The other factor that could decide the Senate contest in Brooklyn

Although both sides have declared victory, the race to replace State Senator Carl Krueger remains too close to call. The preliminary results released by the Board of Elections show David Storobin with a 143 vote lead over Lew Fidler. While several news outlets have focused on the 757 absentee ballots which remain to be counted, recent anomalies with New York's voting machines show that in a very close contest, only a hand count can ensure that the right contestant has won.

Just last month, the results of an upstate City Council race declared to have been won by challenger Augustine Beyer by a single vote were overturned after a full hand recount revealed a two-vote discrepancy that tilted the race in incumbent Richard Slisz’s favor. The voting machine was unable to read one improperly marked ballot where the voting oval had been circled rather than filled in. The hand inspection was enough for election officials to determine that the voter's intent had been to vote for Mr. Slisz. Perhaps more troubling however, was the second ballot with a vote for Mr. Slisz’s that was never scanned or registered by the machine at all.

Furthermore, the results from a Daily News investigation into the exceptionally high overvote rates the Brennan Center uncovered in the South Bronx indicate that these voting machines are far from infallible. The Daily News found that one of the machines used to scan ballots in the South Bronx made errors in reading nearly 70 percent of ballots during the September 2010 primary.

In our report analyzing overvote rates in New York, we listed one of the precincts in Mr. Kruger’s district — AD 46, ED 051— as having Brooklyn’s 8^th highest overvote rate in 2010. Many other precincts in this senate district did not provide any data at all, but given the demographics of the district, it seems likely that there were other precincts with high overvote rates both in 2010 and 2012. In some of those cases, voter intent may be clear to the human eye, but not a machine.

Unfortunately, New York City does not publish the number overvotes — as is done in Rockland County— making it virtually impossible for anyone outside the Board of Elections to identify areas where voting machines have registered high rates of uncounted votes.

A provision adopted by the City Board of Elections requires a hand recount of paper ballots in contests where the margin of victory is less than 10 votes or half a percent of the total votes cast. Given the newness of these machines and recent history, even a margin slightly higher may warrant a careful hand recount to ensure that the actual winner is declared the victor. If a recount does happen, look for totals (and maybe even the declared winner) to change.