How risky is it to manipulate a scoring rule under incomplete information?

This paper is concerned with the manipulability of common voting rules, in particular with the Borda rule and other scoring rules. It is shown that, if one deviates from the assumption of complete information of the voters about the preference profile in the slightest possible manner, the Borda rule may "punish" a manipulator in the sense that the manipulation might lead to a worse outcome for the manipulator than had she told the truth. Voting rules showing this kind of behavior can be considered to be more manipulation resistant than other voting rules especially if we think about sufficiently risk averse voters.