What Actually Makes a Horse Game — And Why RedM Changed Everything

If you’ve ever searched for “best horse game” and ended up disappointed, you’re not alone. Most games that claim to be about horses treat them like motorcycles with manes. You mount up, you go fast, you dismount, you forget they exist until you need a ride again. There’s nothing inherently wrong with that approach, and horses as transportation works fine for games that are about something else. But it’s not a horse game. And for those of us who grew up riding, or being around horses in general understand the difference between a horse that trusts you and one that’s just tolerating you, who know what it feels like to build a partnership with an animal over months or years. And that gap between what we want and what exists has always been frustrating.

Even Red Dead Redemption 2; which has genuinely the best horse mechanics in any video game to date, still treats your horse as transportation first. Sure, you bond with it. You might feel a small hit to your chest when it dies. But at the end of the day, horses exist to get you to the next mission. They’re not the point. And you just go on and buy another one from the stable.

That’s the gap we set out to fill when we built The Tackroom.

Why Red Dead’s Horses Set the Standard

Here’s something most players don’t realize: horses are one of the hardest things to animate in video games. Not just difficult; genuinely, notoriously hard. There’s a reason most games treat horses like reskinned motorcycles, and it’s not laziness. It’s that doing it properly requires an enormous amount of specialized work that most studios can’t justify.

The fundamental problem is that horses aren’t built like other animals. They’re a massive amount of weight supported on four surprisingly slender legs, with most of the muscle concentrated in the upper thighs while the lower legs operate through a complex system of tendons that stretch and snap like rubber bands; differently for each gait. Where a human character model needs maybe two moving joints per leg to look semi-realistic, a horse needs around four per leg, and those joints have to bend and spring in very specific ways or the whole thing looks wrong. Get it slightly off and your horse looks like it should collapse under its own weight. Elden Ring’s Torrent is a good example: fun to ride, but an actual horse moving that way would crumple into a heap.

Then there’s the gait problem. Most quadrupeds share the same basic footfall patterns; walk, trot, canter, gallop, but horses execute these gaits with timing and weight distribution that’s genuinely complex to replicate. The canter alone is a three-beat gait where the intervals between foot strikes are irregular, which means you can’t just loop a simple animation and call it done. The body weight shifts constantly, the spine flexes, the head and neck counterbalance the movement, and all of it has to happen in sync or the horse looks mechanical. Animators have studied this stuff going back to Eadweard Muybridge’s photography in the 1880s, and getting it right in an interactive, real-time game environment is exponentially harder than getting it right in a pre-rendered film.

The honest truth is that for most games, it’s not worth the investment. If your game is about combat or exploration and the horse is just transportation, players aren’t going to notice whether the leg joints flex correctly during a collected canter. They’re going to notice the enemies ahead of them, the landscape around them, the objectives on their map. Studios know this, so they allocate resources accordingly. The horse gets “good enough” and development moves on to the parts of the game players will actually focus on.

Rockstar decided “good enough” wasn’t acceptable.

Their team at Rockstar North spent years studying real horses; filming them in Scotland, motion-capturing a Hollywood stunt horse named Blanco, and working with professional riders to understand how horses actually move. They didn’t just want horses that looked good. They wanted horses that felt like horses. Lead AI Programmer Tobias Kleanthous explained in his GDC talk that their early prototype was “functional, but missing something”; it moved correctly in a technical sense, but it felt like a car, not a living animal. That single observation drove years of additional development.

The result was something no other game had done: lateral movement. Real horses don’t just turn; they sidestep, shift their weight, lean into direction changes. When a horse prepares to change direction, it doesn’t rotate on a pivot point like a car. It repositions its feet, adjusts its balance, and moves its mass in a way that’s fundamentally different from any vehicle. Rockstar built systems that actually account for the irregular timing of three-beat gaits, the way weight transfers through the body at different speeds, and how a horse’s movement changes based on terrain and situation. They applied facial animation technology; normally reserved for human characters, to the horse’s muzzle, hindquarters, shoulders, and ribcage, creating subtle muscle movements that register subconsciously even if you can’t articulate why the horse looks more alive than in other games.

And then they kept going. Horse enthusiasts who’ve dissected the game point to details like visible veins, realistic eyes and teeth, the way sweat appears on a worked horse, and how the animals respond dynamically to temperature changes. Rockstar built horses where you can see the anatomy working under the skin; not because most players would consciously notice, but because the cumulative effect of all those details creates a sense of life that players absolutely feel. When someone spliced real horseback footage with first-person gameplay from Red Dead Redemption 2, viewers genuinely struggled to tell which was which.

Even equestrians who’ve analyzed the game critically agree the foundation is remarkable. The Mane Quest broke down the dressage mechanics move by move and found that while some animations like the piaffe could use refinement, the leg yielding is fantastic; and honestly, the fact that horse people are having detailed conversations about video game dressage accuracy says everything about what Rockstar built. No other game has come close.

The Problem: Rockstar Built a Foundation, Not a Horse Game

Here’s the thing though; Red Dead Redemption 2 is an outlaw story. It’s a game about Arthur Morgan, about loyalty and betrayal, about the dying West. Horses are supporting characters in a narrative about shootouts, heists, and survival. Rockstar’s stated goal was to create “a trusted friend with whom a deep connection is forged” over the course of Arthur’s journey, and they absolutely succeeded at that. Your horse feels like a companion in a way no other game has managed. But for those of us who wanted the horse to BE the journey? The base game only gets you so far.

You can bond with your horse, sure. But there’s a ceiling. You can’t breed. You can’t train specific skills. You can’t shape a horse’s temperament through groundwork or desensitization. You can’t develop a line of horses that carry your work forward across generations. The attachment you form is real and emotional, but the systems supporting that attachment are limited by the fact that Rockstar was making a story-driven action game, not an equestrian sandbox. The depth stops where the game’s actual focus begins.

That’s where RedM comes in.

RedM: The Engine That Let Us Build What Rockstar Couldn’t

RedM is a modification framework that lets communities build on top of Red Dead Redemption 2’s engine. It gives us access to that incredible movement system, those realistic animations, that sense of weight and presence that Rockstar spent years perfecting; and lets us extend it in directions a major studio would never prioritize. We inherit the foundation without inheriting the limitations.

The Tackroom exists because a group of equestrians and horse lovers looked at what Rockstar built and thought: what if we made horses the entire point? What if instead of horses supporting a story about outlaws, we built systems where the horses ARE the story? We’re not a game studio with shareholders expecting quarterly returns or executives who need mass-market appeal to justify development costs. We’re horse people who happen to know how to code; sometimes or break it in the meantime. That means we can build systems that would never survive a corporate pitch meeting; systems that matter deeply to players who actually care about horses, even if they’d be incomprehensible to someone who’s never been around them.

Training That Reflects Real Horsemanship

In most games, “training” means watching a progress bar fill or grinding the same repetitive action until a number increments. That’s not training. That’s waiting with extra steps. Real horsemanship is about communication, trust, pressure and release, reading the animal and being read in return. It’s a relationship that develops through consistent work, not a stat that levels up.

We built training systems that reflect those principles. Round pen training; although it’s just riding in circles – it does give you the sense of training in a round pen. Fear training; although we don’t have courage training in the SireVLC script,  helps them stay calm around stimuli that would spook an untrained animal: gunfire, predators, crowds, unfamiliar environments. Trail training rewards you for actually riding; increasing from 100xp every 5minutes to now 400xp; exploring the map, navigating varied terrain, building that partnership through hours in the saddle together rather than through a menu screen. Your horse changes based on how you work with it. A horse you’ve put real time into handles differently than one fresh from the stable. That progression isn’t a cosmetic stat boost that makes numbers go up. It’s felt in the movement, the responsiveness, the way your horse trusts you in situations where an unbonded horse would fight you.