Update [2005-9-4 12:7:16 by Marc Normandin]: Clemens and Carpenter's VORP totals had not been updated as of the time of this posting, so here are the new ones: Clemens has 81.0 VORP, and Carpenter has 74.8. As far as the Willis/Carpenter issue goes, I think we may have found part of the reason for the WARP discrepancy. Willis has an offensive contribution of +6.3 VORP this season, while Carpenter has hurt his team to the tune of -7.0 VORP. Clemens has added +3.9 VORP along with his superior pitching this year. That does not account for the higher PRAA and PRAR totals by Willis, but it does help explain the WARP discrepancy (I think.) I'll keep thinking about this.
Chris Carpenter won his 20th game of the season last night, which makes Roger Clemens' chances at winning the Cy Young just that much slimmer. How much better than Carpenter has Clemens actually been this year? For most of the first half of the season Clemens was untouchable by every other pitcher in baseball (besides maybe Kenny Rogers, but that was BABIP induced) but Chris Carpenter has done his very best to play catch up and narrow the gap between them. With one month to go, the Cards and Carpenter on fire and Clemens seemingly cooling off due to injury, it is looking more and more like this is Carpenter's award. Let's take a look at a few numbers.
They are much closer than some are letting on. Besides in hits per 9 and homeruns per 9, Clemens does not have a sizeable lead in any statistic besides ERA, which could fly up with one bad start anyways. Carpenter leads in K/9 and BB/9, as well as IP and K's (not to mention less walks in over 20 innings more). Let's take a look at some advanced metrics to see what all of this means.
I threw Willis in figuring that if Clemens was thrown out of the competition due to wins that Carpenter would need someone to be compared to. In doing so, I now see that Carpenter just might be the one who needs to catch up to Willis. It is a good thing he won 20 games at the beginning of September, because Dontrelle Willis is directly behind or ahead of Clemens in everything except VORP, which for some reason he trails both pitchers by a decent total.
His peripherals are nowhere near as impressive as his advanced metric statistics. Maybe that accounts for the lower VORP total, while the WARP and high PRAR totals are due to actual results. Carpenter appears to be the better pitcher, but for some reason doesn't register as that when compared to Willis. Carpenter's homerun rate cannot be dragging him down that much, even though it is the only statistic he is significantly worse than Willis in. His DERA is higher at 2.99, as compared to Willis' 2.96, but that does not account for a 1.7 WARP discrepancy. I'll have to admit I'm completely baffled at the moment. By the way, Clemens is still the best pitcher of the three according to WARP and VORP, although in simply looking at peripherals he is not that far ahead of the pack. Much could change over the next month; I'll have to pay close attention to how Carpenter and Willis win their games in order to determine why it is that Willis is where he is in comparison to Carpenter even though it does not look like he should. Any thoughts?