Reply #76: Lower or higher than expected votes for major candidates are very

[AirAmFan]

difficult to detect. We can detect unexpectedly high votes for minor candidates in ths case because their countywide total was way under one percent.

You're calling into question my simplifying statistical assumption that all wrong vote possibilities in a site were equally likely. Without that assumption or something equally strong, there's no way the problem of detecting miscounts and estimating votes lost to Kerry and to Bush countywide can be solved. So it's important to look for patterns in the data that would be inconsistent with any simplifying assumptions. I haven't found such patterns.