Red Post Collection: Diana Discussion, Mystery Gifting may soon be permanent, Lyte on Team Builder & Feedback Loops, and more!

Posted on at 3:38 AM by Moobeat
This early morning red post collection features Vesh sharing this thoughts on Diana and her "role identity crisis", a note that Mystery Gifting may soon be a permanent shop feature, and Lyte discussing all sorts of player behavior topics , including comments on the ETA of Team Builder, player feedback loops, and the relationship between leavers & loss prevented.
Continue reading for more information!

Recent Posts:


Diana Discussion

In response to a lengthy forum thread about Diana, Vesh chimed in with his thoughts on Diana's current problems, her "role identity crisis", and what the future may hold for the Scorn of the Moon.
"This is an incredibly well done post and you've correctly identified a number of key issues with the champion.
"This is because she has a fighter/assassin design and you guys don't like shields on assassins or burst damage on fighters. You have accidentally given her too much durabilty during the situations where she gets a full shield off which makes her feel unfair to play against even though one crit will probably eat her shield assuming she gets it off and doesn't have to use hourglass immediately or get hit with some form of CC. "
This is pretty much the problem. Diana has a role identity crisis right now where shes pretty good at a lot of things but not really the best at any of them. When she's strong, she's good at way too much. Her vacuum on E also accentuates this problem because it puts a huge amount of her power budget into her team's followup ability (Diana when Rumble and J4 were really strong in competitive was really scary despite her average winrate in soloqueue/normals) 
Let's talk about both the fighter and the assassin role and what some options and challenges are for each.

Assassins generally have one primary goal: Get in, get a kill, and get out. Diana on live currently can get in, kill someone, and then just generally stays alive due to tankiness. She doesn't really have that many hoops to jump through to do this either. Squishier assassins like Akali have to measure very carefully when they go in and then have other elements of evasiveness (her twilight shroud) that gives her play/counterplay with the enemy that's both positioning and sight based.

Diana doesn't have this however. She just needs to jump a guy and hit all her buttons and if she's ahead enough then she wins. This is especially problematic due to her AP scaling shield (get defensive rewards for building offensive) and you correctly identified that as a core problem. Even without it, if she was powerful enough to just nuke somebody from full but then squishy enough to die for it, Diana post-6 might as well have a Q that reads "on hit, both Diana and her enemy die." Clearly not a very exciting ability for either party involved. 
I personally find her prospects as an AP fighter to be much more promising. Taking off some of her crazy burst while giving her tools to do more in fights late game could be a cool approach, especially since she has a cool tuning point in her passive. Increasing base attack speed could definitely allow her to make better use of her 3 hit passive pattern which is generally a little less "bull****" then getting Q -> R -> R'd to death as a carry.

There's other ways to do this as well, and I'm not saying in any way that Diana should be an auto-attack-only champion. Her passive definitely feels pretty good but to add to that, her pattern of making a grand entrance with Q -> R -> E -> Zhonyas is really awesome. After that effect wears off though, we need to ask what Diana should be doing. Should she get to auto burst a carry with another Q->R->R combo, or should she be using a mix of her spells and autoattacks to have a longer time to kill but with more sustained damage? 
I personally believe that this approach would be better for Diana overall and would allow her to have healthier tuning points.

This is seriously a great post though. I'd love to have more discussion about her."
He continued, explaining his thoughts on Diana's Q :
"Q is already an extremely reliable skillshot. it has a large width, moves very fast, and has an AoE explosion on the end. Making her not stop to cast it would accentuate this problem even more. 
Thematics are important, but should not compromise gameplay.

(Part of the reason why the Q being so reliable isn't very good is because it enables her to go R->R right afterward extremely reliably to just kill someone)"

He continued, commenting on his ideal role for her:
"I feel that it would be even cooler if she started off in a more "initiator" type role with her R->E and maybe zhonyas that set up her team to follow up with additional CC and then transitioned into fighter mode after. Diana feels really good when she's hitting people with her epic moonblade.

With the way she functions currently, her assassin pattern would basically just degenerate to an even more bursty Akali with less options and less points of mastery."

As for the idea of allowing Diana's moonlight to remain on other targets after using R, he commented:
"I don't think the core Diana fantasy is spamming R on everything she hits with her Q. Plus, this is an unrealistic expectation of her. When given the tools to jump a squishy, Diana isn't going to make 4 other pit stops on the way there, she is going to go kill the target she can kill immediately. By the time that target is dead, it's not really important if other targets have moonlight on them because her Q is on such a short cooldown. Plus if moonlight lasted long enough for this to relevant the enemy team would basically be permanently revealed and there's a ton of power in the fact that it grants sight.."

In response to someone suggesting Diana's shield scale off health rather than AP, Vesh commented:
"The shield scaling with health instead of AP has been suggested a couple of times. I agree with the goal, but it's hard to say if a specific solution will actually work when it's on paper."

As for how much work he thinks Diana needs to get into a better shape, Vesh commented:
"I don't think Diana needs any sort of sweeping changes. Outside of QRR, I think her core pattern is really fun. If you've ever played Diana in 3s you'll know how fun it is to skirmish with her when you can't nuke anyone out. There's something really great about a passive proc into your second rotation of spells as they come off cooldown. It's really satisfying."

Mystery Gifting May Become a Permanent Feature

In a reddit thread pointing out that the mobile Lunar Revel site accidentally said mystery gifting was here to say as a permanent featued, Hippalus commented :
"We are thinking of making it permanent but decided to first make sure players (both Mystery Gift senders and recipients) are happy with the experience during Lunar Revel before we make the final call. Somehow the wrong messaging made it into the mobile site. That will be fixed soon but I can still say that MOST LIKELY it will be permanent, barring unexpected survey results. Satisfaction was very high in previous surveys."


Lyte on Team Builder, Feedback Loops, and Loss Forgiven

Team Builder
Lyte popped up in a team builder thread to comment that the team is currently toying with the idea of a live beta of team builder using the current builds that have been testing on the PBE:
"We've been doing a lot more PBE tests of Team Builder, and are trying to see if we could run some Live Betas with the current feature set. What do players think? 
The system should be great for your particular situation Raptamei, have you checked it out lately on the PBE?"
He reiterated this in another thread, saying:
"We've been doing weekly playtests on PBE, and are trying to collect feedback to see if players would want to do a Live Beta with the current features. We know a lot of players are excited about Team Builder, and we're excited about it too "

Feedback Loops
When asked about the possibility of getting a notification when someone you reported in game gets punished, Lyte commented:
"I agree with some of the other players here, we have to break down the feature and problem space. 
I think players want strong feedback loops for every action they have in League of Legends (or really, anything you do in life). If you Honor someone, you might want to get some feedback about that action--did that person unlock an Honor ribbon later and you helped them do it? Did the person continue on a 3-game winning streak afterwards? If you Report someone, you might also want some feedback that the action did something. Did the player end up improving their behaviors, or getting a chat-ban? 
However, having feedback for every action is not the same as encouraging vindictiveness or revenge behavior. We've talked about more feedback for Honors and Reports in the past, and we're definitely open to the idea and have a few designs in mind; but, we'd probably never reveal player names in any of the feedback due to privacy issues and the fact that it'd create some unnecessary witch-hunting behaviors.

For now, the player behavior team is heavily invested in trying to get Team Builder in a state that players would enjoy on Live servers. Our research shows that a lot of bad behavior results from good people experiencing triggers or contexts such as having a bad day, having a bad game, etc. We haven't seriously invested time in preventing these contexts or triggers, and believe Team Builder is the highest value to players right now compared to working on things like improving feedback loops for our other systems."

On the topic of feedback loops, he continued:
"This is a pretty interesting story, thanks for posting. 
I've been doing a lot of thinking about online cultures versus real-life cultures, and why they are so different in some cases. In some of our recent research and data (which we'll be presenting in talks in 2014), we're beginning to see more and more that the differences in online versus offline isn't the people--it's the systems and structures in place (or lack thereof). 
As humans, we're all about feedback loops. We commit an action, we receive some feedback, and we learn from it and the result changes some aspect of our behaviors. This is admittedly a super simplification of feedback and learning, but an example would be we study for an exam, do well on the exam, and the exam result reinforces us to keep studying for future exams and we eventually learn that studying = good.

In real life, strong feedback loops are everywhere. Almost every action you do in real life has some kind of outcome--negative or positive, it doesn't really matter. What matters is that every action has an outcome. I'm beginning to see that online cultures lack a lot of feedback. In the early days of the internet, you could be extremely negative or racist, and nothing really happened and you received 0 feedback. But in real life, if you were racist, cultures have learned to step up and challenge racism. There are still gaps in real life cultures though; for example, there are studies showing that if someone needs critical help near a crowd of people, people are less likely to assist the person than if someone needs critical help in front of 1 person--this is known as the bystander effect. Further studies have shown that if you do need help in a crowd of people, you need to point at someone and say specifically, "you, please call 911" or similar. This call to action works the majority of times, showing that people are willing to help and good at heart, but that there are simply social or cultural barriers in place preventing the average person from simply stepping up and acting. But, by pointing at someone, you can create a hero out of an everyday person and now they do something extremely positive such as calling 911 which could save someone's life. 
So what happens when we make it easy for people to be heroes that stand up to negativity, and celebrate sportsmanship? What happens if we add a bunch of systems to real-life and online cultures that make it easy for players to be everyday heroes? Let's try to make being good the path of least resistance in League, because in the end, we all want to just have awesome experiences and fun in League.

Anyways, there's a lot to think about when it comes to the future of online games. I'm hoping Team Builder is one of the first systems to incorporate a lot of our new research and show what a system that makes it super easy to 'do the right thing' can do for the player experience. Through Team Builder, we're giving all the power to the players--we're giving more trust than we ever have before to the community, and so far, the games I've been seeing from Team Builder have been fun to watch. The number of teams trying double jungle strategies also makes me smile with glee :)"

Leavers and Loss Protection
Lyte also discussed the idea of better protection from people who abandon games and leave their teammates in a difficult situation.
"I'm not opposed to taking a second look at LeaverBuster or Leaver systems in general; however, it's all about prioritization on what problems we work on because we simply cannot work on every system at all times. 
If every player in this thread looked at their match histories, on average, they'd see very few leaves. We know that a percentage of leaves/afks are due to ISPs and connection issues or hardware, and there's nothing we can design to fix those problems. After you account for those 'organic' leaves/afks that we can't influence, you have what we call the potential space where we can improve leaving/afk. 
We know that if we reduce or remove the penalties for all teams if a person leaves the game, that a few things happen: 
1) Leaving/AFK actually goes up
2) The number of games that actually completes goes significantly down 
The problem with the above is that although players feel better when they get a leaver in their game and can leave without penalty, they get frustrated over time by the sheer number of games that will never finish because people are just leaving and quitting en masse in games. This 'hidden pain' is difficult to explain, but is very easy to see in the data.

I'm open to thinking of solutions for the specific problem of a person not connecting to a game because I think that's a miserable experience that is completely not the fault of anyone in the game; for example, maybe restarting a match if a person never connected in the first 2 minutes."
In response to the comments on offering loss prevented to players whose team mates failed to ever connect, he noted:
"Yep, we're thinking along these lines too. For this particular case, it's no one's fault, and we really shouldn't penalize players for being in a game where someone just happens to not be able to connect into the game.

What is more complicated is the "no penalty when someone leaves at the 40 minute mark" scenario."

No comments

Post a Comment