I played Thomas Was Alone this weekend – a game made by Mike Bithell (@mikeBithell) narrated by Danny Wallace with wonderfully fitting music by David Housden. Thomas is a visually minimalist game (think: colorful rectangles) with a Bastion-like element of story. Without ruining anything – it’s about AIs that were accidentally born. Each level has you controlling a few AIs while the narrator drops you inside one of their heads. Dan Marshall (maker of Time Gentlemen, Please!) posted that it’s the best indie game he’s played all year. Since I’ve loved Dan’s games, I had to give Thomas a shot.
And I’m left feeling mixed. I think I did my best to leave behind any expectations that Dan planted, but I still can’t help feeling like the ideas in the game had immense potential but stumbled on delivery.
I pulled the trigger on buying the game because I’m a sucker for narration – especially this particular kind that is dripping with personality. It’s one of the reasons why I love The Stanley Parable. So, hats off to Danny and Mike. Past the initial allure of the narration, I stayed because of the potential for telling a story through the mechanics of interaction between the different AIs. Each AI is a rectangle where its shape, speed and jump height define its individual personality. You control multiple AIs as you try to move them to the exit – press a key and you can switch from controlling one AI to the next. You only control one AI at a time, so the gameplay is a platformer, but those rectangles you are jumping on are other characters. So there is an added level of storytelling through interactions. That tall AI feels like a superhuman since he can get anywhere and can lift up the other AIs. And that slow, tiny block that can’t jump is the jerk that needs everybody’s help to do anything. Mash that up with a narrator who fleshes out what each character is doing or thinking, and you can engagingly tell a story.
Braid is another game that has some aspects of this type of storytelling – where narrative elements join with mechanics. In Braid, you play as Tim who has the power to manipulate time. You can rewind everything that just happened – so if you just fell off a cliff, rewind back to a time when gravity wasn’t trying to pull you to your doom. Every world starts off with a bit of story. Here’s an excerpt – “Tim needed to be non-manipulable. He needed a hope of transcendence.” So you’ve got a motif of wanting to feel like you can’t be manipulated. Then the puzzles are themed accordingly – one of the puzzles requires you to realize that there’s a way for you to rewind time without altering yourself. You find a way to escape your own power. Whether Braid succeeded in this or not deserves it’s own blog post, but I think the example helps illustrate the type of stuff I expected in Thomas – a marriage of narrative and mechanics to tell a story about relationships.
And it was, to some extent. The unfortunate bit is that the game has an excess of filler that dilutes the experience. There are pressures from audiences, critics and developers themselves to stretch out concepts to make these extended experiences. But when I jump into a world someone has built, I don’t want or need that. I will pay you the same amount for a more concentrated, and hence, powerful experience. There are 100 levels in Thomas, and yet, only about 10 or 20 felt like they mashed up narrative and mechanics into a cohesive whole. There were plenty of puzzles that just seemed to exist simple because another level was needed.
And these filler levels didn’t stand on their own. The idea of multiple AIs was intriguing, but the implementation was frustrating. There is set of AIs that you meet called Team Jump. They are essentially 5 copies of the same tiny block. Alone, they can’t get anywhere, but together, they can get anywhere. Could be a fun mechanic, right? Well, solving any puzzle with these guys requires moving one guy, switching to another, moving him to the first guy and jumping on top, etc. You have to build a step-ladder to get anywhere. You know exactly what you want to build to solve the current level, but you have to repeat the same tedious actions to get there.
On the other hand, there is a fantastic set of levels involving two characters who turn out to have equal and opposite abilities. You never felt like you were switching to the ‘slow’ guy, and with only two AIs, there was nothing tedious about switching back-and-forth. So the frustration level with the implementation varies greatly over the 100 levels.
In the end, I could have agreed with Dan’s assessment of Thomas as the best game of the year if the content had been concentrated. Cut out plenty of levels, throw out some of the characters, and really squeeze the story elements into all of the mechanics. Maybe even explore some additional territory for interactions. There are two AIs who are in love in the game – why not make it so those two have some special interaction that only they share? Or when they first meet – throw them into a tiny room where they awkwardly bump into each other when trying to solve the puzzle (and let the narrator work his magic)?
But it’s not that I didn’t enjoy Thomas! It’s just that it tapped into this well of storytelling possibilities, and I was sad to see it fall short. But go buy it, decide for yourself and tell me I’m wrong :)