clicking ‘Show Dates” in chat box will cause text boxes to overlap
  9. Friend Request notification only appears when the option is toggled off, and vice-versa
We're also going to deprecate an old Javascript framework (more on JS in a sec) and move the social panel entirely to ember. That transition is a good opportunity to go through and clean up rough spots in the code, leading to more reliability and fewer bugs. 
Client Stability 
We're now going to introduce two new focus areas for improving League Client stability: Client crashes and Javascript errors, which can have any number of effects, including.... client crashes. Rather than focus on a specific portion of the client for this work (ex. champ select, client bootstrap), we're taking a broad approach across the client. Together with the stability work we were already prioritizing, here are the four different buckets:
  • Reducing crashes: In patch 10.25, players across 16 Riot regions experienced 2.2 million client crashes. After our initial investigation, it seems the best option for us to address the majority of crashes is to upgrade our version of Chromium. This is the web browser at the heart of the League Client's architecture.
  • Reducing Javascript errors: In Patch 10.25, players across 16 Riot Regions experienced 36 million Javascript (JS) errors in the client. While not all JS errors result in a crash or performance hit, even invisible errors can eventually stack up, leading to those bigger types of impacts. Driving this number down will lead to a range of benefits across the client.
  • Continue to fix memory leaks: Fixing some of these were a big win for us last year. We want to keep big-winning this year.
  • Continued architecture work to clean up our codebase: To upgrade Chromium as mentioned above, we have to complete a bunch of prerequisite work which, itself, will further improve client performance. We plan to complete this work and kick off a Chromium upgrade by the middle of the year or so.
    • The last time we updated Chromium was December 2019, a few months before the Client Cleanup Campaign kicked off. We saw good results from that upgrade (CTRL+F "Chromium") and we'll highlight the improvements that come with this next upgrade once we're closer to launching it.
We appreciate the feedback we continually receive around the client, and we are listening! We had a good 2020, and are going to continue with the high momentum to improve the client for our players!"


Ask Riot: Where's Rod of Ages?

Here's this week's Ask Riot - "Why we removed some items in preseason, role strength throughout the game, and MMR.":
"Welcome to Ask Riot

Today we’re talking about items, MMR, and how we think about role strength throughout the game. 
Also, please send us all your questions! It’s the only thing that makes the editor smile. 
Why were some items removed in the new season? (i.e. Rod of Ages) 
While we try to avoid removing items that players love, sometimes they violate certain principles which require them to be either removed or reworked. Items in the new season were typically reworked or removed if: 
  1. They violated game health principles for counterplay, like Hextech Gunblade as a point-and-click tool for assassins.
  2. They had overlapping purposes with other items, like Statikk Shiv being very similar to Hurricane.
  3. They were near-mandatory first purchases for certain champions in a way that couldn't be shifted to be more flexible, meaning they'd still end up restricting choice for those champions if turned into a Mythic item (Rod of Ages was a necessity for mana champions who needed health and wanted to scale; Hextech Gunblade’s sustain and hybrid AD/AP stats made it a requirement for many hybrid damage dealers).
Despite being very, very gold efficient, Rod of Ages was only used by a handful of champions due to its stat profile (Anivia, Twisted Fate, Ryze, Singed, etc.). Notably, it left out manaless champions like Mordekaiser, Rumble, Akali, and others who’d be interested in a hybrid damage/durability/healing option. It was also too powerful for the champs who wanted it to consider other options. 
We actually tested several versions of Rod of Ages as a Mythic item, such as featuring a transformation mechanic and cool unique passive like Seraph's/Muramana. However, when balanced, users felt like they were irrelevant for too long while waiting for their Mythic item to scale, and it was a mandatory purchase for champions that wanted to scale (rather than a choice). This ultimately violated two of the principles for Mythic items that we were aiming for—satisfaction and choice. 
The Rod of Ages Mythic iterations eventually became Riftmaker, which opened up the healing/durability niche for manaless champions. Mana users looking to prioritize survival should be able to situationally purchase Everfrost, although that item has been underpowered and is receiving buffs in the next patch or two. 
PhRoXzOn, Game Designer, Summoners Rift Team

Are there any plans for making the ladder system more transparent? The ELO and the MMR make it a bit difficult to understand. 
We’ve discussed this a few times, and the reasons we’ve stated still hold true today. MMR is basically the current estimation of your skill—and for it to be accurate, it will at times over or underestimate your abilities. If you’re only watching MMR, losses can feel like the game telling you you’re getting worse, when the truth is just that the system was intentionally testing you against tougher opponents and learning how to match you more accurately. 
By adapting MMR into Rank, players progress through the system via Tier and Division milestones that put more focus on long-term growth than the result of any one game in isolation. One isolated loss won't knock you out of a rank, and one isolated win won't promote you into the next—your performance over time is ultimately what impacts your displayed Rank. This system also lets us include things like demotion protection, which keeps you from being overly punished for a streak of losses at an unfortunate time. 
We’re going to keep working on ways to make both the Rank and MMR systems feel smooth and accurate, but we don’t want to make either system do a job it wasn’t designed to do. 
Riot IAmWalrus, Lead Competitive Gameplay Designer

What is your goal for individual roles at various points in a match? Should roles have similar strength at all points of the match or should some be stronger? Would love to hear thoughts on what seems like the developers trying to let ADCs have a strength curve more aligned with other roles. 
The goal is for each role to feel like it can meaningfully change the outcome of a match. Due to the shape of the map (dragons spawn earlier and are bot, Baron spawns later and is top, mid lane can roam to other lanes, etc.), exact parity at all points in the game isn’t something we strive for. In fact, we think unique power curves are a good thing. Slight advantages and disadvantages in power at different times make for a wider range of decisions and strategic elements to consider. 
Likewise, different roles do and should index differently as to what kind of strengths they are bringing to a match. A support undoubtedly has less 1v1 combat power than a midlaner, but they often set the vision up for a battlefield and the utility in their kits becomes critically important as the game goes on. Put another way, they have less direct combat power so they get to index higher in strategic capabilities. 
In addition to having some variance in power between roles, we find power-over-time is a healthy way to diversify the feeling of our champion classes. For instance, Marksmen have some pretty clear distinctions between early game bullies (Draven, Caitlyn) and late game hyperscaling champions (Vayne, Kog’maw). This helps players pick a style they enjoy or one that fits the team they’re trying to create. 
For the Marksman item changes in preseason, we thought their late-game-skewed power curve was slightly too sharp. This led to situations where games ended before they could reach and make use of their power spike. So we changed Marksman items to allow them to spike earlier, which was more to match the reality of the games than to sync them with other roles. That directionally has seemed like an improvement, but we still intend to keep their power curve skewed toward late game overall. 
Riot Sotere, Senior Game Designer, Summoner’s Rift Team 
Have a question? Head here, drop your question in the box, and ask away."

 

11.3 Mid-Patch Updates

A small amount of mid-patch updates went out, fixing some bugs & hotfixing some balance changes for ARURF! Check out the details from the Patch 11.3 notes:

"2/5/2021 ARURF Balance Changes 
11.3 General 
  • MANA TO HP CONVERSION RATIO 0.4 ⇒ 0.35
11.3 Buffs
  • AATROX +20% damage dealt & -22% damage taken ⇒ +25% damage dealt & -25% damage taken
  • GAREN -10% damage taken ⇒ -10% damage taken & +5% damage dealt
  • KATARINA +5% damage dealt & -5% damage taken ⇒ +10% damage dealt & -5% damage taken
  • RUMBLE -5% damage dealt ⇒ Normal damage dealt & -5% damage taken
  • VIEGO Normal ⇒ +5% damage dealt & -5% damage taken
  • VLADIMIR -8% damage dealt & +8% damage taken ⇒ -5% damage dealt & +5% damage taken
  • YUUMI -20% damage dealt ⇒ -15% damage dealt
11.3 Nerfs
  • AHRI -5% damage dealt ⇒ -5% damage dealt & +5% damage taken
  • ANIVIA +15% damage dealt & -15% damage taken ⇒ +5% damage dealt & -5% damage taken
  • ANNIE -5% damage dealt ⇒ -5% damage dealt & +5% damage taken
  • DIANA -3% damage dealt ⇒ -3% damage dealt & +5% damage taken
  • EZREAL +5% damage dealt & -5% damage taken ⇒ Normal
  • KAI'SA Normal ⇒ -5% damage dealt & +5% damage taken
  • KASSADIN +5% damage dealt & -5% damage taken ⇒ Normal
  • MISS FORTUNE +10% damage dealt ⇒ +5% damage dealt
  • RYZE +5% damage dealt & -10% damage taken ⇒ +5% damage dealt, -5% damage taken
  • SERAPHINE Normal ⇒ -5% damage dealt & +5% damage taken & -10% healing
  • SIVIR -5% damage dealt ⇒ -5% damage dealt & +5% damage taken
  • SYNDRA -10% damage dealt ⇒ -10% damage dealt & +5% damage taken
  • VIKTOR -5% damage dealt ⇒ -5% damage dealt & +5% damage taken
  • XAYAH -5% damage dealt ⇒ -5% damage dealt & +5% damage taken
  • ZIGGS Normal ⇒ -5% damage dealt

2/3/2021 Viego Balance Change and Fizz Bugfix
Fizz
  • E - PLAYFUL/TRICKSTER BUGFIX Fixed a bug where Fizz would be locked out of casting other abilities until he recast his E - Playful/Trickster
Viego
  • Q - BLADE OF THE RUINED KING SECOND HIT HEALING VS. MINIONS 150% ⇒ 100%
  • PASSIVE - SOVEREIGN'S DOMAIN TOGGLE ABILITIES BUGFIX Viego no longer keeps Gragas' Q - Barrel Roll and Lux's E - Lucent Singularity abilities after having possessed them
  • PASSIVE - SOVEREIGN'S DOMAIN HECARIM BUGFIX Fixed a bug where Viego would cast different, incorrect abilities after possessing Hecarim while his W - Spirit of Dead is active
  • TRANSCENDENCE BUGFIX Fixed a bug where Viego would be granted additional permanent ability haste from Transcendence if he levels to 5, 8, or 11 during enemy champion possession"

 

Lunar Revel Merch Drop

New merch was released on the Riot Games Merch store celebrating the Lunar Revel! Check out the full collection here, or look below for some of the new items:

Miscellaneous

  • Riot Scruffy noted a change that went out in 11.3:
"PSA - in 11.3 we added additional LP penalties for AFKing or leaving games. Repeatedly AFKing will ramp this penalty up over time.  
With a little more skin in the game, hopefully a potential leaver will think twice about ruining a game for their team. 
Leavers never prosper"
  • Reina Sweet also noted some bot AI changes that went out in 11.3:
"One subtle change to LoL's Co-op vs AI that went out with 11.3: 
Bots are a little bit smarter. Bots will use new recommended items now. Which is a system that updates as people build differently. 
So, effectively, the bots are learning from you all. :)"
  • Riot KateyKhaos noted changes to Shan Hai Scrolls Neeko's recall coming with 11.4:
"We’re working to unlink it to effects quality, so the sun should display regardless, so folks playing on lower settings can also enjoy it.

Should be in for the next patch (11.4)!"
  • Here's Riot Tuxedo with some changes for the Match History API:
[1] "Very very very excited to tease the next version of the League of Legends match history API. Bunch of cool stuff: 
* Calculated and assigned position
* Spell cast counts
* Objectives stolen counts
* Turret plating event
* Champion transform event (Kayne) 
More to come on this. 
[2] "Also shout out to the League team that's been working on this. I get to tease it, but they've been doing the hard work. 🙌🏼

rough match-v5 change log

match-v5.getMatch

match-v5.getTimeline
Check out the tentative changes coming in Patch 11.4.

Reminder: Patch 11.4 will be live on Thursday, February 18!
"11.4 Preview with full changes. 
Nearly final but a few things may still change."
"Updated jungle changes (pulled back on a few of the nerfs)"

No comments

Post a Comment