Just eyeballing the numbers, I’d guess that the crowd may have done a bit better at the top of the chart than the AAV difference suggests – because in signing players who are mostly heading into their declining years, a higher AAV for a shorter contract is a reasonable trade-off. The Red Sox signings of Drew and Dempster, for example, don’t really seem to be above expectations at all.
Average error over all these projections isn’t a good indicator of success. Only nine players actually had a projection that was within eight percent of the actual contract. If you compare the average error between Lance Berkman and Michael Gonzalez you have close to a 0% average.