Premature Evaluation: Parkitect

After a couple hours of tinkering and revising, I've almost got a decent amusement park in Parkitect, Texel Raptor's pseudo-remake of Rollercoaster Tycoon. Most of the rides are full or at least half-full, I have a cool rollercoaster or two that are drawing a lot of customers, and I'm turning a handsome profit every month. It's a good-enough park, and the customers who walk back out through the gates feel like they got what they came for, but with a little more care and convenience it could be great. Not unlike Parkitect itself.

Parkitect takes most of its cues from the original Rollercoaster Tycoon, which means it's effectively two games in one. On one level it's a lighthearted, lightweight business-management game about building, maintaining, and expanding a profitable theme park. On another level, which only tangentially relates to that first game, it's a rollercoaster construction kit, where you build a variety of different rides in a variety of different styles, customizing each twist, dip, and dive to make sure it's giving riders the experience you—not necessarily they—want.

There are, of course, the kind of weird, macabre touches you'd expect from a game that works so hard to channel Rollercoaster Tycoon. The first person to ride my Alpine rollercoaster was a little bald kid in a blue shirt and baseball cap that looked almost exactly like that awful Caillou character, which meant I hated him on sight.

Fortunately, I hadn't done any safety tests of the new coaster, which was really just a test of my own ability to design and build my own coaster using Parkitect's layout tools. It wasn't a great ride, limited both by the simplistic engineering of the Alpine design template and my own inexperience with the tools, but it had a few big drops and steadily built-up speed through a series of downward corkscrews on the back-half of the track. Caillou Two started making little cheering noises as he went through one of the long banked turns at over 40 km/h.

Something I hadn't noticed, while I was building my coaster, was a cryptic little rectangular button with a red dot on it. Later, when I hovered over it, I would discover that the button was for "braking" segments to slow the ride down. But I'd just been focused on making sure that the end of the ride connected with the start so that the loop would close neatly. I figured that the ride would simply end when the car pulled back into the boarding station. Which, in a manner of speaking, it did.

I don't know how fast Caillou-Two was going when he entered the final turn and then hit the final, steep dive back into the boarding station (I'd miscalculated my slopes and needed to shed a few more meters of elevation in order to complete the circuit). What I do know is that the car slammed violently into the loading area at full-speed. With a loud pop, Caillou disappeared.

The ride's holy mission now complete, I added some braking near the end of the ride which ensured that all future riders, who were good people and not insufferable toddlers, were able to complete their rides safely and get back to spending money in my amusement park. Editing the ride, however, further underscored the fussiness of the construction interface that had caused my first design to go so lethally astray. It's possible I missed some nuances to the system, but the coarse-grained elevation controls meant that the slightest change in one segment usually throws-off the the rest of the track as you attempt to re-connect the new section with the old. Even when it looks like you've succeeded and the model looks seamless, the game will register a break in the track and you'll watch rollercoaster cars inexplicably derail time and again. I would have killed for some kind of automatic "splice segments" button that could have done the job instead.

Still, there weren't any repercussions for killing a rider. In fact, I didn't get the sense that the game had even taken note of the fact that a character appeared to have perished on one of my rides. I understand that a game like Parkitect cannot and should not punish the slightest error with immediate reprisals, like a fully-justified wrongful-death lawsuit, but I also feel like after watching Caillou Two wink out of existence like he'd just been Raptured, people would have at least thought twice about staying in the queue.

In fact, Parkitect has an overall problem with push-back that I find in a lot of modern sandbox construction games. Again, it's understandable: death-spirals aren't fun (unless they involve a trainload of helpless rollercoaster passengers) and it would be easy for players to paint themselves into a corner.

It is perhaps more satisfying game design for most people to avoid punishing players in favor of lavishing rewards on strong play, but I found the feedback vague enough that it made it hard to assess my own efforts. The people visiting my theme park were only too happy to waste time in mediocrity. It didn't really matter where I was putting the trash bins, or where the mechanics were assigned, or whether there was a decent flow between the rides. When my park really sucked, people were moderately bored and peeved. When it was mildly crap, people sort-of enjoyed themselves. When it was decent, they spent a little more money. In any case, visitors seemed to stick around, and park attendance continued to rise.

I think one reason that Cities: Skylines turned everyone into a planning-obsessed traffic engineer is because gridlock was one of the few ways the game really highlighted failure and inefficiency. For the most part, people didn't seem to much care where they lived or where they worked, and would continue to arrive either way, but you could at least tell when businesses couldn't move their goods out of the industrial districts, or when nobody could reach the shopping districts because congestion was too bad. It wasn't a perfect solution, but it at least gave players something to strive for: effective transit.

Right now, I'm not sure Parkitect really has a similar hook for the perfectionist. Visitors enjoy bad coasters and good ones alike, so there's not much incentive to fine-tune a design unless you're really obsessed with lowering a ride's footprint or unleashing special torment. Visitors will find their way to shops, and employees will find their way to their jobs provided you have enough of them, so the business kind of runs itself. In its sandbox mode, at least, Parkitect is an enjoyable game about planning and design undercut by systems that don't require enough of either.

Parkitect is available on Steam, but your stack must be at least this tall for this ride: £13.59 / $24.99. My impressions are based on build 1154835 on 14 June 2016.