While it isn’t unusual to see developers iterate on a previous release, adding new characters, story beats, or features, you don’t normally see them change up their visual style. Sword of the Necromancer: Resurrection does just that though, taking the well-worn top-down pixel graphic look and moving to a third-person polygonal one. Not content to just give it a visual overhaul, new complexity and game systems were added as well, though in the end I’m not positive they clearly made the experience better in the process.

What has remained the same is the general story, the ability to resurrect your former foes to fight on your behalf, and the general style of play mixing together action as well as some strategy. You’ll be making your way through procedurally-generated dungeons, taking on a variety of monsters, as well as solving puzzles along the way. Armed with 3 weapons, and working with as many as 3 monsters in tow, you’ll have a variety of options for taking down enemies. Where things get more interesting and involved, is the fact that you can essentially program your monsters on a rudimentary level to work through a progression of commands, giving you a pretty impressive degree of control over how combat can play out… at least if you have the patience for it.

Where the game struggles, like its predecessor, is in moment-to-moment excitement and fun. Combat can be very hit-or-miss overall, in part due to the troublesome camera, but also because it typically feels more like going through the motions than exciting. At times it can be hard to follow, as you try to manage which enemies to focus on, but even when things are clicking it feels generic more often than not. In terms of your summoned monster helpers, it’s absolutely a mixed bag. While trying to solve puzzles I’d often need to shoo them away unless specifically needed, since more often than not they’d simply get in the way. While I appreciate the attempt to take on a programming edge to make them more helpful and capable, I’d much rather have seen more focus on the camera and making the combat easier to follow without quite so much trouble. In the end the time it takes to try to program behaviors simply doesn’t tend to yield results that are clearly worthwhile, so while this feature is novel its true value is ultimately questionable.

Reviews like this are always a bit tough to write. On the one hand, there are ideas at play here that feel fresh and like they could be interesting. On the other, in most areas I’d say the implementation is spotty at best, and while the results are unique that doesn’t make them good by default either. Overall, I’d say the move to the third-person perspective wasn’t a clear win. With the new problems it introduced with the graphics having odd artifacts at times and quite a few issues with the camera, I think I honestly prefer the previous version when it comes to visuals. As for the rest there are some interesting ideas at play, I just can’t say they clearly resulted in a more satisfying experience in the end.