I have a deep love for this game, but lately this has been fading deeper and deeper into disappointment.
I bought it on sale a bit over a year ago, I think it was about $10 at the time, which for 350+ hours of time in a game is certainly money well spent. Many of these 350 hours were enjoyable, and if you'd asked me 6 months ago whether to buy this game, I 100% would have told you to go for it.
The conceptual design of the game is amazing. The way the ships lock together through all the different engineering features is just plain awesome. The dynamic nature and wide range of build options will unleash your creative child, and you'll find yourself spending many hours nutting out tricky designs to see what works best.
The community is even better. The people I've met playing SE have usually been mature and fun to play with. There's always a few exceptions .. but hey, we're gamers, right? The mod community is extremely active and you can find a mod for just about anything.
So why won't I recommend this game? Well ...
It's buggy.
And by buggy I don't just mean buggy in an annoying way. I mean buggy in a way that will progressively drive you crazy.
With early access, this is normally forgivable. Early access itself comes with an implicit promise that the developer will eventually fix most (if not all) of the bugs and eventually release the game. So paying cash for something in early access usually involves developing a thick skin for the bugs. I knew this going in, and it didn't stop me.
But after 2 years in early access, a disappointing pattern has been emerging with this game. I don't think that the person responsible for prioritising the work for SE actually plays the game. If they did, they would likely be running things very differently.
At the moment SE is running a weekly release cycle. Mid to late in the week, the developers push out an update containing features and often a few fixes. The root problem seems to be that each update brings too many features, and not enough fixes.
The result is that over time, what initially started as a fairly simple physics sandbox game has evolved into a massively ambitious project full of unfinished features. Almost every time a feature is released, it gets 90% done, then left on a shelf. Sometimes the 10% isn't that important, and the feature is worthwhile. More often the 10% is something rather critical (like ability to work in multiplayer) and the feature simply becomes a point of frustration, as you spend hours trying to make it work or basing designs off it, only to have it fail you in the worst possible ways.
I have very limited experience in games development, but 15 years in the software industry have taught me that letting the bugs stack up is a really bad way to develop a product. The longer a bug remains in software, the harder it is to fix. The more you stack on an unstable codebase, the buggier it gets. The SE forums are completely littered with bug reports, many of them very serious. The developers know about the bugs, but instead of fixing them, they just add more features.
I don't think this is going to end well for SE. At the time of writing this review, I can no longer stand to play the game through the seemingly endless number of bugs that plague it. It's starting to look like one of two things will happen:
1. SE will stay in early access forever. It will continue to sell copies. The developers will keep adding features. The bugs will keep stacking up, until some critical level is reached where people get so frustrated that the review standings and sales plummet. Development then scales back (or stops) and the game is left in early access.
2. SE will leave early access in a buggy state, not with the level of stability expected in a finished product. This will in turn lead to lots of disappointed people, bad ratings, and probably a collapse in sales and support.
I want Keen Software to prove me wrong in this. I would LOVE for them to prove me wrong, and get SE out of early access into a stable state where it can be as awesome as it deserves to me. But with every feature release, I lose a little more faith. This seems to be a constant problem with the approach of funding games using early access - there is no commercial incentive to ever actually leave early access.
So it looks like SE for me will remain a broken dream. To anyone that is considering buying this game, I recommend first taking a look at Empyrion - its build system isn't yet as sophisticated but even in its very early days I found it much more stable than SE.
But if you can't resist and you really want to buy the game, be careful with the following features. They will give you little but frustration:
- Pistons and Rotors (these mess up the centre of gravity in ships, and intermittently cause massive unexplained velocity/damage - especially in multiplayer)
- Wheels (no effective brakes + other physics bugs)
- Small ship conveyors (these don't bridge effectively with normal conveyors. The blocks that connect them are intemittently unreliable)
- Landing gear (hitching and physics issues, especially in multiplayer), connectors are better
- Batteries on complex power grids (when combined with reactors and solar panels, these need to be micro-managed. It's improved recently, but will still drive you crazy)
- Projectors (these will destabilise servers in multiplayer, giving big crash issues. Big loss as there's no way to move ships to survival mode without them)
- Large ships and complex worlds (the game server itself seems to run all ticks on a single thread, so any more than 1 CPU core is wasted. This means there's no way it can scale as ships get bigger and worlds get more complex. Everything just slows down)
- Ships parked on unlevel planet surfaces in multiplayer (jerks around like crazy, can be very disappointing as it often destroys the ship)