I'd read that as 'pick your battles well'. You can work hard at things and get nowhere, you can work hard at other things and make a ton of progress, and these

I'd read that as 'pick your battles well'. You can work hard at things and get nowhere, you can work hard at other things and make a ton of progress, and these things can sometimes be estimated beforehand. You can keep churning out spaghetti code (hopefully the uncooked kind) or you can learn and apply better methods, for example.