Pruning the decision tree - strategic mind set in Armizare.

Oh Decision tree, with thine barren
 and thorny branches, bleak are your
heights, for in you is a forest of 
blows, too many choices, and you
 are home the balorous squirrel,
 Hesitation.

Decision trees.
What are they? What do they have to do with sword fighting?

My instructor recently reminded me that if I am not acting with a plan, I am just cutting pretty patterns in the air. He is not wrong, we need a plan of action in order to close in and land a hit on the other person.

Some say that your plan should keep you two or three moves ahead of your enemy, kinda like a good chess player. My only beef with this "chess" analogy is that chess is not played in real time, the players take turns, so initiative is fixed. Not so in a sword fight; initiative is dynamic and can change hands frequently.

Also, I cannot wait for a few moves to occur before I see how the fighting is shaping up.  A fight develops much faster than chess does. I need to know what the enemy is going to do, and I need to know it before I get hit.

Some others would say planning begins when you meet you opponent before your match. While its all well and good to suss out those details (lords know I have), what I am really talking about is the strategy we employ once the fight is joined. IE, we have turned OFF the targeting computer, and are listening to the force. Yeah, yeah, I know we're not Magic Space Wizards.

My A.D.D. mind has assembled a slightly different analogy, which can be applied to not only sparring, but to various most of my training models.

Let's call this analogy pruning the decision tree.

We begin with the problem: I am trying to solve is that I want to hit my enemy without being struck myself in return.

In my fight, I am not exactly planning ahead. We don't know what the person in front of us will do (not magic future reading space wizards), so we have a large number of potential decisions we may need to make. Too many, in fact, as our foes can do lots of stuff and strike from many angles.

Or can they? Position and range heavily limit what your foe can bring against you at any given moment. The wider the range, the more potential options the foe has, and one of those options is decreasing the range to the next tier. Mind you, though, range can be so great as to make your enemy momentarily irrelevant. If the foe cannot make a threat that means anything, ignore it.

The cone shrinks as we close, reducing the number of meaningful things that can happen, but also robbing you of reaction time, so be careful not to take this visual out of context. If range gets to 0, and you have not won, you are likely dead. 

Position is the other variable control. We cannot know the future, but we can know the current position and range. This is where the pruning begins. Since we can know the targets current position (guard and foot alignment), we can know their options that flow from that position.

If we adjust our position to potentially counteract the things the enemy's position can do, already our decision tree has lost 3/4 of its branches, since those options are divisible into the quadrants through which the first movers sword will pass.


By knowing the enemy's options, and adjusting our own position accordingly we have gone from "every thing can happen" (within the confines of weapon, and range) to "these things can happen", this being 3-5 things that any given position can manage.

Now, by choosing our next action (it does not matter really if you are first or second mover at this level of the discussion) to cover the enemy's shortened list of options, we limit that enemy to 1-2 meaningful actions. (we have to remember that some of the things that can happen, such as getting hit and dying, the enemy abruptly leaving measure, or killing your foe don't matter to the next layer of the process, because the process is halted.)

Instead of trying to remember hundreds of if-then statements that our training has driven into our lizard brains, you have neatly boxed the target's choices, and you only need to have one if-then line in your cued up code. We can handle that, right?






In Summary,
1. If you are not paying attention, anything can happen.

2. By maintaining an effective counter guard, or going on a planned offensive, you can reduce the number of potential out comes.

3. after the first action begins, there will only remain one.  or two valid outcomes, and with training, you can use that foreknowledge against your foe.

4. I deliberately did not talk about initiative. This is because if you are planning well, and reacting in real time to your opponents position, you can choose who has it. One can move first,  or by choosing to move second, assuming you can control your measure, you keep the initiative after a fashion. Also, the model work on both side of the coin, in terms of initiative.

- every move you make is building a coffin, lets just hope the right person goes winds up in it.

Jesse out.