Intensity and pacing in The Last Of Us: Left Behind

By Antony de Fault. Posted

Naughty Dog’s 2013 hit The Last of Us, one of the most highly acclaimed video games of all time, has undoubtedly seen some of the most influential video game writing in recent years. It’s often regarded as complex, subtle, and ‘filmic’ (a problematic term). I’m here to tell you that writer Neil Druckmann’s structuring is completely run-of-the-mill, and that’s fine.

Rising action

In creative writing 101, one of the first lectures you undoubtedly receive is on ‘rising action’. The basic theory’s outlined in Figure 1. The story starts with some explanation of the scenario, then the level of ‘action’ (I prefer ‘intensity’) rises. This might come in the form of increased enemy presence (Tusken Raiders attack, then stormtroopers arrive in A New Hope), or just more heated drama (Romeo and Juliet fall in love at first sight). The stakes are getting higher.

Then comes a climax. Obi-Wan is killed by Darth Vader. After that point in the story, the level of action/intensity slowly falls, until in the final phase, we have a very low-energy conclusion to the story.

However, the astute reader will surely realise that what I’ve just written is total crap. Figure 1 has the highest moment of action slap-bang in the middle of the story. But Obi-Wan’s death is just a smaller climax on the way to the Death Star assault, just as Juliet’s parents forbidding her to see Romeo is just one step on the way to their accidental double suicide.

Figure 1: Known as Freytag’s Pyramid, this theory is foundational but insufficient. Disregard it as soon as you've scrolled past it.

The Fichtean curve

Bear with me, I’ll get to The Last of Us in due time. Now let’s look at Figure 2. Here, we see something much more useful. The Fichtean curve (ignore the fact it’s not actually a curve: we’re writers, not mathematicians) acknowledges that instead of a simple pyramid, people have a taste for stories which grow through a series of mini-climaxes or ‘crises’ until they reach the main crisis in the last quarter of the story.

Figure 2: If you find yourself needing to pace a game, cut this image out and stick it to a nearby wall.

What follows is often then a much quicker fall to calm, low-intensity scenes, not punctuated by crises. There’s often a little kicker in the last few lines when the story’s themes get hammered home, but, y’know, stories very rarely end halfway through a high-octane car chase. They end with a sunset, a conciliatory conversation, or a peaceful celebration.

I really can’t stress this enough. This is the rhythm of intensity people expect to see in all stories, and they appreciate it. It sticks out like a sore thumb when not followed, so unless you know what you’re doing and are deliberately deviating from it for effect (as in The Last of Us Part II), just don’t. All you’ll earn yourself is reviews talking awkwardly about ‘pacing issues’ but failing to really explain themselves.

The Last of Us: Left Behind is a bite-sized slice of excellence, perfect for tearing apart.

Left Behind

So, a demonstration is in order. The main two instalments of the franchise were too long to play just for this purpose, but Left Behind clocks in at only a few hours on a completionist run, so I stuck it on easy and waded in. Having recently played through all parts of the franchise for my own entertainment, I began my Left Behind case study with an idea of the game’s highest and lowest points already. I decided on a 0–10 scale, with 0 representing the lowest-action scenes in the game, and 10 representing the most intense.

I used a timer to pause the game after every five minutes of play. I then used a spreadsheet to record the time elapsed; a subjective judgement of the last five minutes’ narrative intensity; and a score for the gameplay intensity. As the game bounces back and forth between two time frames, one being inherently more violent and pressured than the other, I also thought it would be interesting to record which time frame I had been playing in: past, present, or both.

In Figure 3, you can see my results, and I have included an orange line which charts the average between the two types of intensity (you can find all the raw data here if you're interested). You can see that it maps very nicely to the Fichtean curve. There are a few observations to make here. First, the narrative and gameplay intensities often serve to balance each other. During an intense cutscene, the player is obviously not actually doing anything.

Figure 3: This research isn’t exactly scientifically watertight, but you know I’m right.

The first five minutes are a good example of this: we start with a slightly intense bit of narrative to hook us, but the player’s doing nothing, so overall it’s the least intense point in the game. Keeping the narrative quite intense in the final moments of the game serves to counteract the fact that the gameplay intensity shoots down to zero (it’s all cutscene), so the action drop-off is less harsh.

 During the portion of the game in the past (around minutes 40–55), there’s not much to do other than walk around, so at this point the game sustains its narrative peak rather than allowing the average intensity to tank downwards, preserving the curve. The same effect can be observed in reverse between minutes 70–80.


But for the climax, to ensure maximum general intensity, Naughty Dog quickly cut between intense, zombie-chased gameplay and fast, revelation-filled cutscenes, twice over. Both the narrative and gameplay must be at their peak at the same time, which is a difficult thing to do in non-text-based games, since you need to either switch modes rapidly as in Left Behind, or find a way to deliver your most climactic narrative moment during gameplay, of which the only example I can summon right now is Brothers: A Tale of Two Sons.

So let this be a lesson: if it’s good enough for one of the most critically acclaimed games of all time, and you’re not an experienced writer, then it’s good enough for you. If you’ve already got a story in mind for your game, plot its intensity on the Y-axis against your expected playtime on the X, and make sure it’s not wildly different from this. If it is, please consider nudging events around until it has a more Fichtean look. Until next time, stay safe!


From Wireframe store