Tuesday, December 09, 2008

More zone 2 chess analysis

In my continued attempt to connect some of the techniques in chess strategy and tactics to ultimate, I've evolved the model to the next step. This time, we'll use the chess board and the players on the board to evaluate the merits and flaws of different zones. This seems reasonable based on chess being a warlike simulation.

Figure 1: Shows a 3-3-1 zone.

Figure 1, shows an oversized chess board (that attempts to represent a field) with kings in a representation of the 3-3-1 classic cup zone. The red arrows indicate which throws a king can stop, and the transparent red dots show what discs a defender could, likely, get too. The radius of this distance decreases as the player gets closer to the disc (the deep could arguably have a larger radius).

The interesting thing is the figure shows some of the weak spots in this type of zone. These weak spots are clear squares on the board. Also, we can see the concept of a loose piece. The right mid has no immediate support meaning they can be attacked by a fork.

Figure 2: a 1-3-2-1 zone.

Trying the same thing with a zone we will call the 1-3-2-1 (wall like structure), Figure 2 shows what might be weaknesses against this zone. It is clear that the additional players in the front makes it harder to break the cup as more of the squares around the cup are doubly covered with overlap.

The clear squares show there is some weakness over the wall, but three defensive players approach those points (meaning this will need to be a fast throw). Also, it seems clear that a swing across the field will be hard, but worth a serious gain in distance.

Figure 3: 3-3-1 in a arrow formation

Next, Figure 3 shows the 3-3-1 in an arrow head formation (know as chinch, arrowhead, etc.). Here, it appears that the middle is completely covered, which isn't really the case. Is the model failing? Partially, but this is a static position. Most teams would use this zone emphasizing that they don't want the disc to go around the outside and they want the attacks in the middle. This means that the areas covered would favor one direction over the other, and this would show lower probability to cover short attacks through the middle.

Based on my early view, this tool seems to be reasonable in evaluating zones. I'll try to refine the model over the coming weeks.



rich(ard) / dean said...

Genius. Modelling Ultimate with a chess board: brilliant!

For what it's worth, I agree that the model breaks down with the third example. I think this is because the pitch needs to be wider compared to the 'players'.