But that's not true. August - End of Season 2004: .254/.348/.563/.911 2005: .237/.370/.438/.808 2006: .176/.299/.346/.645 I didn't say *each* of the last three seasons, I said *over* the last three seasons. Weight those three OPS numbers by plate appearances, and you get 797. Ergo the conclusion, over the last three seasons, Dunn's OPS for the months of August and September is below 800. Put it this way -- if we knew for sure that the Cubs would get one of those three numbers from Dunn (911, 808, 645), then the enthusiasm level around here would decline dramatically. You averaged them? I'd say that's not very intellectually honest. Your original point was implying that Dunn was not as good a hitter down the stretch. Given that time-driven splits aren't reliable to begin with, there'd need to be a consistent drop in performance for it to have any merit. That obviously hasn't been the case. To use an average is to put extra emphasis on the outlier of last year. As far as "trend lines" go, given what I just said about the reliability of timed splits, there would have to be some sort of cause for him to somehow be declining in his ability to hit after the deadline. Unless there's something you can point to that would cause him to heavily wear down after July, we have to assume that last year is an outlier. Not very intellectually honest? It was a quick way to boil down Dunn's late-season production over the last three years to a single OPS number. I surely could go to the trouble of tallying up the guy's hits, walks, HBP, TBs, etc etc and get the true OPS number, but it wouldn't be very different from the 797 figure. In fact it might be exactly the same. Nobody seems to dispute the notion that Carlos Zambrano is a notoriously slow starter, and a cursory look at his April stats over recent years confirms this. I don't hear anyone questioning the validity of those time-based splits. It's mostly just become an accepted truth. Zambrano ERA by month, 2004-2007: April 4.86 May 2.59 June 3.12 July 2.91 August 2.94 September 3.03 So why is there a problem with the notion that Dunn is a notoriously slow finisher, when a cursory look at his August and September stats over recent years confirms this? Dunn OPS by month, 2004-2007: April 1.070 May 0.852 June 0.949 July 0.981 August 0.836 September 0.760 August + September aggregate = 0.799 Wha? If I wanted to put extra emphasis on the outlier of last year, I would simply say, "Dunn had a 645 OPS last year," and leave it at that. Averaging that 06 figure with other years actually helps to hide how horrible he was last year. That would be de-emphasizing 06.