The Action Scene is a column exploring the construction of action set pieces, but it also considers “scene” in the sense of field or area: “action” as a genre and mode that spans different cultures and historical periods. By examining these two levels in tandem—one oriented toward aesthetic expression, the other toward broader contexts and concepts—this series aims to deepen appreciation for and spark discussion about action cinema.
With 2020 behind us, we’ve rounded up some of the year's best action scenes. Although the pandemic-related delays to several would-be high-profile action releases invited worry that the year's genre pickings would be slim, such concerns turned out to be misplaced. If anything, the relative absence of blockbusters left more room for less “mainstream” fare to enter the conversation, which is precisely what this piece aims to facilitate.
Structurally, this article approximates my best-of-the-decade roundup from 2019. The caveats and clarifications from that piece apply here as well:
- Not all the best action films contain the best action scenes, and not all the best action scenes are found in the best action films. My focus will be on scenes, which means that some solid actioners may fail to show up (e.g., The 2nd, Bad Boys For Life), and, conversely, some spotlighted films are, in my opinion, overall not very good (e.g., Birds of Prey).
- Because a particularly bountiful action picture may contain multiple set pieces that outstrip all competition, I have, for the sake of variety, limited myself to one action scene per film. The one exception is if a film had two (or more) set pieces of year’s-best caliber that were markedly different from each other.
- Given that I’m based in the U.S., I went with films’ (non-festival) U.S. premiere dates when deciding which titles to include.
And now, to the action. As with the 2019 piece, I’ve organized the scenes into loose thematic groupings, both to enhance ease of perusal and to highlight stylistic/thematic patterns.
Cut to the Chase
Who would’ve thought the year’s best action film would be a straight-to-YouTube, $15,000 short with a three-person crew? And yet, here we are with Ahmet Atalay’s The Cure, a near-perfect diamond of chaos cinema whose compressed runtime feels like the pressurizing force from which emerges a brilliant, hard-edged gem. The overall film is essentially one long action scene, but the highlighted segment (00:06:28 - 00:11:38 in the video) is one of the most thrilling: a borderline avant-garde sensory blitz in which Neveldine/Taylor-style editing (spastic cutting across different camera angles and media interfaces) is combined with dizzying camera movement and vertiginous, Go-Pro-style lensing that makes the onscreen space seem to bend around us, enveloping us in a world where the experience of space-time has been warped by adrenaline and speed. The seemingly contradictory impulses embodied by the scene—which is simultaneously disorienting and intensely oriented, chaotic and attune to both trajectory and geography—generate a singularly kinetic action experience.
The climactic road duel in Guillaume Pierret’s Lost Bullet (01:11:45 - 01:23:14 in the film) grounds The Cure’s physics-straining vehicular acrobatics, replacing it with the car as pure horsepower, burning down blacktop. Evoking the car chases of the 1970s, the bulk of the Lost Bullet scene is bulk: the rippling muscular mass of the car as signaled through the roar of the engines; wide, steady shots that underline the weight and realness of both vehicle and stunts; and the simplicity of straight-line driving that foregrounds the car as thundering physical fact. At the same time, because the cars feel so manifestly heavy and present, all “fancier” mid-road strategizing—e.g., one vehicle snagging another via installed bumper hooks—feels both literally and figuratively weightier, success made that much more satisfying because a beast so powerful had to first be tamed to do one’s bidding.
Boots on the Ground
The action centerpiece of Sam Hargrave’s Extraction (00:34:5 - 00:46:27 in the film) also begins with a (superb) car chase, but it continues into third-person-shooter-style gun-fu; a knife fight in the streets; and then another (albeit shorter) car chase. Crucially, all are linked by what appears to be a single, unbroken, nearly twelve (!) minute shot, which, though comprising sundry cuts artfully hidden within whip-pans, dim lighting, and flurries of onscreen movement, achieves the illusion (and, if not always that, then still the momentum) of visual continuity. As the camera races and jerks with the movements and blows of the characters, the scene becomes triply physical: we feel not only the onscreen fighting (by proxy via the characters) and our own movement “around” the action (through camera movement) but also the athleticism of the camera operator himself, who becomes an increasingly embodied presence with each passing minute. Furthermore, the cinematography is precise in a way that speaks to director Hargrave’s own experience as a stuntman and fight coordinator. The way the action is captured—the push-pull between tighter, immersive framing and slightly scaled-back two-shots that capture a more complete blow; the use of handheld that nonetheless maintains a clear view of the action—suggests a veteran’s intuitive eye and guiding hand.
If the single-take effect in Extraction seems to anchor us in the ground-level experience of pursuers and pursued, the climax of Rod Lurie’s The Outpost (01:07:09 - 01:44:57 in the film) pushes this perspective to the extreme. Based on the book by Jake Tapper that details the 2009 armed assault on an American military camp in Afghanistan, the film’s climax dramatizes the firefight in which U.S. troops attempted to hold their ground against nearly 400 Taliban fighters. Deploying lengthy tracking shots for most of the sequence, Lurie hews so tightly to the embattled soldiers that, most of the time, we can’t even see the enemy. Often, U.S. soldiers appear as if they’re not shooting at anything, and incoming fire seems to rain down out of nowhere. In many ways, The Outpost feels less like an action film—in which combat typically involves two or more, visually and narratively delineated parties—and more like a disaster picture, the smallness and frailty of human efforts dwarfed by cosmic forces whose origins we cannot see. This abstracting of the enemy into an almost mystical, otherworldly ambiance certainly risks exacerbating the Hollywood-reinforced ideology of the faceless Other, but it also folds into the film’s larger vision of powerlessness, of mere men in thrall to the laws of a foreign land, the decisions of their superiors, the institution of war, and the terrifying uncertainty of existence itself.
The other combat picture on this list is Aaron Schneider’s Greyhound, which depicts a standoff between a fleet of Allied ships and a wolfpack of German U-boats. In the featured scene (00:50:20 - 00:56:03 in the film), one of the enemy submarines is forced above water after myriad scenes of shark-like prowling, resulting in the sudden reestablishment of clear spatial relations. In addition to the catharsis of restored tactical geography—which the film highlights through exhilarating overhead shots that map out the relative positions of all the ships involved—the scene (and the film in general) is striking for its talkiness. Featuring instructions and enemy sightings being verbally relayed through the ship’s communications network before any strategic maneuver is made, Greyhound ingeniously introduces a temporal gap between perception and action. Whereas a typical fistfight or car chase hinges on split-second action and reaction, here, the fight-or-flight response is delayed, and even more so due to the fact that a multi-kiloton ship can’t move as nimbly as a human fighter or even a car. In the life-or-death situation of live combat, the palpable clunky hugeness of the ship leads to bracing suspense; conversely, the speed and size handicap means that every successfully executed move feels like a resounding triumph.
If the power of Greyhound lies in the way it exploits the upper limit of human kinaesthetic experience—the way the “body” of the ship feels like both a tactical, tactile extension of the individual soldier and an instrument out-of-step with the (temporal as well as material) scale of his experience—then Shigeaki Kubo’s HiGH&Low: The Worst scales up action in another sense. A spinoff from the film series HiGH&Low and a part of the larger Japanese transmedia franchise of the same name, The Worst reprises the series’ most iconic feature: improbably expansive brawls featuring scores upon scores of extras swarming across massive sets whose size is periodically underscored by soaring tracking shots that sweep over and through the melee, convincing us that all this prep work and coordination actually did take place.
The film’s climax (01:36:13 - 01:48:46 in the film) contains the most ambitious fight of the series so far, which involves both horizontal and vertical planes of action; committed stunt work; and a balance between handheld camerawork and clear center-framing. When the tracking shot centerpiece drops, it’s jaw-dropping. Beginning atop an apartment walkway, the camera shoots laterally out into the open air; races alongside the building exterior; drops down with a plummeting body; and then careens through the fray on the ground. No matter that the “hidden” cuts are less than subtle: the momentum and lightness of the “continuous” shot feel genuinely epic, allowing us to be both viscerally involved with the earthbound conflict and whisked above it as if by the wind.
If one strain of action cinema dwarfs the human figure beside structures and events that surpass what is sometimes called “peripersonal space,” or the tactilely manipulable elements immediately surrounding the human body, some set pieces work in the other direction, compressing the field of potential action to the point of claustrophobia. These instances of close-quarters combat, typically occurring in tight spaces that restrict fighters’ range of motion, are innately visceral because the physical constraints of the set encroach upon characters’ peripersonal space, forcibly limiting their options for strategic maneuvering. Watching fighters navigate this spatial constriction, we kinaesthetically feel constricted as well, and not only through our embodied identification with the characters. A tight profilmic space typically also constricts the cinematography, forcing tighter framings and closer views so that we almost feel like we’re as squeezed up against the characters as they are with each other.
One effective close-quarters fight occurs near the start of Dave Wilson’s Bloodshot (00:07:23 - 00:08:16 in the film), in which the iconically big and burly Vin Diesel fights two assailants within a cramped bathroom. Even before the action starts, the film confines shots to close-ups and medium close-ups, implicitly establishing the physical space as being very small. This tight framing continues into the fight itself, which occurs mainly through snatches of action viewed through a flurry of quick shots. Despite the limited visibility that occurs as a function of limited space, however, Bloodshot is strikingly precise with what is captured. We don’t see much, but what we do see (e.g., face, face, needle in wall, stabbing motion) is enough to track what happens. Throw in a general adherence to continuity editing principles—match-on-action editing, maintenance of the 180-degree rule—and the scene becomes perfectly legible even as the fragmented editing lends the action a percussive force.
Another close-quarters brawl appears toward the end of Diao Yi’nan’s The Wild Goose Lake (01:26:26 - 01:27:30 in the film), in which our hero, boxed into a small apartment with a trio of thugs, fights his way out in a series of blunt, brutal blows. The brawl opens with a series of jarring close-ups: hand on wrist, foot on ground, shoulder into chest, then a match cut to a tighter view of the body slam’s continued trajectory. The editing—which is in some ways reminiscent of Sergei Eisenstein’s famous montage style—eschews the impression of smooth, continuous space in favor of sharply splintered views, enhancing the felt weight and staccato rhythm of blows through the jolt of broken visual continuity (in this sense, the scene overlaps stylistically with the Bloodshot fight). When the subsequent shot opens up to a wider framing, it feels like the violent release of compressed air, corresponding to the completion of the physical assault: the victim crashes back onto the floor, knocked out cold. The rest of the scene, despite visually pulling back, nonetheless continues to punctuate the action with close-up views—an umbrella in the stomach is a particularly striking example—delivering the sensory shock of sudden, tactile proximity.
Various set pieces from this year are visually structured around medium and medium close-ups, with the camera positioned largely between waist- and eye-level vis-à-vis the characters. It’s a very common scale of framing that reflects Hollywood’s long-standing proclivity for analogous framing during conversation scenes. The concept of “peripersonal space” is once again germane: in-person social interaction tends to involve gestures above the waist—the face, head, and hands, mostly—and, hence, scenes that reproduce this scale often invite some form of visual-kinaesthetic recognition from the viewer, subtly easing them into the onscreen drama by appealing to peripersonal familiarity. Action scenes thus framed often make a similar, peripersonal address. Fights tend to be an extension of social interaction and, in the same way, typically rely on social cues and maneuvers from the waist up: reading the opponent’s face for signs of aggression, for instance, or throwing or deflecting blows with the arms and hands (which, though obviously not the only combat-capable parts of the body, are usually the most dexterous and tend to be used the most). As such, “peripersonal framing” in action scenes makes sense because it compels viewers to imagine themselves as being physically at the scene of the fight or, especially when over-the-shoulder shots are used, even in the position of the character in the foreground.
This type of framing appears in every action scene on this list, but I wanted to highlight two that are structured around this stylistic approach. The first is from the climax of Jesse V. Johnson’s The Mercenary (01:18:55 - 01:21:13 in the film), in which a former hired gun seeks vengeance on his erstwhile associates. The second is a scene from Gina Prince-Bythewood’s The Old Guard (00:12:06 - 00:14:49 in the film) in which the immortal warriors of the title resurrect from the dead and lay waste to their would-be killers. In both scenes, shots tend to cleave tightly to heads, trunks, and hands; in both, the wobble of handheld camerawork enhances the feeling of physically being “there” among the fighters. The framing is such that many blows occur off-frame or are obscured by other bodies. Visual blocking literally “blocks” a lot, often hiding the point of contact behind a character’s body in the foreground. The resulting hits look convincing enough, but largely because so much of the movement is obscured, preventing us from telling either way.
Some might consider this visual framing to be lazy and obscurantist. The thing is, “action” isn’t just about seeing people fighting but also feeling ourselves to be involved in the fight. To this end, both scenes are exemplary. They seem to place us physically in the shoes of fighters, in the blurred rush of movement and the tangle of limbs. Like with the Bloodshot scene, continuity editing helps us track geography and trajectory even within this destabilized visual field. Occasionally, these continuity techniques are used to generate thrilling lines of force that cut through the fray, like when a shot of Charlize Theron’s character ramming an opponent match-cuts to the flipping shotgun of a fellow immortal sprinting in the same screen direction, the synergy of the team made palpable through editing.
Despite the immersive powers of peripersonal framing, there’s a reason that seeing a stunt in its entirety remains so compelling. No strategy of cinematic immersiveness can overcome our felt knowledge that we're not actually Dominique Vandenberg from The Mercenary or Luca Marinelli firing a gun in The Old Guard. Direct, “first-person” identification can only go so far. When films seek to systematically emulate a through-a-character’s-literal-eyes perspective (e.g., Hardcore Henry ), the spell of embodied identification dissipates, since the onscreen view too obviously belongs to a physical position we could never inhabit. Indeed, the effectiveness of the previous section’s action depends on the evocation of “first-person” perspective being modulated by the “third-person” nature of the images. Even as the camera places us behind Marinelli almost as if we’re the ones pulling the trigger, the “almost” remains operative.
The simultaneity of first- and third-person modes of address lies at the crux of what makes much action cinema so engaging, blurring the line between literal embodied identification and distanced observance of an idealized action star who physically does what we cannot. Furthermore, the latter feeds back into the former to an extent. When we see another person’s body, gait, and comportment, we tend to subconsciously map their visible physicality onto our own embodied experiences and vice versa; we often try to imagine what it might feel like to be them and what it might look like to be us. Seeing an action star execute a stunt triggers in us the ability to almost feel and imagine ourselves to be doing the same. We project ourselves onto them and vicariously experience a feeling of physical empowerment.
A scene that straddles the “peripersonal framing” of the previous section and a more stunts-centric visual style can be found in Debt Collectors, the sixth collaboration between director Jesse V. Johnson and actor/martial artist Scott Adkins. Although this scene (01:06:06 - 01:15:28 in the film) has some stylistic overlap with The Old Guard, shots run a beat or two longer than is customary for mainstream Hollywood; tend toward wider two-shots; cleanly deploy continuity editing so that a sense of movement and geography across cuts is maintained; and emphatically center on key actions such that, even when shots are brief, we still see the gnarliest maneuvers completed in their entirety. Many shots are over-the-shoulder, but a few moments nail the trickier profile shot, in which actors lack the luxury of the other person’s torso to mask a feigned blow. In such cases, fighters have to hit much closer to the body, risking injury but also making hits look more convincing. All these choices enhance the scene’s overall impression of strain and fatigue, of a fight that continues for longer than it has to and was comically counterproductive to begin with (in this sense, the scene evokes the famous brawl between Roddy Piper and Keith David in They Live ).
Another scene that splits the difference between peripersonal framing and emphatic stuntwork can be found in Christopher Nolan’s Tenet (00:32:09 - 00:33:20 in the film). As our hero (John David Washington) takes on thugs in a kitchen, much of the over-the-shoulder visual blocking is angled such that the body of the foreground fighter almost fully occludes that of the background one. The result: it is manifestly clear that combatants’ blows likely aren’t actually connecting. The way Tenet generates a compensatory sense of physicality is, firstly, through tactile sound design; when Washington’s Protagonist takes a cheese grater to a man’s face, we can hear the bladed holes snagging skin. Secondly, the film intersperses moments of undeniably “real” stunt work within the scene as a way to forcefully restore a sense of weight and solidity. Bilge Ebiri, writing on The Raid 2 (2014), equates that film’s “neck-snapping” and “bone-crunching” with the arabesque that might punctuate a series of ballet moves, the “impact” without which “the movement is nothing.” In the case of Tenet, the “impact” comes in the form of shattering glass, breaking plates, and crumpling shelves—the visceral “something” that staves off the threat of “nothing.”
A director who more systematically structures his films around stunts is Chad Stahelski. Despite not directing any 2020 releases, he and his stunt team completed second-unit work on Cathy Yan’s Birds of Prey, and the effort shows. Ubiquitous in the featured scene (00:49:43 - 00:53:41 in the film) is one of the key stylistic fortes that have made Stahelski’s John Wick films (2014-2019) so iconic: wide, steady framing and long takes that capture a whole combination of moves in their entirety, such as when Harley Quinn (Margot Robbie, doing her own stunts) slides between a thug’s legs, topples him using her momentum, rolls across his back, and then cracks him over the head with a bat, all before a cut happens. A veteran stuntman himself, Stahelski has brought his practice to bear on his action direction, crafting a crystalline style whose sense of heaviness, exertion, and force turns on the choreography and athleticism of onscreen fighters.
Lost Bullet’s second appearance on this list is for a scene (00:28:49 - 00:32:59 in the film) in which protagonist Lino (Albert Lenoir) violently attempts to escape a police station. The lead-up to fisticuffs sets the tone of the scene: amusement at Lino’s attempts to free himself shifts to concern and then panic as what seemed to be just another day on the job turns into full-tilt chaos. As the fight travels from room to room, there’s a continual sense of disrupted routine; as a cop is kicked into a printer, we hear a phone ringing elsewhere in the station, suggesting that the pandemonium hadn’t yet upended the humdrum office work happening next door. The sense of the brawl’s emergence from the mundane is central to the scene’s effectiveness. Within the context of the everyday, violence carries a special kind of physical and psychic shock. It’s like being forced to sprint sans warm-up: with muscles still cold from being unprimed for action, the sudden onset of danger feels especially jolting. Within this banalizing narrative and aesthetic frame (which is aided by flat, muffled sound design that feels more “realistic”), the foregrounding of stunts in this scene (e.g., flipping over tables, crashing onto the floor) has the effect of helping us feel the jolt of real and relatable bodies in real and relatable pain.
Adrian Teh’s Wira was one of the year’s breakout action hits, and its best fight involves a brother-sister duo (Hairul Azreen and Fify Azmi) facing off against a final boss played by Yayan Ruhian, the sinewy phenom of screen combat from The Raid (2011) and John Wick: Chapter Three (2019). The scene (01:28:22 - 01:34:57 in the film) is a hodgepodge of stylistic techniques: shaky cam and tighter framing predominate, but flashes of the baroque occasionally appear—a camera that spins with a body being slammed onto a table, for instance, or a switch to slo-mo when two combatants punch each other’s faces simultaneously. Sometimes, there are bursts of intense clarity, like when a single shot captures an entire combination of moves; other times, a clear sense of geography and trajectory will be momentarily scrambled, such as when the film toes or crosses the proverbial 180-degree line. Mileage will vary on the effectiveness of this erratic, eclectic style. I felt some of the stylistic fumbling-around to be a smidge distracting and sloppy, working against the flow of the choreography, but the choreography is exceptional, and the energy of individual stylistic beats—handheld camerawork that whips with a particular hit, for example—ensures that the scene consistently astonishes even despite overall unevenness.
If Wira’s impressive choreography seems to exist in tension with its style (at least vis-à-vis certain norms of martial arts filmmaking), the two dovetail more smoothly in Wong Jing and Kenji Tanigaki’s Enter the Fat Dragon, a loose remake of Sammo Hung’s 1978 film of the same name. Every one of the film's main set pieces is aces, deploying tried-and-true techniques of Hong Kong action filmmaking—a foregrounding of spectacular and painful-looking stunts; intricate, whip-fast choreography; and the precise, steady capture of individual moves such that, even when the pace of cutting quickens, the larger action remains legible—all while updating them for contemporary audiences with a glossy visual finish and elaborate camera maneuvers. The spotlighted scene (01:14:35 - 01:16:35 in the film) observes as a fat-suit-clad Donnie Yen takes on a legion of yakuza thugs in a fight that begins on the ground but quickly progresses to the rooftops. As Yen (or his stunt double) leaps from building to building, dodging projectiles and trading blows with assailants, the placement and movement of the camera are thrillingly varied but always seem to complement fighters’ maneuvers. Whether the camera’s tilting down to track the entirety of a henchman’s fall; going wide to observe our hero whacking four guys with a metal pole; or going wider still, sweeping over the expansive set to convey a sense of scale, the scene effortlessly keeps us oriented to characters’ positions and trajectories.
Our final entry is the only animated film on this list, and it engages one of the chief challenges of animated action: how does one convey physical weight, solidity, and force within a medium that is so visibly made “by hand”? The featured scene from Ethan Spaulding’s Mortal Kombat Legends: Scorpion’s Revenge (00:05:07 - 00:06:59 in the film), which depicts the eponymous Scorpion slaughtering the assassins who massacred his wife and clan, executes the task with gusto. Motion blur was added to convey speed, and a lower frame rate was used to generate visual judder, making movements feel more halting and effortful. Elsewhere, more “cinematic” methods are deployed, such as simulated camera movements that implicitly make it seem like a cinematographer had been physically on-site, capturing the action. This physicalizing of the virtual camera also subtly physicalizes the space “in front of” it, an impression strengthened by moments in which the depicted space is rendered sharply three-dimensional. Last but not least, the orgy of anatomically detailed bloodletting functions much like the flashes of palpable stunt work in Tenet or the limb-snappings from The Raid 2: Scorpion’s gruesome finishers are the “arabesque” that puncture a string of movement. Moreover, the sheer extremity of the gore functions to literally “flesh out” the bodies of the characters, overcoming animation’s inherent two-dimensionality through a compensatory excess of tactile bodily spectacle.
The body is, ultimately, the grounding element of the films on this list and action cinema in general. An effective action scene is intensely physical, and it is a physicality complexly mediated by vision: action seen on a screen, bodies viewed from a level of physical remove and yet, through different filmmaking choices, to which our own bodies are brought viscerally near. The dynamics of action cinema are rich beyond measure, and the action scene is the most concentrated, visible, and spectacular embodiment of the genre’s proclivities and potentialities. It is toward a deeper and vaster appreciation of action cinema that this column strives, and the action scenes from this year have beautifully helped us move in that direction.