strategy – Julian Wise

This post requires knowledge of the game Starcraft: Brood War. For those unfamiliar, skip to the bottom for an AI-assisted translation.

Even if you’ve studied the deepest recesses of a bad strategic pathway, it’s still a bad strategic pathway.

In season 10 of the Starcraft Broodwar ASL tournament, Protoss player Bisu cannon-rushed his Zerg opponent, Shine. The Zerg had fast-expanded despite the mineral structure in the natural expansion making cannon rushes incredibly difficult to block and Bisu being literally the person who invented the forge fast-expand (so he’s almost definitely going to have a forge!).

After the cannon rush kills Shine’s expansion, Shine switches into lurker drops, where he sneaks two lurkers into the mineral lines of Bisu’s main base’s before Bisu has detection.

Bisu micros for a hectic ~4 minutes, but holds this counter and wins.

The kicker: maybe Shine intended this all to happen. He’s the sort of player who finds a weird timing and takes his opponent down this unusual pathway. The sort of player who would “bait” Bisu into attempting a cannon rush so he could catch Bisu off-guard with the lurker drop.

But even in that case, Shine would have had to throw away 300 minerals, his first set of zerglings, and a small handful of drones. Not worth it!

Don’t get yourself into a bad position because there’s a chance you can win it. Get yourself into weird positions because they’re actually superior.

—

AI-assisted translation:

Two players face off. One of them, Bisu, is the most famous practitioner of a particular safe, standard opening — an opening that involves building a defensive structure early. Everyone knows this about him. It’s his signature move.

His opponent, Shine, chooses to expand his economy early anyway, on a map where that expansion is unusually hard to defend against exactly the kind of aggression Bisu’s signature structure enables. It’s like dropping your guard in a duel against the man who invented the thrust that goes exactly there.

Bisu does the obvious thing. He attacks the expansion and kills it. Shine loses the investment, the units he’d built to protect it, and part of his workforce.

So Shine pivots. He sneaks two units into the heart of Bisu’s home base — units that Bisu, at that moment, literally cannot see, because he hasn’t built the thing that lets him see them. For about four minutes, Bisu is fighting blind against something that can kill his entire economy. He survives it by hand, one frantic decision at a time. Then he wins.

Here’s the interesting part: Shine may have wanted all of this. He’s known for finding strange timings, for dragging opponents onto unfamiliar ground. It’s completely plausible that the undefended expansion was bait — that the whole point was to make Bisu commit to the attack so that the counterattack would land on an unprepared base.

And even if that’s true: the bait cost him a large chunk of his economy, his first defensive units, and several workers. He engineered a situation where he had to play nearly perfectly, against one of the best players alive, just to get back to even.

A trap that costs more than it catches is not a trap. It’s a loss with a story attached.

Don’t take a bad position because you can imagine winning from it. Take a weird position because it’s actually better.

In which chair-sitting is frog-boiling.

A coworker once taught our company how to sit in a chair. The problem: humans are very adaptable. So when we sit in a chair, we adjust our bodies to fit the physical circumstance. This is bad. We should instead adjust our circumstances to fit our bodies. (The desk doesn’t care if it’s adjusted to be higher or lower. Our bodies do prefer we don’t slump.)

The rules of the game change your play. That sounds obvious, but its effects often go unrecognized.

Take a simple rule – like the football rule that the clock stops when a player runs out of bounds – and imagine the changes to the entire game that could result. Obviously the end of the game is faster: more hurry-up plays, less pre-defined set-ups.

Now consider how different technology was when this rule of this game was established – at some point before 1909 (citation: pg 214 here). Was this rule intended to play out the way it is? No – no way – not really – it can’t be. But it shapes how today’s entire game is played.

We often accept the slight changes in our environment, in the rules that govern our games. But adjusting our behavior to maximize our desired outcomes is not easy. Do you think second-order effects (Since A happened, B will happen) are hard to predict? Third-order effects (A, therefore B, therefore C) are even harder!

Eg: If we changed the clock-stop rule, would Quarterbacks make more in-the-moment decisions? Become more skilled at rapid decision-making? Would we select for quarterbacks who are more tacticians and less strategists? Would that change lead to the rest of the team being more strategic (to fill the gap) or less (because their leader is less strategic)? Is this even the right pathway to follow, or would quarterbacks actually become more strategic because they would plan their whole series of plays ahead of time for those low-on-time situations? Would timeouts become so incredibly valuable in the endgame that they’d never be used otherwise? How would that impact how strategic a quarterback needs to be?

It’s really, really hard to tell. Those who can see the second-order effect in a very complicated situation are often highly-prized experts.

A chess grandmaster can sometimes see 10+ moves ahead. On the other hand, one former chess world champion is commonly crediting as saying, “I see only one move ahead, but it is always the correct one.”

Which would you rather do? And in what areas?

–(Oh, and GO BEARS!!!)

Tag: strategy

Bad-with-possibility is still bad

Clown School Break Day 50: Seeing Ahead