Per Curiam.

Respondent, Alfred E. Santangelo, and petitioner, Herman Badillo, were contending in the past primary elections for the party position of Democratic District Leader, Male, in the 16th Assembly District, Part A, New York County. Santangelo appeals from an order at Special Term confirming the report of a Special Referee and directing that a new primary election be held for this position.

The essential facts are not in dispute. Voting machines were used and there were four inspectors of election, two Democrats and two Republicans, at each board. The part of the Assembly district in which the two parties were candidates comprised 15 election districts. The vote at the close of the day showed 1,472 votes for Santangelo and 1,397 for Badillo — a margin of victory for Santangelo of 75 votes. This result was confirmed by an official recanvass of the machines taken by the Board of Elections.

Thereafter, an examination of the records at the Board of Elections by petitioner revealed that 90 persons not registered as Democrats had been permitted to enter the voting booths reserved for the Democratic primary. Badillo does not charge Santangelo with fraud, nor with in any manner procuring or encouraging registered Republicans to vote in the Democratic primary. Badillo had watchers at each polling place; and there is evidence that Badillo's adherents in one instance caused one nonregistered person to vote and in another unsuccessfully attempted to persuade the Election Board to allow another nonregistered person to vote. (This vote was not included among the 90 in dispute.) There do not appear to have been any other unusual incidents, nor any claims that the entire election was permeated by fraud.

There is no evidence as to whether any of the 90 non-Democrats voted for Santangelo; nor, for that matter, as to how many voted for the office of District Leader, Male, at all. The absence of such proof must be viewed in the light of the fact that 3,282 persons were recorded as entering the voting boths, but 413 of that number did not vote for the position of male leader.

In the three election districts in which no irregularities appeared Santangelo won by a vote of 83 to 15. In the two election districts where 40 of the 90 nonregistrants voted, Badillo won.

It will be recalled that no claim is made of fraud in procuring the nonregistered persons to enter the voting booths. Therefore, it is appropriate to consider how many of these voters who were not influenced improperly could reasonably have been expected to have voted for Santangelo. As we compute it, only if Santangelo had received 83 of the 90 invalid votes would the results of the election have been changed, as in such event Badillo would have had a margin of one vote of those qualified to vote.

Bearing in mind that Badillo secured almost half the votes cast for male leader throughout the electoral unit, and that 413 of those persons who entered the voting booths did not vote for that position, it is highly unlikely that 83 of the 90 disputed votes were cast for Santangelo. An election will not be overturned upon a mere mathematical possibility that the results could have been changed, when the probabilities all combine to repel any such conclusion ( Matter of McGuinness v. De Sapio, 9 A.D.2d 65). Neither the statute nor this court's opinion in the McGuinness case projects such a strained and unrealistic formula to the practical affairs of politics.

Accordingly, the order appealed from should be reversed, upon the facts and the law, and the petition should be dismissed, without costs.

BOTEIN, P.J., BREITEL, STEVENS, EAGER and STEUER, JJ., concur.

Order entered on January 29, 1962 unanimously reversed, upon the law and upon the facts and the petition dismissed, without costs. Any findings of fact and conclusions of law made by Special Term herein that are inconsistent with the Per Curiam opinion of this court are hereby reversed.