@Lyte - Player Behavior, Matchmaking, and Life as a Scientist

First Riot Post
Comment below rating threshold, click here to show it.

YOUR CHAMP IS OP

Senior Member

07-28-2013

I'd love to hear your thoughts on the fact that negative reinforcement has been repeatedly shown to be not only ineffective, but also detrimental.

Any ideas on stopping the pointless and ultimately ineffective negative reinforcement, and using positive enforcement for perceived "good" behavior, instead?


Comment below rating threshold, click here to show it.

Jaxon

Member

07-28-2013

Are there any pet projects that you've personally wanted to tackle but haven't been able to justify compared to other priorities?


Comment below rating threshold, click here to show it.

Kodoku

Senior Member

07-28-2013

Quote:
In the near future will there be any downside for loosing while being on division V of any league? Becuase not loosing anything is a huge reason many people in division V decide to troll.
You lose MMR, so you play with and against increasingly bad players. It's simply not true that you lose nothing.


Comment below rating threshold, click here to show it.

스르륵

Senior Member

07-28-2013

Quote:
Originally Posted by YOUR CHAMP IS OP View Post
More like players who are positive (more sportsmanlike) just won more games and thus weren't angry.

Like your loading screen tip says about players who are more negative - they win 23% less games right ? Guess why they're more negative AND lose more games. Yes, one causes the other, but guess what? It's the reverse of what you claim.

When someone is throwing your match, of course you are more negative.
welcome to Lyte's toy 'experiment' where he throws around meaningless (and/or often wrongly expressed) 'statistics' but nobody's smart enough to challenge the absurdness.

*edit: nobody smart enough to challenge the absurdness is left unbanned.


Comment below rating threshold, click here to show it.

YOUR CHAMP IS OP

Senior Member

07-28-2013

Quote:
Originally Posted by Š르륵 View Post
welcome to Lyte's toy 'experiment' where he throws around meaningless (and/or often wrongly expressed) 'statistics' but nobody's smart enough to challenge the absurdness.

*edit: nobody smart enough to challenge the absurdness is left unbanned.
I really like Lyte. I like his attitude, and I like how he tries to engage this crappy community. Having conversations with him about player behavior and reading his interactions is a MASSIVE improvement over beating your head pointlessly into the sneering, gloating, unwarrantedly self-important brick wall that is Pendragon's head.

But I think he is all wrong about how to approach player behavior. You aren't going to change human nature, and banning people is never going to do anything but bring someone back with a REALLY good reason to just do their worst: they just lost access to their account for, most likely, some words in chat that could've been handled with an ignore. It's just pointless.

So while I like Lyte, I wish he would get with the program, realize he's not turning the internet into Mr. Rogers' Neighborhood, and stop with the bans for harmless chat and start rewarding people who behave the way he (purely subjectively) feels is "right" and "good".


Comment below rating threshold, click here to show it.

The Aviatrix

Senior Member

07-28-2013

First of all, please do not remove or punish players playing 4-man premades.

I have a few friends I play with and end up in 4-mans a fair amount. But our lone teammate rarely knows it... In the majority of my premade games, only one person will really talk in chat to the solo player. So people who say "every 4-man premade is toxic"... may not even realize they're playing with a 4-man premade most of the time that they do so.
The only times people have asked if we were a premade, and these are rare:
1. Organized in champion select without talking. (One person chooses an ADC, then with 15s left, switches as another wants ADC - that sort of thing.)
2. Playing in a fashion that seems a bit more organized than is normally seen in solo queue. (Nothing fancy, just playing cohesively)
3. The random person is raging and I will defend the choices of whoever is being raged at. This is not 'we all gang up on the random person'. This is ME explaining why my teammates did X. Which I do when playing solo as well. And by defending a person - people assume you went into queue with them often times.
3b. If we had two or more ragers in a row, the next neutral-positive person may end up getting much adoration by the team and receive 4 honor for just being a decent guy. Even if they play poorly.



Anyhow, my question for Lyte...

Are you going to do a player behavior panel at PAX? If so, what day?
This would be on the top of my list of things to do if it's on Sunday or Monday (only two days I'll be going).


Comment below rating threshold, click here to show it.

41noob

Senior Member

07-28-2013

"mid or feed"

"you called mid before me and you're first pick? gtfo im going mid, you're going support ahri or else I feed!"

"lol u didn't ban x champion and other team picked x champion gg I troll u all"

^ fix the "toxicity" in pregame lobby and you're going to eliminate a ton of problems.


Comment below rating threshold, click here to show it.

SlightlySuicidal

Member

07-28-2013

Quote:
Originally Posted by DreamsOfGrandeur View Post
That's not an elephant at all.

The main reason for smurfs is due to toxic players making new accounts when permanently banned. (They're not even "high-elo", just more experienced than genuine newbies and casuals.)

All of Lyte and his department's current efforts are to reduce toxicity and lean away from permanent bans. Thus, smurfs will naturally reduce in number as they are more successful.

The "elephant" you see, is merely a by-product of a much bigger prevailing issue, toxicity..
That wasn't what I was talking about at all.

Challenger tier players (scarra, doublelift, chauster, voyboy, chaox, etc. etc.) were given smurfs with all unlocked content from riot at unranked status. They were stomping a bunch of bronze players on stream. The rumor is that this was a player behavior study. I was asking if that was true or not since it seems pertinent to ask the player behavior science guy about an alleged player behavior experiment.


Comment below rating threshold, click here to show it.

hargtard

Junior Member

07-28-2013

Hi Lyte,

I think what you guys are doing is really interesting and not just from the perspective of science. In particular, the tribunal is pretty interesting from the perspective of my area of professional interest, the law.

But before I go full-on TL;DR here's a suggestion - an honor category for people who are the targets of toxicity but don't respond by becoming toxic themselves or respond constructively. Maybe call it "Stoic" or "Cool-headed" I think it would give people an incentive to take the high road during the game and then after the game, when they see that little +for "cool-headed", it would give them a nice little boost to offset that residual sense of bitterness that comes from feeling like you were unfairly attacked and didn't defend yourself. Maybe that would help make toxicity less infectious.

Now on to the pretension!

1. It seems like much of what you are doing is influenced the "nudge" approach to government intervention of Cass Sunstein and others. Is this the case?

2. Since tribunal seems to have an explicit justice motivation, have you explored various philosophies of justice and/or various countries' legal traditions?

If so, I would love to hear more in depth discussion of what philosophy of justice/approach to law guides the organization of the tribunal (and player discipline generally). Particularly with regard to the focus on tribunal member discretion, Riot's decision not to create explicit rules for the tribunal to apply and the how players defend themselves from accusations/appeal judgments.

These characteristics seem different from the approach of the Anglo-American legal tradition (I can't speak for other traditions). For example, the Anglo-American approach requires the accused be given meaningful opportunity to respond to the accusations (what makes the opportunity "meaningful" depends on the severity of the proceeding). Additionally, one reason why the law is based on rules that are publicly available is to maintain predictability and consistency - people need to know what they can and can't do to modify their behavior accordingly to avoid punishment.

Now there are obvious pragmatic problems with making things too transparent - the an ability to confront your accusers (i.e. the people who reported you) would be a first class ticket to Toxicity City on the express. But I wonder if this leads the judgment of the tribunal to seem less legitimate to players who have an intuitive understanding of justice based on the legal tradition in which they were raised.

Which leads me to question 3:

3. Since an effective justice system needs to be viewed as legitimate by the people subject to it, have you done any research into how legitimate the tribunal seems to players generally and, more specifically, to players who are punished by it? Questions that pop into my mind are - Do opinions/reactions to discipline vary across country in a way that might be attributable to different native legal traditions? For those people who don't view the judgment as legitimate, what is it that they object to (e.g., the process of decision making or the standards applied?

Again, I really like how much thought you guys put into trying to create a situation where people will be naturally sportsmanlike and constructive and I'm always interested hear more about how you and your team think these things through.


Comment below rating threshold, click here to show it.

Windstorm

Member

07-28-2013

Quote:
Originally Posted by Lyte View Post
I have a pretty heavy Psychology and Neuroscience background from Bachelors to PhD; however, my type of position in the industry is getting more popular and common with time. As the industry moves more and more towards online multiplayer games, player behavior is simply going to be a problem that needs to be challenged and addressed in every online game.

Psychology is one of those fields that can be useful in a variety of positions. If you are an expert in Industrial or Organizational Psychology, you might have a position in Human Resources or Facilities figuring out the best way to organize desk layout to maximize interactions between talented people. The thing about Psychology is that it isn't really about the facts or raw knowledge you learn--it's about learning how to break down problems, and critically measure things that are difficult to quantify such as human behavior.



This is a really interesting idea, and we agree that toxicity tends to spawn a bit more after a death--part of this is because a player has very few options other than chat available to them during a death.

I'll have to think more about the idea of restricting chat during death.



Everyday is different for me these days, but the most important thing is that I'm learning something new everyday. One day, I'll be working in the trenches with the player behavior team trying to work on the latest project. Another day, we'll be spinning up a research project with a university lab to see how insights and research from League of Legends can inform and change the world of science. Or another day, we might be trying to fix a unique player behavior situation in Southeast Asia that we haven't seen anywhere else.
This. As a recent Human Factors Psychology graduate, the amount of different places and positions where applied psychology is needed can be a bit surprising to some people. For instance my particular area of specialty is User Experience design, though a friend in the same program now does large-scale Training & Development stuff.

Sounds like you have the kind of job I want to have, someday.