November 16, 2006

Prospectus Matchups

To Leave These Lolling Stats, to Get You VORPside

by Jim Baker

Printer-friendly

In light of Brandon Webb being awarded the National League Cy Young Award on Tuesday, I thought I'd address a few things about the Cy Young voting over the past decade or so.

Voters are faced with three possible scenarios when they go to cast their ballots:

• There is no clear-cut leader.
• There is a clear-cut leader and they choose him.
• There is a clear-cut leader and they do not choose him.

Of course, my definition of leader is based on looking at VORP, something--I'm guessing here--less than a quarter of the voters do. What seems like a cloudy VORP picture might actually be an easy decision on a vote based on more traditional metrics. For the purposes of today's discussion, though, we'll stay mostly VORPside, going trad only when trying to search for reasons for enigmatic votes.

The 2006 National League Cy Young "race" fits into the first category shown above, both in terms of VORP and trad stats. In the absence of an historically great relief performance, there was really no wrong choice provided the voters stayed within the handful of VORP leaders. In fact, since 1996, the gap between the VORP leader and the average VORP of the next four finishers was the smallest it's ever been in this year's National League. Roy Oswalt's 72.4 was 6.5 points higher than the average of the four men in the two through five slots: Webb, Chris Carpenter, Bronson Arroyo, and John Smoltz.

With that kind of tight bunching, there is simply no cause for outrage when the second-highest finisher gets the award. One can puzzle over the high man finishing fourth, I suppose, and the absence of Arroyo on any ballot might seem surprising, until one considers that he tallied 14 wins, and wins still count out there in the award world. It was that kind of year in the National League, though--nobody lapped a field that was relatively lap-able.

How lapa-ble? In the 22 full league seasons of the Post-Stoppage Era (1996 onward), the 2006 NL season ranks 15th in average VORP of the top five finishers. In contrast, the highest average came in the American League in 1997, when the top five men averaged 83.5. This was pulled along greatly by Roger Clemens having one of the best seasons ever (at 116.3). However, Randy Johnson's 85.9 was the best second-place finish of the era. There's a good case that the previous year's AL showing was even better in that Clemens and Charles Nagy had the best fourth- and fifth-place finishes of this eleven-year era by far. If I had to vote, I would say that the 1996 American League quintet of Pat Hentgen (who deservedly won the Cy Young), Alex Fernandez, Ken Hill, Clemens, and Nagy had the strongest combined output.

Going back to the three scenarios, let's look at some examples of each:

No clear-cut leader

1997 National League, 0.9 VORP: Greg Maddux had a slight VORP edge over Pedro Martinez. Maddux had the better won-loss record, but Martinez had a somewhat better ERA and 305 strikeouts, and that won the day.

2001 American League, 1.1: Joe Mays over Freddy Garcia. The voters resolved this conflict by going well outside the top five for their winner. Mays has done nothing but struggle ever since, but didn't get a single vote, while Garcia finished third. Mark Mulder led the league in wins and was fifth in VORP; he finished second in the voting. Instead, the voters turned to the man with the eighth-highest VORP, Roger Clemens, the beneficiary of superior Yankee firepower. Clemens didn't even pitch the best season by anyone on his team that year; that honor went to Mike Mussina who finished third in the league in VORP but ran into hard luck, finishing at 17-11 and getting just two points in the Cy Young voting. I suppose it's hard to ignore a 20-3 record; 20-1, 21-2, or 22-1 would have been unprecedented, and even I, a person who usually has no earthly idea what a pitcher's won-loss record is at any given moment, wouldn't have argued with the Clemens choice had he won 20 while losing two or less. Splitting hairs? Yeah, sure, but at least it can be said that nobody had ever done that before. Instead, he lost two of his last three starts to the last-place Devil Rays, and got a no-decision in the other against a mediocre Baltimore team. So, with a best case of 23-1 and a worst of 20-4, he closed out closer to the worse. The ironic thing is that four and five years later, he's pitching light years better than he did in 2001, but getting poor run support and little award recognition for his trouble.

1998 National League, 1.7: Greg Maddux over Kevin Brown. Two years earlier, Brown was 13.9 points ahead of the second-place finisher in VORP and 15.5 points ahead of the winner, John Smoltz. In this season, he lost to another Brave, but it wasn't the man who edged him in VORP. Instead, the voters chose Tom Glavine who was not all that far behind Brown in the VORP count. Brown finished third, ahead of Maddux but behind teammate Trevor Hoffman who, as in 2006, finished second.

There is a clear-cut leader and they choose him

The biggest gaps between first and second place since 1996 are Pedro Martinez in 2000 and 1999, Roger Clemens in 1997, and Randy Johnson in 1999. All four won the Cy Young; Martinez unanimously both times. It should be noted that a few voters didn't quite get it, though:

• 1997 American League: Two voters went for Randy Johnson instead of Clemens, and one shot his bolt on Randy Myers. Is it possible to revoke credentials retroactively? Among relievers, Myers did have the highest WXRL that year, but Clemens had one of the best seasons in the history of baseball, and that's got to trump everything, doesn't it?

• 1999 National League: Mike Hampton's 22-4 record earned him 11 first-place votes in spite of the fact that Johnson had struck out 364 men to Hampton's 177, and that Hampton walked over 100 men. Perhaps more puzzling still is the voter who thought Kevin Millwood had had a better season than the Big Unit. Still, though, the voters and VORP agreed one-through-three.

There is a clear-cut leader and they do not choose him

1996 National League, 13.9 VORP: Kevin Brown over Greg Maddux. As mentioned above, Brown lost this race to third-place finisher John Smoltz. Their expected won-loss records were 18-6 for Brown and 19-9 for Smoltz. The ones the voters paid attention to, though, were more diverse: 17-11 for Brown and 24-8 for Smoltz. He had twice as many strikeouts, and without an attendant boost in wildness. Nevertheless, Smoltz's numbers were just too loud for all but two voters who saw the subtleties in Brown's advantages.

2005 American League, 18.2: Johan Santana over Mark Buehrle. Last year's AL vote is the most egregious example of overlooking the better candidate in the Post-Stoppage Era. Santana out-pitched Colon at every turn, and had an expected won-loss record that fairly well matched his real W-L. On the other hand, Colon pitched at a level you'd expect to generate a 14-10 record, but good support got him up to 21-8. Fortunately for Santana this year, there is no such danger lurking to entice voters away from acknowledging him as the best pitcher in the league.