I noticed that some of the tryouts this weekend cut very mechanically. They might fake, but then their actions do not depend at all on what their defender does. They don't even appear to be watching the defender at all.
Idris talked about this in his Oh you had 'em blog entry (Idris has this annoying practice of making everything point to idris.org/ultimate, even when you link to another blog from his page). Players just don't seem to realize when they're open. I commented there, "Bad players either plan too many fakes or else they get so caught up in trying to read the defender that they misread him. Instead, do a simple fake, expect that it's going to work, but be ready to do something else after 2 or 3 steps if you see that it hasn't worked."
Maybe the way to drill this is to have them watch real cutters and defenders and attempt to identify the exact moment at which the cutter simply needs to move in a straight line to get open. Anything a cutter does after that is at best inefficient and at worst the first part of a miscommunication turnover.
We sometimes say that a cutter has several seconds and several options before he has to clear, but that includes the setup time, which should occur before the disc is live.
1. The setup. As the disc is in the air to the new thrower, the cutter moves into position and might do a little bit of juking, but is basically trying to force the defender into a repositioning error.
2. The cut. Make a final sell and then go hard in one direction, making a commitment. THEN you evaluate whether you'll be open. When you get good, you'll know as you're making that hard move whether or not you're successful. If not,
3. The 2nd cut. Turn 90 or 180 degrees and go hard that way. If that's not open, clear. The only exception is when you're in an iso situation with a lot of field and the defender overcommits to the 2nd cut, and you are 100% guaranteed to be open in a good place if you return to your original direction.
I guess basketball players work on their fakes by themselves, repeating until they've internalized the sequences, but they will still need the feedback of whether their defenders are going to buy the fakes.