Featured Blog

How A Traumatic Experience with a 1998 PC Game Led to a Big Game Design Mistake

How can a critique become a template for improvement? A critical review of new monster boxing game Beast Boxing Turbo made the lead developer reconsider some assumptions he traced back to a bad experience with seminal PC game Thief.

Beast Boxing Turbo for PC/Mac launched to some quick jabs from the press, but instead of dodging the issues, Lead Developer Gordon Luk saw them as valid points that pointed to a deep-seated but correctable mistake with the game design.


I woke up a few weeks ago to be greeted by a 3/5 star review from Gamezebo’s Mike Rose. It was tough to see criticism for my game, and it was even tougher to admit that he might be right.

Mike praised for the way that Beast Boxing Turbo felt to play, but wrote that it didn’t have enough variety because there was a single strategy that worked all the way through the game.

It was true – and it was an artifact of a thought process I went through when I was originally designing the gameplay for Beast Boxing. To understand why I’d keep the core mechanics relatively unchanged throughout the game, I should really explain why I went from loving to hating the 1998 PC game Thief: The Dark Project.

In Thief, gameplay in the first level was amazing. It started with creeping past, hiding from, and pouncing on guards to stealthily move through medieval environments. I was complex, tense and exciting, and I utterly loved it. Then I got to the zombie level.

Zombies. The pinnacle of game AI.

There, an effective strategy was to throw caution to the wind and run past hordes of slow-moving, near-invulnerable monsters. It was very upsetting to lose my connection to the fun aspects of the game when all the awesome things I had been trained to do were suddenly replaced with outrunning zombies.

With disappointment, I quit the game and never came back. Right or wrong, I was twenty years old and felt like someone had turned Robotech into GoBots. I couldn’t understand why something that was so much fun was suddenly changed so drastically. That memory seared itself into my sensibilities about game design.

When developing Beast Boxing, I single-mindedly focused on the boxing gameplay. Once it was fun, I intentionally made sure the entire game played just that way. Looking back on it now, I can see a direct trail from this decision leading back to my fear of ruining a good experience from my time with Thief.

Beast Boxers each get their own set of spreadsheet formulas.
Beast Boxers each get their own set of spreadsheet formulas.

In the original release of Beast Boxing Turbo, I ratcheted up difficulty slowly by tweaking variables and changing attack patterns, but once I found a fun AI behavior, I applied it across the board to all opponents. At the time, it seemed anything that made a match more fun should just be present in every level. That’s how the gameplay ended up being so similar from opponent to opponent, and why a strategy that worked in the beginning also worked in the end. I didn’t stop to consider why this approach might be too extreme until I saw Mike Rose’s review, and then the mistake was clear.

I see now that I missed an important middle ground in my desire to hold tightly onto successful game mechanics. Spreading out the introduction of new stuff isn’t just useful for in-game tutorials or smoothing difficulty. It’s also really important for keeping the variety and pace of learning that modern players expect in a “real” game. I still think my experience with Thief was informative – in that one should take an incremental approach when introducing new types of gameplay, and take care not to vary things too drastically in the beginning.

Treimann is the first character with all of the new AI behaviors. And he's a doozy.
Treimann is the first character with all of the new AI behaviors. And he’s a doozy.

However, the Gamezebo review forced me to look for a better approach – which I instantly recognized as traditional gameplay element progression (duh). It’s not all about in-game tutorials – you need to keep your player learning the entire game if you want to fulfill the needs of the modern gamer’s brain. That is, to either start with the ideal core gameplay and layer on top of that, or slice it and recombine it throughout the game. In the case of Beast Boxing Turbo v1.1, I followed the former method by keeping what I already had and rolling out new opponent behaviors after each section. After every major accomplishment, the player is asked to adapt to new circumstances of behavior, expectations, or pacing.

The good news? These changes didn’t have to be content-based, and I got stuff done by hacking on the gameplay.

Improving the Game

For this update, I had multiple mini-goals – to eliminate the degenerate strategy of just blocking-then-punching, to introduce new behaviors throughout the game, and also to create a less difficult option for casual players. I added brand new enemy AI behaviors as you advance through the leagues. I tweaked the way that blocking works to make it ineffective as a permanent strategy, and introduced an “Easier” mode that offers itself up when the player loses a match. These changes are intended to keep players learning on one end of the spectrum, and prevent them from giving up on the other end.

The new progress / loading scene.
The new progress / loading scene.

For fun, I also added in a progress screen to replace the old static Loading image, so players can get a sense of how far they’ve gotten in the game at each fight, and give them a bit of context. It’s a small improvement, but it’s something I meant to do in the original release and now had time for.

In Conclusion

My final understanding of this whole experience is that players crave learning. We’re built to adapt to changing conditions, and so static conditions (even difficult ones) get boring without some kind of metagame or competition going on. It’s important for developers to question and explore industry standards, and perhaps do their own work in understanding why those standards exist.

In closing, it was tough to get that harsh review, but ultimately I have to express my gratitude for helping me understand what went wrong and improve as a game designer (Thanks, Mike!). It’s my sincere hope that the update will result in a much better game and a bit of rejuvenation for Beast Boxing Turbo.

P.S. I was an impatient 20-year old when I played Thief. It would be great to go back and see whether I feel differently about it as a more mature gamer. :)

Latest Jobs

Xbox Game Studios

Redmond, Washington
Technical Lighting Artist


Hamburg, Germany
Game Designer - Elvenar

Six Foot

Houston, TX
Six Foot Director, Player Relations

Hometopia Inc.

Lead Engineer
More Jobs   


Explore the
Subscribe to
Follow us

Game Developer Job Board

Game Developer Newsletter


Explore the

Game Developer Job Board

Browse open positions across the game industry or recruit new talent for your studio

Subscribe to

Game Developer Newsletter

Get daily Game Developer top stories every morning straight into your inbox

Follow us


Follow us @gamedevdotcom to stay up-to-date with the latest news & insider information about events & more