[sanddrag]

Originally Posted by Colin P

Have a mentor modify a version of the code to go berserk at random. Don't tell the drivers, just say that they need to do a speed run of something and it will be timed. Eventually, the robot will stop working correctly, and they'll be distressed when their time is compromised. This will help to see how they react under pressure and when things go wrong.

My team is using mecanum drive, we're modifying the code to drop the tube whenever the robot strafes to the right. We put a number of strafing obstacles on the course so it will eventually happen, and they will have to adjust to it.

If you do this and you find out that your driver can't handle it and has a fit, it's better now than in front of thousands of people in a match.

Wouldn't the time better be spent to minimize/eliminate failures, rather than preparing for the expectation that it will fail?

There's always a chance something wierd will happen, and who knows how (or if) you'll be able to react to it. To me, it seems more beneficial to practice the actual intended functions.