Deconstructing Scott Diamond

Warning: Stats ahead! If you prefer to learn about Scott Diamond's unique place in baseball history, you can read it here. 

2012 was a completely lost season for the Minnesota Twins. I don't need to rehash all the details, as we all lived it together. One of the few bright spots was a 26-year-old former Rule 5 draft pick in Scott Diamond. Diamond shined brightly for the Twins in 2012. He won 12 games, posted a 3.54 ERA and dazzled hitters and fans alike with his pinpoint control. Diamond will go into the 2013 season as the only certainty and likely would have been the Opening Day starter, had he been healthy. As it stands, the expectations for Diamond are high. Should we expect to see a repeat of 2012, or will he revert to pre-2012 Scott Diamond? Let's use any and everything FanGraphs.com has to offer to find out.

The very first thing I see when I look at Diamond's stats is his below-average strikeout rate. Diamond posted a 12.6% strikeout rate in 2012, where league average was around 19%. This stat alone means only one thing - Scott Diamond doesn't strike out a lot of batters. Not exactly breaking news. However, the rate is so far below the league average, that it bears a closer inspection. Here is a list of "successful" pitchers with a career K rate of 13% or lower (from 1993-2012). I used an ERA- of 100 or less as the gauge for success, but within that number there is much fluctuation. A 100 ERA- is considered league average. I used ERA- instead of FIP- because I wanted to reflect actual performance and not peripheral skill. I only looked at pitchers who were considered starters through the majority of their career.

Note: ERA- is a stat that standardizes ERA for park factors and era. It is a great way to compare ERA from players in multiple seasons and eras. A smaller number is better and 100 is always the average, much like IQ. 

Aaron Cook

Chien-Ming Wang

Mark Gubicza

John Lannan

Jon Garland

Paul Byrd

Dennis Martinez

Tomo Ohka

Butch Henry

9 names and no one that really jumps off the page. We can add players who posted a 13 to 14% K rate and then these names appear: Mark Buehrle, Kenny Rogers, Orel Hershiser, Ken Hill, and Jarrod Washburn. The names are a bit more impressive, but we are also talking about a 10% increase in K rate. It is possible to have success with such a low strikeout rate, but it is pretty rare. However, K rate is just one part of the puzzle. What else should we look at?

Diamond's walk rate was 4.3% last season, almost half of the league average of 8.1%. How rare is a walk rate that low? Looking at the same sample as before, here is a list of pitchers who posted a walk rate as low as Diamond's 4.3%?

Bret Saberhagen

Bob Tewksbury

Kevin Slowey

Greg Maddux

Scott Sanderson

Josh Towers

Brad Radke

7 names, and all but Towers and Slowey won at least 100 career games. Sanderson doesn't really fit, as the sample was from the tail end of his career.  Towers and Slowey couldn't keep balls in the park, something Diamond did at roughly the league average. Maddux, Saberhagen and Radke are the class of this group. Each had a significantly higher K rate than Diamond. However, it is clear that Diamond had elite control last season. If he can maintain that control, he should remain effective. 

Limiting walks as Diamond does really keeps runners off of the bases. The low walk rate masks his low strikeout rate a bit. His 2.9 K/BB ratio last year was well above league average. in our 1993-2012 sample, only 75 pitchers had a K/BB ratio higher than 2.9. Of those 75 pitchers, only 6 posted an ERA- over 100: Kevin Slowey, Ricky Nolasco, Bob Tewksbury, Josh Towers, Scott Sanderson and Steve Woodard. We discussed Slowey, Towers and Sanderson in the prior paragraph. What is the deal with the other three?

Let's talk a bit about LOB% or left on-base percentage. This is basically the percentage of batters that pitchers put on-base, but then strand. Low strikeout pitchers often have troubles with LOB% because they can't just rear back and get a strikeout when they need it. This seems to be why Tewksbury was only league average. He had a 11.6% career K rate, even lower than Diamond. Nolasco and Woodard have low LOB% as well, but did not have the same issues with strikeouts, as each approached league average. Their low career LOB% could indicate that they pitch worse out of the stretch, or perhaps they just pitch poorly with runners on base. Whatever the reason, this seems to be hurting their overall performance. Scott Diamond has a LOB% above league average. Now, that could regress a bit, but if he can maintain that rate, he could continue to succeed as a low-strikeout pitcher. 

Another reason for Diamond's success was a significantly improved ground ball rate of 53.4%. Only 58 pitchers in my 1993-2012 sample posted a rate that high. High ground ball rates mean fewer line drives and fly balls, therefore leading to fewer extra-base hits. Here is a list of pitchers from my sample who posted a ground ball rate comparable to Diamond's with a walk rate in his neighborhood: Roy Halladay, Kevin Brown and the oft-injured but always effective Brett Anderson. Not bad. Each of these pitchers gets more strikeouts, but no one limited walks like Diamond did in 2012. 

So, why was Diamond more effective in 2012? Why did he get more ground balls? Why did his walk rate dip? Let's look through some PitchF/x data and see if we can figure it all out. 

A couple factors jump out at me. Diamond has a great curve ball. He gets far more swinging strikes on that pitch than is considered normal. Batters swing at roughly half of his curve balls outside the strike zone (30% is about average), but only make contact on those half the time (68% is about average). This is pretty impressive. 

His fastball doesn't generate many swings and misses at all, but the fastballs that are put in play against him are mostly ground balls. This improved ground ball rate on fastballs pretty much explains his overall increase in ground ball rate. Is this something he is doing differently with his fastball? The heat charts are not perfect because the sample was much smaller in 2011, but it seems that he is working away from left-handed batters more and working in on right-handed batters a bit more. It also seems that there is a trend toward him working lower in the zone, which certainly could explain the extra ground balls. 

There are a couple other items I learned looking at his PitchF/x data. First, he almost exclusively uses his change-up against right handed batters. He keeps it low and away and works that pitch mostly out of the strike zone. He induces mostly weak contact with that pitch, making it an out pitch of sorts. He likes to put his curve ball down and away from left-handed batters and down an in on right-handed batters. This indicates a match in approach with his fastball. Consistently keeping the ball down is a great way to get ground balls.

Looking at two specific games shows the two sides of Scott Diamond. On June 24, against the Cincinnati Reds, Diamond was efficient. He went 8 innings, struck out 7 and walked just one. PitchF/x shows that he worked the outer parts of the strike zone, kept his curve ball down, but not too far down, and ran his fastball in on righties. He kept the ball away from lefties as well. On September 16, the White Sox tagged him for six runs in 5.1 innings. Against righties, His curve ball was all over the place and his fastball was out over the middle of the plate. Against lefties, the ball was up more over the middle of the plate. 

A few other items. Diamond didn't have much of a platoon split in 2012, showing relatively equal effectiveness against lefties and righties. Lefties actually made better contact, but Diamond gets more strikeouts and walks against lefties as well. Home/road splits don't vary much either, although he did get hit a bit more on the road. Diamond's walk rate increased and his ground ball rate decreased as the year went on. This would help to explain his ever increasing ERA.

Diamond's FIP (fielding-independent pitching, which measures peripheral skill) was pretty comparable to his final ERA. His extra-base hit rate was in line with league average, as was his home run to fly ball ratio. His BABIP was around league average as well. Basically, when Diamond was effective, he wasn't lucky, he was good.

2012 Scott Diamond started hot and cooled off as the year went on. He gets a lot of ground balls with his fastball and he has a really good curve ball. He doesn't get a lot of strikeouts, but his curve ball generates a lot of swings and misses. He has truly elite control.. He can overcome the lack of strikeouts when he keeps his pitches down. He is an efficient pitcher who can go deep into games. If 2012 Scott Diamond is the real Scott Diamond, the Twins have found a uniquely effective pitcher who can defy a lot of widely-held notions about starting pitchers.