Building Self-Managed Teams: A Case Study from Riot Games
codingsans.com
codingsans.com
So simple and so true. It's insane how many companies hire and build teams around visions their teams don't understand well at all. That's certainly where most killer companies win. Their force doesn't need to be extremely proficient and efficient to move the needle because everyone understands what moving the needle means and has a clear path to take small decisions without looking at a spec or consulting another party.
The best managers are certainly the ones that make an effort to enable learning around what is being build and why. That allows people to build agency and criteria to make decisions that are in the best interest of the company.
I've worked at precisely 0 companies in my time that have offered career opportunities that offer the kind of incentive that I would need to go all in and give a job all my attention.
If your definition of "full effort" means: 50+ hour weeks with no end in sight, arbitrary deadlines disconnected from reality that exist only to give some a-hole exec an illusion of lower risk, make-believe agile, inability to disconnect after work, and countless meaningless hoops to jump through, that's NOT something which will reap rewards in the long term and it is far from a healthy relationship.
"Full Effort" should mean a solid day of work, with enough time to have a life outside the office. Sure, sometimes there's a crunch and one needs to stay late. If the relationship is healthy that will happen a handful of times a year, max. Anything more than that, and you're getting used.
You might get lucky in the current job, but odds are against you.
A bit of a lightbulb moment for me as i didn't want to go back into people management. I now use this term to set out what i have and what i can do for the project and how that has lead to greater success/value of a product.
It also highlights to management what happens if i was to leave.
It's their rules. I'm just playing the game.
As someone who have experienced British managers in practice, this explains a lot.
I don't think your expectations are realistic. Most peoples' salaries don't increase monotonically throughout their career, they top out around 40, the day you get your last big raise. For most people that's the day they become a senior/top/lead/etc, whether they're a chef or a welder or a doctor or whatever.
I've seen several developers work towards management roles, make the switch and then end up moving back to IC roles after realising the grass isn't always greener.
There's so much to unpack here. Some doctors make 400k. Some make 700k. Some make 300k. Some are stressed and work a lot of hours. Others aren't and don't.
Some lawyers want to be on partner track at a big firm. They might be stressed and working a lot of hours so that they can make 400k or 500k or $1,000,000 annually. Some make $100k and don't work much, aren't stressed.
Most software developers make less than $150k. Most software developers I know are stress balls until they check out and then start working 24-30 hours a week and the rest of the office time is bullshit.
On the west coast of the US and in Fintech in NY some developers make 300-400k.
In VHCOL areas those numbers are much higher of course.
'full effort' might mean a variery of things, but if we are talking about working overtime to meet a sudden new requirement, or studying new framework over weekend, then thats a cancelled family trip and dissapounted spouse.
And if you do that repeatedly it will result in unhealthy relationships at home, which is much worse than at work
You miss all the shots you don't take.
I tend to put effort into my jobs and I've gotten very large promotions or bonuses multiple times now at jobs. Other driven people I know at both large tech companies and startups have had similar experiences.
Your original post did mention "freedom from the quarterly demand for metrics" which I don't see as driven from a company perspective. Companies and teams have complex sets of overlapping goals and generally you want to align what you do with those goals. Many of those goals are disjointed from what one would consider necessary for company success. If your boss wants quarterly metrics to go up for their own promotion then you're not going to get promoted by not making those metrics go up.
Salary? Isn't the above definition of employment?
What I mean is that I really don't think of my co workers as family at all and any company actively trying to promote that is going to find me very annoyed with that. It's not my family. Don't force me into parties and such.
What I am looking for though is peers that care about doing a good job. I'm tired of working with people that don't know how to do their job. That don't care about clean code. That follow the letter of agile practices but not the spirit.
I couldn't care less about the mission. Every company I've been at so far has had a completely different domain and many of them I couldn't have cared less about.
I belong to a social group that does barbeque and goes for beers. What would group succeed at if I go "above an beyond"?
Local sports groups dont win championships, most of the time they don't even compete. I struggle to find an example where this analogie hold.
Well, I suppose this would mean that instead of just being a passive participant, you go out of your way to look for that cool new pub that your friends might enjoy, or a way to make the hamburgers more juicy. You'd take some pride in your ability to provide this value to the group and enjoy the improved social standing of being 'the one who really knows this stuff', and of course just enjoying a better culinary experience overall.
But I've spent years in both startup and non-profit contexts, and I think intrinsic motivation is much more powerful, especially for work that requires creativity. The best people I've worked with, even in startup contexts, weren't focused on the money. They liked making things and solving problems.
And I don't think it's just my experience. McConnell's Rapid Development has data from surveys of developers; they were much more motivated by things like responsibility and challenge than they were by cash.
I think it's mainly a matter of training. People who want us to do things for them train us in responding to extrinsic motivation. That's not terrible; some amount of it makes the economy work, after all. But for me one of the essential characteristics of adulthood is deciding how to train ourselves. If you want, you can reconnect with your intrinsic motivations.
I don't think it's impossible for cash-maximizers to be good developers, but I do think it's hard. A lot of software is essentially a long-term investment, so the short-term thinking of people hustling for a buck works against it. More than once I've had to clean up code bases put together by contractors who were clearly in it for the money and it showed in their work. There might be some theoretical way to bridge the long-term needs via a complicated short-term incentive structure, but I doubt it. It's too prone to things like the "code me a new minivan" problem: https://english.stackexchange.com/questions/488178/what-does...
Don't get me wrong, I am in it for the money in the sense that I need a proper salary but I couldn't care less about a 'bonus' being dangled in front of me and putting metrics in place that I'm supposed to hit to get a larger or smaller raise and percentage of the measly bonus (and no, making the bonus larger doesn't make it better. Makes it worse).
I do my best every day just because I enjoy my job. I won't try to do it better because of some metric someone is judging me on. I might actually do a worse job if I find my employer trying to use these metrics to deny me proper raises. Worse in the sense that you will probably see the metric reached. Actual output or overall results for the company will suffer.
Conversely, it’s also insane how many companies hire and build teans around visions the teams know won’t work. The knowledge is there (perhaps distributed), but it never congeals, and everyone just plows ahead.
On the plus side, it keeps me employed.
I've seen so many managers cargo culting what they read somewhere, and force their teams to do it.
Riot's obviously far from perfect, but this advice seems reasonable to me. Do you think following it is likely to lead to problems similar to Riot's? If there's some sort of causal link between empowering teams and failing to build an inclusive culture, I'd really find that interesting.
I don't know. Listen to yourself. There are a lot of companies out there that manage people well, and also didn't have such pervasive sexism problems. Let's talk about those instead.
And before you say no such company exists, listen, I personally just don't say and do obviously sexist shit. It's really easy! A lot of people go their whole lives without farting on someone's face, unlike that Riot exec. So of course if you can't get these basic things right - if you don't know you can't fart on people's faces - why are we elevating your opinion on "self managed teams"?
The bigger takeaway here is that their sexism problem was so much more devastating than their executives could have ever anticipated. It has miscolored every single aspect of what they do, a complete trashing of years of hard work by thousands of people. Why in the world would you ever want to work for these people? There are a lot of shitty MOBAs out there, truly, there is nothing they have besides a player count, that will someday be zero that no one else has or ever will have.
> Advice always depends on the context
It's not clear to me whether that's true, or what it means exactly. It seems to mix at least two senses together, one (A) where it's self-evidently true, one (B) where it's not true. (I believe that's called equivocation, and it's a very common way of going astray, in philosophy, informal arguments etc)
(A) Yes, when someone tells you something, you should take into consideration who is telling you, like when you assess the credibility/reliability of what an unfamiliar website says. Not doing that would be very foolish.
(B) The sense of the original "quality of advice" quote: You can read a quote somewhere, not knowing where it came from, or it's some author you know nothing about, or anonymous etc, and think it's good advice, and follow it. Where it came from doesn't really matter, i.e. the quality of the advice is separate from its source. Good advice can come from a "bad" source, or bad from a "good" source.
League of Legends remains the premier e-sport, dethroning all competitors in viewership rates and numbers.
Valorant is rapidly rising, and threatens the long-incumbent direct rival Counter-Strike
Teamfight Tactics, while not an e-sport, still remains in the top 10 streamed games on twitch.
3 releases, all (at least by twitch streams) major successes. From the product point of view, they're clearly doing something right, and whatever work culture issues they're claimed to have, they're certainly outperforming.
Note: As most of riot's releases target the e-sport crowd, I'm using twitch viewership numbers as an analogue for success. Sports, after all, are judged by popularity amongst viewers above all.
Riot is a prime example of this, but unfortunately they are hardly the exception - it is a fairly common pattern amongst the AAA game companies and publishers.
The specifics are not the same from company to company, but arguably the end result is. People are unfortunately willing to endure a lot just to get into their "dream industry".
League of Legends is a premier e-sport more so because of marketing than game design or engineering. In fact, their game client remains worse than dota 2's in every way. It's actually pretty hard to take them seriously considering their client is still such a buggy mess after about 10 years.
While the viewership of casual Valorant is greater than CS-GO, it's also not "rapidly rising" - https://sullygnome.com/game/Valorant/365/summary
And the esport viewership of CS-GO is much larger compared to Valorant.
And while Teamfight Tactics was an success, Legends of Runeterra wasn't. So they are 3/4... which is still pretty good - but can also be explained by good marketing and IP handling. Similar to the new Activision-Blizzard.
Indeed, it's debatable because it sounds weird
I don't play LoL for close to decade due to their marketing (is it even marketing? or actually their Esport stuff?)
For me LoL is great game because it changes often, so you don't get bored fast, also it's highly competitive with decent ways to express your skill
It's when the money dries up that cracks in the foundation are evident.
It’s been said to death at this point, but technical shortcomings can be orthogonal to a video game’s success on the market.
[1] https://technology.riotgames.com/news/taxonomy-tech-debt
https://github.com/TerryCavanagh/vvvvvv (my favorite piece: https://github.com/TerryCavanagh/VVVVVV/blob/master/desktop_...)
(There was also discussion about it on HN before as well, if one is interested: https://news.ycombinator.com/item?id=22011358)
That said, refactoring for later gains might be wasted time if you have only a certain amount of scope planned for the rest of time, counter to the typical heuristic.
[1]: https://technology.riotgames.com/news/determinism-league-leg... [2]: https://technology.riotgames.com/news/future-leagues-engine
they pushed significant changes in backbone infrastructure in e.g EU
https://technology.riotgames.com/news/fixing-internet-real-t...
https://technology.riotgames.com/news/fixing-internet-real-t...
https://qz.com/790208/how-the-company-behind-league-of-legen...
Saying the game doesn’t appear to be “competently programmed” is analogous to the “coded badly” that gets thrown around, not taking into account that all successful consumer software has tech debt and inefficiencies.
(I don’t work for Riot and have no affiliation with them)
- https://kotaku.com/inside-the-culture-of-sexism-at-riot-game...
- https://kotaku.com/current-and-former-employees-sue-riot-gam...
- https://kotaku.com/top-riot-executive-suspended-without-pay-...
- https://www.espn.com/esports/story/_/id/26686991/riot-games-...
- https://www.vice.com/en/article/dy8g87/league-of-legends-fak...
- https://www.polygon.com/2020/7/29/21347587/riot-games-saudi-...
> COO, Scott Gelb [...], as a comedy bit, [...] has repeatedly touched subordinates’ balls or butt or farted in their faces. [...] his punishment—two months of unpaid leave and training. [1]
From OP's interview with CTO Mike Seavers:
> Every time we sat down to go through the past quarter and plan the next one at Riot Games, we came up with a leadership theme for the quarter. For example, we weren’t going to make decisions for others for a month. This put all of us on the same page, and we got everyone to think about it and talk about it. When my direct report came to me asking, “Can I do this thing?” It was easy for me to say, “I’m not going to answer that question. You know the theme of the quarter, so you need to at least think it through before I give you an opinion.”
At least they are consistent.
[1] https://kotaku.com/top-riot-executive-suspended-without-pay-...
I have work with a few ex rioters and have interviewed a few (at the time) current ones, and have heard personal confirmation of many of these problems. It's a big company though, so some departments are better or worse in this respect.
But yeah, it's not really a great place to emulate if you're trying to build a healthy culture, at least IMO.
There would be a plan to accomplish that goal, but it was pragmatic in that if we knew what good looked like, constant management wasn't needed for motivated troops.
You see a similar thing in Pragmatic Marketing's "Enlightened Teams" and other frameworks.
I dont have empirical data to back it up, but I believe that you get the results because of all the small things that teams take to get to good because of that higher level objective. I think Ben Horowitz's talk about these small here is golden: https://a16z.com/2011/11/13/lead-bullets/
I think culture plays into this quite a bit. E.g. it's probably a lot easier to build these kind of teams within a Scandinavian culture, due to extremely strong belief in egalitarianism.
It kind of begs the question; how does one hire for this?
https://lanepushinggames.com/moba/defense-of-the-ancients
There were numerous versions, iterations, and frankly better games IMO.
Even Eul's game was an iteration on Aeon of Strife maps.
I honestly prefer https://lanepushinggames.com/moba/tides-of-blood
and it's ilk way, way more.
Unless you mean the original DOTA mod for warcraft 3, in which case, yeah I mean why wouldn't you make a full game out of a mod that people loved?
LoL is a far more forgiving game for newcomers. Easier return to base and self-minion hitting are the big differences.
It also has a more cartoony art style which attracts a wider audience.
Wtf is Juggernaught? Something big? Wtf is sandking ability, quicksand?
When I played it at launch, the heroes were just reskins of the DOTA heroes. The art style was less cartoony than LoL, but the graphical fidelity also meant that it was more difficult to pick out the HoN heroes from the background than it was in LoL, and the graphical clutter was a fundamental part of why HoN failed.
"I don't like a game so it is bad"
"I don't like it so it sucks"
LoL is still one of the most popular online games now than a decade after launch. HoN was a critical and commercial failure at launch.
Popularity and quality have a very tentative relationship at the best of times. When differences in barriers to entry and network effects come in, then quality has to at most just clear a watermark.
So I'd say it's probably more likely "not bad enough to lose popularity" than it is "so popular because it's so high-quality"
Cigarettes are more popular than cuban cigars and McDonalds is more popular that a Michelin star restaraunt.
LoL and HoN are essentially different versions of the same game, so popularity is a good indicia of the market perceptions of game quality. LoL has been popular since it was launched over a decade ago, and remains popular to this day, outlasting dozens of other e-sports games, indicating that the player base thinks the game is still good enough to keep playing it even in the face of competition from newer games. DOTA2, similarly has a large player base, indicating that a large number of people also find the game to be of sufficient quality and enjoyment to spend years playing. Both Riot and Valve earned hundreds of millions from their respective MOBAs, and continue to make large amounts of money from these games. In contrast, the abject failure that was HoN almost bankrupt S2 games (their followup MOBA, Strife, finally killed them off.)
On a further note, I find it quite humorous that the parent to which I was replying accused LoL of being just a clone of DOTA, despite Riot actually putting a significant amount of effort into differentiating the heroes and combat. At launch, the HoN heroes were just reskins of the DOTA heroes.
IOW, HoN was entirely unoriginal and players might as well have just stuck with the original DOTA. And people did...DOTA on a given day had more players than HoN during its entire run.
LoL may have simplified the mechanics of DOTA, but it did something new and fundamentally set the stage for e-sports as we know it. Complex mechanics in themselves aren't to be celebrated; a lot of what LoL did was to eliminate the pointlessly complex mechanics. Even the edit:[lead developer] of DOTA thought the mechanics had gotten too complex to be enjoyable for more than a very niche audience, which is why he streamlined a lot of them when given the opportunity to develop DOTA2 for Valve.
https://kotaku.com/inside-the-culture-of-sexism-at-riot-game...
How can managers hold back to allow their employees to turn into self-managed teams?
honest question, but what justifies having a manager if your going to have "self-managed teams"? what justifies keeping the (now redundant) position around?I’m not sure what their vision is, but product wise there is a big discrepancy between their customers and the company.
> 1.9 million people played LoL in April 2021.
> LoL is the third most watched game on Twitch.
> In March 2021, 117 million hours of LoL gameplay were watched on Twitch.
> LoL is the second most popular eSports in the world.
[1] https://techacake.com/league-of-legends-player-count/