For Science: Crit-Rate is another survey-driven attempt to match gamer personalities to games


I have to admit I almost deleted the press release for this story without reading it, but my finger twitched, and let me tell you why: One of the most common Reddit threads I see and one of the most common emails we get here at MOP can be boiled down to “what MMO should I play?” It’s not that there are no games; it’s that there are way too many, and very few companies are working on helping people find the right ones in a world where they are paid to show you others.

Of course, whether this new take on finding games is actually successful at that task is another story. It’s called Crit-Rate and is debuting at PAX East this week as the brainchild of marketers, podcasters, and press; participants take a gaming-personality test, and then the AI pools the data from likeminded gamers and tries to recommend the most appropriate next big thing.

“A robust digital recommendation engine powering Crit-Rate aims to provide video game suggestions like trusted friends. A pool of analytical data is first gathered from a detailed personality test about how the user enjoys video games, including challenge preferences, problem solving processes, and even what immerses them. Users are then sorted into Houses based on these dimensions. Once their profile is made, users can explore their House to find highly curated recommendation scores from other gamers with similar preferences and playing tastes. The most renowned, popular, and high quality games aren’t always the correct fit for everyone. Crit-Rate assists users in finding the right game for them by focusing solely on ‘fit,’ rather than quality, through the aggregated ratings of like-minded gamers. The unique score given to each game is read not as percentage of quality, but the likelihood that this game will be a critical hit for the user. Hence Crit-Rate.”

It actually looks like a mash-up of the old Bartle Test, Quantic Foundry’s Gamer Motivation Model, and some cutesy online quiz, with the added element of actually suggesting specific games based on the survey – and I had a feeling it was likely to have a lot of the same problems, too, and perhaps even more since it’s run by marketers rather than academics (unlike Quantic).

So – for science – I went ahead and let them collect my info to see how it would go. You can take the test or just choose your house yourself, but I wanted to see what the survey would do. Here are a few of the survey questions that leaped out at me (you’re supposed to mark out whether you agree or disagree or neither).

If there’s a competitive aspect in a game I tend to enjoy it more.
I don’t enjoy a game that’s mind-numbingly difficult.
I actively seek out the most challenging aspects of a game.
Games that focus on planning a town or raising an army are hands-down my favorites.

We’ve previously discussed how words like competition, difficulty, and challenge are not helpful when describing wildly different types of mechanics and systems in video games, since someone could feasibly consider twitchy combat difficult and annoying but live for the thrill of a challenging competitive economy, while others might be terrified of learning an absurdly engrossing crafting system but prefer to spend all day in cheeseball battle royales. These words are not only unmeaningful but misleading. (I had the same complaints about Quantic’s model.)

After you complete the couple dozen questions in the survey, the site asks you to rate a bunch of games quickly. Literally none of them was anything I would ever buy or play. Not one single one. I don’t think the options were drawn based on the survey, for what it’s worth, but I also can’t see how that offers much nuance to my score. I was able to at least pick three out of the list of games I was looking forward to! One of them I won’t be buying, but at least it’s representative of a genre I like. Of course, nothing listed (for me?) was even remotely an MMO.

In any case, the test slotted me into the Curionaut category, which is somewhat accurate, somewhat inaccurate, but overall so vague that it could describe a huge percentage of MMORPG players. That’s about what I would expect from a survey this short and limited, honestly – there’s an obvious Forer effect at play here.

The real problem is that after the site assigns me to the Curionaut role, it recommends games to me that are, again, not something I would ever buy or play. So at least for me, as a core sandbox-leaning MMORPG player who leans heavily on questing, crafting, and building, the algorithm simply doesn’t work effectively. Maybe it’ll work for a different type of mainstream gamer or console gamers (I see a lot of console games in these lists), or maybe it’ll keep being refined with more games and better recommendations when more people have filled out the survey and pushed the algorithm in the right direction.

Is it going to find the perfect game for the average MMORPG player? I doubt it, but it’s worth a look anyhow on a rainy Thursday. If you give it a try, let us know how you’re slotted!

Previous articleWisdom of Nym: What will a new producer mean for Final Fantasy XI?
Next articleCasually Classic: How permadeath runs are revitalizing WoW Classic

No posts to display

oldest most liked
Inline Feedback
View all comments