Legibility, Confusion, Prediction and Enjoyment

As I have mentioned before, I’ve been following the magician Dani DaOrtiz ever since his appearance on Penn and Teller. One of his oft-repeated lines (in lessons) is that “Confusion is not magic.” But there are subtleties. Obviously the audience must be unaware of “the secret.” If you catch the magician secretly palming the card, or if the assistant has a twin then the trick unravels. So the audience must not know certain (key) facts. The effect must be clear, not confusing. The method should not be obvious.

That person was over there, tied up, in a bag. A second later they are over here, untied. Crystal clear.

When talking about Horseless Carriage I called the game state “illegible,” but now I realize that was imprecise. The current state is perfectly legible. A player has a score ($), a factory layout with lines that can produce X cars with certain spec, Y cars with different spec, Z cars (or trucks or sports cars) with another spec. That player has research positions, an order on the focus track, generates so much research/gannt charts per turn.

It may take a few minutes to “read,” but its all there. No cards. Nothing hidden.

The issue with H.C. is that game state changes massively (and simultaneously) during the “Build a factory” phase. You add parts and specs and whatnot and suddenly that player can generate cars with new specs, etc. The “build” step contains zero-to-many substeps of “add something to the factory floor.” What can a player do in that phase? Lots. Improve R&D, or leave it and make much better cars. Perhaps open a new dealership, or improve marketing, or second line. It’s similar to the Bin Packing Problem, and it’s hard.

I worry that there are optimal solutions. I don’t worry that people can find them at the table. But when I called the game “illegible,” I was fumbling around the following idea “When this step starts, it is very difficult to predict where the game will be at the end of this step. You have a wide variety of options and even deciding what you should be optimizing is difficult, much less optimizing it. But even if you optimize it, you may do well or not based on what other people are optimizing, if they are in conflict with you, and the like.”

(Of course, how well you optimize is also a big deal, but let’s skip it).

“And even if you are correct, other players may not see the solution the same way and their choices may greatly benefit/harm you based on their views.”

So …. Horseless Carriage is legible. But it’s not easily predictable.

In some ways, this is the same as a dudes-on-a-map game where the winner may simply be decided by who attacks who. That’s predictable (X is a jerk, everyone attacks him, given a choice, etc), but it’s not necessarily a function of the game rules. This is (sometimes) considered a flaw with the game. Similarly, one game of H.C. was decided by two players fighting Sports Cars, two players fighting for the high end market, and one player left alone in trucks (winner!).

But you want some level of predictability in a game. Chess has lots of predictability …. good players will all look at the same small list of moves in a given position. You still need to do the hard work of calculating to show that a move is good, and spectators delight in surprise moves, but the move is a surprise because of the superior skill. You thought g6 was bad, the grandmaster played g6! What did he see that you didn’t?

Interesting.

If Chess let you move all the pieces each, it would be a less interesting game, partially because predictability would fall (but more because the constraints makes tactics more interesting. This is not a theoretical example, looking at you Fuedal).

I got to thinking about this because of last night’s game of Darwin’s Journey. (Which I didn’t enjoy.) Darwin’s Journey is legible. You can glance at a player’s board (and the board) and see what spots are open to his workers, explorers, position in turn order, etc. It might (again) be difficult to read, but its open. There are some tiles drawn each round that add uncertainty, but OK. It’s much more predictable than Horseless Carriage, because each turn is a “simple” one … place a work and resolve an action. You might still have the same problem of “That player misjudges and takes a space that doesn’t help him, but hurts me,” but that can happen in most games. In fact, Darwin’s Journey is predictable enough that I could often say “I will take this spot, I think A/B/C will do X/Y/Z and then I’ll take that spot.” I could see many ply into the future (which is basically impossible in H.C.).

But I didn’t enjoy the experience. Perhaps because the game state had too much information? When a chess player makes a move you hadn’t considered, it’s exciting! (Not necessarily good for you, but exciting!) Maybe they blundered. Maybe I missed a nuance. But something important happened.

When a player makes a surprising move in Darwin’s Journey, well … it’s a point salad game. They are still going to get a few points, maybe a few more than you thought, maybe a few less. And there is so much going on in the game state, perhaps its a wash.

(“This game lacks focus” was my comment at game night.)

You want some predictability. Not too much. Not too little. Tic-Tac-Toe is predictable (once you are no longer a small child). So it’s boring. Chess is in the sweet spot (for many people). People who play Bridge or Go are self-selecting into an amount of predictability that they want. (Bridge is interesting, because the first part of becoming a good bridge player is learning how to read the game state, which is a non-trivial task in hand evaluation and understanding a bidding system, as well as the card play. For a novice, the game is mostly illegible, and the experience involves ‘learning to read’).

For those games, surprise is exciting, but you need the framework to be legible, so you have enough information to be surprised, and not just bewildered and confused. And as we’ve seen by Go and Bridge, learning to read a game state is often a valid (cherished!) part of the experience. I suspect that most of my irritation at Darwin’s Journey is that the learning to read the game state is complex, but it’s just “kitchen sink” complexity, not organic.

Anyway, I still feel like I’m not quite at the point, but I think I’m moving closer.

And moving back to magic, I see a parallel in the showmanship. Dani talks about “ending on a snapshot” where even if someone just wandered into the last second, they could get a rough idea of what’s going on. “Ah, there’s a pile of cards and a face up card. He must have found the card in the right place” or “He flips up two piles of cards and all the red cards are in one pile, and all the black cards are in another.”

The trick should not be confused. The effect should be simple. You need focus.

Advertisement