Laralyn McWilliams: The problem with "best practices" in free-to-play

I actually agree with him, but I come from an environment where I recognise how challenging it is. One of the frustrations among the people in that space - people who have been there for years - is that people only now coming in assume that we're doing things the way we are because we're greedy or we're evil. That's the whole point of the Evil Game Design challenge - it kind of makes fun of that.

But, in fact, a lot of what people suggest we do has been tried, and has failed. It doesn't mean that it can't be done. A couple of times I've had a go at monetising at least partly by story, using the same sort of mechanism that gets you to tune into a TV show every week.

Yeah, pay because you don't want to wait to find out what happens. That's behind people paying for Netflix. If the whole story was all just sitting there, and you could either get it a week at a time for free or you could pay for it all now. That could really work.

But the times I've tried it's been mired in 'best practices'. I talked about this at GDC Next, and I really believe that, to a large extent, the best practices that we lean on... you may as well just perform a ritual, right? Every game is different, and the market is different, and the timing is different. Even if it's just a clone, the circumstances will be different.

But that's the thing: it would be the same Bill of Rights. When you think about what motivates people, there's a psychological principle called "systematic desensitisation" - basically, the ability to change the way someone feels about something by [controlling] how they're affected the moment that they experience it. It's the same thing as getting really drunk on Midori - which is what I did in college - and then not being able to stand the taste of it.

You can't stand it, right? It's at the root of your brain; a function that deep. It's Clockwork Orange shit, but it works

No, I'm talking about us doing this right now. Think about Candy Crush Saga, which I've played a lot and I really like, but it's a great example of this. The moment that you monetise in Candy Crush you're probably extremely frustrated. You want to get past this level you've failed to complete 40 or 50 times, and that's the moment you spend. But mixed into that moment where you spend is that frustration. It's building a bad connection. I'm not monetising at a positive moment. An example I've used before is the person leaving the Apple store with a new iPhone after having stood in line all day, and they're like, "Yes!" Isn't that how we want people to feel when they spend money? They should feel awesome.

Part of the problem is that those methods work, at least sometimes. And then they become 'best practices' and the assumption is that they always work, and there's nothing better.

"The moment that you monetise in Candy Crush you're probably extremely frustrated...It's building a bad connection"

You have two kinds of companies. You have big companies, where there's no room to not use best practices; you have to have data to get them to try something new, which, when you think about it, is stupid. If it's new, how can you have data? Those companies will always lean on best practices because they yield the most revenue.

And then you have the smaller companies, who are desperate to survive. And even if they experiment they might still go out of business, and when they see that an experiment isn't working they can rapidly fall back on those best practices to start monetising. It's the same kind of rut.

It's really a matter of having the space to try something and have it fail, then try something else and have it fail, then try something else and it works a little, and then continue to iterate on that until you can get it to a place where it's profitable. Mostly, the exceptions had that room.

There was a really great monetisation talk at Steam Dev Days, where Valve talked about Team Fortress 2. They have a company culture that says if at any point the player is unhappy with or regrets their purchase it is a fail. If they see that happening - if there's any negative consequence to the player - they won't even continue testing it. Right there, it's done. Happy players are number one in terms of importance, and Valve will not mess with that for any amount of money.

"It's really a matter of having the space to try something and have it fail, then try something else and have it fail"

It was really interesting to hear that. But it was a little frustrating to come away and hear so many people say, 'Well, they're Valve, They can afford to do that.' I think that misses the message. For me, the message is that a company with bucketloads of money is doing the research for us, and are now sharing what they have done that works. You don't need deep pockets to make that work. They needed deep pockets to experiment long enough to figure that out.

It is. You can see with both Free Realms and Clone Wars, which is also closing, that it's just really hard to monetise kids. And for good reason - I don't think it should be easier to monetise kids. I haven't spoken to [Sony Online Entertainment] about the reasons for closing them, but I could put some pretty good bets down, and one is the really high production values. That's just really hard to maintain.

With the game I'm working on now, I use Free Realms as an example of setting a quality bar so high that you can't sustain it. Unless you hit the lottery, you're not going to go from 5000 paying players to 5 million paying players overnight. A lot of people think you have to launch with really high production value and players will just flock in, but if that's not the case you're kind of stuck. If it takes a lot of effort to make content, you're toast.

That's the conundrum. Free Realms, when it first launched, did not monetise, and there was no example of how to do it in the West. We spent a year at least tuning and making nitpick changes to the game. It was really fun, people liked it, we had great retention... About a month in, I was at a conference and I met somebody who said, 'My daughter loves Free Realms! I play it with her every night for an hour, and I can now get her to do her homework because I only let her play once it's done. I love it - I've been playing ever since launch and I haven't spent a penny.'

"It's just really hard to monetise kids. And for good reason - I don't think it should be easier to monetise kids"

I was like, 'No, no, no. That's our problem. You love it and you're not spending a dime.' The easiest way to get money in a free-to-play game is to put pressure on, but do you want to really want to pressure kids that way? It's a core problem.

It depends what you want out of it. When I was going through my last job hunt... I guess if you've read my blog you'll know that I had cancer. That changed me. I came out of that and I really did a lot of soul-searching. I was going through the interview process and there was this pivotal moment. I was working on a Facebook game, trying to monetise by story - I was going after that again - but it was small company and it needed the revenue. So along with that attempt we had a lot of 'best practices' around friction-based monetisation.

It was a hidden-object game, and one monetisation method for a hidden object game is that the player doesn't want to play the same scene hundreds of times. 'I really want a new scene, so I can either keep grinding away on the scenes I've already played, or just unlock the new scene now.' I believed that quests, which have little bits of story, would give a reason to replay old scenes, and kind of soften the grind of having to play them again and let us re-use our content a little better.

And it worked, but then a designer came to me and said there was a spot where it got really rough; there weren't enough quests, and the grind was really terrible. He wanted to add five or ten quests to make it feel better. Now, adding those quests was, I don't know, four hours of his time and a couple of mine to do the writing; I mean, a trivial amount of work, no new art, nothing.

But when I looked at our numbers that was the spot where we had our best monetisation. The awful feeling of that grind was getting people to spend money, so I had to say no to something that would make players happy because it would cut our revenue. At that point I said, 'Nope,' and I got out of social games.

Well, we know a lot more about how to get money out of people [laughs]. We've learned a lot about that.

Oh, I don't know. There was an interesting moment at Steam Dev Days, where I got into an argument with someone who isn't from this space. The project I'm working on now is the first free-to-play game from [The Workshop], and I got into this argument [with another attendee] because he couldn't believe that the big free-to-play companies don't really care about retention in a long-term way. I mean, they do care because you need to hold players in order to monetise them, but as far as I know there's no real way to understand the gradual erosion of goodwill that might come about as the player is monetised.

Right. And so much of their tuning is around whales that, every day, they choose between making the game more fun and lowering the potential for monetisation, or raising that potential and slightly reducing retention. Or increasing monetisation in one group of people that makes another group who don't spend much churn out of the game. There's no good way to measure what results from those decisions.

You can hold an A/B test for maximum a month, and then you hold A/B tests two, three, four and five, all out in the wild and you're making decisions all along the way. So, six months down the line, if you start to lose players it's really easy to say it was because of that last update, but it could be the one six months ago, or it could be all of them. At that point, you have no way of knowing why that player is leaving. Every player is some mishmash of A/B tests that they've been through, what their experiences were, and all of the other changes you've made over time.

To be honest, that's why I had that sinking feeling in Peter's talk. When he said that he believed a game that is delightful and good will win out, I thought, 'Oh no dude. Let me type up a list of fantastic games that totally failed.'

[Long pause] Only a little bit.

I would say it depends on where you work. Even though I'm working on a free-to-play game now, it's at a console/PC house. It's a different vibe. The Workshop is not, and will never be, the kind of company where if I want to do something for the benefit of the players I have to justify that to some bean-counter. It's been nice, because it proved to me that you can make a free-to-play game in a place where that isn't the case.

Another part of the free-to-play world is the concept of the "Minimum Viable Product". I understand the need to fight against the instinct of artists and the designers to polish something forever, but I feel that focusing on the MVP gets unfinished thoughts out there, and a lot of the time, unless you're at the right company, that thought gets killed before you even have a chance to do anything with it. I miss the opportunity, in retail game development, to finish your thought.

There are some games where it can work; any game where it's building on a self-contained experience. Don't Starve would be fine at that stage, Angry Birds would be fine at that stage. There are some games that you could put out in an early state and they would be fine, but not if it's anything involving complex mechanics, or much of a story.

If you're making a free-to-play game and you can just pull out a big feature, or pull out the way it monetises, and it still works then you haven't designed it well. To design it well it all has to be interwoven, and an MVP makes that problematic.

Yeah. Our MVP is going to be a lot more robust than most for that exact reason.