Actually, your observational bias isn't as bad as that. It's true that a majority of Arcia's HR have not come with the game essentially out of reach. But that's because relatively few of his plate appearances come in such conditions. So the counting stats don't come out as extreme, but the rate stats do: From B-R.com, and not counting tonight's game: Lead within 4: 231 PA, .203 BA, 7 HR, .653 OPS Lead > 4: 37 PA, .351 BA, 4 HR, 1.130 If he could get a full season of PA at the rate of that second line, he'd hit 75 HR or so. Needless to say, the first line isn't a very big sample size, and the second line is woefully small. League-wide, the rates in such pairs of lines tend to be more close to each other, and likely with more chances Arcia's numbers would trend more like that too. But, if one is remembering very good performance from Arcia this season (and to a lesser degree last season) in blowout situations, it's not observational bias per se. Unless your observational bias is that the Twins are hopelessly behind in 50% of their plate appearances. Source: http://www.baseball-reference.com/players/split.cgi?id=arciaos01&year=2014&t=b