Rethinking User Input: A Pseudoscientific Approach

No peers to review this one.

It was an encouraging week over here at Spieler HQ, folks. We had a rather dicey several days in the wake of my altered employment status, but the bureaucratic headaches are now resolved and opportunity stretches out before us. The world coquettishly beckons me toward a marvelous, intoxicating destiny, and I will rush to meet it! Or, in any case, that’s roughly how I felt when my tax refund check cleared and it sank in that I had another three days of free time each week. As I write these words, most of this newly acquired free time is being reinvested in The Spieler. I have some great stuff planned for the very near future.

Today, I’ve got one more scalp-camera-gameplay-adjacent piece for you. It marks the final newsletter whose production was directly affected by real-life shenanigans, but it turned out to be a lot of fun to produce nonetheless. From here on out, though, expect a gradual increase in ambitiousness. My near-term goals for the publication largely boil down to “more and better output” and “find more high-minded creative types who’d want to consume it.” I have plans for both, of course, and I’m superlatively grateful to everyone who’s shared, Subscribed, and left comments thus far. Big things are ahead.

Alright, let’s get to the fun part. An unignorable opportunity suddenly dropped into my lap last week: my non-gamer spouse asked to have a go at 2012’s most violent video game, and a choir of muses began to sing in my ears. A chance to hasten the trajectory of ludophenomenological science and to stroke my ego as (presumably) 2025’s most prolific writer on a low-budget indie game from nearly thirteen years ago! And they said God was dead.

---

For context: my wife has lots of experience with Stardew Valley, some experience with Nintendo’s oeuvre, and approximately zero experience with any other video games. I was surprised at first when she asked to give Hotline Miami a try on my PC, but I guess I’m just that magnificently compelling of a writer. Then again, it’s not like she didn’t already have a taste for blood — she’s messily brutalized her fair share of Slimes in the caves under Pelican Town, after all.

I was happy to oblige, of course. I wouldn’t miss the precious opportunity to study the effects of scalp-camera gameplay on an individual with no prior exposure to top-down action games. Hotline Miami probably isn’t the best introduction to that style, but it wasn’t a terrible choice, either: hardcore shooter gameplay turns out to be surprisingly approachable when instantaneous retries are built into the core mechanics. My first shooter was the original Counter-Strike, and I used to have to wait minutes at a time in between being sloppily killed and getting another try. Technology marches ever forward in spite of itself.

Let me tell you how it all went down.

EXPERIMENT ONE

The first experimental procedure consisted of sitting her in my desk chair and pressing the “Start” button. Whereas subject was new to the paradigm of 360-degree aiming, the tutorial stage took several minutes longer than anticipated¹. This certainly wasn’t due to a paucity of reflexes or hand-eye coordination because, as I say, this woman can breeze through the Skull Cavern in Stardew without suffering a single creature to live.

By the time she was midway through the first mission in the Metro, I was convinced that Hotline Miami’s controls were showing their age even at the most fundamental levels of character movement and aiming. Its implementation of octodirectional movement looks frankly stone-age from the spectator’s point of view — the acceleration curve is so brief that the protagonist breaks into a sprint almost instantaneously upon pressing a movement key, and it only starts to feel predictable after significant practice. “This must’ve been what it looked like the first time I played in 2012,” I remarked aloud.

“Ssh,” responded subject, who was trying to focus on killing two men with a briefcase. I was serious, though — for as much as I praise Hotline Miami’s Tarantino-esque narrativization and legendary spritework, the gameplay itself is pretty inconsistent. When it’s working correctly, you find yourself choreographing ad-hoc fight scenes that would give Tarantino a priapism. But when it’s not, you find yourself struggling to wrangle those goddamn doors. In subject’s case, the most notable source of friction was aiming under pressure. This wasn’t surprising, given the tiny cursor sprite and crunchy, low-resolution character artwork that make target acquisition very difficult in the heat of the moment.

We eventually finished the mission, and she got a good laugh out of being evaluated based on the brutality of her slaughter. I gather it was a refreshing change of pace from the archetypal performance review. She politely declined further wetwork at approximately T+9:30, so I drew the first experiment to a close. My foremost observation: the cognitive load produced by aiming in 360 degrees while moving in two dimensions is unexpectedly burdensome. Other games have come along in the years since to significantly refine these core mechanics, but what if we’ve been missing the forest for the trees? What if there exists a better, more accessible paradigm for scalp-camera action gameplay that doesn’t sacrifice its cathartic appeal?

But before I got to thinking about that, I needed a more rigorous approach to data collection. Conveniently, we’d had enough fun that subject suggested a second attempt the following day.

EXPERIMENTS TWO THROUGH SEVEN

Subject was installed in a Järvfjället seating apparatus and equipped with peripherals necessary to facilitate man-machine interface. Scalp-camera action subroutines were invoked and subject was reexposed to wetwork simulation. Subroutines equipped subject with one (1) IUPAC-standard crowbar and presented subject with two (2) Russophonic antagonist analytes R₁ and R₂. Crowbar was used to introduce a large bolus (8.164662 x 10⁵ g) of high-carbon steel to the scapular region of R₁. An aqueous, crimson-colored solution (approx. 3,000 mL) rich in erythrocytic hemoglobin precipitated, causing the temporal expiration of R₁. R₂ was axially reoriented to face subject and a parasympathetic reaction was effected in subject’s nervous system. Subject attempted secondary administration of high-carbon steel to R₂. Bolus was intercepted by inert atmospheric medium and no reaction took place. A mechanical impulse (approx. 31 kN) was applied by R₂ to the cranial region of subject’s avatar using a Louisville athletic bludgeon, concluding the experiment.

Experiment was repeated five times under same initial conditions. Results were consistent with those of first trial.

EXPERIMENT EIGHT

Initial conditions of Experiment Two were again recreated. A reduction in higher-order hippocampal activity was observed in subject. Avatar was spatially modulated in the approximate direction of R₁ and R₂. Subject rapidly actuated input mechanism associated with crowbar-administration procedure until R₁ and R₂ were mechanically neutralized. Subject experienced an excitatory dopaminergic reward event and emitted a vocalization consisting of an elongated close-front vowel sound².

Final experimental yield included two (2) structurally compromised humanoid cadavers, a moderate quantity (approx. 200 g) of loosely aggregated neuronal cell bodies and myelinated axons, and a quantity of mammalian blood too abundant to measure under laboratory conditions.

ANALYSIS OF EXPERIMENTAL RESULTS

So, what did we learn from this absurd farce? Well, I have three central takeaways:

1. Computer scientists can’t do chemistry

2. Hotline Miami’s controls are severely mediocre by modern standards

3. There was no observable correlation between repetition and skill improvement, suggesting that the latter must come down to one or more exogenous factors.

Those first two conclusions are already well understood by the mainstream scientific establishment, so my results simply confirm what was already known about Hotline Miami’s interface and about my own dearth of natural-scientific rigor. What replication crisis?

That third bit, however, is a straight-up discovery. Let me explain.

See, I embarked on this experiment under the assumption that an action-gameplay neophyte would struggle tremendously at first, adapt to the basics after sufficient repetition, and exhibit roughly logarithmic improvements in skill from there — that’s how I learned Hotline Miami, after all, and I suspect it’s how most gamers learn to play difficult action games with tight focuses on player skill.

But that hypothesis was not borne out by my data. Instead, I observed the expected repetition taking place (attempt combat encounter, fail, try again within seconds), but each attempt looked very similar to the one before, and significant progress came only when the pretense of deliberate control was abandoned altogether in favor of maximizing entropy within a given encounter. Why the disconnect? What external influence makes cognitively demanding action gameplay easy for some to pick up but next-to-impossible for others?

Still intoxicated by the astringent spirit of discovery, I plugged in an Xbox controller in order to halve the number of objects being manipulated. Subject noted that it felt more natural in the hands than keyboard-and-mouse, hypothesizing that the controls would be more intuitive with only one peripheral to manage. She confidently strode backwards into a combat encounter and swung at the air a few times, whereupon a bat-wielding mobster disproved said hypothesis. We kept at it a few minutes longer and eventually concluded that the central issue was not the input devices themselves but the burden of simultaneously managing different actions. After a bit of practice, she could effortlessly move about or aim at enemies, but synthesizing the two into a discrete move-and-aim unit proved intractably difficult. Given that, Hotline Miami’s primary gameplay challenges become effectively insurmountable by the intended means.

After much reflection, my best guess is that these are all consequences of the contemporary two-handed control paradigm³. In a certain sense, this is perfectly intuitive: if you need to perform different actions with each hand at the same time, then you’re at the mercy of the brain’s ability to coordinate between its two halves. It’s like patting your head while rubbing your stomach — one can get good at it with practice⁴, but the human brain lacks the evolutionary capacity to perform such complex multitasking on demand. Hotline Miami — like most action games — places different demands on each hand, generally in unpredictable sequences. The cumulative effect for the new player is one of tremendous and persistent cognitive load until they develop second-nature intuitions or deliberately abandon strategy and go hog-wild on the controls, hoping that the combat encounter will resolve itself by coincidence.⁵

What about me, then? Am I good at Hotline Miami because my brain is just naturally more capable of binary multitasking, or because I’ve played hundreds of high-octane action games and simply have very well-developed intuitions? My instinct is that it’s almost certainly the latter, particularly since I can barely multitask well enough to breathe and sit still at the same time in the absence of amphetamines. But such a thorough background, while clearly sufficient, is obviously not a necessary condition for adapting to the demands of action gameplay. Otherwise, the franchise wouldn’t have millions of dedicated fans.

COULD/OUGHT WE DO BETTER?

So, I ask again: what if there exists a more accessible yet equally effective paradigm for controlling scalp-camera action than that which currently prevails?

When I say “accessible,” by the way, I very specifically mean “reasonably straightforward by normie standards” rather than “accommodating of all skill levels, period.” I make this distinction because I don’t think the latter strategy would work for Hotline Miami or other games like it, even though I’m strongly in favor of making games more playable for more people.

By way of example, a blunt solution might simply be to add “assistive” gameplay options that make the game easier or even outright unloseable, cf. 2019’s Control — by those means, players for whom the game’s challenges are more frustrating than stimulating can still enjoy the narrative, art design, and the general vibes. I think this strategy worked fine for Control because that game was essentially a work of SCP fan-fiction that Sam Lake presented as a video game instead of a Wiki entry, so its appeal is (by my estimation, anyway) grounded principally in its narrative and only secondarily in its gameplay challenges. A fine approach for the plucky Finnish auteurs at Remedy Entertainment.⁶

Perhaps not so for Hotline Miami. My wife suggested that it could benefit simply from difficulty options, and while I agree that it’d make the game more playable, I worry it’d cheapen the overall experience. One of the reasons I love the game so much and won’t shut up about its glorious legacy on modern video gaming is that it marries narrative, art design, and general vibes with its core gameplay, and I can’t imagine how that balance could realistically be maintained if the designers couldn’t count on every player facing precisely the same challenges. I even wrote a whole-ass newsletter considering this phenomenon:

In other words: in order for the quintessential Hotline Miami experience to track, the player has to feel like a hilariously outnumbered, drug-addicted maniac whose every inch of progress is won after a desperate struggle at unsustainable expense. If the enemies take unnaturally long to react or if their savage attacks don’t kill the player, that sensation necessarily breaks down.

As I said in the abovementioned newsletter, one of the game’s greatest strengths is its commitment to ludonarrative synchronicity, or the reinforcement of narrative elements through gameplay obligations. My central contention here is that the cognitive burden inherent to controlling the game is not fundamental to that synchronicity and could be reduced to profitable effect. The ideal version of Hotline Miami would be one in which every single failure comes as a fair consequence of the player’s decision-making, and I reckon that’s a suitable goal for just about any game predicated on challenging the player.

Let me spitball a few ideas for how we might theoretically go about achieving this:

• Invent a mind-machine interface that produces a perfect, one-to-one map of conscious intentions to their associated gameplay effects (I worry this might take awhile)

• Invent new, more intuitive input peripherals that reduce the cognitive burden of manually translating intentions to gameplay effects

• Invent new, more intuitive paradigms of user input that expect fewer mental gymnastics of the player without sacrificing any important aspect of the gameplay experience

I am of course cognizant of the fact that video game input has been treated as a solved problem since the advent of twin analog sticks and that these are relatively idiosyncratic ideas. Then again, it was also being treated as a solved problem forty-plus years ago when Coleco decided to iterate on the Intellivision controller instead of starting over from scratch:

Above: A controller for the ColecoVision, a home console from the early 80’s. The competing Intellivision controller, released by Mattel a couple years previously, included the goofy keypad and tendonitis-inducing side buttons but lacked the syrupy joystick.

Point is: there’s been a relatively stable paradigm of two-handed control for the better part of video gaming’s history, but that hasn’t stopped folks from relentlessly trying to innovate. Wii-style waggle controls had little staying power, and the less said about Microsoft’s Kinect, the better. VR isn’t much healthier from a broad commercial perspective, but it certainly represents a fundamentally more immersive style of control for those who aren’t brought to paroxysms of vertiginous nausea by the rest of the experience. There must be some way of meeting traditional input in the middle, and I intend to elucidate it.

Just not right this moment — we’re starting to run long. I’ll let you know when I have more thoughts worth sharing on the matter. In the meantime, I welcome your perspective if you have any to offer. Another truth of which my experiments kept reminding me: after one spends nearly three decades playing thousands of video games, one tends to forget what gaming felt like before one could effortlessly mow down twenty Slavic bangers in a dance club.

In Hotline Miami, I mean. My IRL record for bodies caught at the club is a comparatively paltry fourteen.

See ya next time <3

1. Incidentally, didja know that it’s possible to die in the tutorial stage? I didn’t, and I’ve been playing this damn game for over twelve years.

2. To wit: “Ee!”

3. I should once again thank Jim for helping me refine these observations with some extraordinarily thoughtful comments on the most-recent Hotline Miami-related newsletter.

4. ...as indeed my wife has. She demonstrated after proofreading that line, profoundly humbling my hardcore gamer instincts.

5. I think it’s most likely possible to beat Hotline Miami like this, although I don’t suppose that so doing would reflect the intended experience.

6. I swear it’s just a coincidence that my newsletters keep specifically offering praise for Finnish game-devs.