How good were the Mets pitchers in 2006?  We are going to use some standard baseball statistics and some Sabermetrics in order to analyze just how good the Mets pitchers were last season.  Before we jump into the analysis we need to understand how the players are grouped and what stats we will be using.

1) The Groupings:  The pitchers are split into three groups base on their roles and how much they pitched during the regular season (no playoff data).  The first group contains the starters that threw more the 90 innings (Glavine, Trachsel, Pedro, El Duque, and Maine).  The second group consists of the relievers who threw more then 50 innings (Heilman, Oliver, Wagner, Bradford, Feliciano, and Sanchez).  The third group consist of every other pitcher who took the mound for the Mets in 2006.

2) The Stats: We are going to look at 7 statistics from 2006.  They are IP (innings pitched), K/9 (Strikeouts per 9 innings), W/9 (Walks per 9 innings), WHIP (Walks + Hits per 9 innings), ERC (Component ERA – also known as CERA), DICE (Defense Independent Component ERA), and ERA (Earned Run Average).

Finally, we need to know why we care about some of these statistics.  IP, ERA, WHIP, K/9 and W/9 are simple enough and most baseball fans are comfortable with these stats.  Everyone knows what to look for in these statistics.  We all like a pitcher to have an ERA under 4.00, a WHIP under 1.30, and a high K/9 coupled with a low W/9.  But what about ERC and DICE?  How are they calculated and what do they mean?  Let’s take a quick look:

ERC: Component ERA (aka CERA) was invented by Bill James.  This statistic attempts to forecast a pitcher’s ERA using only hits and walks allowed.  The ERC formula basically takes pitchers HITS and WALKS given up and multiplies the total by a PTB (Pitcher’s Total Bases estimate).  That number is then divided by the Batters Faced in a season TIMES Innings Pitched.  This gives an estimate of how many runs a pitcher will give up in an inning.  The number is multiplied by 9 to make it a 9 inning average (like ERA).  The -0.56 is to adjust to real ERA.

ERC=((H+BB+HBP)*PTB)/(BFP*IP))*9-.56

Where PTB = .89*(1.255*(H-HR)+4*HR)+.56*(BB+HBP-IBB)

Once criticism about ERC is that HITS are too dependent on your team’s defensive ability (a pitcher’s home ball park can also be a factor).  Critics say that HRs, walks and strikeouts are the only factors that are completely controlled by the pitcher.  That brings us to DICE.

DICE: Defense Independent Component ERA was invented by Clay Dreslough in 2001.  DICE takes out the defense factor in ERC.  HRs have a multiple of 13, Walks a multiple of 3, and Strikeouts a multiple of -2.  The formula is basically saying Strikeouts help you ERA, Walks hurt your ERA, and Homeruns absolutely kill you ERA.  Since these are the only real categories pitchers have total control over, they are the only categories factored in.

DICE = 3.00 + [(13HR+3(BB+HBP)-2K)/IP]

Lets now take a look how the 2006 Mets did on the mound.  The players are sorted by IPs and the criteria mentioned above in the Groupings section.  I hae attached my spreadsheet at the bottom of this post for those who want to work with the data.

The Starters:

#1) Pedro Martinez was the best Met starter when he was healthy.  He led the starters with an ERC of 3.20 and a DICE of 3.91.  His ERA was 4.49 last years, but it was inflated due to his last 4 outing when he was pitching injured.  Pedro’s ERA before his injury was 3.42.  Martinez had the highest K/9, lowest W/9 and the best WHIP among all the starters.  Quite simply, Pedro was on fire before his injury woes, and the numbers suggest he will be the Mets best pitcher when healthy in 2007.

#2) El Duque?  I was pretty surprised about this one, but take a look at the numbers for yourself.  High K/9, low W/9, 1.24 WHIP (we love to see that number below 1.30) and a 3.62 ERC and 3.90 DICE.  El Duque’s ERA was 4.09 in 2006, a bit higher then his ERC and DICE.  This suggests El Duque may have been a bit unlucky with his defensive support.  Supposedly Orlando Hernandez is now 40 years old. If he can have another solid year I think he can have an ERA under 4.00 with some help from the Met fielders. 

#3) Tom Glavine is our #3 pitcher for the year but you could ABSOLUTELY make him your #1.   Glavine led the team in IP, started 32 games, and was the only consistent start for the Mets all season, going 15-7.  Tommy deserves to be the pitching MVP because of his productivity and durability, but his statistics were not as good as many people may think.  They may have suffered because the Mets had to lean on him all season, so we should keep that in mind, but lets break out the numbers anyway.  Glavine had a relatively 5.95 K/9 but a solid 2.82 W/9.  His WHIP was a tad high at 1.33.  While Tommy had an ERA of 3.82, he had a 4.01 ERC and a 4.15 DICE.  This is suggesting Glaving may have had a better season then he deserved.  However, it might also suggest Glavine is a seasoned vet and knows how to get out of jams.  Regardless, don’t be surprised if Glavine has an ERA north of 4.00 in 2007.

#4) John Maine is our #4 starter but he pitched well enough to be considered a #3.  He only threw 90 innings but he did have a decent 7.1 K/9, a 3.3 W/9, and a very impressive 1.13 WHIP.  Maine had a 3.60 ERA but his DICE and ERC were worlds apart.   Maine’s ERC was 3.22 but his DICE was a surprisingly high 4.76 (I had to recheck my calculations 3 times).  Why is Maine’s DICE so much higher then his ERC?  It is due to the high percentage of HRs Maine gives up.  Since Maine is still young and inexperienced, I hope his HR numbers get better with age.  He is averaging 1.5 HR/9.  Glavine and El Duque are around 1.0 HR/9 and Pedro and Trachsel are around 1.25 home runs per nine innings.  When DICE give HRs a multiple of 13 you know it is going to have a huge impact for pitchers that give up a higher percentage of HRs.  Maine seems to have a lot going for him.  If he can limit his mistakes (and the long balls) he should have a solid year.  I like him for 15 wins, a WHIP under 1.20 and an ERA under 4.00 in 2007.

#5) Steve Trachsel is our #5.  Steve had a nice 15-8 season and was able to start 30 games in 2006, but he only had 13 Quality Starts.  Trachsel only had a 4.3 K/9, a very high 4.28 W/9, and that 1.60 WHIP is down right criminal.  His 5.36 DICE and 5.57 ERC was higher then his 4.97 ERA, which suggest the defense came through for Trachsel.  You could make the case Steve is a crafty veteran and should be more adept at getting out of tight spots, which would explain his low ERA compared to his DICE/ERC.  However, take a look at Glavine & Trachsel’s stats.  You wouldn’t think they had almost identical records.  I’m not surprised the Mets did not resign Trachsel nor am I surprised thiere are not a lot of teams interested in him.  I was a big fan of his last season but the numbers don’t make a compelling case for him to be on the team in 2007.

The Bullpen:

#1) Billy Wagner was the best bullpen pitcher, which should not be a surprise.  In 72 innings he had a 11.73 K/9 and a 2.62 W/9.  Outstanding.  Throw in the 1.11 WHIP and you know why he is a dominant closer.  Wagner had a 2.24 ERA, 2.84 ERC and a 2.69 DICE.  Wagner’s ERC and DICE are a bit higher then his ERA, but since he throws so hard he does give up the long ball a little bit more then the rest of the Met bullpen (except for Oliver).  Wagner give up 1 HR per 9 innings pitch, which can explain the slight elevation in DICE and ERC.  I don’t think there is anything to worry about here.

#2) Chad Bradford was the man in 2006 with a 6.53 K/9, a sweet 1.89 W/9, and an impressive 1.16 WHIP.  His 2.48 ERC and 2.30 DICE were lower then his 2.90 ERA because Bradford only gave up 1 HR in 62 innings last season.  That is 0.15 HRs per 9 innings pitched.  I was sorry to see him go to Baltimore.  Chad will be missed.

#3) Aaron Heilman racked up 87 innings for the Mets in 2006 and pitched very well.  A 7.55 K/9, 2.90 W/9, and a 1.16 WHIP are sweet numbers.  Aaron had a 3.62 ERA but his ERC was 2.76 and his DICE was 3.14, both significantly lower then his ERA.  Heilman only gave up 0.5 HRs per 9 innings in 2006 and I expect another solid season out of him whether it be as a starter or out of the bullpen.

#4) Feliciano gets the nod over #5 Duaner Sanchez.  Feliciano is just a little bit better in the K/9 and W/9 ratios, and their WHIPs were about even.  While Pedro had a low 2.09 ERA, the ERC and DICE says his ERA should be more in the 3.20 to 3.30 range.  That was interesting.  We’ll have to see how 2007 plays out to see if Feliciano got lucky in the field or if he really is that good.  Duaner had a 2.60 ERA and the numbers suggest it should be more towards 3.00.

#6) Somebody had to finish #6 and it happens to be Darren Oliver.  Oliver’s biggest problem was his high 1.44 HR/9.  The rest of the bullpen averaged around 0.50 HRs per nine innings.  While Oliver had a solid 3.44 ERA, the numbers suggest his 2007 ERA could be more in the 4.00 neighborhood.

Everyone Else:

I included these just so you can take a look at the rest of the team.  Keep in mind that the fewer the innings pitched, the less reliable the numbers.  For example, I would not read much into Humber’s stats.  However, I was encouraged by Oliver Perez and Mike Pelfrey’s numbers.  I love the strike outs Perez can generate, and his DICE and ERC suggest his ERA can be better (keep in mind the playoffs are not included in these stats).  It looks like control is the issue with Perez, and if the Mets can get him going he could be a very solid pitcher.  Pelfrey’s numbers were not that great, but his low number of HRs is nice to see, and it is reflected in his DICE.  If he could get those walks down he could really improve his game.

On Deck:

I’m on vacation through President’s Day.  Unfortunately this means I will be away from the site for a while.  I’ll be back in 2 weeks for more SaberMETrics.  In the mean time be sure to check back daily to see what Regis is cooking up.