Good Robot was co-developed by Shamus Young, best known for his work at the Escapist Magazine (Stolen Pixels and the recently ended Experienced Points) and his own personal website, Twenty Sided. I’ve been a regular reader of Young’s work for about two years and consider him to be one of the biggest influences of my writing, both in terms of style and content. While standard video game review sites tend to be concerned with driving pre-release hype machines (and aggregating the superficial aspects of new releases into point systems so they can be further aggregated on Metacritic so they can serve as some sort of proxy to evaluate a game’s critical success in the most superficial way possible), Young tends to write long-form analyses of games which situate their qualities within the video game medium as a whole. He is less interested in describing why a game is good or bad, or whether or not someone should buy it, than in figuring out what makes video games work and not work, how narrative and mechanical properties can be skillfully integrated, and how the medium as a whole can be improved over time. (He describes his commentary approach here.) Along with pundits like George Weidman, Campster, and Yahtzee, Young is one of the few video game commentators today who makes real intellectual contributions towards moving video games forward as an artistic medium.

(It’s also worth noting that Rutskarn worked on Good Robot as well, and I like a lot of his stuff too.)

So of course I was excited to get my hands on Good Robot, even though I honestly wasn’t really sure what to expect of it. I didn’t know how Young would or could incorporate the insightfulness found in his hilarious take down of Fallout 3’s moronic story or his seemingly endless walkthrough of the Mass Effect series (which is up to Part 42 as of writing this) into a relatively simple game about shooting a lot of robots. I was also concerned about how my perception of the creator would affect my evaluation of the game. Would I be overly-disappointed at every minor error because I expect more from a creator I admire? Or would I unfairly ignore faults because I want the game to live up to my expectations?

Ultimately I have no idea if my evaluation of the game is skewed, but regardless, I’m happy to report that after seven hours of playtime, I am loving Good Robot.



Ironically, a video game made by a guy who is great at describing video games is extremely difficult to describe.

The problem is that Good Robot just feels so damn good, and it’s hard to describe why it feels good without just telling someone to just play the game. Game-feel is an incredibly important but elusive aspect of all game design. It bypasses most explicit descriptions and comes from a realm of intangibility. Describing why it feels good to fly and shoot in Good Robot, run around and sneak with Snake in Metal Gear Solid V, or fight friends in Super Smash Bros is like trying to describe what it is about a comedian’s timing that makes a joke funny.

The most I can say is that Good Robot’s mechanics achieve a feeling of freedom and control rarely seen in shooters. The titular Good Robot is an adorable floating sprite which glides around beautifully colored levels with 360 degree movement and an equally wide shooting arc. The early scattered enemies soon swell to massive hordes which fire even larger barrages of projectiles at the player’s general vicinity while he scrambles to dodge shots, shoot down missiles, or otherwise mitigate incoming damage through wild improvisation. The only way for the developers to make this system work is by carefully balancing the player’s control with enemy power. If the enemy projectiles were too weak or the player too fast, the game would feel patronizing. If the projectiles were too powerful or the player too slow, the game would be excessively punishing. Fortunately, the player’s control over the Good Robot is so crisp and precise, and the enemy units are so well balanced, that it threads the needle between those two positions to create level, satisfying game-feel.

The result is that victories and defeats in Good Robot feel earned. With rare exceptions, I feel like it’s my fault whenever an enemy hits me, and that it’s to my credit when I destroy a group of enemies. This should be the ultimate goal of any mechanically oriented game, especially roguelikes. If my death is going to make me lose an hour or more of progression, I had better not feel cheated if the game wants me play again. Good Robot nails those triumphant highs and depressing lows with expert precision.



What’s weird is that it took me a little while to get into Good Robot. Actually, I thought the game was pretty mediocre for the first hour or so. Not that it was bad, just that it was underwhelming, repetitious, didn’t warrant more than a few hours of play, and therefore wasn’t justified by its default $9 price tag. What turned my evaluation around was that the mechanics just clicked in my head at some point. What started as rote “fly in range of the enemies and shoot them” transformed into a smooth process of sweeps, angles, and circles, largely owing to the game’s unusual level design.

The way it works is that the player enters procedurally generated caverns consisting of interconnected maze-like tunnels, at the beginning and end of which are sanctuary areas with upgrade shops. Enemies rarely initially attack the player, but rather sit and wait in static positions or emerge from set locations within the labyrinth. As a result, the player is forced to navigate through the labyrinth and take on clusters of enemies in particular spots. My unwise early inclination was to charge through the tunnels and inevitably end up fighting multiple clusters from different angle simultaneously, thereby trapping me in a crossfire. I soon learned that the way to deal with the tunnels is to perform surgical sweeps of segmented areas to divide and conquer the general mass of opponents.

At times Good Robot weirdly resembles a cover-based shooter. After all, floating out in the open is a good way to make your Good Robot a Dead Robot. I found myself subconsciously developing a pattern of sweeping into a new tunnel to reveal an enemy cluster, firing off a barrage, fleeing to the nearest barrier, and then circling in and out of cover to fire more barrages. In the early levels this strategy can feel a bit gamey (and to be honest I find it a bit annoying to have to go through the slow paced early game after every restart). But as the player progresses and faces tougher enemies in more convoluted cavern structures, adaptation becomes a necessity, both in the form of raw improvisation and more complicated cover-based maneuvering.



Fortunately, the core combat mechanics are wonderfully integrated into a diverse weapon system. I haven’t played long enough to figure out how balanced the various grenade launchers, ion pulses, laser beams, and machine guns are, but the sheer diversity of weapon types is impressive. Good Robot has the type of weapon array which makes me fall in and out of love with another gun every hour. First it was this ion pulse thing which shoots ricocheting orbs, then it was a shot gun thing which reminded me of my beloved flak canon from FTL: Faster than Light, then it was this device which sent a burst of ion balls from my center mass in every direction, then a bouncy grenade thing, etc.

Good Robot is not a game with deep combat strategy where the player stops to consider how to approach critical junctures. Rather it’s a game based on flow and rhythm where the combat chunks meld together into a coherent progression. Thus the key to its weapon design is to play with trade-offs which passively nudge the player into subtly different play styles. The primary trade-off seems to be “precision vs. spread.” Weapons like the machine guns fire along a wide angle which reduces the need for player accuracy and makes short work of diffuse enemy clusters, but also makes it difficult to target specific enemies or hit anything hard at long range. Of course focused laser weapons have the opposite strengths and weaknesses. Other important trade-offs include “power vs. speed,” “splash damage vs. friendly fire,” and “range vs. projectile size.”

The ultimate effect of these trade-offs are to incentivize the player to use different combat styles based on movement and positioning. A player using the ion repeater, which rapidly fire ricocheting orbs, will tend to move around a lot and fire wildly in the hopes of using walls to create multiple simultaneous lines of fire. Meanwhile, a player using the ballistic rocket, which fires slowly but has long range and causes splash damage, will tend to keep a distance and carpet bomb enemy clusters. With weapon placement and type being procedurally generated, Good Robot forces the player to adapt to multiple styles for ever-dynamic combat.



Shamus Young is known for ruthlessly nitpicking other people’s games, so it’s only fair that I do the same to his game (hence, Bad Robot).

First, while I love how lethal the combat is (enemies do a lot of damage very easily), occasionally that can result in bullshit deaths. I think I lost my last three games because an enemy turned a corner in a narrow corridor and immediately blasted me with a full-strength shot. I understand that there is a lot of randomness in combat systems like this, and as I said before, Good Robot generally does a great job of making the player feel like victories and defeats are earned, but there are exceptions to this quality. I guess “getting suddenly blasted with a full-strength shot by an enemy randomly coming around a tight corner” is Good Robot’s equivalent to “getting crushed by a boulder while running up a staircase” in Dark Souls.

Then there are the exploitable boss fights. Overall, I think the boss fights are great, with lots of variety, cool designs, and especially awesome music thrown in. Unfortunately, a lot of bosses are unfairly vulnerable to long distance attacks. The enemies in Good Robot won’t fire at the player until he gets within the screen’s frame, and most enemies won’t pursue the player if he retreats, including bosses. So for a lot of boss fights, the player can simply walk into the center of the cavern, trigger the fight, then retreat back to the start of the level and launch unreciprocated fire off-screen until the boss is dead. Hopefully this issue can be patched down the line.



More importantly, Good Robot is too opaque for its own good. I think I get what the developers were trying to do but it ultimately doesn’t work. Good Robot has that Dark Souls style of exposition and progression where the player is supposed to figure out how the game works through trial-and-error and small details. Unfortunately, that process works better in big game like Dark Souls where the player has more time to develop his understanding of the game world through ample sources of subtle information. In contrast, Good Robot’s system comes off as needlessly confusing, if not a bit annoying.

For instance, at the end of each cavern the player is given the choice to progress through one of three doors to the next cavern. There are symbols on each door to denote the contents of the cavern behind it, but after seven hours of play, I still can’t figure out what half of the symbols mean. Two are for bosses (main level boss and sub-boss), one is for electrified walls, and another is for money. I have no idea what the rest are. Maybe I’m just slow, but the symbols are lost on me, and I can’t see how that benefits my experience of the game.

Then there’s the story, which is really damn cool. Due to rampant pollution, all of humanity abandoned the earth’s surface to live in underground caverns built by a shady corporation and managed by robots. But then all of the robots flipped out and murdered everyone. The player controls the last ostensibly “good” robot, in the sense that it is still loyal to its corporate creators. So in accordance with its programming, the Good Robot embarks on a heroic journey to exterminate all of the bad robots to cover up the whole mess and save its masters from lawsuits.

That’s a hilarious backdrop to a fun game. The problem is that the only reason I know that story is because it’s on the game’s Steam page. There is nothing in the actual game to suggest this setting besides short news blurbs and advertisements at the upgrade shops. Honestly, I stopped reading the blurbs after my second playthrough. By that time the combat had “clicked” with me and I was too caught up in the flow of the mechanics to stop and read flavor text.

Maybe that’s me being a bad player, but I’d speculate that most players would share my sentiments. Plus from what I did read, while the short clips were funny, they were hardly enough to form some sort of coherent story. Which isn’t to say that a mechanics-focused game like this needs much of a narrative; a simple setting will suffice. But if the Steam page is going to set up the game as if it does have an interesting story to tell, I expect it to deliver.

(Also, I haven’t actually managed to beat the game yet, so maybe there is some narrative closure at the end that I don’t know about.)



But these critiques are mere nit-picks for an otherwise excellent game. Good Robot isn’t earth shattering and doesn’t redefine the roguelike genre, but it does offer an extremely well-designed combat system which will keep players coming back for more long after they’ve felt they got their $9 worth. Anyone who likes roguelikes, fast paced shooters, or even just fun visual/audio design should check out Good Robot on Steam.