Red Post Collection: The Making of Yasuo's Wind Wall, Conclusion to Team Builder Experiment, and more!

Posted on at 8:41 PM by Moobeat
Tonight's red post collection features CertainlyT with a dev blog on Yasuo's Wind Wall, Lyte a conclusion on the recent Team Builder experiment that required Captain's to invite at least one friend, Riot Martlet with info on how can get involved with LoL Viewing parties, and much more!
Continue reading for more information!

Recent News


Dev Blog: The Making of Yasuo's Wind Wall


Our next developer blog is up and this time it's CertainlyT with an entry on Yasuo's Wind Wall - including the origins of the missile deflecting attack, why it was put on a melee carry, and how the team feels about it now that Yasuo has had plenty of time to settle in since his release in December. 

"Hey all, 
Welcome to another design blog entry! Today we'll be talking about Yasuo and, specifically, his unique ability: Wind Wall. How did Wind Wall come about? What unique challenges did we face in designing Wind Wall? Why does a melee fighter like Yasuo have Wind Wall? All of these questions and more will be answered in today's Yasuo Wind Wall design blog. 
I'll reiterate here that the design blog is still under heavy iteration, so if you've got any feedback on the layout, the topics you'd like to see, or even the name of the blog (please!), then let us know! 
Anyway, I'll pass over the mic to Champion Designer CertainlyT, the man behind Thresh, Zyra, Darius and, most recently, Yasuo. Onward!
Pwyff


When we introduce new abilities to League of Legends, our main goal isn’t about making them interesting on their own – instead, we want to make interesting play experiences that work in the overall framework of the game. In other words, ability design serves kit design (overall cohesiveness of the champion’s abilities), and kit design serves game design (overall cohesiveness of the game). This blog post is focused on how Wind Wall came about, and what makes it appropriate for Yasuo as a champion. 

Background

Projectile Blocking is one of those effects that's both obvious and intuitively appealing -- I’m sure many players wished they had the ability to bat a Nidalee Javelin out of the air as they saw it glide toward them or a wayward teammate. It was almost a certainty that it would make it into League at least in some form. The challenge – aside from creating the tech for it – was finding the right champion/kit for the ability. 
One of my first paper kits (basic ability descriptions that conceptualize a champion’s role and core play pattern) at Riot was for Darius, where he had the ability to swat his axe in a direction, knocking enemies to the side while also destroying any missiles on the way. It didn’t make sense on Darius, but I’ve thought about the ability a few times since - namely on what sort of champion the ability would work on. 
Fast forward to our Season 3 pro playtests and the spell came up again. I polled the players as to what sort of champion they'd like to see this kind of play pattern on, and most responded “support” without hesitation. Only one player gave a different answer: HotshotGG suggested the ability would be best suited to a melee AD carry. 
We agreed. 

Why a melee carry?

After reviewing our melee carries, a few things become clear: 
Melee carries have extremely dangerous combat patterns, which means they need equally powerful self-protection abilities. By virtue of being melee, they expose themselves to damage from the enemy backline of mages and marksmen: damage tuned to be high enough to kill 200 armor / 3000 health tanks. The speed with which a melee carry needs to kill – without defensive tools – limits the amount of in-combat decision-making we could put in their kits which, in turn, also removes the ability for their victims to react defensively (aside from staying far, far away). 
In the past, we’ve resolved this problem through some powerful and iconic defensive abilities, usually involving invulnerability/untargetability (Trynd’s Undying Rage, Yi’s Alpha Strike Chaining, Fiora’s Blade Waltz). We gate these afore-mentioned abilities with long cooldowns (typically on the R button), but that conflicted with another of Yasuo’s design goals: to create a light fighter who felt fair to fight on a moment-to-moment – rather than game-by-game – basis. This meant trying to avoid cooldown gating, where the champion was weaker than their opponent until they used their long cooldown abilities, in which case their opponent would become weaker than them. That back and forth is fair in the overall pace of the game, but doesn’t offer a singular moment of parity. 
Finally, both to make room for these powerful self-protection abilities and because we assume that melee carries are unlikely to appreciate altruistic power, we've historically not put utility skills onto our melee carries. This contributes to wide variances in their performance across games, something we call in designer-speak the "feast or famine" effect. In the context of League, it means champions feel immensely satisfying when they're ahead (feast) but they're almost useless when they're behind (famine). 
Wind Wall sought to reconcile all of the above problems by replacing temporal invulnerability (invincible personal states) with spatial invulnerability (invincible zones). This approach let us do a few things:
  1. It allowed us to tailor Yasuo’s defenses to function against the enemy backline without depriving the mid and front line of the ability to fight back against him.
  1. It enabled unique counterplay through enemy movement. Where Undying Rage requires you to outrun Tryndamere, Wind Wall tests the opponent’s skill to reposition relative to a static zone and the Yasuo player’s skill in creating that zone such that his opponents can't work around it. Thus, Wind Wall is less of a dominating ability than Undying Rage.
  1. Wind Wall protects teammates. A Yasuo player who has performed poorly throughout the game can adapt his playstyle to protecting his teammates using the spell. By comparison, a Fiora who has underperformed can only execute a lackluster version of the same play patterns she has when she's ahead.

How did it all pan out?

Now that Yasuo’s live, we have a lot of data to judge the ability. Overall, I would say that Wind Wall has been partially successful by the metrics laid out above. 
It's a powerful tool to deny the backline the ability to focus Yasuo when he engages, but Yasuo is also too potent a diver to make fighting in the midline an optimal strategy. Last Breath's synergy with allies like Vi, combined with the late game power of his shield made him far better as a backline dive-assassin. We’ve iterated substantially on Yasuo since, but only time will tell if his play pattern evolves to a point where Wind Wall is used as a midline protection tool. 
The counterplay of re-positioning to avoid the wall has played out well, but the instant nature of Wind Wall often leaves opponents feeling like they have no way of preventing their most important spell from being blocked. One drawback here is that the successful counterplay against Wind Wall can only feel satisfying well after the initial bait. Generally speaking, the clearer and more direct the link between action and result, the more satisfying the play. Baiting Wind Wall with a Nami Bubble and waiting for the wall to dissipate before tossing in your nukes is less rewarding than immediately knocking Katarina out of her Death Lotus, for example. 
Finally, the team utility element of the spell is where I think Wind Wall has exceeded expectations. Repeating the feast or famine paradigm above, when melee carries get behind they're typically forced into split push strategies because they don't feel they can meaningfully contribute to a team fight. With Yasuo, we've seen a lot of Yasuo players team fighting (from solo queue to LCS) even when behind. A great deal of that is due to the fight-turning power of Wind Wall (in addition to the extra crowd control capabilities on Last Breath). Regardless, it's been rewarding to watch players improve with Yasuo in their approach to team fighting.
CertainlyT""

A conclusion to the Team Builder Captain requirements experiment!

Lyte has posted up the results of the recent Team Builder experiment that required players who wanted to queue as a Captain to invite at least one friend.
"Hey everyone, 
When we first launched Team Builder, we saw an overwhelming number of Captains and not enough Solo players to fill out teams. For the past week, we've been testing an experiment in Team Builder that requires players to invite at least 1 friend to their group to become a Team Builder Captain. 
The experiment saw some queue time improvements; however, we also received a lot of feedback that Solo players enjoy being Captains, and were willing to have longer queue times to have access to that feature set. We also uncovered data showing that a large number of queue time issues were due to players who were high (or low) MMR not being handled well by the matchmaker. 
In light of the data and feedback, we're going to stop the experiment and go back to allowing anyone to be Captains for the time being; however, we'll be working on the following features to improve queue times:
  • We'll be working on a way for Solo players to see what roles and positions are currently in demand, allowing players to "fill" for teams.
  • We'll be working on a way for the system to suggest more Captains or Solos depending on what's currently needed.
  • We'll be optimizing the matchmaker specifically for Team Builder, to solve the issues with high and low MMR players having trouble getting paired onto teams (or finding teammates for their teams).
As a quick reminder, Team Builder is a living system. We want to continue working on improvements to the system, whether it's optimizations or new features but because of the constant experimentation, there's going to be rough patches here and there once in a while--please be patient, give us feedback and help us make this the best system it can be. 
See you in Team Builder,
Lyte"

When asked about the recent change of having to be level 25 to queue for Team Builder, Lyte replied:
"We're running a lot of experiments in Team Builder at all times, and messaging for this one slipped a bit. 
Right now, we're trying to see different ways we can improve the Team Builder experience, and we're seeing some queue time issues and match quality issues with lower level players. The vast majority of Team Builder games are Level 30 games, and because there are fewer low level players playing Team Builder, queue times are longer for low level players and the matches tend to be pretty lopsided (and sometimes stomps!). This is ultimately a poor experience for brand new low level players getting their feet wet in League of Legends.

We'll collect some data for awhile with Team Builder set to Level 25 while we brainstorm ways in the future to make Team Builder accessible for all level ranges without the decreases in match quality. For example, maybe it's something as simple as allowing you to be invited into a Team Builder game by someone who is Level 25 or higher. But, we haven't figured out what we want to do here yet."
He continued:
"As far as we know, the Team Builder flow or user experience is a pretty positive one for new players to League of Legends or MOBAs in general. So, I agree there's immense value in giving a Team Builder option for lower level players; however, right now, the majority of Team Builder games are at the Level 30 range. If a low level player has a high-level friend invite them into a Team Builder match, they usually have no issues with queue times because they end up playing higher-level players. But, if you're a brand new low level player queuing up to Team Builder alone, the long queue times and poor match quality is a concern that we'd like to address.

We want to modify Team Builder to be accessible for all levels, but we want to take a moment to gather the data from low level play in the first two weeks and see how we'd like to iterate on the low level experience."

In response to some strongly worded feedback regarding letting Captain's now queue without friends and the new level 25 restriction, Lyte commented:
"I'm all about feedback, but let's keep it constructive. 
The experiment saw improvements in queue times--but not enough to justify removing the ability for Solos to become Captains entirely. In fact, if we dig into the data at a higher resolution, some demographics of players saw improvements, while others did not--that's not a slam dunk. We're going to work on a feature that suggests more Captains or Solos depending on what's in demand, which is going to be the best of both worlds. 
That's not back-pedaling, that's just making a tough call, but one we think is the right one. 
--- 
About the restriction at Level 25 for now, that's an experiment. One, 95% of players playing Team Builder aren't below Level 25. In fact, the majority of players playing Team Builder are Level 30, which is why we're seeing some queue time issues and match quality issues for low level players. We want to provide a Team Builder experience to players of all levels, but we need to take a moment to assess the data and see what the right next steps are."

When asked about giving Marksmen the ability to choose their supports specifically ( i.e I choose Caitlyn but I only want a Taric support ), Lyte replied:
"Right now, Team Builder only looks at Position and Role when matchmaking players onto teams. If you think about it, if you choose to play Karthus-Mage-Middle, the Team Builder system is currently only trying to find teams that are looking for Mage-Middle, and not specifically looking at your Karthus proposal.

If we were to add more champion-specific filters like "I want to play Caitlyn, but only with a Taric," we'd 100% increase queue times and not optimize them! But, if we keep optimizing the system in other ways, and more players keep enjoying Team Builder (and reduce queue times through player volume), there's always potential to expand the capabilities of Solos or Captains in the system."

Lyte also replied to someone curious if and when Team Builder will replace other queue types, such a Blind normals.
"I think there's potential in expanding Team Builder to possibly replace some queues; however, it's too early to tell. We're still very much iterating on Team Builder and running experiments every few days, and it wouldn't be responsible (or fun for players) if we made Team Builder the main queue when it's going through so much iteration.

Let's revisit this question down the line, when we've created a polished, awesome system."

He also commented on the idea of incentives for queuing for certain roles or positions:
"We're working on ways for players to see what's currently in demand by teams, and choose options to fill for those teams; however, I don't think we're at the point where we need to give players rewards for playing specific roles.

If you think about the psychology behind Team Builder, a lot of the reduced negativity / increased teamwork comes from the fact that you made a choice and got to play your champion, your position, your role, when you want. Your fellow teammates made the same choices, and together as a team you can now collaborate far easier than before. If we start paying players to play a specific role, you lose a lot of the benefits of Team Builder; this is related to research on the psychology of motivation, and how giving people a reward could change the motivation for why a person is doing something. We want players to play a Mage because they want to play it, practice it, or enjoy it--not because they are being paid to do so."

As for Team Builder games not currently adding to summoner's total games played, Lyte commented:
"We'll be fixing Team Builder stats to count towards Normal Stats in the EOG and Summoner Profile in the future."

To accompany all this, Riot Eglorian posted this announcement in the service status forums:
"We hope everyone has been enjoying Team Builder so far!

We have reduced the minimum number of players needed to start a game as a Captain. You can now start building your team as a Captain without having to invite another player.

For more information straight from the team behind Team Builder, check out this forum post!"

Find Viewing Parties Near You

Riot Martlet has posted up an announcement about League of Legends community viewing parties and how you can get involved!
"What’s more fun than watching League of Legends? Watching with friends! What’s more fun than watching League of Legends with friends? Watching and making new friends! Players all over the world have been organizing events in all kinds of venues to view everything from LCS matches to community videos. 
Whether you’re looking to find when and where events are going on, or hoping for some help getting the word out about your event, the Viewing Party hub is here for you. Check out every event we know about, including information on the venue, dates, times or costs to attend.
Or, if you’re the organizing type, you can submit the time, place, and what’s on the viewing slate along with some other basic info and we’ll be in touch to get your event on the map and help out in the other ways we can. Not every event is approved, so visit the FAQ to make sure you’ve provided everything we need. 
The interactive map is dynamically updated with events as they register, so you’ll always know what’s happening in your area, no matter where you happen to be. For more information or to submit your next event, check out Viewing Parties today!"

State of the 'Dinger

20thCenturyFaux swag walked on to the forum to share a few comments on Heimerdinger and the numerous, small balance tweaks he's been receiving since his VU.
"ヽ༼ຈل͜ຈ༽ノ Heimerdinger thread ヽ༼ຈل͜ຈ༽ノ 
To my mind, any change to Heimer should keep in mind the massive disparity in power by skill and coordination; if the game is either plat+ or a ranked 5s, he sufffers. I feel Heimer's recent competitive play was quite healthy and enriched the games he was in--but I don't want to buy that viability with bronze blood xD

His power disparity has been the subject of a lot of my attention since the rework (before it, too :P ). It seems to come down to a few key factors: respect for the threat level of turrets, having to learn their counterplay, and turret AI being a help for beginners but a hindrance for experts.

I'm quite confident the recent buffs and nerfs moved Heimer in a better direction--better skill expression lategame, more difference between RQ and RW use cases, weakening the 'caught alone but coughed up RQ before dying' play, etc. My current Heimer strat is another period of data-gathering focused on understanding the power disparity, potentially followed by a second berf.

To that end, what factors do you folks think are most impactful at the extremes of skill, both high and otherwise? What about coordination, solo vs 5s?"

No comments

Post a Comment