HOUND (PC) Preview

Neural networking has been used to teach an AI to play games such as Super Mario World from the ground up, and now it is used as an engine of a game. HOUND is truly an experimental title beyond the usual, in which the player is given their own world to play around and build lifeforms in that can evolve on their own given some time. Some control is needed, though, otherwise havoc will roam.

On paper, HOUND might be a really interesting title, and after a quick look at it, that truly seems to be the case. However, there are currently some problems that hopefully will get solved for launch. The biggest one is that there is very little tutorial or other guidance in the game. Normally, this isn't that big of a problem, but here it truly is, as nothing quite like this has ever been made before, and working with HOUND is similar to learning how to work with an engine itself. In other words, not an easy thing to grasp without some guidance.

The game does also feel very barren on goals or objectives, and leaves the player blindly trying to stumble through the learning process on their own. There is a help section, but for something like this, that is far from enough help make the average person learn a system as complex as the one in HOUND.

While the purpose of the game means that there is a great advantage if the players are roaming the system on their own, it must not go to the point that most people will be unable to even understand what is happening, like it seems to be the case at this point in time. This is an issue that is said to be addressed at launch, and it will be the thing that decides the fate of the game.

Whether or not HOUND will end up being an enjoyable title for consumers is extremely dependent on if the developer creates a good tutorial and help section to get the player started. Either way it goes, HOUND is certain to become a truly interesting title both from a consumer and scientific perspective.