People want to know how to write action scenes, and advice on this tends to range from “adopt a specific writing style” to “give up, people only read books for dialogue/inner monologues/flowery prose descriptions of sunny hills.” Now, no one is ever going to be satisfied with a book that’s trying to be the novelization of a hypothetical Jackie Chan film. You can’t carry a book on action-comedy with an excuse plot the way you can a movie, because not being a visual medium does have some problems.
But it’s still totally possible to write good action scenes, and the secret to that is not in how long your sentences are. It’s in reversals, limited resources, and compounding problems.
Reversals of fortune just means that whoever is currently winning starts losing. The person currently losing pulls out some special tactic, unveils a new weapon from their arsenal, receives reinforcements, whatever, and now suddenly they are winning. Then their opponent does the same thing. Constant reversals keep it up in the air who’s going to win. It’s worth noting that “winning” is relative here. A hero badly outmatched by a powerful villain might be trying to escape, in which case reversals in the hero’s favor can be about the hero creating distance and hiding rather than doing any kind of damage. You can also have reversals where the person currently losing tries something, it looks like it’s going to work, and then it doesn’t, leaving the situation back where we started. The villain unveils a new death ray, powers it up, points it at the hero, and then it blows up in their face. The reversal in the hero’s favor is just that the reversal in the villain’s favor turned out to be a dud, but it still works.
Limited resources means setting up how many resources your hero and villain have to burn through, and then using the remaining resources to track how close they are to defeat. These can be literal resources, like bullets in a gun or gas in a tank, and they can also be abstract resources, like layers of defense on a castle. You can have a battle that goes from the outer walls, to the streets of the city, and then into the keep at the center, and your audience will get that the defenders are getting closer to defeat as they run out of places to retreat to. Limited resources need to be significant enough that running out of them serves as a reversal, or better yet ends the action in favor of one or the other entirely, and they should not generally be something that you can just get more of. Although it’s possible to have a battle that bounces back and forth between the city streets and the city wall, you don’t usually want that to happen, and instead let defenses topple one by one. Measure the attacking army’s strength separately, with something like the number of active commanders. This way, attacker and defender are not swinging back and forth in a stalemate that could go on forever, but instead both are slowly depleting their resources and whoever hits zero first is the loser. You can also have both sides of a fight have separate pools of the same resource. The most straightforward one is bodily health. Unless you’re writing LitRPG, you don’t want to quantify this with literal hit points, but you can have both sides taking serious injuries that meaningfully impede their ability to fight, and your audience will get that eventually one side or another will be totally incapacitated.
Finally, compounding problems. You don’t want an action scene to revolve too heavily around a problem coming up and then immediately being solved, as this gives no time for tension to build. Instead, bring up a problem, have the protagonist(s) attempt to solve it, and while they’re trying, increase the pressure and the stakes. While one protagonist tries to hack door controls, the other has to fight off waves of bad guys, and the bad guys do things like call upon elite reinforcements, start coming up through the walls, and start putting snipers into place to shoot the hacker, with each problem compounding the last: The elites are still there when the vents pop open and start a second flank to the fight, and both of that is ongoing when a sniper starts lining up a shot. The pressure keeps rising until the door pops open and our heroes can make their escape. Instead of problems coming up and being solved one by one, creating the impression that our heroes could do this all day, compounding problems can create the impression that the next bit of straw could always be the one that breaks the camel’s back.