« March 2004 | Main | May 2004 »

April 18, 2004

Notes on Chess

It's pretty foolish of me to think I could say something about chess that nobody has said before but here goes anyway. I've been playing it a lot lately and it's my new favorite game of all time. (It's also one of the reasons I've been posting less frequently.) Chess has a number of interesting characteristics that I find appealing:

- orthogonal unit design

Chess is the prototype, possibly the first game ever to demontrate this. Checkers (or did they call it Draughts back then?) and other games of past centuries tended to have just one unit or counter. Although go may have more depth than chess--which is why computers can solve chess but can't solve go--the orthogonal units give chess more breadth. It makes chess more tactically rich than go. (As your board gets smaller, the number of possible patterns in go falls off faster than the number of possible patters in chess. I'm not sure at what point - it's less than 8x8 - chess takes over but at some point it does, which is rewarding for people who like to micromanage and less rewarding for people who like to stay on top of the big picture.)

- "Chess is brutal" - Tom Henderson

Small advantages in chess get magnified. If you're ahead by just one pawn, as long as you don't make any stupid mistakes you are likely to win. This is an example of positive feedback. Also, because there are no random die rolls or cards or the like, the effect of luck is minimized. (Richard Garfield once pointed out that if he was lucky enough, he could beat Kasparov at chess. It's true: I, for example, have a ranking of about 1100, and Kasparov is, I don't know, 2700 or something - that means out of 10,000 games I can expect to win one.)
This means players are greatly differentiated. Meaning: good chess players are more likely to beat bad chess players than with most other games, and it takes very few games of chess to establish who's better.
(In contrast, Mark Nau has told me that it takes 70,000 hands of Texas Hold 'Em to prove with a reasonable degree of certainty that you're a good player.)
For people like me, that's great. I know exactly where I stand, and I can watch myself get better. (I was ranked 1000 a month ago.) But for people who don't like the humiliation of knowing how mediocre they are, this isn't so good. (And I may not think much of it once I plateau, depending on where that plateau is...) Which is why they might turn to a game that isn't so brutal, so they can pretend to themselves that they're better than they are.

- Chess is popular in my culture

I started learning go a couple of years back; because it's a game that computers suck at, it was interesting to me. I didn't stick with it because none of my friends play it. But almost everyone I know plays chess, or knows how to play chess. So a rich-get-richer phenomenon happens. Maybe go's superior, but switching to go would be like switching to metric.
(And there is something nice about playing games where the computers are so good; it's good for training.)

- Iterative design

When playing chess, I'm amazed at how well it seems to work. The rules have a coherence to them that encourage interesting dynamics. The way a bishop can block half a knight's field of moves...the way the bishops complement each other, white and black...the slow but maneuverable king...the promotion of pawns.
I think that the main reason it works is because it's been tweaked and tuned for centuries. Take en passant; this is no doubt a kludge added to the rules to compenate for another, earlier, kludge, the kludge that allows a pawn to move two spaces from the second rank. (A kludge no doubt added just to get the game going quicker.)
And this century has brought us chess clocks.

- Speaking of quick

If you don't overthink, chess is a quick game. (I am very guilty of overthinking if I don't have a clock to keep me honest.) I like the under-half-an-hour variety; this gives me enough time to do a little due diligence to check that the move I'm contemplating isn't heinously stupid. Which makes chess addictive for me; just one more game. Just one more game.
Without chess clocks, sandwich gameplay becomes a problem.
(Speaking of sandwich gameplay and Texas Hold 'Em - since statistically you should be folding on eleven out of twelve hands, Texas Hold 'Em has to take the cake for the most boring game ever. Mark Nau tolerates it by playing it online at four tables at once.)

The big problem with chess, of course, is that it doesn't have a singleplayer campaign with a compelling storyline and dramatic CG sequences. There's no reason it couldn't - I'd be interested in a singleplayer chess game that was a sequence of puzzles designed to gradually teach you the game. It could begin with your white king being exiled from his home, with only his trusty rook by his side. In the forest, they come across a lost enemy general. And the first "level" begins: mate a black king with a white king and a rook. As the game progresses you recruit more pieces, learn their properties, then endgames, middle game combinations, then openings. Once you've finished the white campaign you play the black, now on the defense against the whites who want to recapture their throne...

April 14, 2004

Better Late Than Never

A new Manager In A Strange Land is up. Has been for a while, actually.

April 11, 2004

Bridge On The River Kwai

Saw this movie for the first time last night. It is a morality play about project management, something I suppose movie makers know a fair bit about. I'm not the first to notice this. I couldn't help rooting for Alec Guiness as he stood up for his beliefs, the founding principle of which was the officers shouldn't do manual labor, they were needed for leadership. Leaders must lead. I've said the same thing myself, and Tom DeMarco said it before me. How many times have you seen upper management tell middle management that they better get in there and do some of the work-work-type-work themselves? That's why I found Alec Guiness's situation so compelling. Eventually upper management, in the form of Sessue Hayakawa, gave in, and the project turned around.
(Still, in the movie, they did hit a point, near the end of the project, where the officers volunteered to do manual labor. At some point on your project, the planning is done, the course is set, and a lead will have some free time to pitch in and fix some bugs or light some texture maps. Don't count on it, but use it if you've got it.)
Let's see, what else: schedule tracking. Sessue Hayakawa was tracking the schedule but it didn't do him a lot of good: all he knew was that the bridge was late and becoming later. Schedule tracking is useless unless you take action when the schedule slips - it's hard to reduce scope when you're building a bridge (or filming a movie, I bet. Scope reduction is a secret weapon game developers have that I imagine most project managers don't get to use), so you either have to add resources or push your ship date back.
But sometimes it's not a shortage of resources but a shortage of resourcefulness. Sessue Hayakawa, by (finally) allowing Alec Guiness to take command of the project, made the resources more effective; also, Alec Guiness found extra resources by asking for volunteers from the infirmary to do some of the "light work." Sometimes necessity is the mother of invention, and you can get more done by simply requiring it, but I think what you typically will see happen then is like from the Michael Keaton movie *Gung Ho*, where you do manage to ship on time but it's crap.
When a ragtag band turns into a well-oiled machine, a la *Major League* and *Bridge*, it gets me every time. Haven't actually seen it in real life, though: all I've seen is a ragtag band become slightly less ragtag, and enough less ragtag than all the other ragtag bands out there that we still get work. But hey, that's something, right?