A LitRPG combat scene lands when the reader always knows what is happening, what it costs, and what it means, and the numbers make that clearer rather than interrupting it. The failure mode is not too much action or too little, it is a fight that reads like a patch note: skills fire, damage rolls, HP drops, and none of it lands emotionally because the prose stopped to recite the character sheet. Write the fight as a scene first and a system readout second.
Combat is the load-bearing wall of most LitRPG and progression fantasy, and it is also where I see the most manuscripts quietly fall apart. Running a web fiction platform, I read a lot of fight chapters, and the ones that do not work almost never fail for lack of action; they fail because the action stopped being a scene and became a transcript of a battle the writer could see but the reader could not feel. The stats were all there, the skills all fired in the right order, and somehow the whole thing had the emotional temperature of a receipt. So this is a guide to the opposite problem, not how to add more crunch, but how to make the crunch you already have carry weight, because in this genre the system is supposed to sharpen the fight, not sedate it. Everything below can be, and has been, successfully contradicted by writers better than me, but these are the patterns I keep coming back to.
What makes a LitRPG combat scene work?
A LitRPG combat scene works when the mechanics and the emotion move in the same direction, so that every number the reader sees also tells them something about the danger the character is in. The genre's whole promise is that the system is legible, that fights are governed by rules the reader can follow, and that legibility is a gift most action writing does not get. Brandon Sanderson's first law of magic holds that an author's ability to resolve conflict with magic is proportional to how well the reader understands that magic, and it is doubly true here, because in LitRPG the reader can literally see the rules on the screen. When you honor that, a fight becomes a puzzle the reader is solving in real time alongside the protagonist, and the tension comes from watching a known system get pushed to a place it was not built for.
The trap is treating "the reader can see the rules" as permission to show all of them, all the time. My feeling is that the system should behave in a fight the way a good referee behaves in a good match, present, decisive, and almost invisible until the moment it matters. Matt Dinniman's Dungeon Crawler Carl is a masterclass in this: the game layer is relentless and often blackly funny, but during the actual violence the numbers arrive in punches, one at a time, each one landing like a consequence rather than a footnote. That is the target. Not less system. System that hits on the beat.
Why do stat blocks and system messages kill the momentum?
Stat blocks kill momentum because they force the reader out of the body of the fight and into the interface, and every time you do that mid-swing you reset the tension to zero. A fight is a sequence of physical actions with rising stakes, and prose momentum lives in that forward motion; the instant you drop a six-line damage readout into the middle of a lunge, the lunge stops, the reader's eye recalculates, and the adrenaline you spent three paragraphs building leaks out through the table formatting. I think of it as the difference between a character checking their watch during a conversation and a character checking their watch during a car chase. Same action, completely different cost.
The fix is not to cut the mechanics, it is to change where and how they arrive. Show the effect in the body first, in the prose, and let the system line confirm it a half-beat later only if the confirmation raises the stakes. So instead of [Fireball deals 340 damage. Enemy HP: 1200/1540] interrupting the action, you write the heat cracking across the creature's hide, the smell of it, the way it barely slows, and then, if you want the number at all, a curt not enough from the interface as the thing keeps coming. The prose carries the hit; the system delivers the gut-punch. Andrew Rowe does a version of this in Sufficiently Advanced Magic, where the mechanical precision is always in service of a specific tactical problem the protagonist is trying to solve, never a stat dump for its own sake. If you are writing your system messages as their own craft problem, I wrote a whole piece on how to write a LitRPG status screen that pairs directly with this one, because a status screen that reads well standing still is the raw material for one that reads well mid-fight.
How do you keep the action legible?
You keep the action legible by giving the reader one clear image at a time and making sure each one changes the situation, because confusion, not boredom, is what makes readers skim a fight. A combat scene is a chain of cause and effect, and the reader has to be able to hold the geometry in their head: who is where, who has the advantage, what just shifted. When a writer tries to convey chaos by writing chaotically, cramming four simultaneous actions into one breathless sentence, the reader does not feel chaos, they feel lost, and a lost reader skips to the outcome. Will Wight's Cradle books are worth studying here because even at absurd power levels the individual exchanges stay clean; you always know what technique is being used, what it is meant to do, and whether it worked, so the escalation reads as thrilling rather than as noise.
The practical move is to write the fight in short causal beats and let the sentence rhythm do the pacing. One action, its result, the reaction, the next action. You can vary the length, and you should, but the causal spine has to stay unbroken, because the moment two things happen that the reader cannot connect, the scene goes blurry. This is also where the number of named skills matters. A protagonist who fires nine differently-named abilities in a single exchange is asking the reader to memorize a spellbook mid-brawl; a protagonist who wins with two skills used cleverly is legible, and legible is tense. The same discipline that keeps an underdog fight fair keeps a combat scene readable, which is why I treat this as a companion to my piece on writing underdog fights without breaking your own system. A fight you cannot follow cannot surprise you.
How do you keep a fight tense when the reader can see the numbers?
You keep a fight tense by putting the real stakes just outside what the stat screen can measure, so that a full health bar still feels like thin ice. Health points are honest about exactly one thing, how close a body is to zero, and if that is the only variable in your fight then the reader can do the math and the suspense evaporates. The best combat in the genre understands that the interesting stakes are almost always the ones the interface cannot itemize. In He Who Fights with Monsters, the mechanical clarity is high, but the tension routinely comes from cost and constraint, from afflictions and resource management and the price of a given victory, not from a simple race to drain an enemy's health. The number is visible; the danger is elsewhere.
So when you plan a fight, ask what can go wrong that the HP bar will not warn the reader about. The mana that runs dry two exchanges before the enemy does. The cooldown that will not clear in time. The ally holding the flank who is about to break. The skill that works but leaves the protagonist exposed for exactly the beat the enemy needs. Progression fantasy fights get their best tension from resource pressure and tactical desperation, from a character who is winning on paper and losing everywhere the paper does not look. If a reader can watch the health bar and know how the fight ends, you have not written a fight, you have written a status update with quotation marks around it, and that is the single failure mode this whole genre has to fight hardest against.
How do you pace and end a combat scene in a serial?
You pace a combat scene by giving it real beats and a genuine turn, and in a web serial you end it, or break it, on the worst possible moment. A fight needs the same architecture as any scene: an opening that sets the terms, a middle where the situation reverses at least once, and a resolution that costs something. What web fiction adds is the chapter as a physical unit, because your reader consumes a serial one installment at a time, and the shape of the fight has to respect that boundary. A fight that sprawls unbroken across five chapters loses its silhouette and starts to feel like the middle of a treadmill; a fight compressed into three paragraphs never earns the stakes it is asking the reader to care about. My advice, then, is to size the fight to the installment, contain it in a chapter or split it across two, and if you must break it, break on a reversal, not on a lull.
The ending is where LitRPG combat most often betrays itself, because it is so easy to resolve a fight with arithmetic, to let the bigger number simply overwrite the smaller one and call it a climax. It reads as hollow every time, because a number beating a number is not a story, it is a calculation. The victories that land cost something the character sheet cannot record: a plan that barely holds together, a skill used in a way it was never designed for, a sacrifice, an ally who steps in at the price of their own safety, a win that leaves the protagonist worse off in some way the stats will not show until three chapters later. Let the system confirm the outcome, never author it. If you want to feel the difference in your own draft, the fastest way is to write the fight in a real editor and read the chapter back end to end at reading speed, which is part of why I keep pointing writers at the no-account editor where you can draft a combat chapter in about thirty seconds and actually hear its rhythm. And if you want the wider frame this fits into, my complete guide to writing web fiction is the hub that this and the rest of the craft pieces hang off of.
The through-line under all of it is simple enough to keep on a sticky note. The system is the lens, not the drama. Point it at a fight the reader can follow, a cost the health bar cannot show, and a victory that had to be earned in a way arithmetic could not have predicted, and the numbers stop being the thing slowing your combat down and start being the thing that makes it hit harder than any prose without a system ever could.
Common questions about writing LitRPG combat scenes
How do you write a LitRPG fight scene without it reading like a spreadsheet?
Treat the numbers as consequences, not as narration. Instead of printing a full damage calculation mid-swing, show the effect in the body and the world first, then let a short system line confirm the stakes only when it raises tension. A fight reads like a spreadsheet when every exchange pauses to recite stats; it reads like a scene when the mechanics arrive in the same rhythm as the punches, one clear beat at a time, and each number tells the reader something has changed.
How much should you describe skills and stats during combat?
Less than you think, and only the parts that matter to this fight. Readers do not need the full stat block every time a skill fires; they need to understand what the skill costs, what it does, and why it might not be enough. Name the ability once, show its effect in the prose, and reserve explicit system messages for the moments where a number genuinely changes the outcome, like the mana running out or the cooldown that will not clear in time.
How do you keep tension in a fight when the reader can see the HP bar?
Make the danger about something the number cannot show. Health points tell the reader how close a character is to zero, but they do not tell you whether the mana will hold, whether the ally behind the line breaks, or what the protagonist has to sacrifice to win. The strongest LitRPG fights put the real stakes just outside the stat screen, so even a full health bar feels precarious because losing is not the only bad outcome on the table.
How long should a combat scene be in a web serial chapter?
Long enough to have real beats and short enough to end on a pull. Most web serial fights work best contained within a chapter or split across two, because a single installment is the unit your reader consumes in one sitting. A fight that sprawls over five chapters loses its shape; a fight compressed into three paragraphs never earns its stakes. Give the exchange a beginning, a turn, and a cost, and if you have to break it, break on the worst possible moment.
What is the most common mistake in LitRPG combat writing?
Letting the system do the emotional work the prose should be doing. When a fight resolves because a bigger number beat a smaller number, the reader feels nothing, because arithmetic is not drama. The fix is to make victory cost something the character sheet cannot itemize, a plan that barely holds, a skill used in a way it was never meant for, an ally who steps in. The numbers should confirm the tension, never replace it.
by Jacob Tam · July 28, 2026
I run IlorisNovel, a platform for web fiction writers. If this kind of craft writing is your thing, the rest of the blog lives here.