top of page
Search

Throw Away the Script: Improvise or Become Irrelevant

Writer: Matthew Dyet
Matthew Dyet
Nov 11
5 min read

Originally posted in SAE Imaginings Volume 2.


When the comedian steps out onto the stage, how much have they prepared for the moment? Did they script every punchline, planning the act minute-by-minute? Were all the jokes written ahead of time, memorised to be delivered with immaculate comedic timing? Did the comedian research the audience demographics based on ticket sales and write a 10 page action plan outlining the path to the highest rate of return on laughter to each joke they delivered?


Does any amount of planning guarantee that they’d actually be funny?

Concord was Firewalk’s first ever game, and they did everything possible to play it safe while developing it. Their publisher Sony Interactive Entertainment was keen to bring more live service titles to their library, and directed Firewalk to make one. Live service games were (for a time) considered safe investments in the games industry. They encourage players to engage daily to unlock rewards, ensuring ongoing income for the studio and publisher through the use of systems like microtransactions and battle passes. Firewalk looked to other live service games for influence, particularly those with a positive reception and solid ongoing player numbers.


Development of Concord started in 2018 with the design of the game heavily inspired by Destiny and Overwatch, both of which had large ongoing player numbers at the time. Artistically they went for a retro sci-fi aesthetic, and their characters appear to have taken heavy inspiration from the designs seen in the Marvel movies.


But after 6 years of development the gaming landscape had shifted. Both Destiny and Overwatch had received sequels, and those sequels started to see declining or drastically lower player numbers. Marvel movies had seen continuously declining audience numbers. Sony Interactive Entertainment had cancelled most of their in-development live service titles.


Concord was released in August of 2024 and endured only two weeks on store shelves before it was unceremoniously pulled from sale entirely. Not due to game breaking bugs or issues, but because of its exceptionally poor critical reception and even poorer sales (MacDonald, 2024). It was the only game Firewalk ever made; Sony shut down the studio shortly after.


An over-prepared comedian is a bad comedian. They ignore the audience and rely on a big book of jokes, failing to truly engage or respond to the reality experienced from the stage. They live inside their heads enacting a plan, relying entirely on being proactive rather than reactive. Market analysis and audience demographics won’t teach you how to be funny or perform a good stand-up routine. Sony instructed Firewalk to read from a joke book, and punished them for it when the audience didn’t respond to a comedy routine half a decade old.


But when you get given a bad script, you could just throw it away.


The first Borderlands game was released in 2009, a fairly experimental multiplayer first person shooter that used roleplaying and loot mechanics from games like Diablo. It was developed by Gearbox Software, an at the time relatively small studio that had seen some success with their Brothers in Arms series, a game set around World War II. Borderlands was risky due to its experimental nature, there was little evidence that it would work. So to try and draw in more players Gearbox played it safe with their art style, opting for the gritty realism of popular franchises like Call of Duty. There was an impression in the studio that they needed to follow that trend in order to capture a market for their game.


In the final stretch of development playtesters made it clear that they hated the art style.

The team at Gearbox was faced with a difficult decision: release with the art style they have, or take on the herculean task of changing the art style of the whole game at the 11th hour. The art team went away and performed experiments based not on popular styles in the market, but on what felt right. What they came up with was wholly unique and incredibly risky for a first person shooter, a bold style inspired by comic book art with thick lines, bright colours, and strong silhouettes. They improvised, trusted the artists on the team, and followed their heart rather than the trend (Shea, 2025).


Borderlands released to critical acclaim and commercial success, establishing a franchise that has since received numerous sequels, spinoffs, and a film. The Borderlands movie would fail in all the ways the Borderlands game succeeded, panned by audiences for its inability to take any risks at all, wasting a safe star-studded cast on a script unwilling to risk meaningfully engaging with the source material.


The fear of failure and uncertainty is insidious. It kills off the joy of improvisation so slowly that none of us notice it happening. Nowhere is this desire for control more visible than in our current obsession with generative AI.


The push for the inclusion of AI tools in every facet of life is not because these tools improve our lives in a substantial way, but because they offer certainty in terms of the reduced costs associated with people. Bringing out the robot comedian is not a response to a demand from the audience for more robot performers: it’s a response to the fear of the show having a poor financial performance. Even if the reviews are poor and the attendance low, it will still have saved money over the cost of hiring a human stand-up.


We become reliant on plans because they offer certainty, because we trust it more than we trust ourselves to make the right decision in the moment. But it’s corporate astrology, it’s a tarot card reading by a well-informed manager. It cannot actually tell us how we should act in the future, it exists primarily to give us a sense of direction and purpose.


It would be unreasonable to say that planning is the problem. I have worked as a Producer in the games industry for over a decade, in a job all about making accurate predictions and plans for the future. But while I may have always ended up close to objective reality, I was never perfectly accurate. Circumstances change in ways that none of us can ever predict. Players' tastes change, team members come and go, technology moves rapidly. The plan offers comfort in the sense of control it provides, but it’s an illusion. It cannot predict that a genre popular in 2018 will feel tedious in 2024. That the plan cannot predict the future isn’t the problem, the problem is when we choose to follow the plan at any cost, even when our heart tells us it’s not the right call. This is the pivotal moment a project succeeds or fails, when the team makes a choice on how to react in the moment the plan fails them.


A plan is just a map, only a human can choose how to navigate it. When the map turns out to be inaccurate, we don’t question our objective reality. The cliff is right there, we see it even if the map didn’t predict its existence.


We don’t plan to improvise because that would be an admission that we do not have all the answers. But aren’t we lying to ourselves if we pretend otherwise? Improvisation is scary. It’s also an absolute joy. The best art feels inevitable in retrospect, but was often chaotic in its creation. Stop trying to write the perfect set before you get out on the stage. Read the room, go off script and explore, pivot when a joke bombs, and master your improv.


The script isn’t going to save you. Learn to trust that you can perform without it.


References

MacDonald, K. (2024, September 11). Sony’s big-budget hero shooter Concord failed spectacularly – here’s where it went wrong. The Guardian. https://www.theguardian.com/games/article/2024/sep/11/sonys-hero-shooter-concord-failed-spectacularly-heres-where-it-went-wrong


Shea, B. (2025, September 1). The behind-the-scenes story of the art-style swap that saved Borderlands 1. Game Informer. https://gameinformer.com/2025/09/01/the-behind-the-scenes-story-of-the-art-style-swap-that-saved-borderlands-1

 
 
 

Recent Posts

See All
The Joy of Figuring Shit Out

Modern games have been having a crisis of trust with their players. Yellow paint (for its excessive hate) is a great example of how this often comes through in the way we make our games. Level designe

 
 
 

Comments


WebsiteBookends_HeaderT.png
  • BlueskyPaddedWhite
  • LinkedIn
  • Youtube
  • ItchIoPaddedWhite
  • Instagram

© 2025 Matthew Dyet. All rights reserved.
The content, design, code, and graphics on GreatGames.dev are the intellectual property of Matthew Dyet and are protected under international copyright laws. No part of this website may be reproduced, distributed, or transmitted in any form without prior written permission, except as permitted by fair use provisions.

bottom of page