HOUND (PC) Preview

By André Eriksson 31.03.2017

Review for HOUND on PC

Nowadays it is difficult to create something truly different and unique, but HOUND is a title that has never been seen before in gaming. The player creates different lifeforms, and the AI uses neural networks to memorise and learn patterns that evolves them so that they are able to survive on their own, creating an evolution simulator unlike anything that has come before. Cubed3 takes an early look at this very experimental title.

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.

Screenshot for HOUND on PC

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.

Screenshot for HOUND on PC

Final Thoughts

For people interested in evolution and/or neural networks and how they can be used to create better AIs in the future, HOUND is certainly a title that is worth taking a peek at. It is an experimental game that touches zones very few projects have ever touched. Otherwise, it fully depends on whether or not a good tutorial is created in-game to help the player get started in what might be one of the most complex simulator games on the market, which is not a light feat.

Developer

Nick Shesterin

Publisher

Meridian4

Genre

Simulation

Players

1

C3 Score

Rated $score out of 10  n/a

Reader Score

Rated $score out of 10  0 (0 Votes)

European release date Out now   North America release date Out now   Japan release date None   Australian release date Out now   

Comments

There are no replies to this preview yet. Why not be the first?

Comment on this article

You can comment as a guest or join the Cubed3 community below: Sign Up for Free Account Login

Preview PostPreview Post Your Name:
Validate your comment
  Enter the letters in the image to validate your comment.
Submit Post

Subscribe to this topic Subscribe to this topic

If you are a registered member and logged in, you can also subscribe to topics by email.
K-Pop Korner - The Best of Korean Music
Sign up today for blogs, games collections, reader reviews and much more
Site Feed
Who's Online?
oroblram

There are 1 members online at the moment.