Starting Pitchers 2000-2009

Over the course of the past few months, I have done many statistical delving into the world of Game Score and I decided to do it once more while looking at the past decade of pitchers instead of this year. As the title so cleverly claims, I’ll be looking at starting pitchers from 2000 to 2009 but before I get too deep, I’d like to set some ground rules. First of all, I decided to only look at pitchers that qualified for the ERA title. I know that some teams play fewer or more than 162 games over the course of a season but that was my cutoff; 162 innings pitched. The number came out to 862 individual seasons so I think our sample size should be just fine.

Secondly, some of the pitchers on this list pitched in relief as well as starting and to solve this dilemma, I had two options and I went with the one that was much easier. One, I could go through the game logs for the past ten seasons and adjust their statistics to reflect only the games that they started. Two, I when determining their average pitching performance, I could divide by the number of games instead of games started (which is what I did). This does not accurately reflect how well Tim Wakefield pitched as a starter in the 2002 season when he made 45 appearances but just 15 starts but considering the small number of pitchers that made a significant number of relief appearances, I’m comfortable with any inaccuracies that might be present. For the basic analysis I’m going to be doing, it won’t harm anything.

This brings up another question? Why am I only looking at pitchers that qualified for the ERA title? Simply because if I use that cutoff, then I am looking at a group of pitchers that threw a large number of innings every year. For instance, of the 862 seasons in my sample, 860 of them made at least 20 starts and 705 of them made at least 30 starts. If we assume that only 9 innings were played in every game for the last ten years (and yes, I realize the faults in that assumption) then these seasons account for 36.7% of the total innings pitched over that time span. In other words, I wanted to look at the guys that were pitching lots of innings whether or not they were pitching well.

THE GOOD

Since 2000, a starting pitcher has averaged a GS of better than 60 34 different times. To give you an idea of what exactly that looks like on a per game basis, the average line for those 34 pitchers was 7.0-5.5-2.2-2.0-1.7-7.7 for an average GS of 61.4. In other words, those pitchers averaged 7 innings per starts, gave up 2 earned runs (0.2 unearned runs) on 5.5 hits, walked 1.7 and struck out 7.7 batters per start. Not exactly a no-hitter but averaging that over the course of an entire season is very good.

The most interesting part of this grouping of 34 pitchers is that only 11 of them actually received a Cy Young Award during their respective season. The best season not to be awarded with a Cy Young was put forth by Jason Schmidt, who in 2003 went 17-5 with a 2.34 ERA and an average GS of 65.0. Unfortunately for him, he had his best season at the same time that Eric Gagne decided to rewrite a good portion of the record book for closers.

Also intriguing are the eight starters who did not average a GS of 60 or better and still managed to win the Cy Young. In 2000, the top starter (by average GS) won the Cy Young in both leagues (Pedro Martinez and Randy Johnson). In 2001, Randy Johnson won again and once again he had the best GS of all starters that year. Roger Clemens won the AL award and he not only didn’t lead the AL in average GS (he finished 4th), he didn’t even lead his team (Mike Mussina had a GS of 57.5 to Clemens’ 55.0).

Here are the Cy Young winners for the rest of the years in the sample with their respective ranks (within their own league) and the leader if it wasn’t them.

2002
AL – Barry Zito (3rd, 57.7) – Pedro Martinez (63.5)
NL – Randy Johnson (1st, 66.3)

2003
AL – Roy Halladay (4th, 57.4) – Pedro Martinez (61.9)
NL – Eric Gagne – Jason Schmidt (65.0)

2004
AL – Johan Santana (1st, 63.1)
NL – Roger Clemens (6th, 58.5) – Randy Johnson (64.9)

2005
AL – Bartolo Colon (6th – 53.8) – Johan Santana (62.8)
NL – Chris Carpenter (3rd, 60.9) – Pedro Martinez (63.1)

2006
AL – Johan Santana (1st, 60.6)
NL – Brandon Webb (1st, 57.7)

2007
AL – CC Sabathia (3rd, 57.7) – Erik Bedard (60.0)
NL – Jake Peavy (1st, 61.2)

2008
AL – Cliff Lee (2nd, 59.7) – Roy Halladay (60.0)
NL – Tim Lincecum (1st, 60.5)
Both – CC Sabathia (61.2)

2009
AL – Zach Greinke (1st, 61.6)
NL – Tim Lincecum (1st, 64.5)

Note: in 2008 CC Sabathia led the entire league in GS but because he made 18 decent starts in the AL and 17 incredible starts in the NL, he didn’t garner enough votes to win the Cy Young Award.

What does all of that tell us? Well, for a very long time the Cy Young voters have been infatuated with a set of statistics that don’t necessarily say who actually pitched the best that season. Number one on that list of stats is wins. In 2001 Roger Clemens finished 4th in GS and still won the Cy Young award. That season he was 20-3 with a 3.51 ERA. Barry Zito (17-8, 3.49), Mike Mussina (17-11, 3.15), and Freddy Garcia (18-6, 3.05) all pitched better according to GS, but none of them won 20 games.

The same thing happened several other times with the most notable snubs being 2004 and 2005. These were the following lines for the Cy Young winner and another notable starter (IP-H-R-ER-BB-K-GS).

2004
Pitcher A – 6.5-5.1-2.3-2.2-2.4-6.6-58.5
Pitchers B – 7.0-5.1-2.5-2.0-1.3-8.3-64.9

I understand that sometimes average Game Scores can look very much alike but I have to say, this is not one of those times. Pitcher B allowed the same number of hits, fewer earned runs, walked one less batter per start, and struck out 1.7 more batters while pitching an extra half inning every time toeing the rubber. So what gave Pitcher A the Cy Young that year? Perhaps it was the fact that Roger Clemens went 18-4 for a Houston team that won 92 games and made the playoffs while Randy Johnson went just 16-14 for a Diamondbacks team that went 51-111, one of the worst records of the past decade.

Let’s try again, the very next year…

2005
Pitcher A – 6.7-6.5-2.8-2.6-1.3-4.8-53.8
Pitcher B – 7.0-5.5-2.3-2.2-1.4-7.2-62.8

Again, I understand that sometimes Game Scores can be very close and two stat lines can look very similar, allowing more hits and fewer walks, so on and so forth… 9 full points though? This example becomes even more laughable the more I look at it. Bartolo Colon had a good season, going 21-8 for a 95-win Angels team. Despite the fact that Johan Santana won “only” 16 games (16-7) and the Twins only won 83 games, I would think that the stat lines above should tell you that Santana (Pitcher B) pitched much better.

Ok, one last example before we move on…

Pitcher A – 7.0-5.3-2.2-1.9-2.1-8.2-64.5
Pitcher B – 6.9-5.9-1.9-1.7-1.5-7.3-61.6

This is actually a bit of a trick. Pitcher A is NL Cy Young winner Tim Lincecum’s line in 2009 and Pitcher B is Zach Greinke’s Cy Young winning line the same year in the AL. Why would I ask this trick question? To point out that Tim Lincecum was 15-7 and finished 4th in the NL in wins and still won the Cy Young while Zach Greinke was 16-8 and finished a distant 7th in the American League in wins. The two of them combined to beat out four 19-game winners (some with very good lines). However, they led their leagues in average GS and pitched the best throughout the year. Maybe there’s hope for the voters yet…

THE BAD (WITH THE UGLY THROWN IN)

In the past ten years, only six times has a pitcher averaged a GS under 40 and still threw more than 162 innings in a season. Out of those, only one player managed to do it twice (for two different teams no less). Jose Lima never managed to pitch “well” in this decade but he was decent in 2004 with the Dodgers (13-5, 4.07). However, in 2000 with the Astros and 2005 with the Royals, he didn’t do quite as well. In 2000 (the “better” of the two seasons), Lima was 7-16 with a 6.65 ERA. His average stat line was 5.9-7.6-4.6-4.4-2.1-3.8, for an average GS of 38.3. Five years later he managed to surpass himself with a stat line of 5.3-6.8-4.4-4.1-1.9-2.5, a GS of 37.8, a 5-16 record, and a 6.99 ERA which is the standard for futility over the past ten years.

Yet somehow, the lowest average GS doesn’t belong to Lima, or even the second lowest. In 2008 Livan Hernandez pitched his was to a 13-11 record despite a stat line of 5.8-8.3-4.2-3.9-1.4-2.2-37.5 and an ERA of 6.05 (he had a good offense supporting him, apparently).

Before we move on, we have to throw a shout out to the Colorado Rockies, who spent absurd amounts of money on front line starters in the late 1990’s and early 2000’s, hoping to raise the quality of starting pitching a mile above sea level. In retrospect, the humidor was a much cheaper alternative. In 2002, Mike Hampton was paid $9,503,543 and produced a WAR rating (wins above replacement) of -1.5. He was 7-15 with a 6.15 ERA that season. His stat line was 6.0-7.6-4.5-4.1-3.0-2.5 for a GS of 37.0. Yes, you read that correctly, in 178.2 innings that Hampton pitched that year, he walked 91 batters and struck out just 74.

Lastly, here is a list of the pitchers that finished last in the league in GS year by year. Included are their win-loss records and ERA’s along with their stat lines.

2000
Jose Lima
7-16, 6.65
5.9-7.6-4.6-4.4-2.1-3.8
38.3

2001
Dave Milcki
11-11, 6.17
4.9-6.0-3.6-3.4-2.2-2.9
39.6

2002
Mike Hampton
7-15, 6.15
6.0-7.6-4.5-4.1-3.0-2.5
37.0

2003
Mike Maroth
9-21, 5.73
5.9-7.0-4.0-3.7-1.5-2.6
41.3

2004
Brian Anderson
6-12, 5.64
4.7-6.2-3.5-3.0-1.5-2.0
39.3

2005
Jose Lima
5-16, 6.99
5.3-6.8-4.4-4.1-1.9-2.5
37.8

2006
Joel Pineiro
8-13, 6.36
4.1-5.2-3.1-2.9-1.6-2.2
40.6

2007
Scott Olsen
10-15, 5.81
5.4-6.8-4.1-3.5-2.6-4.0
40.8

2008
Livan Hernandez
13-11, 6.05
5.8-8.3-4.2-3.9-1.4-2.2
37.5

2009
Livan Hernandez
9-12, 5.44
5.9-7.1-3.6-3.6-2.2-3.3
42.3

THE LUCKY AND THE UNLUCKY… SORT OF…

After looking at the good, the rewarded (not necessarily the same as the rewarded), the bad, and the positively ugly, I want to shift gears a little bit.

What I wanted to look at next were the pitchers whose win-loss records were out of whack with the rest of their stat lines. Yes, I understand the problems with using wins and ERA to do any sort of serious analysis but the results are interesting, especially when viewed in a vacuum. If you were to delve into these seasons I found, I’m sure you could find some fascinating reasons why these pitchers had losing records with very good ERA’s or vice versa. I encourage you to do this and if you want any of my data, let me know and I’d be happy to dump it on your electronic doorstep.

Before I get going, I wanted to include a quick note about win-loss records. For them, I compared the pitcher’s individual win-loss record against the win-loss record of his team without him. For instance, if a pitcher went 20-10 for a team that went 100-62, I’d be comparing the pitcher’s .667 winning percentage against the team’s .606 (80-52) winning percentage.

The pitcher that most outpitched the rest of his team who deserved it the least was Scott Elarton in 2000 for Houston. While the Astros went just 72-90, Elarton somehow managed to go 17-7 despite a 4.81 ERA and average GS of just 47.0. Of the small selection of pitchers that I found who outperformed their pitching, Elarton was the only one who actually played on a losing team. Most of the others, such as Paul Abbott, who went 17-4 with a 4.25 ERA, had the good fortune to play on good teams. Abbott played for the 2001 Mariners who went 116-46.

On the flip side of the coin, nobody deserves more sympathy than Matt Cain of the Giants. In 2007 and 2008 he had the misfortune of playing on bad teams and still pitching well. For the 2007 Giants who had a record of 71-91, Cain was just 7-16 but had an ERA of 3.65 and an average line of 6.3-5.4-2.6-2.5-2.5-5.1-54.3.

The next year, his line stayed similar (6.4-6.1-2.8-2.7-2.7-5.5-52.9) and unfortunately for him, the results did also. He was 8-14 with a 3.76 ERA for a Giants team that went 72-90.

If that doesn’t shake your confidence in the use of wins and losses to judge starting pitchers, I’m not sure than anything ever will.


As always, thank you for your time and get ready because October is just 24 days away…