Reply #26: Anyone care to check my math? [View All]

[MarkusQ]

If...

• It takes someone about 3 minutes to vote on average (call this T_v)
  
• Typical lines at polls ran 0..4 hours (call this T_l)
  
• The polls were open for 12 hours (call this T_o)
  
• Anyone in line at the time the polls closed got to vote, so the effective open-time was T_o+T_l
  

A single voting machine could handle

N_v1 = (T_o+T_l)/T_v = (12..16 hours)/ 3 minutes * 60 minutes/hour = 240..320 voters

Similarly, precinct2 with more than one machine could handle

N_v2 = 480..640 voters
N_v3 = 720..960 voters
N_v4 = 960..1280 voters
N_v5 = 1200..1600 voters

If any of the numbers are off (say, it takes more like five minutes to vote instead of three) it would affect the exact values, but not the general pattern. The key part of the pattern is that there is a gap between the ranges for N_v1, N_v2 and N_v3, meaning that if a smaller precinct (with less than four voting machines) is "resource bound" it should show up as a quantization of the total number of votes cast.

Thus it looks like it should be possible (at least in theory) to detect if resource constraints were used to "rig" the election by looking only at the total number of votes cast, per precinct, and then looking at the same data split by party.

--MarkusQ