June 14, 2009
Prospectus Idol Entry
Every year when the PECOTA projections come out, I like to sort through some of the statistical leader boards to see which players PECOTA likes for the upcoming year. One of my favorite categories is ERA because it gives me an idea of how well certain pitchers may do within the context of their teams. I am also on a sort of Easter egg hunt. I figure that PECOTA has data on hundreds of starting pitchers from professional baseball, and when it projects ERA (or anything else), it will find the usual CCs or Johans, but it might also like a rookie or another player that I'm not expecting as much.
This year I found someone who was not only projected to be the best rookie, but also, one of the more successful pitchers at any level of experience. This was neither Tommy Hanson nor David Price. It wasn't the Brett Anderson or Trevor Cahill that people might expect either. Rather, it was a far less famous Toronto prospect: Brett Cecil. PECOTA has him at 3.97 and even that is probably understating the case. PECOTA doesn't exactly give away ERAs under four like they're government bailouts. That is the 29th best ERA for a starting pitcher and it gets even better when you consider that some guys ahead of him-like Sheets and Smoltz-aren't even around.
Scouts don't tend to share PECOTA's enthusiasm. He isn't even ranked as the 29th best prospect. He lands at 72 in Baseball America's top 100 and 90 in BP's. These sources peg him as a "no. 3" or a "middle-of-the-rotation starter." I should mention that I'm not interested in starting a scouts vs. stats argument here. I could cherry-pick hundreds of players that make one or the other look good. Furthermore, the scouts are basing their estimation of him on a lot more than this year's possible ERA in Toronto, and PECOTA is accounting for a lot more in its projection than his raw ability. However, there is still enough of a disconnect here that it is valuable to explore who the real Brett Cecil is.
Brett Cecil played at the University of Maryland before being drafted, and even while in college, his "body, arm action and stuff improved significantly" according to the Baseball America Prospect Handbook 2008. This praise is even more impressive when you realize that it was for a kid who couldn't even legally order a drink in a bar at the time. The most important thing to remember about his collegiate career-and it may still affect him today in many ways-is that he mostly performed as a closer in college. Toronto drafted him as a supplemental first round pick in 2007 and quickly converted him to a starter as they've done with Shaun Marcum and Dave Bush.
Cecil had two plus pitches, a good slider and a sinking fastball, which he used to get plenty of groundballs, but once he was a pro, he learned that he could better compete by adding a change-up to his arsenal. As he told BP's David Laurila earlier this year, "I worked on about 15 different grips, and this past year I finally found one that I've been able to strike guys out with." This is perhaps the very reason he was able to strike out 56 batters in 49.2 innings playing Low-A ball in 2007, elevating his status as a prospect.
In 2008, he continued to K guys at a rate that translates to a major league 6 or 7 strikeouts per nine, which explains PECOTA's projection of 6.5 K/9 for 2009. Similarly, he had good walk rates that translated to about 3, again leading to a comparable PECOTA projection.
It's great that he can strikeout his share of batters and isn't hurting himself with the free passes, but this is an average pitcher's skill set. As Kevin Goldstein noted, "While he has a deep arsenal, he doesn't have a true wipeout offering, and he'll always need a good defense behind him." We'll get to that defense in a moment, but first let's talk more about his arsenal and what it is that PECOTA sees differently.
Scouts view him as a less than dominant pitcher, which is absolutely true, but what he does while not dominating still leads to some very successful results. Cecil keeps the ball on the ground. Well, great, we've heard that groundballs are good before, but let's examine exactly how valuable they are. Cecil has had groundball rates between 60 and 70% throughout the minors. Based on this, PECOTA projected a 55% rate for the majors. Typically, a 55% success rate will lead to a rate of about 0.8 home runs per nine innings. PECOTA likes his potential even better than that and gives him 0.6 HR/9 but let's err on the side of conservatism and go with the 0.8 for now.
The average pitcher gives up around 1.06 or just over one HR/9. According to our projection, Cecil will allow 0.26 HR/9 less than an average pitcher. To know how valuable this is, we first need to know what a home run is worth. This is not as simple as seeing how many men are on base when the typical home run is hit because some of those runners would have scored anyway. In The Hidden Game of Baseball, Palmer and Thorn found that home runs in the modern era are worth about 1.42 runs. Their study only ran through the 1977 season, though, so the more recent emphasis on OBP probably leads to a higher figure. However, because we are concerned with the impact on ERA, some runs won't count due to errors so the figure isn't going to be too much higher.
Comparing starting pitchers over the last 8 years and using methodology similar to that which I used here, I found that each HR/9 affects a pitcher's ERA by 1.44 runs. In the case of Cecil, the 0.26 HR/9 better than average we found earlier would drop his ERA by 0.37. In recent history, pitchers who have been good enough to pitch at least 150 innings have averaged walk and strikeout rates very similar to what we projected for Cecil and the typical home run rate of 1.06, while managing a mean ERA of 4.17. Subtracting the 37 points, Cecil's projected ERA comes down to 3.80.
That is just looking at Cecil's raw ability. Scouts also look at a player's long term potential and durability. We'll get to those a bit later. Meanwhile, our initial PECOTA projection looked at Cecil through the lens of context. Therefore, it's important to look at Cecil's ballpark and his defense. According to The Bill James Handbook 2009, the Roger's Centre increases home runs by 13%. When we increase his home run rate by 6.5% (cutting the 13% in half due to Toronto's away games), it comes up to 0.85 and the corresponding ERA stands at 3.87.
At the same time, the Toronto defense should help Cecil. Last year they finished third in the majors at .704 and John Dewan's +/- system had their infield at a +50 while the outfield broke even at 0. Remembering that Cecil is a groundball machine and that Toronto is bringing back that same infield, projecting their defense to be at .710 in his 2009 games seems very reasonable (their team rate is .707 at the moment). Applying my defense to ERA conversion, which I outlined in my first article, it improves his ERA by another 0.17. Using these methods, we get an even better projection than PECOTA: Cecil's Toronto ERA projection for 2009 pulls up at a handsome 3.70.
After looking at these numbers, we should assume that Cecil went out there and provided a terrific performance for the Jays in 2009, right? Well, not quite. Cecil has run into some problems so far this year. After being named a candidate for the Jays rotation and then pitching exceptionally well in the Grapefruit League, Cecil wasn't chosen for the rotation. I'll resist the urge to go on my rant about spring training 'competitions' that some of the 'candidates' have no chance to win, but in the case of Cecil, it may have been the right decision; his early season numbers from Triple-A were less than desirable, including a 9:8 K:BB ratio. Nevertheless, the Jays rotation had been decimated by injuries by the end of April, and they called up Cecil. After a six day layoff, he made his major league debut on May 5th.
Cecil came bursting out of the gate just as our previous analysis might have suggested. In his first two appearances, he logged 14 innings and only gave up one run without any jacks. The next start might have been chalked up to just not having his stuff. He gave up two homers and had an unimpressive K:BB ratio of 3:2.
His fourth start was against the Red Sox. He pitched relatively well for the first four innings, but then he was nailed in his non-pitching arm by a come-backer in the fourth. The fifth inning, if not historic, was at least something Cecil will never forget. His big skill was supposed to be killing gophers, but Jason Varitek led off the inning with a long ball. Four batters later, after a walk and a double, Ortiz hit a shot to deep center-his first homer of the year. The fun didn't stop there. Bay and Lowell also hit homers before Cecil was taken out of his misery.
After the game, Cecil was sent back to the minors. My initial reaction was that his arm injury caused the problems, but further analysis demonstrates a bigger problem that scouts have always had with Cecil: a lack of stamina. Looking at his PITCHf/x data, while his cupcakes that led to the jacks were up in the zone, he had the same movement on his pitches as he did in the other innings. The bigger problem seemed to be that his velocity had steadily decreased throughout the game. This holds true for all of his major league starts. The following chart, courtesy of brooksbaseball.net, shows Cecil's velocity throughout his first start against Cleveland:
You can see that while Cecil's fastball reaches as high as 95 mph to start the game, it slows down by 4-5 mph over the course of the start. That is not the only way Cecil shows fatigue. Will Carroll spoke to me about how he often throws slower in the second inning of consecutive starts (using the second inning to "remove 'nerves' and adrenaline from the equation as much as possible"). Indeed, Cecil regularly threw 93 in the first few innings here (after the six-day layoff), but he was only topping out at 92 in the following start, and he never again threw as fast as 95 during any of his other starts in the majors.
Toronto has tried to combat his stamina issues with hard pitch limits, but Carroll suggested they use a different approach: "Cecil would be the perfect type of pitcher to use a logical progression development on, adding pitches slowly to his pitch limits as he shows that he can pitch successfully and recover. At some point, he'll max out-that may be at 120 [pitches] or 50, but they'll know."
Baseball America has praised Brett Cecil for his work ethic and his "bulldog mentality", but he may have his back up against a "genetic maximum" that likely exists, according to Will Carroll. The Blue Jays could easily make him into an excellent relief pitcher/closer, but he probably has more value as a starting pitcher. What may be in order is some creative usage, like pushing him back whenever the team has an off day. However they do it, if the Jays can get him in the rotation without the stamina issues, he has the skills to become one of the premiere pitchers in baseball.