Why Your Fight Scenes Look Like Bad Choreography
I spent three years working on a comic where the action panels kept getting rejected for looking "static" or "confused." The problem wasn't anatomy or composition. It was that every pose I drew had the figures standing in neutral stances before the impact happened, which made the reads feel weightless and directionless. Once I started actually studying combat reference footage with intent instead of just skimming it, everything changed. Most people don't realize that the stance and gesture of a fighter before and during a strike is what communicates weight, intent, and consequence. The discipline here is mostly about understanding that a fighting pose isn't something you invent. It's something you pull from real biomechanics and then compress for readability. I learned this the hard way when my editor told me a whole issue looked like people doing interpretive dance in a ring.
The Hidden Discipline Of Pose Reference Fighting How Stance And Gesture Dictate Combat Outcomes
This isn't a formal system with a name. It's something that happens when you actually study how fighters move and then apply that knowledge to static illustration or animation reference. The "discipline" part is just the practice of being deliberate about every angle of a hip, every shift in shoulder weight, and every finger tension before you put it on the page. When you look at a fighter mid-combo in video, you'll notice their center of mass is always moving forward, sideways, or dropping. A jab isn't thrown from a standing position. The back foot rolls, the hip rotates, the shoulder leads, and the supporting leg absorbs recoil. That chain of events is what makes the pose readable as a fight instead of a greeting. The gesture tells you exactly what came before and what's coming next. I hit a wall with one particular sequence where I needed a character to pivot out of a clinch and throw a knee strike while off-balance. Every version I drew looked clean and composed, which defeated the purpose entirely. Off-balance means asymmetry, compression on one side, extension on the other, and a center of gravity that's clearly outside the support base. I spent about forty-five minutes pulling frames from MMA footage, specifically a couple of videos of Demetrious Johnson's movement, and I noticed his rear heel always lifts first during a pivoting knee. That detail is what I added to the pose. It made the moment feel violent and controlled simultaneously, which is exactly what I needed.
The workflow most people skip is the stance analysis step. I go through three phases when building a combat pose reference sheet. First, I pick the specific strike or defensive move I need. Second, I pull reference from real footage and film the key frames in slow motion, usually pausing at the point of maximum tension. Third, I map the skeletal structure onto a mannequin or rough stick figure before even thinking about rendering or detailing. The skeleton map is where the actual discipline lives. I'm checking things like: is the weight foot grounded enough to support the force being generated, are the scapulae positioned to allow the arm to reach the target, and is the spine rotated in the same direction as the hip? If any of those are wrong, the pose will read as floating or unconvincing regardless of how well it's drawn. I've found that gesture drawings done in thirty seconds capture more usable fight energy than ten-minute rendered poses. The reason is that gesture forces you to commit to a line of action, and fight scenes live and die on that line. You can fix proportions later. You can't fix a pose that has no directional intent. When I built my own reference library over time, I ended up with probably two hundred gesture sketches labeled by type of strike, defense, and recovery. The library took about six weeks to compile, but it cut my daily work down significantly after that.
Get the Full Details
There are some limitations worth acknowledging upfront. This approach requires access to good reference material, which means footage or live observation. Stock photo sites have very little of actual combat dynamics. Most images online show posed fighters, not fighters in motion, and those are almost useless for what we're talking about here. You'll want to pull from MMA streams, fencing footage, historical European martial arts demonstration videos, or even stage combat reels. Stage combat is actually underrated for this because it's designed for readability at distance, which is exactly what panel composition requires. Another bottleneck is that static references can only capture one moment. A real fight is a continuous flow. If you only draw the landing frame of a punch, you lose the windup and the follow-through, both of which communicate power. The workaround is to create pose sequences rather than isolated images. Even just three frames, like the loading phase, the impact phase, and the recovery phase, will give your work much more credibility than any single perfect pose ever will. I also ran into a specific problem where my characters' hands kept looking too relaxed after landing strikes. In real striking, finger tension is extremely high during and immediately after impact. The knuckles clench, the forearm tendons lock, and the wrist stays braced. I noticed this pattern in nearly every fight video I pulled. Once I started emphasizing hand tension in my reference poses, the strikes read as much heavier without me changing anything else about the drawing.
For people who want to build a reference library, the easiest starting point is to pick one stance. Southpaw orthodox boxing is a good baseline because the mechanics are clean and widely documented. Study the lead hand jab, the rear hand cross, the hook, the uppercut, and the defensive slip. Draw each one from multiple angles. Do this with gesture first, then add bone structure, then layer in clothing and detail if needed. One person can build a functional library of about thirty solid poses in a single week if they focus only on the core set. Stance dictates outcome because it determines what's reachable and what's exposed. A fighter standing too upright gives up the leg. Leaning too far forward leaves the back vulnerable. A properly angled stance splits the difference by keeping the hips squared to the opponent while allowing the lead shoulder to drop for defense. This is why you'll notice that in any well-choreographed fight scene, the character's hip line is almost always at a different angle than their shoulder line. That separation creates the torque that makes the pose look like it could generate force. Gesture adds the emotional layer. A relaxed overhead pose reads as casual or preparatory. A compressed crouch with the torso twisted reads as explosive. The same basic stance with a shifted shoulder line and dropped elbow can switch between those two readings without changing anything else. That's the leverage gesture gives you, and it's why seasoned artists spend so much time on it before touching detail.
If you want downloadable reference material, several sites host free pose packs that include fight-specific gestures, though most are generic and need to be filtered. The most useful sets tend to come from animation communities and martial arts demonstration channels. I also keep a folder of screenshots from slow-motion fight breakdowns that I've labeled by category, which I share with collaborators through a shared drive. It's not a product, just something I accumulated over time. The discipline here is boring because it's mostly repetition and observation. You watch, you map, you draw, you adjust. There's no shortcut around actually seeing how weight shifts happen in real bodies. But once you internalize that, your combat poses stop looking like decorations and start looking like moments where something real just happened.
