A common
logical fallacy where someone makes a
statistical analysis based on only
one datapoint. Particularly when that one datapoint is the person making the
statement!
This is a problem because a statictical analysis is more valid with large sample sizes. But you could do worse! Consider using completely made-up data or using data you know is wrong.