I love visual proofs, but the problem with them is that they rarely work without some explanation (and the ones that seem to work without some explanation are o

I love visual proofs, but the problem with them is that they rarely work without some explanation (and the ones that seem to work without some explanation are often those that re-express theorems you already sort of understand). What the visual proof does is eliminate excess words so that you have multiple streams of information. For instance, I have published a cool visual proof of the amortised analysis of array resizing,[0] but even in animated form it would take at least a little bit of explanation for someone who didn't already know the basic proof to understand what's going on with it. The elegance of it is that I only need a few words to explain what the visuals mean. [0] The long abstract is at http://www.blahedo.org/papers/dblaheta-iticse09-amort.pdf http://www.blahedo.org/papers/dblaheta-iticse09-amort.pdf and the slides from my presentation---which from page 4 or so serve as an animated proof-with-few words---is at http://www.blahedo.org/papers/dblaheta-iticse09-amort-talk.pdf http://www.blahedo.org/papers/dblaheta-iticse09-amort-talk.p...