Designer's Notes: Proteus

by Francis K. Lalumiere

"And Proteus began at once with his old tricks, first changing himself into a lion with a great mane. Then suddenly he became a dragon, a leopard, a wild boar; the next moment he turned into running water, and then, finally, he was a tree."
--Homer, The Odyssey, Book IV

I designed the basic mechanics of Proteus while waiting in line for the Back to the Future ride at Universal Studios, in California.


For some time I had been toying with the idea of playing a game with dice for pieces and the rules of chess for movement, but I had never sat down to figure out how such a game might actually work. I don't really like chess; I find the frame of the game too restrictive (and I don't mean the edges of the chessboard, here), but I'm fascinated by the movement possibilities of chess pieces. As a result, I oftentimes contemplate chess, but I rarely play it. I had tried my share of chess variants, but somehow I felt that my "dice chess" idea would create something interestingly different. If I could only figure out how to play the game.

Things stayed pretty much status quo for about two years. My subconscious might have been polishing up some concepts, but I didn't consciously work on the game until I found myself stuck in line at Universal Studios. I'll try to make a long story short . . . I was alone in Los Angeles on business, and I found myself with a free afternoon -- the temptation of the Universal Studios demon was too great to resist. So I went. At some point I ended up trapped in line at the Back to the Future ride, and since I had no one to talk to and hadn't brought a book (water rides oblige), I thought it might be worth my while to start laying down the foundations for my dice-chess game. In any case, it would be a lot more productive than staring at the Hawaiian shirt in front of me for an hour. By the time I got on the ride, I already had a very good idea of the inner workings of the game. And when I stepped out of my rigged Delorean (great ride, by the way), I couldn't wait to get back to my hotel room and write everything down. I used the rest of my in-line waiting sessions to run a few mental tests and do some basic math.

At the end of the day, I drove back to the hotel with my head full of sketches and design notes. I knew that my game, like chess, would be a two-player game and that players would alternate turns (I know this sounds very basic, but at some point you do need to give that stuff consideration). I also knew that each player would start with eight dice, that the dice would show a different chess piece on each face, that the goal of the game involved capturing several pieces (as opposed to neutralizing one particular target), and that a player would perform two actions on his turn: move a die and rotate a die (to change the identity of that piece) one step at a time. Most importantly (hey, we all have our fixations), I also had a name for the game -- Proteus, the Greek god of ever-changing form.

I thought I would stay up all night but, much to my surprise, I was content with simply jotting down what I had devised during the day and going to bed. All right, I'll be honest; I had some problems falling asleep. Morpheus was always bumped by Proteus who thought some of his needs still had to be addressed.

And he was right.

First of all, I didn't know how you were supposed to win the game. I had a vague notion related to taking out as many of the opposing pieces as possible, but somehow that felt wrong. Also, I still hadn't found a balancing mechanism that would prevent powermongers from rotating all of their dice up to queens. There was one most delicate matter: I didn't know what to do with the king. Since I was dedicated to the disappearance of the checkmate concept from Proteus, was it still worth it to keep the king in there?

I eventually managed to fall asleep, without having resolved anything. I did try to refine and complement my notes on the flight back, the next morning, but somehow I didn't make much progress until I got home. Then, in one evening, I solved practically all of my problems.

For the sequence in which the chess pieces would appear on the faces of the die, I chose to adhere to the traditional chess value sequence, with one exception: I placed the bishop in front of the knight, because I felt it would open up the game faster. That gave me pawn, bishop, knight, rook and queen. Still no idea about the king. However, it became obvious pretty quickly that a player would have to move a die and then rotate a different die; infiltrating the enemy camp as a bishop and immediately transforming into a knight smelled of overpowered tactics. Interestingly enough, the questions of game balance and victory conditions were solved at the same time, and with the same gimmick. Back then I was playing the game with standard six-sided dice, remembering that the 6 was the queen, 5 the rook, 4 the knight, 3 the bishop and 2 the pawn (1 could be the king, but I felt queasy about that). I thought it might be fun to win a game of Proteus by having the most points, with each piece being worth a certain number of points. In assigning point values to the pieces -- with my endless originality, I chose the number of dots that represented each piece -- I discovered that such a scheme would automatically balance the game, because the stronger pieces were worth more points when they were captured by the opponent. I had done it!

Or so I thought.

I moved on to the problem of the king. Since a game of Proteus was won by having the most points, there was little incentive to keep the king in the game. Without his vital quality, the rules of movement make the king into a limping queen, and that's nowhere near appealing given the morphing nature of game pieces in Proteus. Plus, where would the king fit in the die sequence? I could simply have thrown the king away, but here was an opportunity to inject another new concept into the game. So instead of eliminating the king, I would replace it. After considering different ideas for a while, I chose the simplest one: the new piece would not move at all, but would also be impossible to take. I named that piece the fortress.

With these problems apparently resolved, I could sit down and play a few "real" games against myself. Everything was going rather smoothly. The game was simple, played fast, yet retained its possibilities for deep strategy. And then I hit a brick wall with an evening dress and a crown.

The queen.

Because of the mathematical progression of the value of pieces, and because of the similar progression of their respective powers, the game was exquisitely balanced -- except for the queen. The problem laid in the power of the queen being much greater than its point value. Between a rook and a queen, the difference in power is immense; yet the different in point value was only one. That couldn't work. One solution was to raise the point value of the queen. But that didn't appeal to me because it implied assigning the queen some ridiculous value, say, 12 points. This meant that a player capturing a queen would probably win, and I wanted to avoid that like the plague. I therefore chose the only alternative, which was to make the queen less powerful. But instead of toying around with its movement and capturing rules, I thought it better to curse the queen with a weakness. I had the idea of making the square directly behind the queen vulnerable to attacks (much in the same way the queen herself can be taken on her square), but I couldn't help feeling that the right solution couldn't be that easy, or that elegant: one could capture the queen by stepping on her gown. So I played some more games and, surprisingly, the queen seemed to behave. I was delighted with the fortress, and overall Proteus worked. It was time to take the game out for a spin.

I took Proteus to the local game store where I always went, and a couple of friends expressed interest in the game. We played over and over again, and I kept thinking that some obvious flaw would blow up in my face; but no such thing happened. Back home, I put Proteus to the test with a few more friends, and the game still held up. We did try a to increase the number of pieces each player starts with (up to 12), but we only succeeded in creating quite a mess of a traffic jam, with games that would take more than an hour. And I was determined to keep play time to 30 minutes or less. We also experimented with fewer pieces for each player, but the game quickly became boring and predictable. Eight seemed to be just the right choice.

At that time I had just finished translating both Illuminati and Tile Chess into French for Steve Jackson Games, so it was an easy matter for me to tell Steve -- between emails, as it were -- about Proteus. He was immediately interested, asking for only one change: that the fortress become the pyramid. I should have thought about that myself. And before I knew it, my game was on the list of upcoming Steve Jackson Games titles.

Then I started to worry.

You see, I hadn't really planned on selling Proteus to anyone; I was really just playing around with the game for my own enjoyment. But now it was going to be out there, and I realized it had to be as good as I could make it. While I didn't think anything needed to be changed in the game itself, I thought the door was wide open for a series of intriguing variations on the basic rules.

So I set out to have Proteus itself morph into half a dozen different forms, much like its components could do on the battlefield. I designed a handful of variants and thoroughly tested them all with my (very patient) wife Stephanie. Four of them ended up in the finished game: Trade-Off, Russian Roulette, Wall Street, and Polarity. That's five if you count the two different setups for Wall Street. Steve suggested an additional variant, which we ended up calling Warhorses -- for a total of six variants. But where did the other ones go? Right here.

Random Setup

This was the shortest-lived of the variants I came up with. Rolling your dice to see what pieces you would start with was the dumbest idea this side of the sun, and it only survived for about half a setup phase -- I realized pretty quickly there was no way I could make this work. Still, the concept of rolling pieces (because they are dice, after all) fascinated me, and that eventually became the Russian Roulette variant.


The idea here was that pieces could only capture opposing pieces that were of equal or higher value. So a pawn could capture everything (except the pyramid, but that's a given), and a queen could only capture other queens. (Can you see the pyramid pattern? You're quick.) This didn't work because no player wanted to upgrade to stronger pieces. So what if you have a knight? I'll just stay with bishops and the only thing you'll catch is a cold.

I tried playing the Inverted Pyramid (capture downwards) but it only succeeded in destroying the effectiveness of the weaker pieces and handing a flak jacket to the stronger ones.

Pyramid didn't evolve into anything else; it got buried in the back yard.

Black & White

Whenever you capture a piece, its point value is the standard one if it was captured on a black square, but is whatever's on the bottom face (or 7 minus top face) if it was captured on a white square.

Pretty inventive, right? This is one I actually sent to Steve, but in the end it was not included with the rest because it didn't really add all that much to the strategy. Plus, the configuration of the Proteus dice would be different from that of regular six-siders in order to facilitate the upgrade/downgrade procedure, and that would throw the point balance out of whack.

There were other variant possibilities, but most of them involved a complicated set of special rules, and I didn't want to burden Proteus with unnecessary weight. There was only one problem left to solve, and that involved my asking Steve for a favor: I wanted to dedicate the game to one of my oldest friends, whom I've known since we were both nine years old (I was 28 last time I checked), and with whom I've shared a universe of close games, hot matches, and gaming disputes. I gave Steve two versions of the dedication, and they ended up going with the short one. Here's how the longest one read:

"This game is dedicated to my dear friend Alexandre "Le Brown" Boivin, who started bugging me about designing my own game 12 years ago and who hasn't stopped since. I'm as happy to see my first game published as I am to finally shut him up. :)"

Unfortunately, we ran out of space.

And speaking of running out of space, here's a fun little variant suggested by Steve right after he played his first game of Proteus. It didn't make it on the rulesheet, but since there's no limit on electrons (yet), here we go:


This is played with regular chess pieces and two Proteus dice. Each turn, roll the two dice. You must move one of the two pieces showing on your dice (pyramid=king). If you can't make a legal move with one of the indicated pieces, you don't move.

If this is too hard, roll three dice to give yourself more options. Or play with a handicap: the stronger player gets only two dice, and the weaker player gets three.

Should you develop your own Proteus variants, I would be delighted to see them! Feel free to email them to me at We might even compile the best ones and post them on the official Proteus homepage, at

My next game? I do have an idea in the back of my head. I'm trying to arrange a trip to Bush Gardens to sort it out.

Article publication date: October 5, 2001

Copyright © 2001 by Steve Jackson Games. All rights reserved. Pyramid subscribers are permitted to read this article online, or download it and print out a single hardcopy for personal use. Copying this text to any other online system or BBS, or making more than one hardcopy, is strictly prohibited. So please don't. And if you encounter copies of this article elsewhere on the web, please report it to