Felix Hernandez and the AL Cy Young Race

As you are no doubt aware, the leading (sabermetric) candidate for the American League Cy Young Award this year is “King” Felix Hernandez of the lowly Seattle Mariners. The problem is that this year his win-loss record is a very blemished 11-11. However, he currently leads the league in ERA (2.39), innings pitched (225.2), strikeouts (214), and Wins Above Replacement (5.2). The question is simply will he win the award?

Standing in his way are two other aces that have much more gaudy win-loss records despite peripherals that don’t stack up to Hernandez’s. Tampa Bay’s David Price is currently 4th in WAR amongst AL pitchers (4.6), has a 17-6 record, and a 2.87 ERA for a very good team. CC Sabathia is in the discussion mostly because of his 19-6 record. His ERA (3.14), and WAR (4.0, 11th in AL) ratings are nowhere near those of Price and Hernandez but he’s winning games for a playoff bound team.

Before we go too much further, I should mention that the metric Wins Above Replacement (WAR from now on) is going to feature very prominently in this posting. In a nutshell, WAR was developed to determine what players were the most valuable assets to their teams. It compares their numbers against particular teams in particular ballparks and compares them to what they call a “replacement level” player. I guess the best way to imagine a replacement level player is imagine that your starting pitcher gets hurt and nobody else on the active roster is able to fill his spot in the rotation (unlikely yes, but certainly possible). The replacement level player would be the one called up from AAA to fill that turn in the rotation. The higher the value of WAR, the less the team can afford to lose the player.

For a little more clarification… I’ll be using the WAR values from baseball-reference.com. I admit that it’s a fairly glaring omission that I didn’t even know there are multiple ways to calculate WAR but I don’t really care. A good rule of thumb to keep in mind that is provided by b-r.com is this.

<0 – Replacement Level
0-2 – Bench Player
2+ - Starter
5+ - All-Star Level
8+ - MVP caliber

So, onto the analysis!

Will the voters be able to overlook the margin of wins between Hernandez and everyone else? Historically speaking, probably not. Since 1990, 20 of the 40 Cy Young winners led their league in victories, which doesn’t bode well at all for Hernandez. Furthermore, of the 20 Cy Young winners who did not lead their league in victories, only five times was the discrepancy in wins greater than 3 and two of those were seasons in which closers won the Cy Young.

4 – 2009 – Tim Lincecum (15) over Adam Wainwright (19)
4 – 2008 – Tim Lincecum (18) over Brandon Webb (22)
5 – 1999 – Randy Johnson (17) over Mike Hampton (22)

The next question is whether or not it’s unprecedented for the Cy Young winner to not have the highest WAR rating in his respective league for that season and the answer is of course not. For whatever reason, voters vote according to what they think is the best way to choose the best pitcher for that season. As one might expect, with 28 individual voters, there will be 28 different methods for determining who the best was that season. According to b-r.com, there have been 12 instances of a pitcher not having the best WAR and still winning the Cy Young Award, or once every other year on average.

First of all, I have to eliminate two of those seasons from this discussion. Dennis Eckersley and Eric Gagne may have had otherworldly years in 1992 and 2003 respectively, but the baseball community isn’t quite together on how we want to deal with closers and saves so for ease of analysis, they aren’t going to be considered.

In the past twenty years, pitchers have put forth a WAR of 4.0 or greater 192 times and only 9 times did the pitcher in question record more than 30 saves. Out of those the highest of the past two decades was Mariano Rivera’s 2004 campaign when his WAR was 4.8. To put that in perspective, in 2009 only one of the ten pitchers who received any Cy Young votes at all failed to reach a WAR of at least 4.8.

THE EXCEPTIONS

If we accept the premise that WAR is a good measure of who the best pitcher was in a given season, what happened in the ten seasons when a starting pitcher won without having the best WAR? Well, when you look at the numbers it’s actually pretty easy to see why they were chosen.

Of those ten seasons, the Cy Young winner led the league in victories 7 times and in several cases the difference was quite profound. In 1993 Greg Maddux finished second in WAR but won the Cy Young with a 20-10 record. Jose Rijo led the league in WAR but was just 14-9. In 1996 John Smoltz finished 3rd in WAR but was the only 20-game winner in the NL (24-8). Ahead of him on the WAR rankings were Greg Maddux (15-11) and league leader Kevin Brown (17-11).

The biggest discrepancy in the voting results occurred in 1990 in the American League. Bob Welch won the Cy Young with a sterling 27-6 record, a 2.95 ERA, and a 1.22 WHIP. However, his WAR was just 2.5, by far the lowest by a Cy Young winner in the past twenty years. Dennis Eckersley was the only other award winner with a WAR under 4 and his was 3.0 in 1992. Meanwhile, Roger Clemens had the third highest WAR rating since 1990 (9.5), still went 21-6 with an ERA of 1.93 and a WHIP of 1.08. I believe in this case we can say that the voters became very enamored with a pitcher who won 27 games and didn’t look too much past that.

THE ROCKET

The most interesting part of this little statistical exercise is how often Roger Clemens’ name comes up on the exceptions list… on both sides. He led the league in WAR three times in seasons where he failed to win the Cy Young award; 1990, 1992 (7.9 vs. Eckersley’s 3.0), and 2005 (7.2 to Chris Carpenter’s 4.8). He has also won the award twice when he did not lead the league in WAR. He accomplished that feat in 2001 (5.4 while teammate Mike Mussina’s WAR was 6.5) and again in 2004 (5.2 to Randy Johnson’s 7.4).

The 2004 season in particular is a great example of why wins were, for a very long time, extremely overrated as a pitching statistic. In some corners, they are still overrated but word is getting around about the problems with judging pitchers by W’s and L’s. That season was Clemens’ first in the National League and he pitched extremely well, with a 2.98 ERA and a 1.16 WHIP. However, Randy Johnson’s ERA was 2.60 and his WHIP was 0.90 and he pitched 31.1 innings more than Clemens so why did Clemens take home the hardware? He was 18-4 for an Astros team that went 92-70, took the wild card, and came within one game of the World Series. Randy Johnson was 16-14 for a Diamondbacks team that only managed a 51-111 record that season.

THE OTHER END OF THE SPECTRUM

Any of you who frequent my blog know that I can’t escape a topic where I talk about the best without also talking about the worst but in this case, for the most part, it’s not the fault of the individual pitchers, but rather the metrics that have been created to judge them. Of the 11 pitchers who have received Cy Young votes and managed a WAR under 2.0, only Bill Gullickson of the 1991 Tigers was a starter (and while his 20-9 record was nice, his 3.90 ERA and 1.33 WHIP were less than stellar, leading to a WAR of 1.9). The other ten seasons on that list were all done by closers who averaged 41.1 saves and (more importantly for the WAR rating) only 63.0 innings pitched in those seasons.

THE VERDICT

So what does this all mean for the 2010 Cy Young Award season? Well, let’s look at just a few more numbers for the 40 awards handed out:

- The leader in wins have won the Cy Young Award 20 times
- The leader in average GS has won the Cy Young Award 21 times
- The leader in WAR has won the Cy Young Award 28 times
- No starting pitcher has won the Cy Young Award while finishing farther than five wins behind the league leader

When you add all of that up, I would say that right now, with two or three weeks to be played this season, we’re heading towards David Price (or maybe CC Sabathia) in the American League and Roy Halladay in the National League.


Maybe…