The following image compares my bootstraps for the b1 (h_c) data to Jim's. We were using the same priors here, so points should overlap. Unfortunately I don't have errorbar information for these points.

Sorting the bootstraps (first 100 of Jim's) and plotting the central 68% gives the following:

The effective mass plots of the local, 1S, and 2S smearings for the h_c. Dark blue, red, and green are the local, 1S, and 2S respectively. Lighter variations are where the effective mass could not be determined (would require the log of a negative number).

A version with just the local and 1S....

I'm somewhat suspicious that the above bootstrap comparison plot wasn't to the same data, same priors, or something, because all other tests are very good. Here are examples:

The following plots show what happens to the chi2 when you vary each parameter. The x-axis in each case varies the parameter by 3 sigma (chi2 errorbars). The red line is the prior, the blue is the data, and black is their sum. These were done with my choice of the priors (so relatively narrow).

First, the a 3-state fit to the b1 local-local data: