The Direct Attack drill has been one of my favorite games since we have started doing game-centric classes at Bucks, for several reasons, some of which I have stated in previous articles. However, as I have been thinking about the ideas of specifying and non-specifying information, I think it’s time to revisit this long used game, and see if we can do better.
The Direct Attack drill is a very constrained game with very little choice available for each player, and as such it hones in on a few specific things. The first is striking mechanics, your opponent is trying to parry you before you hit, which affords a fast, efficient strike. I also use this to teach the ballistic passing step footwork, but you don’t have to do that, you can do it with a lunge or other footwork if you want. The second is distance, you learn that you have to be quite close in order to have a chance of landing a direct attack. The third is timing, if you get into a predictable rhythm of striking, your opponent will parry you every time even if you are very close. The one I want to fix here is distance, and for that I need to talk a little bit about specifying and non-specifying information.
I talk a little bit about specifying information and control laws in this article, but I want to briefly go over it again with a different example that I learned from Rob Gray. Imagine you are trying to cross a street. You see a car coming 200 yards away, is the street safe to cross? This question is not possible to answer without more information, we need to know how fast the car is going, or more specifically the time it will take the car to reach us. The distance in this case is non-specifying information, because we can’t use it alone to decide whether or not it’s safe to cross.
Now imagine that you are trying to cross a street where all cars go exactly 25 mph all the time. In this case, you would be able to use the distance of the car to determine if you can cross, because all of the cars will have the same time to reach you from the same distance. However, if you ever try to cross any other street where cars move at different speeds, you will be in trouble if you use the same perception strategy you had been using on your 25 mph street.
Let’s take this information back to the Direct Attack drill. Usually when we play this game, like with all games, we do ten reps with the same partner, then rotate, usually in groups of three. So you end up getting 20 reps as the attacker and 20 as the defender, with two different partners. Usually what happens with people who are experienced at this game is they will find their ideal distance within a couple of reps, then alternate hits and misses between a longer and a shorter distance. What this could potentially allow people to do is get used to the distance that they can hit this specific partner, and train themselves to use that perceptive information in order to act in other games or in sparring. The Problem with this is, similar to the car analogy, distance is non-specifying, when you fence different partners or in different circumstances, the exact distance at which you can attack will vary. This may be mitigated by the fact that you’re doing it with two different partners, but it’s still a lot of blocked practice in a row with the same partner, so let’s see if we can do better.
The solution that I tried was doing the game “king of the hill” style instead of the previous ten reps and rotate style. You are still in a group of 3 or 4, one pair does a rep of the game, whoever loses steps out, and whoever wins stays in. The attacking side still changes distance as normal after each rep, whether the attack switches out or not. The idea is that in this format, every rep will be different, with a different partner, at a different distance, and whether you are attacker or defender. We tried this in class, and I was happy with the results, this is probably how we will play Direct Attack drill from now on. The mechanical optimization aspect of the game is still preserved, but now it is harder to use the distance as a perception strategy since your opponent is always different from rep to rep. Additionally, this format is suited by the self scaling aspect of the drill, because it prevents one person from staying in for too long. This tends to happen with other activities when doing king of the hill style rotation, which causes more skilled and experienced players to get more practice than players with less experience.
I may start mixing in the king of the hill format with other games as well, but I will use it sparingly. The format allows for more variation from rep to rep, but in doing so does not afford adaptation – which is a feature in this case, but a bug in other cases. With more open games that have fewer constraints and more paths to success, I want to encourage adaptation and exploration of the limits of the rules, and for that you need to do multiple exchanges with the same opponent. In games like that, using specific distance as a perception strategy is less of an issue, because there are often many different ways to attack and defend.
Overall, I think this format improves the Direct Attack game and makes it more useful long term. I may still have first timers do it in a regular blocked style in order for them to get their bearings and develop the mechanics of the action, but after that we’ll probably always do it king of the hill style. I think there is still an important distance lesson to be learned from this, but it’s a general lesson, essentially “you need to be closer than you think to do a direct attack.” This is a lesson that you will learn regardless of the format of the game, as it is a general idea, and not any overly specific distance.