Crash Bandicoot, 40 Million unit sold: Thinking out side the box literally.

The Crash Bandicoot series was a big commercial success as a video game for the Playstation console. The series altogether sold over 40 million units worldwide, It remains one of the most popular video game series of all time.
There is a phenomenal love and admiration for the game, which is reflected in the sales figures.
This success is the result of brilliant engineering and smart business decisions.

In this article, we will discuss why did the developers chose the Playstation as the release platform and how did they manage to get next-gen graphics quality on their limited hardware.

Why was crash bandicoot graphics so much better than the other games released on the same hardware at the time?

Going with Playstation not Sega or Nintendo

In 1994, Naughtydog inc., the famous game development studio was just two roommates back then, Andy Gavin and Jason Rubin.
They worked on a title called way of the warrior and sold the publishing rights to Universal Studios and things were going pretty well, then they wanted to make a 3D platform game.

3D platform games were all the rage back then, Titles like sonic the hedgehog, Mario were very popular back then.
So they wanted to make a 3D platform game that had cartoony graphics and a unique look and feel.

Game development, in general, requires a huge investment upfront, it’s a product that needs expensive hardware, programmers, designers, testers… all before actually selling the game and you don’t really know the market reaction to the game, especially if this is a new type of game like 3D platformers back in 1995–6.

One of the big decisions in game development is the platform on which you will develop and distribute the game, so then came the question which game console will they use.

At the time, there was The Sega Saturn, Nintendo was preparing to launch a new console and Lastly, Sony was preparing to launch their all new console the Playstation.

At the time, Sega had Sonic the hedgehog as a mascot and was a big selling game for the console, Similarly Nintendo had Mario, but playstation didn’t have that big game that basically represented the console.
Adding to that, Playstation was the first console Sony releases which meant there wasn’t a lot of competition at the initial period of the console release.

business decisions like this are very important, because you can have a great product in a very competitive market and get crushed by competition.
Releasing Crash Bandicoot in a small game market like the playstation meant having some big advantages.
For instance, There is a large room for growth and a chance to acquire a large chunk of the market size, try competing in a big market with a market giant and you will find it nearly impossible, you are outnumbered, have limited and lower resources and lack the experience the market competitors already has.

When the game shipped and the sales figures went crazy, a big immediate success, Sony bought the licensing rights to the Crash bandicoot series, and basically making it the de-facto official game of the console, even shipping the game with console in the same box.
So this decision of choosing the platform ended up to be a very important one.

Crash audience

One point we will discuss briefly is the nature of the game.
Crash bandicoot is cartoony,silly and looks colorful. This made it very appealing to a wide audience, It was not a 1st person shooter game, or a racing game, it was not trying to be super realistic. The developers chose carefully their target audience, It was a game that all people could have fun with.
As with any startup, identifying the target audience and the correct market is huge, it could make or break the company, if you are targeting the wrong audience,you could totally fail.

Going sketchy graphics — Stiff animations

They wanted to do sketchy fluid animations that needed a lot of math and calculations, and honestly they did a remarkable job, the game animation was very fluid compared to other games released on the same console.

So How did they do this?

At the time, designers used “Bones“ structures to make the character animations.
Essentially, using the position of the bones, the hardware will calculate the next position of the bones on each frame.
That sort of worked with other games but they were kind of stiff and the fluidity was limited by how powerful the hardware was because each bone position was calculated in real time.
Also, they wanted to add facial animations, which was super hard with the bones structure.

So they knew they needed a different way to make their animations, they had this idea that instead of calculating the position of each bone and each vertex on each frame, why don’t they calculate each vertex and each bone point in the character and store them in the game, so the game can just draw them as polygons, without the needed calculations or math.
This was called “Baking” the animation, pre-calculating the vertices positions and storing them.

BUT, the problem with this strategy was that you got to store the position of each vertex and each point of every single object on the screen.
That’s a lot of data, So the problem now is how to fit this huge amount of data in the game, which at the time was much larger than what the storage units allowed on the playstation.

The lead programmer had a brilliant idea, He saw that the animation data can tolerate some data loss, in mathematical sense, the animation data did not have big dynamic range.
Just like the JPEG compression, In JPEG, you move the data to the frequency domain, remove all of the low frequency junk, and end up with a picture that has the most important parts of the original image but 1/20th the siz.

He was convinced that the animation data could tolerate some compression like this, So he calculated the dynamic range of the animation data, and developed an algorithm to compress the animation data to 1:80 compression, 1/80th the size.
No one else had the guts to do this, as while this was easy to render, it required inventing some totally hardcore assembly language vertex compressors. They made 5 different vertex compressors each with different efficiency until they were satisfied with the results.

It is easy to see how much better the graphics were in Crash game in comparison to Tomb raider for example.
This algorithm was phenomenal to the point that it was patented and used in other games.

It was that opening sequence, when Crash pulls his flat face out of the sand, shakes it off, looks confused, leaps up, looks at the camera and does his great big goofy smile that SOLD Crash as a character. No 2d game could afford the art, and no other 3d game had the facial animation that our vertex system brought. And thus the main character transformed from emotionless “vehicles” to an emotive friend.

The second thing that vertex animation allowed is total warping of the character beyond bones. If they wanted Crash to become a balloon, they just animated a keyframe of him wrapped around a sphere (shoes and face usually un-stretched!) and the game tweened to it.

Conclusion

The development of Crash bandicoot games has more interesting stories and software tricks that made it successful and remarkable.
Maybe later we can discuss how they hacked the software to have great detailed levels while the other software developers couldn’t do.
It all boils down to good business decisions and brilliant software tricks to get the most out of the hardware, Eventually netting them 40 million units sold.

Feel free to leave your comments.
Cheers.

Further reading :
https://www.youtube.com/watch?v=izxXGuVL21o&t=1444s&ab_channel=ArsTechnica

https://all-things-andy-gavin.com/2011/02/02/making-crash-bandicoot-part-1/