Train the Class, Not the Move

I tried a boulder problem last week that I’d first tried two years ago. In two years I’ve gotten a lot stronger and climbed a lot more, almost all of it indoors. I was barely any better at the one move that matters.

The problem’s called Watermelon. There’s a move into a right-hand gaston that I couldn’t do two years ago and can’t do now. The first move — a throw to a left crimp — went from maybe fifty-fifty to landing every single time. That one got better. Everything I’d done since transferred to it. The gaston didn’t budge.

That’s the interesting part. The standard advice is climb a lot of stuff and build a broad base. It works — for the move that was limited by strength. It did nothing for the move that was limited by skill, because nothing I did was aimed at it. You can’t get better at a thing you never practice, no matter how much fitter you get around it.

So the obvious fix is to go work that one move. Grind the gaston until it goes. I don’t think that’s right either. Drilling one exact move is memorising one rep. It’s solving a single maths problem and calling it understanding the concept — you pass that problem and learn nothing you can use on the next one.

What I think actually works: treat the move as a test, then build a class around it. Not one similar problem — five rough ones. A crimp above my head, a crimp at my shoulder. A heel close in, a heel far out. A good heel, a bad heel. Get good at all of them. The thing they share is the thing I’m bad at, and your body finds it without you having to name it.

This is how you’d learn a tennis backhand. Nobody drills one swing — every ball comes in at a different height and speed, so you hit a thousand different ones until the backhand itself is the thing you own. The specific shot takes care of itself.

The nice thing is you don’t have to be precise. You don’t need to correctly diagnose what’s wrong — that’s hard, and I’m often not sure. You just need the variations to feel like the move. Make five rough versions and the ways each one is wrong cancel out; what’s left in common is your problem. Volume buys you the precision you don’t have. If you’re really good you might nail it with one. The rest of us throw five at it and cover the bases anyway.

And there’s no way to lose. Either the gaston goes next time I’m on Watermelon, or it doesn’t and I got stronger at five adjacent things. The retest is the honest bit — it tells me whether I trained the right class or the wrong one. If I trained the wrong one, that’s a worse use of my time, not a waste of it.

This cuts against how most people seem to think about getting better, which makes me wonder what I’m missing. Maybe the reason everyone says “just climb more” is that diagnosing your own weakness is hard enough that broad volume is the safer bet. Could be. But I have footage, I know roughly what the move feels like, and I can make a decent guess at five versions of it. So next session I’m building the set, getting good at all of it, and going back to the rock to find out. The rock will tell me if I’m wrong.