“Game Design Gems” salutes a game for its outstanding game design. We look at the kind of games that makes us drool in awe, shake our heads, and go on a two-month “why-didn’t-I-think-of-that” drinking binge. This time I’m getting owned by Plants Vs. Zombies from PopCap Games.

SPOILER ALERT – Read on at your own risk!

During my first play-through of Plants Vs. Zombies (PvZ), I got completely absorbed, and entered the flow-like state where the real world fades away and time passes with incredible speed. Being a game designer by trade, this is very rare for me. Though I try not to, parts of my subconscious always has on its “analyzing spectacles”, and constantly takes mental notes of the game’s design choices, features and mechanics.

It was only during my second play-through (which – by the way – is a new game mode in itself) that I “woke up” and started asking myself questions like:

• Why do this game rock so much?
• How did PopCap put this together?
• How many goddamn features did they cram into this thing?

Too many or too few features?
Designing any game, there has with certainty been hundreds of design-ideas, which have been looked at by the developers, but found either too unrealistic, crazy, difficult to implement, conflicting with other existing material or otherwise unfit for the final product.

Then again, most games usually contain features which should have been cut, since they don’t add to the game or weren’t given the love and attention they need, and appear (semi-)broken to the player. Worst case scenario is a game where the parts don’t seem to fit at all, or major parts seem completely “tacked on”.

This usually happens because the individual features are not well integrated with the OTHER features of the game. I’ve found that games rise in complexity the more features they contain (well, duh!) – What may be surprising is that this increase in complexity isn’t linear.

One new feature isn’t ONE (1) new feature, it’s usually more around something like (n/2) new features (n being the number of other, existing features). For example: If you have a game with 10 core features, and add a new one, more often than not you’re really adding somewhere around 5 or more “special cases”, many of which could be considered features in their own right!

This is because new features need to interact with some existing features, to actually work and have any impact on the game. To further exemplify: If you add a gun to your game and it doesn’t interact with ANYTHING, then it’s not really a feature. It must DO something (like “damage enemies”).

… but that spawns the question: “What does the gun do to other things in the game?” (like windows, doors or even dead enemies?) More often than not the consequences of introducing new features are much, much more far reaching than the game designer has initially imagined. Each new special case has to be handled in some way, and sometimes conflicts so much with the rest of the game, that the designers have to cut the entire original feature.

Thus a game, like PvZ, that literally contains +100 features, and where everything interacts smoothly and is well balanced is extremely rare, not to mention mind-bogglingly difficult to produce.

How many features is that?
So how many features is in PvZ? The amount obviously differs depending on the “granularity” with which you look at the game. How “big” does something have to be, to be considered a feature? My personal definition is “something which changes the rules of the game in a non-linear fashion”

I.e. obviously “guns” is a feature (of some games), but a second gun with increased firing-speed or damage is not a proper stand-alone feature. However a second gun that slows enemies, or make their heads turn into pineapples is a feature (or two, if it does both!)

PvZ contains something like:

• +40 different plants (if you don’t count very similar ones)
• 20-or-so enemies
• Money & shops
• 20 wildly different mini-games
• 4 playing field variations
• Zombie head variations and other easter eggs
• Water
• 4 different zen gardens with AI actors
• A handful of Puzzle Modes
• Plus change

Let’s just make that an even “100 features”. Please don’t take the following “guesstimations” as science, but just to get into the heads of the designers at PopCap, that gives us some 50 “special cases” per  feature, or around +5.000 special cases – many of which may be considered features in their own right!

Like: “What happens when a snorkel-zombie is hit by fire while being underwater?”. The developers had to think of that, it might be no biggie, or it might be such a big thing that it can be considered a feature in it’s own right. Now there’s only 4.999 special cases to go!

I may be wildly over exaggerating with my estimations here. PvZ has a pretty basic foundation (2D tower defence based game, on a 6 x 9 grid), which makes most special cases easier to handle than epic 3D shooter-rpg-MMO’s(?!?), but I still the developers must have counted their special cases in the thousands.

I’m in ur gaem – Exploitin ur rulez
An obvious question for game designers: Can the player (with a combination of the tools available to him) somehow destroy or exploit the game? Unfortunately, the chance of this being possible rises exponentially with the number of features!

The more kinds of tools the game gives to the player, to help him battle the undead hordes, the bigger the chance that some combination of these will enable the player to completely circumnavigate the basic rules and break the game.

Somehow PvZ manages to walk this minefield, and comes out on the other side reading a newspaper with a cocktail in the other hand. The developers have managed to cram an insane amount of content into the game, while leaving, no feature “unsupported” by the rest of the game or easily explotable. The coherence is quite stunning, and can only be the result of incredibly talented game design and project management.

I won’t say much more about the content or game design of PvZ (you really should play it yourself.. like, right now!), but leave with a small story of my attempt to exploit the game. This is where I had my behind handed to me.

My life as a Marigold Farmer
Completing the game was so much fun I decided to play through a second time, straight away. I was thrilled to find that that I was allowed to use all my hard-earned tools (plants, mushrooms and other goodies) I’d collected during my first play-through. To make things more challenging the second time around, the game decided to auto-select some of the tools I had to use, on each level. Still, I had a pretty easy time on the first few levels (as was also the case on the first play-through).

So, I’m on top of things. I could handle those brain-eaters no problem. Let’s use the lack of resistance to FARM SOME GOLD! Yearh!

Defeating most of the zombies in a level, I decided to allow a single one to survive, and keep him at bay by letting him slowly chug away on a huge “wall-nut”. Meanwhile, I started planting a crazy amount of gold-generating (and aptly named) “marigold” flowers! Then I would just sit back and relax, while harvesting vast amounts of gold from my marigold forest.

Or so I thought…. But alas, no gold would grow on my flowers! It seemed that they would only yield at an incredibly slow level, while only one zombie was alive. As soon as more zombies appeared they started acting “normal” again. Some damn code inside the darn game had anticipated my clever attempt at debasing the currency through botany. Blasted!

All Hail Our New Botanical Overlords
The above example, might be an awkward scenario, not worthy of elevating a game to any mythical state. Still, this kind of attention to balancing is rare, and it shows how deep the game has been tested. And this example is not alone. There are lots of small ingenious balancing tweaks at place everywhere you look. Add to this, literally more than a hundred fun features, oodles of charm and attention to detail.

I rest my case: Plants Vs. Zombies is truly a Game Design Gem.

Feel free to agree or disagree below. Your comments are most welcome, as this is a completely new blog, and it gets pretty lonely in this corner of the intertubes.