Well, in my personal experience, low scoring defensive robots can work very well. In team 25's history we have focused alot on defense. In 2000 we had a bot that reached into the other teams goal and stole the balls away and in 2003 our sole purpose was to push robots out of the way. Last year we had changed our theory to offense, and it came out poorly during the season. However, we did manage to get to two quarterfinals in regionals. In our first offseason, we had a major problem with our arm and could only drive around. So we decided just to play some good defense. We end up winning 5 qualification rounds just playing defense, and then going to a thrilling 3 match quarterfinal vs 222 and losing by just one ball(ironically having no offense came back to hurt us
). So we decided to keep this strategy for the rest of our offseasons. We ended up winning 3 and placing 2nd in the other, playing pure defense.
We did have help from offensive robots though. So it really depends on the game, and on the balance of what is now a 3 robot alliance. In the past games you could get away with one main offensive robot and one defensive, now, I think you will have to blend the robots better. A pure defense with no manipulator, may not be very practical, because some teams will be able to break through the defense easily. It is basically vital to have something to move the tetras around this year, no matter where you place them.