Last week, someone asked me a question:
What percentage of wins do you think you should have against your ideal sparring/training partner in order to improve? Do you want someone at your exact level, someone better than you to beat out your bad habits, or someone worse in order to test things?
Great question! The answer is, of course, that there is no ideal training partner, for exactly the reasons stated in the question itself: you need different people so you can practice with different mindsets. If you don’t have a variety of training partners at different skill levels, you won’t be growing and improving as quickly as you should be. There’s a concept that shows up a lot in martial arts training called “gradual resistance” that I think is highly relevant to fighting games as well.
Experienced martial artists know that learning a new technique isn’t a one-time event, it’s a gradual process. You start by learning the basic series of movements that compose the technique or combination, and you practice that against a cooperative partner until you can do it without having to think about it step-by-step. Next, you get your partner to slowly escalate their involvement in the drill; at first they might just put a little bit more weight on you, then they might actively resist your technique at 40% effort, and so on. After that, you try the move out in live friendly sparring, and then eventually you feel comfortable enough to do it in a tournament. Only once you’ve executed your move against a resisting opponent in a competitive context can you really say you “know” the move.
The idea here is essentially that when you’re learning new tech under gradually escalating levels of resistance, you’re learning different things about the tech. First, it’s just working out what you need to get your body to do, but eventually you’re learning how you might need to wait until your opponent’s weight is in just the right place, and the kind of speed and timing you’ll need to successfully catch someone with it, and then you’ll start recognizing common answers to your move and figuring out how to counter them. If you skip a step, you’ll have an incomplete understanding of the move, which means you’re not going to have as much success with it in live competition — and that means you probably won’t use it much, because you think it sucks.
The same applies for fighting game tech, whether it’s mastering a setup, or a combo, or the ideal spacing for a poke, or even a new character. Starting out, you learn how to do it in training mode, then you try it out against the CPU, and then against gradually better and better opponents until it becomes tech that you’re confident in using as part of your Real Game because you’ve vetted it against a range of players. If you bring a strategy straight from training mode to a set against your local top players, it’ll probably get exposed so badly that you’ll just think it’s junk and not use it again — but that’s not necessarily fair to the tech, because you haven’t spent the time trying it against other people to find the best possible version.
So when you’re playing against people you can beat for free, use that as an opportunity to try new stuff out. When you’re playing against people who blow you up, use that as an opportunity to get your A-game ripped apart so you know what you need to fix. And when you’re playing against your rival — the Ken to your Ryu — take the tech you’ve been working on for a little while and see what you need to do to get it into your A-game.
Of course, if you don’t have a variety of training partners to go up against, you’ll have to make do. Try introducing the concept of gradual escalation into your practice sessions by playing against each other in training mode for 30 minutes specifically to test your new tech before diving into Vs. Mode. And if you have other tips for experimenting against opponents at different skill levels, share ’em in the comments!