Mike Pelfrey – Season Analysis 2008 

Basic Stats

Besides Johan Santana, we had another starting pitcher who was a pleasant surprise during the 2008 season, Mike Pelfrey.  Struggling in the beginning of the season, Pelfrey really turned the corner in June, posting a 3.53 ERA that month.  He went on to post ERA’s of 2.70 during July and 2.93 during August before struggling quite a bit in September with an ERA of 4.06.  The biggest contribution to his hot July and August was the fact that he only walked 11 batters in 83 innings during those two months.  Pelfrey continued to have his struggles against left-handed batters, who posted an OPS of .816 against him (compared to righties OPS of .654).   

Looking at Mike’s peripheral stats, his K rate actually decreased from last year from 5.57 K/9 to 4.93 K/9.  The lack of strikeouts in troubling and in order to become an even better pitcher in the future he will have to improve upon this.  Any guy that can throw a fastball 95+ mph should strikeout more then 4.93 K/9.  His lack of off speed pitches still has a major effect this. Luckily for Pelfrey, he does make up for this somewhat with the ground ball.  Mike’s GB rate was 49.6% (13^th among all NL Starters with at least 100 innings pitched).  As noted above, what helped Mike the most however is the lack of walks.  Mike has shown incredible control this season (probably cause of the high percentage of fastballs thrown).  His BB/9 is 2.87 in 2008, down from 4.83 from last year.  The lack of walks has reduced Pelfrey’s WHIP from 1.71 in 2007 to 1.36 in 2008.  His FIP went down from 5.06 to 3.96 as well.  When opponents did put the ball in play, they hurt Pelfrey far less than last year as well.  His BABIP for 2008 was .302 compared to .331 from last year.  BABIP is primarily a result of luck.  Most pitchers will hover around .300 for the season so perhaps we could have predicted Pelfrey was a bit unlucky in 2007.  Being a ground ball pitcher has also helped Pelfrey big time as far as giving up Home Runs.  He only allowed 12 home runs in 200.2 innings, a HR/9 of 0.54 which is the 6^th best rate among NL Starters. 

Pitch FX

1) What Does He Throw? 

Using data from MLB Gameday application, we can look further in detail to what Mike Pelfrey threw in 2008.  Mike threw 3323 pitches during the 2008 season, of these gameday has tracked 3281 pitches.  The following table lists the data: 

 Any analysis of Mike Pelfrey must start with the fastball.  Mike throws his fastball over 80% of the time.  According to Fangraphs, only Aaron Cook and Daniel Cabrera throw a higher percentage of fastballs.  Before I go any further, there was some talk about telling the difference between the 2-seam fastball and the 4-seam fastball in the other thread.  I use K-Means clustering analysis to cluster the pitches together and I think it does fairly well but it’s not perfect.  I decided to do a 2-step clustering analysis on Mike Pelfrey to see if it gave me different results.  The two clustering algorithms agreed very well on all off-speed pitches.  On the fastballs however, they disagreed on nearly 100 pitches on whether they were 2-seam or 4-seam.  The two-step clustering gave me a split of 51.33% on 4-seam compared to 29.50% on the 2-seam.  It’s about a 4% difference from the chart I have up there.  Outside of actually sitting down and watching each pitch to see if it’s a two seam grip or a four seam grip, I’m not really sure we will get better telling the two apart.  Also, I decided to break down the data on a game by game basis.  Some parks still do not have their camera calibrated correctly, thus giving me some incorrect results.  For example, in Colorado, the data states Mike threw 88% 4-seamer as opposed to 1% 2-seamers.  Looking at the starts at Shea stadium, they pretty much agree with one another and I did not get anything extreme.  Having said all this, the overall numbers match up fairly well with Fangraphs.  Fangraphs data comes from Baseball Information Systems.  I believe they watch video to determine pitch type so seeing similar numbers that I have gives me confidence it’s pretty accurate. 

Back to the fastball, Mike Pelfrey had huge amounts of success with this pitch in 2008.  Since Pelfrey has entered into the Majors, most have felt he needed to develop a secondary off-speed pitch to be successful.  Instead of adding a reliable off-speed pitch, he began throwing his 4-seam fastball and 2-seam fastball more often.  His average fastball is around 92.7 mph….almost a full 3 mph better then average.  I do think that as the season went along, teams (most notably the Florida Marlins) began to sit on the fastball.  So although Mike had a lot of success with the fastball, he must continue to work on his other pitches. Another thing of note, if you recall from Johan Santana’s thread, his average 4-seamer vertical movement was 10,70 and his 2-seamer 8.14.  As you can see here, Pelfrey gets much more sink on his fastball compared to Santana with averages of 8.48 and 5.96 respectively, hence the reason why Pelfrey gets a lot of ground balls. 

The next pitch Mike uses is his slider.  Combined with the fastball, this makes up 93.7% of all his pitches.  His slider is exactly league average in speed, horizontal and vertical movement.  His slider gets more horizontal movement then Santana’s as Mike’s slider breaks away from the right handed hitter .

Mike’s third pitch is his changeup.  He only threw his changeup 5.27% of the time and I feel this will have to be the pitch Mike will have to develop in order to be effective against lefties.  His changeup actually has more movement then league average, however Mike has trouble locating this pitch (as we will see below).  Also the difference between his changeup and fastball is about 8 mph, which is the league average difference, however this is a far cry from the 12 mph average difference that Johan had. 

Starting on June 27^th against the Yankees, Pelfrey began throwing a curveball.  Like the changeup, Mike had trouble locating this pitch and only ended up throwing 35 during the entire season.  He didn’t get nearly enough movement on this pitch and I will be interested to see if he continues to use this pitch next season. 

2) Who Does He Throw It To? 

Next lets breakdown Mike’s pitches between Lefty vs Righty 

Mike will throw his fastball 82.4% to the lefties compared to 78.2% against the righties.  Looking at the data, Pelfrey feels confident mixing in his slider against the righties, throwing that 20.3% of the time.  However, as you can see, he does not have another pitch to go to against the lefties.  This is a big difference between the two pitchers we have profiled.  While Johan had three pitches to go to for each side (his 2 fastballs and one off-speed pitch), Mike does not have that luxury against the lefties.  Judging by the way he threw the curveball mostly to lefties, it seems like Mike does not have much confidence in his changeup at all right now and is looking for anything to help him against left-handed batters.  Whereas righties have to at the very least think about the slider, lefties can sit dead red on the fastball.  This is why I believe, he will eventually need to develop either the changeup or curveball to be more effective.

 3) What Happens When He Throws It? 

Here is a breakdown on what happens when Mike throws each type of pitch.  I’m just going to show the raw data here.  Last article I calculated BABIP on each pitch type but I ran into all sorts of problems.  BABIP is not an official stat.  As such, different websites calculate the stat very differently.  To me, I just figured the proper way to calculate BABIP would be 1-DER (ie. If the teams’ DER behind pitcher X is .700, then his BABIP would be .300).  Since a lot of sites calculate it differently coupled with the fact I do not have all pitches in my data set, I decided to leave it out for now.

A couple things jump out to me immediately looking at this graph.  First, his changeup is not good at all.  Pelfrey gave up 10 singles in 24 balls in play on the changeup.  Also, the curve was not good either as 2 of the 4 balls in play went for singles.  Eight of the 12 home runs that Pelfrey gave up was off the 4-seam fastball.  The majority of the extra base hits were off the 4-seamer.  Right now, with a lack of off-speed pitch, its necessary for him to keep throwing it at a high percentage but perhaps he should be throwing more 2-seamers then 4-seamers.  4-seamers are thrown harder generally but will yield more fly balls and extra base hits, 2-seamers will yield more ground balls.  Pelfrey isn’t strikeout out batters at a high rate anyways so IMO he should focus on the ground ball.  He should strive to get the ground ball rate to where guys like Webb are.  He will do this by throwing more 2-seamers and less 4-seamers.  Of course, this is all predicated on Pelfrey’s ability to develop his off-speed pitches.

 4) Percentage of Strikes Thrown Per Pitch Type 

I didn’t do this with the Johan article but I think this is important to point out.  As stated above, Pelfrey dramatically reduced his BB/9 rate this season.  Without looking at the data, I assumed this is more because of his high percentage of fastballs and less because he threw his off-speed pitches for strikes.  Let’s see how Pelfrey did compared to league average with each pitch.

Here are the MLB Average’s for each pitch type.  The data is from 2007 but I don’t think there will be much difference in this data from 2007 to 2008.  I’m including balls in play as strikes as well as Fangraphs does.

Here’s Mike’s data for the 2008 season:

As you can see, he had trouble locating all of his off-speed pitches; however he made up for it by being able to throw his fastball for a higher strike percentage than league average.  Since he threw over 80% fastballs, this indeed helped him reduce his walk rate.

 What’s In Store for 2009 

There is no doubt that Mike Pelfrey has taken a step up this season in becoming the pitcher the New York Mets had envisioned when they drafted him in the 1^st round of the draft.  There will be concerns next season as Mike threw an awful lot of innings and pitches this season.  As pointed out in threads countless times, pitchers who receive such an increase in workload typically perform worse in the season following the heavy workload.  Also, he needs to gain better command of his off-speed pitches or else teams will begin to just sit on the fastball.  Even when you can throw in the mid 90’s, I do not believe you can get away with throwing fastballs 80% of the time.  He may have been able to do this in college, but I do believe the rest of the league will catch up and I think they will catch up soon.  He will also continue to struggle against lefties as well if he continues throwing the same percentages of pitch types in 2009 as he did in 2008.  Having said that, Mike took big steps this season and I look forward to seeing him improve even more in 2009.