The three candidates in today’s Ward 5 ward councilor race were separated by 54 votes first (Bill Humphrey) to third (Rena Getz), out of 1413 votes cast. The difference between first and second (Kathy Winters) was just 8 votes. And, second and third were separated by 46.
Looking at the results of ward aldercritter/councilor races from 2001 on, that’s historically close (though there aren’t a ton of races to compare). In the 2013 Ward 1 preliminary, Alison Leary led with 434 votes, beating second place Allan L. Ciccone by 171 votes and third place Roger Marrocco by 331 votes. In 2001, Stephen Linsky won his Ward 2 preliminary with 369 votes, 161 more than second place Robert Antonellis and 185 votes more than third place Elizabeth Salvucci. Also in 2001, Pauline Bryson got 922 votes, 248 more than Jay Harney and 886 more than William Aldrich in a Ward 4 preliminary. Note, not all wards are the same size, so tough to compare the absolute vote totals ward-to-ward.
The turnout figures are not available yet, but we can do a little estimating of magnitude. In 2013, there were 1249 votes cast in the Ward 1 general election race v. 791 cast in the preliminary: preliminary was 63.3% of the general. In 2001 in Ward 2, there were 761 votes cast in the preliminary and 1695 in the general: 44.9%. In Ward 4, it was 1596 in the preliminary and 2541 in the general: 62.8%. (The general election required a recount and Bryson beat Harney by 27 votes!) Both 2013 and 2001 were mayoral years, which increases the vote totals in the general election. There was an uncompetitive mayoral preliminary in 2013 (6,139 votes cast v. 15,841 four years earlier).
For comparison, there were 2346 votes cast in the last contested Ward 5 general election race — 2009, John Rice v. Adyanthaya, a mayoral year — and 1620 votes cast in 2007 — Bill Brandel v. Christine Snow-Samuelson, a non-mayoral year. If 2007 is a good indication of what we can expect for general election turnout for a contested Ward 5 ward seat race in a non-mayoral year, the turnout for today seems pretty high: 1413/1620 or 87.2%.
Obviously, the sample size is wicked small. And, who knows what other factors come into play. You go to battle with the numbers you have, not the numbers you wish you had. Take from the analysis what you will.