Free-to-play whales more rational than assumed

Ubisoft researchers find heavy spenders more driven by long-term planning than impulse, suggest the term hurts devs as well as players

Analytics are great at telling developers what players are doing in their games, but the numbers aren't quite as adept at telling them why. For insight on that, Ubisoft turns to a pair of research scientists it brought on board a year and a half ago: Nicolas Ducheneaut and Nick Yee. At last month's Game Developers Conference, the pair sat down to speak with GamesIndustry International about their latest research into the heaviest spenders in Ubisoft's free-to-play game Ghost Recon Online.

"It's an interesting point in time in the industry because everyone is talking about big data," Yee said. "I think a lot of companies inside and outside the game industry are getting access to these big pools of data and they're starting to get analysts to look through that data. But oftentimes what happens is there's a behavioral finding from the data point of view that they can't understand because they don't know what the player was thinking when those behaviors occurred."

Or as Ducheneaut put it, "It's one thing to have the data. It's another thing to make sense of it."

Sometimes developers may be baffled by the data they see. Other times, they might pull the wrong lessons from it. The pair's research suggests that's what happened with the free-to-play sector's assessment of its heaviest spenders, the so-called "whales." To better understand the motivations of those players, Ducheneaut and Yee surveyed a targeted mix of monetized and active Ghost Recon Online players.

"It's one thing to have the data. It's another thing to make sense of it."

Nicolas Ducheneaut

"One thing that came across was this concept of 'whales' was really framing how developers and our marketing folks were thinking about what drives high-value spenders. [The assumption was] it's impulsive, more irrational, kind of hedonistic behavior," Yee said. "What we found was almost the exact opposite. Instead of being impulsive, they were long-term thinkers, cool-headed, methodical, and they really supported the game."

It turns out the heavy spenders weren't spending money on impulse purchases. They valued long-term learning and mastery of the game, so they focused their purchases on new gear and items to help them master different aspects of gameplay, try new tactics, or unlock new classes to become proficient at. As for how developers could turn that knowledge into concrete changes to make in a game, Ducheneaut said they could start by giving rational customers the basics they need to make rational decisions with their money.

"If you want to make an informed purchase decision, you have to be able to compare items in the store very easily and understand what it's going to bring you in terms of added gameplay value," Ducheneaut said. "So you see how you could reframe the design of your shop such that those things are easier to do for those people."

That would seem to fly in the face of some frequently used tactics in successful free-to-play games like mystery boxes and limited-time offers. However, Yee said the revenue spikes those provide may not be as attributable to impulse purchases as some would think. He said there was one game he looked at where the developers ran discounts on in-game currency every few weeks. It turned out many of the heaviest spenders weren't actually making impulse buys at all; they had spotted the pattern and waited to make their big purchases in anticipation of the reliable sale period. Yee said developers' assumptions about their players' behaviors were turning into self-fulfilling prophecies.

"You're confirming your own bias but you're not testing the alternative," Yee said. "And at first we kind of had these biases too, but it helped us rethink, and helped our designers think, 'Ah, we're seeing this pattern, but maybe it's happening for a different reason.'"

One of the things perpetuating those biases is the terminology of the industry itself. The term "whale" is loaded with negative connotations, and Yee said its continued use hurts gamers and developers alike.

"For the gamers, [the word 'whales'] frames them as being pathological. And for the developers, it frames us as being psychologically exploitative."

Nick Yee

"For the gamers, it frames them as being pathological," Yee said. "And for the developers, it frames us as being psychologically exploitative. And it's bad in a third way in that it constrains how we design games because we assume gamers to have a certain mindset."

Ducheneaut prefers to think of them not as whales but as hobbyists. He referenced a talk from Kongregate CEO Emily Greer in which she described an expensive single-player game with competitive multiplayer aspects in which she spent thousands on gear and months engaged and levelling up, only to reveal that it was not a free-to-play game she was hooked on; it was figure skating.

"It's really understanding those people as hobbyists," Ducheneaut said. "They're committed to a hobby. They invest resources in their hobby, just like someone would in model trains, figure skating, or whatever. It's no different than that."

Ducheneaut and Yee can relate, as they have their own sometimes costly hobbies. For Ducheneaut, it's sailing. As for how Yee gets rid of pesky disposable income, he said painting miniatures helps. Of course, they're no strangers to gaming either, and hope that their work will help make the industry a better place for anyone who counts gaming among their hobbies.

"Speaking more as a gamer, I often wonder about the games that should exist but don't really exist because we've kind of clung to really strict genres. And it's getting worse, seemingly, over time. I think when we understand gamers more, and understand their personalities and their motivations, we can design truly engaging experiences that may break a little free of the mold these genres cling to.

While Ducheneaut said this research-oriented way of interpreting analytics will be increasingly important to the industry as time goes by, there are limitations to how quickly it could happen. Even if the demand for social scientists in the game industry takes off, Yee said there's a bit of a bottleneck on the other side of the equation.

"Right now we're really limited by the supply of people who have the right skill set, who are gamers and understand game mechanisms, can deal with big data, and have some kind of social science framework for understanding human behavior," Yee said. "And it's really difficult to find people who have that overlap."

More stories

Ubisoft reportedly unlikely to sell for less than €60 per share

Some shareholders are said to be holding out until the Assassin's Creed publisher's stock reaches €100

By James Batchelor

Ubisoft+ headed for PlayStation

In addition to rollout of full subscription service, a partial offering will be included with PS Plus Extra/Premium tiers

By Brendan Sinclair

Latest comments (25)

What we found was almost the exact opposite. Instead of being impulsive, they were long-term thinkers, cool-headed, methodical, and they really supported the game."
sounds to me like they are just paying to win.
9Sign inorRegisterto rate and reply
Jakub Mikyska CEO, Grip Digital8 years ago
I have never played Ghost Recon Online, so I may be wrong... but purchases like new classes or weapons can be made by rationally thinking, planning gamers, because that is the way F2P is supposed to be. You pay more - you get more.

Impulsive behavioral patterns come to play when you have to pay in order to continue getting what you already have. The need to buy diamonds, because you need diamonds to build a house that you built many times before, when you still had your free diamonds.

I bet that GRO does not have mechanics that take away all your gear and XP for 24 hours, unless you pay... I still don't understand why so many people cannot see this simple, yet fundamental difference between various F2P implementations.
4Sign inorRegisterto rate and reply
Nicholas Lovell Founder, Gamesbrief8 years ago
That sounds like fascinating research. Perhaps we should start calling people who love your game, and spend lots of money in it, superfans.
3Sign inorRegisterto rate and reply
Show all comments (25)
@Nicholas - now where have I heard that term before ...?
0Sign inorRegisterto rate and reply
Tim Spencer Level Director, TT Games8 years ago
I've got a huge amount of respect for Yee after his work on the Daedalus project, I did wonder where he'd disappeared to! Great to see he's still out there delivering quality research data for us developers.

This research does indeed sound incredibly fascinating - and valuable.
0Sign inorRegisterto rate and reply
Daniel Mesonero Studio Manager, Toadman Interactive8 years ago
Really interesting research, but I also really consider it hurtful to apply findings to "gamers". There are different types of people, playing different games, and my guess is that GRO appeals to more core-oriented players who possibly are less impulse-driven.
It's likely that games that cater to long-term users end up with higher monetization, but there will still be devs that make impulse-driven games for impulse-driven players. Hopefully the market matures, but since we see exactly the same business models in the physical world, i doubt that it'll ever go away.
1Sign inorRegisterto rate and reply
Don't want to jump to conclusions but without seeing the underlying data this sounds to me like someone is complimenting the people who spend a lot of money in their game by telling them: "Hey our research proves, if you regularly give us a lot of money, you must be a really, really smart person".
Additionally I don't think that "irrational hedonistic behaviour" and "cool-headed, methodical" behaviour are contradicting character traits. E.g. heroin addicts can develop quite cool-headed and methodical strategies in order to satisfy their next irrational hedonistic behaviour.

Although being a gamer and investing a lot of time in this hobby, I still see a big difference between aquiring a skill like "Figure Skating", and investing in a free2play game. In those real life hobbies you can actually resell the gear you aquired and train your body whereas in a free2play game - if you're out of luck - the plug can be pulled any time by the provider and you're left with nothing.

Edited 1 times. Last edit by Daniela Grebe on 1st April 2014 5:04pm

9Sign inorRegisterto rate and reply
Brendan Sinclair Managing Editor, GamesIndustry.biz8 years ago
Ah, thanks Barrie, that's been fixed. And I hope there aren't too many developers out there trying to entice players with their "mystery boxers."
6Sign inorRegisterto rate and reply
Brian Lewis Operations Manager, PlayNext8 years ago
I guess I am going to have to ask the question. Did anyone here (implying those in the industry) actually believe that whales were spending thousands of dollars a month on simple impulse buying? I have a hard time believing that anyone actually involved in these type of sales would be that out of touch with reality. I very much respect the work done to analyze the data provided... but I have a hard time believing that the preconception was that large spenders were primarily impulse based.
3Sign inorRegisterto rate and reply
Daniel Laughlin Research Scientist/Education Researcher 8 years ago
Is that "mystery boxers" the fighters or the dogs? Or is that the mystery? :)

Yee is known for doing thoughtful research, so I would not credit the idea above that the results are biased to suit the industry.
0Sign inorRegisterto rate and reply
Jamie Madigan Psychologist 8 years ago
"Right now we're really limited by the supply of people who have the right skill set, who are gamers and understand game mechanisms, can deal with big data, and have some kind of social science framework for understanding human behavior," Yee said. "And it's really difficult to find people who have that overlap."

So, where should I send my resume? :)
0Sign inorRegisterto rate and reply
Nick Wofford Hobbyist 8 years ago
I'm not currently designing anything for a mobile device, but I'd argue that the views of some developers on the subject can be seen in their games. I feel like a lot of F2P games are targeting this idea of rich, ADD players who will impulse buy some cheap item as opposed to what this study is finding. This error in marketing could be one of the reasons that F2P has gone wrong in the past, and could offer a nice way to turn that around.
1Sign inorRegisterto rate and reply
Neow Shau Jin Studying Bachelor in Computer Science, Universiti Sains Malaysia8 years ago
@Brian Lewis

I won't comment on the data provided on free-to-play games as these fully digital data are not tracked by a neutral 3rd party and most like manipulated to suit the purpose of whoever release it. I can, however, point to another industry which exploits whales that offer physical goods as a product. These sales data are much more reliable and tracked by third party.

AKB48 is Japan's million selling pop-idol group, They've consistently selling over 1 million units of their singles within the first week in the past few years. However I like to point to the singles that is particularly released annually in May. You see, this group holds an "Election" every year, where their fans will buy the single released in May, which contains a digital code, then go on the website and use it to vote for their favorite member. These has gone so big, that there are overly fanatic fans who bought thousand of copies of the same single, just so that they can use the code to generate thousands of vote for their favorite members. These fans even admits that they skipped buying the releases in other times of the year just so that they can focus all they funds in one release. These whales, their purchase the definitely methodically planned, but it doesn't mean the reason is rational.

Having a hard time finding any English source that gives you a better picture, just a blog here:

or a youtube video of one of the fans unboxing:

Edited 1 times. Last edit by Neow Shau Jin on 1st April 2014 8:03pm

5Sign inorRegisterto rate and reply
Brian Lewis Operations Manager, PlayNext8 years ago
@Neow Shau Jin

"These has gone so big, that there are overly fanatic fans who bought thousand of copies of the same single, just so that they can use the code to generate thousands of vote for their favorite members. These fans even admits that they skipped buying the releases in other times of the year just so that they can focus all they funds in one release. These whales, their purchase the definitely methodically planned, but it doesn't mean the reason is rational."

I do agree that there could be some debate as to the rationality of spending large sums of money on any form of entertainment, be it games, your favorate bands, or even fine wines. However, none of that has been questioned. What has been questioned is whether this spending of large amounts of money is the result of impulse decision making. Even in the example that you have given, many of could not rationalize the behavior, but there is no indication that this is impulse behavior. In fact, the general assumption would be the exact opposite (and you have also provided supporting facts).

If I were to do a study to prove that people dont actually believe that the world is flat... it would imply that there was a belief that this was true. Running analysis to prove that large scale spending in games is not generally impulse driven is only useful if there is a preconception that it was. I am questioning whether this was a common belief in the industry, or just a straw man.
1Sign inorRegisterto rate and reply
Tim Ogul Illustrator 8 years ago
I play a few games with cash shops, and I spend some money in them, but I try to do so rationally. I spend money as a sort of "compensation" for the new experiences I've gotten for free. If I am enjoying the game, then I think it's only right to give something back. I tend to try to get the best deal for my money though. I tend to wait until "bonus with tokens purchase" sales to pick up tokens, like when you get 10% more coins than usual, or get a bonus prize for buying a certain amount. Then I wait for sales on the items I want before picking them up. And of course if the prices are always too high, I don't bother, and if they try to charge for things that are pay-to-win, I don't bother, but I tend to be fairly generous to developers that are generous with me.
0Sign inorRegisterto rate and reply
Bonnie Patterson Narrative Designer, Writer 8 years ago
I was completely prepared to spend large sums to find out what is in the mystery boxers. Now I am disappoint.
7Sign inorRegisterto rate and reply
Matt Jeffries Senior Producer, Telstra8 years ago
I'm wearing mystery boxers as I type this comment.
9Sign inorRegisterto rate and reply
Eyal Teler Programmer 8 years ago
@Brian Lewis, just take a look at the comments and you'll find enough people who believe that (at least some) gamers don't behave rationally.

Besides, I think you're barking up the wrong tree. The point IMO is more that spenders are fans, rather than addicts, which a lot of F2P games seem to treat them as. F2P games often put obstacles in front of gamers, and this can only come from the assumption that the gamer, wanting to continue to get his or her fix, will immediately go and buy the option to get rid of this obstacle.

What this research suggests is that spenders don't work like that, they have patience and they buy things which enhance gameplay, rather than things which allow it to continue. They buy these things because they like the game and want to explore all that it offers, rather than just pay to make progress in it.

I'm not a big spender, but when I spent money on F2P games (MMO's mainly), that indeed was the pattern, I saw something which I thought would be interesting to add to my character or allow me to create different characters, or provide me with more content, and I bought it. I did pay at least once to overcome an artificial limitation (number of bags in EQ2), so it's not like this tactic doesn't work at all, but I certainly paid more for other stuff (races, classes, expansions).
3Sign inorRegisterto rate and reply
Manny Soetje Junior Business Development Manager, Gamigo AG8 years ago
I've never heard of "whales" being considered impulsive. Sounds like something a person who's never played a F2P MMO would say. We follow each and every one of our high paying users and they are without comparison our most hard-core, dedicated and passionate players. Hundreds of people who at the pillars of your entire community and dedicate their lives and money to your games. These are people who have scribbled post-its all over their desk, wall and computer screens with game design values, dmg output/input, heal procs, and bug abuse step by step guides on them.

In a Theory of Fun for Game designers, humans are described as "pattern finders", gamers are especially good at pattern finding (as corroborated by a few studies now), of course they would spot pattern in item sales and calculate the percentage change of lucky box win/loss ratios. That's part of the gaming experience to them! We've got millions log files lying around here waiting to be worked with, you're welcomed to them Yee :P (also a big Daedalus fan here)
3Sign inorRegisterto rate and reply
Andrew Goodchild Studying development, Train2Game8 years ago
They tested this on Ghost Recon, which presumably attracts the gamers for which this is true.

However, they idea that this is a universal truth really doesn't seem to gel with Candy Crush Saga's success. I finally gave it a look this week, and virtually every chance to spend money is an impulse purchase:
Nearly finish a level but run out of moves, spend 69p.
Run out of lives, wait 20 mins or buy lives now.
Struggle beating your mate's score, buy a power up.

That big spenders are never impulsive would suggest that one of the biggest FTP titles has no 'whales'. I find that difficult to believe.
2Sign inorRegisterto rate and reply
So the finding is that the people paying the most for the product are not irrational and lacks self control but are smart and dedicated gamers. Well, I call them hardcore gamers and have the biggest respect for them as I think you should with all your customers and especially the highest paying ones.
0Sign inorRegisterto rate and reply
Brian Lewis Operations Manager, PlayNext8 years ago
@Andrew Goodchild

It could also mean that big spenders in Candy Crush are NOT seeing this as impulse spending. I would think that they are simply seeing the more difficult levels as P2P, and that they consider longer play sessions as paid play time. I would think that big spenders would make even LESS impulse decisions, and would instead plan out the purchases as part of their strategy.
0Sign inorRegisterto rate and reply
Andrew Goodchild Studying development, Train2Game8 years ago
The problem there is that most of these pop up at the spur of the moment when you fail a level. You don't know you were going to fail the level before you started, or how close you would be, so it is not like you could plan to spend money on it. For instance, if you can complete a level if you have two more moves, there is no way you could have been planning for an hour that you were going to fail with only those two moves left. As you have to pay there or lose progress, it can only be a spur of the moment decision.

Unlocking the levels without 3 friends help may not count as that, but I've only seen that once so far in the four or five areas I've played through, so that would not count as whale behaviour.

The point is, if a game is designed around impulse spenders, the big spenders it attracts will be that sort of person. If the game caters to investment, like League of legends, then that is who that game will attract. The problem with only measuring behaviour in Ghost Recon Online, is that whilst you have some interesting data, you only have data for Ghost Recon players. This can not be then applied to people who are attracted to a different game that attracts people for different reasons, anymore than you can work out why people like Pulp Fiction by asking what they like about Eastenders.
0Sign inorRegisterto rate and reply
Brian Lewis Operations Manager, PlayNext8 years ago
@Andrew Goodchild

I do agree that with candy crush, you can not spend in advance. You only ever get the option to spend money, at the point of resolution. However, it is something that can be predicted (very easily). This is just like putting another quarter in the arcade machine. If you are good at a game, you will know the exact point when you will need to do this in order to move through the content as you have intended. It is only the inexpert player that would believe that they could get through without needing another quarter.

The game may have been designed around impulse spending (by not offering the spending upfront). However, it in no way means that players, especially players who spend lots of time and/or money would not choose to not be limited by this. I have yet to meet any publisher or developer that has any actual experience it the market (has already published a major F2P game) that has any belief that the large spenders are impulse buyers. We all know that they are in fact the most strategic buyers, and the most savy shoppers over time.
0Sign inorRegisterto rate and reply
Emily Greer CEO, Kongregate8 years ago
Thanks to Nick & Nicolas for doing and sharing this research. While as people have commented, and they also called out in the actual GDC talk, research on one game is not comprehensive, though I do think it's very strong directionally. I'd be quite interested to see it on something like Candy Crush, though I think even there purchases are less impulsive than you would think. I know when I've spent money there it's been because I've played a particular level many times, and can see that a win is only 2-3 steps away.

Someone also commented that there's a big distinction between hobbyists and people spending on f2p because they can resell their gear I don't think that's true, nor do I think it changes the psychology. Only a very small % of my figure skating spending goes to anything durable -- most of it goes to ice time (and specifically freestyle ice time, where the # of skaters is capped and you can move any direction, about 2x the cost of public sessions), coaching, and fees & travel for competitions & tests. My skates were custom made (typical for higher-level skaters) and are therefore not terribly resellable, and if you amortize the cost across the 5+ years I'll use them they represent <5% of my annual costs.

I spend all this because I love the feeling of skating and the process of getting better, of unlocking new "achievements" in the form of a new spin or jump, literally leveling up when I pass a test to go up a level. I compete for the challenge and the thrill, and for the camaraderie and connections I make with other skaters. Because I get so much out of it and have the means I spend on what will realistically make me better, and am not terribly price sensitive about it as long as I see a tangible benefit. I don't think there is anything very fundamentally different about my emotions or spending decisions than those of a big spender in an MMO.

If you're interested in my original talk that Nick and Nicolas reference you can see the video online here:
0Sign inorRegisterto rate and reply

Sign in to contribute

Need an account? Register now.