Writing a Game Script Is Not a Project Step

Ian Thomas
5 min readApr 19, 2024


Let’s talk about writing for — well, anything really, but in this case for games.

Sometimes producers, schedulers, or game designers optimistically rough the process out as a part of what seems to be a linear pipeline. For example, “Step 1: create core mechanics, Step 2: white box level flow, Step 3: placeholder text, Step 4: write script, Step 5: record voice”. It never works like that. Or if it does, it ends up with underdeveloped script and lots of cutting and rerecording.

Writing the script is not a ‘step’ in game development process. It’s a whole lot of steps. Or loops. Or parallel streams. Or a chaotic marble-run of content drip-fed through an improbability drive while charred slightly by a blowtorch.

Think of how you would approach putting a level together for gameplay. You would implement mechanics and flow, test it by playing it, refine it, throw some things away and add others, play it again, test it again, iterate until it’s fun, test it again (perhaps with a wider group), take in that feedback, rework, retest.

That’s how writing works too. It’s relatively easy to slap together a first draft, but we won’t know whether even a single line is working without a lot of testing and iteration, dealing with all sorts of things:

  • Does the line fit, time wise, in the space that the player is progressing through? Is it a super-short distance and a super-long line? Some teams test lines with text-to-speech tools, and that will never give you a proper answer to this question — best to use scratch recorded audio.
  • Does the line make sense of the level mechanically, now you hear it in place? Does it give the player useful information or direct them to the right goal? Is it easy to understand? Does it have to change because the level-design has changed or the mechanics have changed?
  • If the line is delivering meaningful information (and it should) then will the player even notice it in the game? Does the line clash with other things going on at the time? Are there multiple attention focuses? Do we need to move the trigger for it to a difference point? Reword it to make the point obvious? Change it to a different character? Cut it completely?
  • Is the line pleasing to hear? Is it hard to say or to parse? Does it match with the action that’s happening and the mood the player is in?
  • Does the line fit with the character who’s saying it, now we can experience a test version along with their art and animation?
  • Have we primed the player with the other info they’ll need to understand the line? “Now frangify the flobbler!” doesn’t make sense if you didn’t know — or didn’t remember — you were looking for a flobbler, or how one frangifies it, or even why you might want to. Did we brief them about flobblers already but it was about thirty minutes of game play ago and now they’ve forgotten? Was it only five minutes ago in gameplay but they put down the controller for a week while they went on holiday to Barbados? How do we cope with that?
  • Does the line make emotional sense in the narrative flow? Now it’s experienced in its proper place after the preceding scene, is it a sudden jarring change of tone, and should it be rewritten?
  • Does the line help progress the story? Does the line make sense in the longer story? Is it correct in continuity?
  • Do we need line variants depending on earlier choices the player has made? Could the current situation be reached in more than one way, and if so does that change what needs to be said?
  • A level has been cut or rearranged for mechanical, progression, or budget reasons — now level A is followed by level C instead. Does it all still make sense?
  • The script was written before the gameplay in a level was finished. We thought it was solid, but now it turns out the gameplay wasn’t fun that way and we iterated so the script no longer makes sense. Please throw away all those beautifully crafted lines and start again.
  • When you add up all the lines in the game including this one, does the story that they tell result in something emotionally satisfying, with the right peaks, troughs, setups, and payoffs, now we’re actually able to play it through?
  • How about now, after budget issues have rearranged the whole game?
  • This all sounds great to you and the team, people who are intimately familiar with the gameplay and story and who have tested it 1000 times. How about a first-time player – do they understand it?
  • We’re fed up of this line, could we cut the line entirely?

You get the idea. And that’s just one line. Let alone whole scripts or scenes.

Hopefully you can see that writing a script as a step to hit a deadline and then immediately recording it is potentially a very costly and demoralising mistake. Like any other game assets, scripts need to be roughed out, developed, worked on, tested, iterated on, and eventually prised out of the hands of the writers and narrative designers just before the latest possible point for recording to happen. Typically when everyone thinks the script is probably good enough to kind of work, hopefully. Maybe.

And even then there will be testing with “final” recorded audio, feedback from all levels of testers, publishers, IP holders, Dave from the canteen, your mum, and pick-up recording sessions when its realised that one of the criteria above (or multiple others) got missed for critical lines.

I prefer to think of creating a game script like moulding something out of clay — get the big lumps of it in early, reorganise them as the game develops, flesh it out more, then refine, test, rework, refine, test and so on.

Do not expect game writers to deliver high art in the first draft. If they try to, they are spending their effort in the wrong place.

I’d suggest you treat game writing as craft rather than art. If you’re making a table, make something you can put other things on first (and do that fast) and make it beautiful later. Because you never know when you might need extra legs, or space for another twenty guests, or something that’s easy to get through the front door.

Remember that if you agonise over the polish for your first draft, you will be devastated when half of it gets thrown away. And it will get thrown away.



Ian Thomas

Ian is narrative director, coder, and writer of video games, films, larp, books, and VR/AR experiences. He has worked on well over 100 titles.