From WORLD GAME REVIEW, Issue 8, July 1988:

What’s Wrong With Ultima

by Robert Abbott


WGR presented a three-part series by Paul Yearout on my game Ultima. Yearout praised the game, saying it had “about the same complexity and fascination as chess.” I would like to take an opposing view here and describe what I think is wrong with the game. You’ll probably think it’s strange for an inventor to attack his own game, but some interesting points are involved, and perhaps someday I, or someone else, will create an improved version of Ultima.

It will make more sense if I present my thinking about this game in chronological order. I first got the idea for Ultima in 1961 while reading Murray’s History of Board Games. Murray classified board games by their method of capture. This classification was possible because throughout history, each game used only one form of capture, though some games used more than one form of movement. I thought, why not break this pattern and have a game with many forms of capture? After only two months of experimentation I came up with Ultima, which has seven different types of pieces, each with a different form of capture. Four types of pieces, the pawns, the long-leapers, the withdrawer, and the king use captures that were taken from older historical games. The other pieces used captures that I invented.

Ultima was first published, without a name, in the December 1962 Recreational Mathematics Magazine, and in 1963 it appeared in the hardcover edition of my book Abbott’s New Card Games. That is the version of the game that Paul Yearout presented in WGR. Incidentally, Recreational Mathematics Magazine had a contest to name the game. The winning entry was “Baroque,” which I still like. However, the publisher of my card game book insisted on calling the game “Ultima.”

I now see two major flaws in Ultima. I came to understand the first flaw around 1964. It is that defense in the game is more powerful than offense. This may not seem to be too terrible, but what it means is that the player who develops his pieces and who launches an attack is at a disadvantage. Usually attacks fail, and after a while that gets to be discouraging.

I now think of this as a major principle of game design: offense must generally be more powerful than defense. If the defender usually wins in an exchange, then a player will realize that the best strategy is to do nothing and wait for the other player to attack. If both players decide to do nothing, than the game obviously falls apart. Surprisingly, this flaw can be found in many games. I suspect it exists even in a successful game like Stratego.

In Ultima, defense is stronger than offense because of the actions of the pawns, the long-leapers, and the co-ordinator. These pieces are generally more effective against enemy pieces that have moved out into the center of the board. In addition, the pawns and the long-leapers can usually launch their attacks from their original positions. They don’t have to be ‘developed’ to gain in power.

When the paperback edition of my card game book was published in 1968, I made a change to Ultima that I thought would cure this flaw of the weak offense against strong defense. I said that a piece on its first rank could move only one square, a piece on its second rank could move one or two squares, a piece on its third rank could move up to three squares, and so forth. I reasoned that this would give the players motivation to move into enemy territory. I now realize that this rule was just a sort of patch. It did alleviate the flaw, but it made a real mess out of the game.

In the early 1970’s I was still puzzling over what was wrong with Ultima, and I was finally able to understand that Ultima had a second and more serious flaw: it lacked clarity. At the same time, I was able to work out this concept of ‘clarity,’ and I wrote an article on the subject which was published in the May 1975 issue of Games & Puzzles. [That is the article Under the Strategy Tree, which is now on my web site.] This article turned out to be pretty influential and has been referred to many times. Unfortunately, almost everyone who wrote about the article completely misunderstood what I was saying. I still haven’t figured out why this happened.

In the article, I said that what we perceive as ‘depth’ in a game has little to do with the size of the strategy tree for the game. Instead, ‘depth’ is how far a player can see down the strategy tree, and how far a player can see depends on the clarity of the game.

Applying this concept to Ultima, it turns out that certain pieces are clear and others are not. The pawns are the worst for clarity. Whenever I play the game, I’m often subject to sneak attacks by the enemy pawns. Obviously, if I can’t see an attack coming one move ahead, than I can’t see very deep into the game. I don’t know why the pawns cause such a problem, since their method of capture isn’t that complicated. Perhaps there are just too many of them running around.

The withdrawer is a piece with great clarity. It’s very obvious when the withdrawer has a piece under attack. The immobilizer also has clarity, since it freezes the action around it. The immobilizer also simplifies and creates a focus for the game. The chameleon is also pretty clear, at least when it is attacking another clear piece. This is rather surprising, since the chameleon is a pretty complicated piece.

Further Notes:

Back in 1988, none of the readers of that article sent me suggestions for new Ultima pieces, but then, the magazine only had a circulation of about 100.

I did come up with a couple of ideas of my own about how to alleviate the problem of defense being stronger than offense. A lot of this problem is caused by the pawns, which capture by creating the following configuration:

A  B  A

where each A represents a friendly piece and B represents the enemy captured. With this type of capture, the pawns are powerless against pieces on the edges of the board or against undeveloped pieces at the end of the board. The new idea I came up with is to let the pawns capture if they create this configuration:

A  A  B

With that form of capture, the pawns could capture pieces that are just about anywhere. For this idea to work, there would have to be a lot of experimentation, especially concerning how the pawns would move.

About the long-leapers: I would just get rid of them. And about the coordinator: I’d replace it with a similar piece called the “triangulator.” Actually there would be two triangulators. If one of them moves, they capture any enemy that is on an intersection of the diagonal lines that run through the triangulators. I figured that looking along the diagonal lines instead of the orthogonal lines would have this advantage: you can now capture pieces that are on the end of the board, and you can also more easily get at pieces on the sides of the board.

Recent developments, March, 2004: has presented two variants on Ultima: Maxima and Rococo. I thought Rococo was very good, especially the “cannon pawns,” which are an improved version of my AAB idea presented above. If you scroll to the bottom of their page and click on “View all comments,” you’ll see that I submitted a suggestion for Rococo. They didn’t think much of the suggestion, but that’s okay. In the future, they promise yet another Ultima variation, called “Supremo.”

A final note: I’ve been saying that any day now, I’ll work on making improvements to Ultima. The problem is I’ve been saying that for the last 40 years, so it’s obvious that I’ll never do this. Therefore, if you’d like to try improving the game, please do. You have my permission to make any changes you want.

This section is printable.

Back to the Ultima page

To the games home page