I’ve always had a problem drawing wheels. I eventually tracked down the problem: my understanding of how circles are effected by linear perspective was flawed. I had studied the problem, but when I applied what I thought I knew, my wheels never turned out quite right. No amount of practice with my flawed principles would have made the wheels I drew look correct. There were only two ways out of the problem for me:
1. Discovery: Over time, I may have stumbled upon the correct principles all on my own. To do this I would have to have made better observations and experimented with any theory I constructed to see if there were any improvements.
2. Learn from someone who has an understanding of the correct principles: This is what I actually did. I’ve talked about it before, but Marshall Vandruff’s DVD series on perspective is pure gold. I listened to the instruction, then experimented with the a few drawings to see if my perspective improved, which it has.
Practice without experimentation is merely repetition. If you want to improve on something, you have to try to do it differently somehow than you did it before. That is why experimentation is so important.
Here are a couple of my latest experiments:
Gannon Beck's Blog 