Can Prehistory Go Digital? A Comparison between the “Stone Age” App and Physical Game

During the last FIMS Gaming Club meet-up we went prehistoric, and played the classic euro-game Stone Age. Stone Age is a worker placement game for 2-4 players, designed in 2008 by Bernd Brunnhofer, and currently published for distribution in North America by Z-Man Games. Worker placement, in case you’re unfamiliar with the mechanic, involves a placement round where players put workers on designated “workstations”, which will then  collect resources from, or perform actions on, the workstation in the “work” round. In Stone Age, your prehistoric workers are attempting the build up their villages and collect the most points by the end of the game, and do this by collecting resources, building buildings, and investing in specialized trades and culture. If you want to learn more about the game, check out Dice Tower’s instructions and review:

This is a game that, before this session, I had only ever played digitally with the Stone Age app, as I’d never had access to the physical game. Stone Age is generally a good game to make an app out of, because it doesn’t rely on human interaction. Like many Euro-games, it’s almost a solitary game as you zone into your own world of strategizing, until someone just happens to take the spot you wanted. Thus, creating computerized artificial intelligence (AI) players is relatively easy, and you don’t really miss out on socializing because it’s not a key component of the game anyway. Or so I thought.

In reality, though the in-person game is still quite solitary, the game is made more interesting by people’s ability to adapt their strategies, compared to the computerized player’s rather static strategies. Each of the four AI difficulty levels in the Stone Age app has a distinct strategy, with the easiest settings focusing on gathering the “essentials” (fields, tools, etc.) and making quick rewarding buildings, and the more advanced ones focusing on a more risky strategy of accumulating cultural items which don’t come into play until the end-game scoring round, but can have more value. There was a hint of similar behaviour in our play session, with more novice players focusing on the “safe” strategy of collecting buildings. However, they were able to learn from watching more experienced players, trying out new strategies and making more unpredictable decisions. The game is more fun with this unpredictability, and competition as well. Even though it is a typical Euro-game, where there is no way to directly attack opponents, we still felt like we could purposefully block other people who were getting too far ahead. For our group, at least, that competition brought about some good laughs and inside jokes, unlike with the passive AI’s.


So far, we’ve been really good at creating AI’s that can beat perfect information games and games with little social interaction, like Stone Age, because they lend themselves to algorithmic processes. However, I think it will be a very long time before we can see AI’s interact with game players on a more intimate level and really engage with them like humans can. And do we really want to anyway? Unless you’re playing games strictly for the mental exercise, games are really about creating comradery and shared experiences with old friends and new. That’s missing when you play with AI (unless there’s a point in the future when AI’s are accepted as basically human, but that’s a whole other can of worms). Regardless, I think apps are still a great way to get involved in games when you aren’t able to pull a group together to play the real thing, and I’ll continue to hone my Stone Age strategies this way, in between live game sessions.