One of the many appealing aspects of the game of baseball is its rich statistical record. By looking at a box score, a fan can recreate what happened over the course of a game without having seen it. By looking at seasons of data, players from different eras can be compared and bar-stool debates become all the richer.
Over the last thirty years, through the development of sabermetrics, our understanding of the statistical record of baseball has been advanced, and the analysis has become more granular. RBI and pitcher wins are less meaningful numbers, while FIP and BABip, among others, have been added to the lexicon and elevated in status.
With the reams of available data and the development of these new statistics, the predictive nature of player analysis has become much more refined. Now a variety of programs can, with some degree of accuracy, forecast each and every player's performance based on their past history and that of other similar players. Certainly there will be outliers, but in general these player forecasting models do a pretty accurate job of projecting performance.
All of which is a roundabout way of getting to Pittsburgh Pirates starting pitcher Jeff Karstens. As Karstens rolled through the Houston Astros on his way to an 83-pitch, five-hit shutout last night, a win that moved the Bucs into first place in the NL Central, the oft-seen debate again flared up on Twitter. Is Jeff Karstens really this good or is this just an incredible hot streak which too shall pass? The two positions are pretty clearly delineated:
Smoke & Mirrors:
ESPN.com baseball analyst Keith Law has become the evil personification of this position in the eyes of many Pirates fans. During last night's game Law tweeted this, this and this among other things.** Law's opinion is based on history and scouting. Certainly the 28-year old Karstens' major league performance coming into the season did not portend this type of success. In 50 starts and 89 total appearances he posted an ERA of 5.07, an ERA+ of 83, and a WHIP of 1.441. He had a K/9 of 4.5, not exactly Gibsonian, and a BB/9 of 2.8. Throw in his 12-27 W/L record and "Cy Young candidate" does not jump off the page.
**Law didn't endear himself to Pirates fans just before the All-Star break either. When asked if the Pirates winning record was for real, he tweeted that no, they were 38-40 against all teams not called the Astros. Fans rightly pointed out that he was cherry-picking his stats, ignoring the Pirates 8-4 combined record against the Phillies, Red Sox, Cards & Diamondbacks, all first-place teams at the time.
Watch Him Pitch:
That is pretty much the best way to sum up the other point of view. Last night's game would be Exhibit A in the courtroom.
9 IP, 5 H, 0 R, 0 ER, 0 BB, 2 K. 30 batters faced. 83 pitches-64 strikes, eight swinging strikes
Advocates of this position argue Karstens has become a much better "pitcher." He has shown much better command and an ability to vary his pitches and pitch speed due to a new approach on the mound crafted with new pitching coach Ray Searage.
The great thing about statistics is first and foremost, before being predictive, they tell us what actually did happen. There is no room for debate here, so let's look at what Jeff Karstens has done thus far this year. His full player page is here.
When analyzing Karsten's numbers against the top 20 qualified National League pitchers based on ERA, we find some pretty interesting things. All twenty have made at least 16 starts and none more than 20. Kartens has made 16 and is the only one of the group to have appeared in relief, which he has done four times.
Of the group Karstens is 18th in IP, 18th in K/9, 20th in HR/9, 19th in HR/FB% and 20th in WAR.** His FIP-ERA is almost double any other pitcher on the list.
**Fangraphs has Karstens at a WAR value of 0.4, putting him ahead of only James McDonald among Pirates starters and 52 out of 60 qualified NL starters. B-R has him at a value of 2.5, first among Pirates starters and 17th in the NL.
On the flip side he is 3rd in ERA, 4th in WHIP, 1st in pitches/IP, 2nd in BB/9, 2nd in BABip and 1st in LOB%.
These numbers generally tell you everything you need to know about the debate. The Smoke & Mirrors camp points to the fact that Karstens doesn't strike anyone out, gives up tons of home runs and has been very lucky with an unsustainably low BABip and LOB% which are currently much better than league average. The argument says those numbers are due to revert and poof, the Karstens of old will reappear.
The Watch Him Pitch camp can point to Karstens' pinpoint command, as demonstrated by his amazingly low walk total, his low WHIP and the fact that he goes after hitters efficiently. They point to the fact that all his home runs allowed as a starter have been solo shots, and batters are rarely squaring him up and making good contact. His LD% is a touch below league average and his HR/FB% should revert back toward the mean, counteracting his BABip reverting the other way. His swinging strike percentage of 13% is just off the league average of 15% and is above his career average, while is K% is a career-best 14.2% and is a more reflective number than K/9 because he has faced so many fewer hitters.
Either way, it is virtually impossible to argue that anyone has been better than Karstens since June 1. Since that time here are JK's numbers:
8 GS, 57.1 IP, 39 H, 7 BB, 0.802 WHIP, 20 K, .554 OPS, 1.26 ERA, 5-0 W/L
While analysts and fans can point to numbers that may revert, past events don't change. This statistical record is in the books. It is what it is. And while some may not like how it was achieved or who it was achieved against** or that Karstens didn't strike out enough batters, I'm certain you won't find a pitcher who was better than Jeff Karstens the last six weeks. That's just a fact.
**Detractors will point to the fact that three of Karstens' last six starts have come against the Astros, one of the worst teams in baseball. While the Astros are bad, their lineup isn't that bad and as a team they have an OPS+ of 94, not much off the league average. Also, facing a team three times in 31 days generally works heavily in favor of the hitters, not the pitcher.
The $64,000 Question:
If you're a Karstens fan, take heart in the fact that the Keith Laws of the world have been wrong before. An analyst who relies heavily on statistics and scouting reports is certainly likely to be right way more often than wrong, but it is very hard to ever believe in any player showing a dramatic change, particularly later in his career. Law was spectacularly wrong on Jose Bautista, even at the start of this year, so it happens.
The reality is that few players do make a quantum leap later in their career, and at 28 this would be a remarkable performance shift if Karstens were able to maintain it.
I've seen every game Karstens has pitched this year. As late as the beginning of June I was expecting and advocating for the Pirates to call up Brad Lincoln to take his spot in the rotation. Karstens' inability to get deep in games and career track record were the reasons. Has my opinion changed? Absolutely. What I've seen on the mound is a different guy in terms of his command and pitch selection. While the comparisons to Greg Maddux sound silly, I can certainly see why somebody watching the last eight games would make them. Batters never seem to square up balls, he keeps hitters off-balance and his command is outstanding.
While it would be foolish to expect Karstens next 15 starts to be as good as his last eight, I do think Karstens is capable of continuing to pitch at a high level down the stretch. It's taken a long time for me to come around, but this is doesn't seem to be an illusion.
I'll just enjoy watching him take the ball every fifth day and see how it plays out. Maybe the Pirates have found themselves a pitching version of Jose Bautista, and other teams can kick themselves that they didn't pick him up off waivers when he was DFA'd off the 40-man roster less than two years ago. Maybe the baseball gods really are smiling on Pittsburgh for a change.