Wednesday, July 9th, 2014
As we watch prospects begin to develop or even see new players acquired via trade or free agency, when it comes to starting pitchers, everyone is always quick to classify the player into a spot in the rotation. We hear things like “He projects to be a low end # 2 or high end # 3 starter” or “He’s an ace.”, but what exactly does that mean? I decided to do a small bit of research as part of my Sunday afternoon and dive into numbers to see what exactly the spots in the rotation actually look like. There were a few things I found that I felt were fairly interesting.
If you’ve not ever played with it, the Baseball Reference Play Index Tool is an amazingly useful tool whenever you feel the need to dork out and immerse yourself in numbers. It requires a subscription, but it’s well worth it. For my study, I used the tool to gather data from 2000 – 2013 in a quest to find out what exactly an ace pitcher looks like.
To put a qualifier on the data, I filtered to only pull pitchers each year who qualified for the ERA title, which means they pitched at least 162 innings in the season. This basically gave me a group of pitchers that, essentially, pitched the entire year. They may have missed a few starts here or there, but for the most part it weeds out guys who either split time between the pen and the rotation as well as guys who pitched a small amount of time and either got hurt or replaced due to inefficiency. To make this list, you really had to be a starter all season long. All of my data was then collected and dumped into a spreadsheet where it could be massaged and summarized into useful nuggets of information.
If you’re like me, you feel a little frustration when your team can’t see five consistent starters in the rotation all season long. If you’ve been reading my blog at all over the last 10 years, I make no secret that I hate bullpens with a passion and believe strongly in filling them with guys from your system and spending money on other components of the team. However, exactly how feasible is that request when it comes to the rotation? I was a little surprised by what I found. Looking at the chart, you will see the number of players who pitched at least 162 innings in each season.
What this data showed me is that, to put it simply, there really aren’t enough guys who can give you 162+ innings to fill out each rotation in baseball. Over the last 14 seasons, the average number of guys who met the requirement per season was 87. That’s less than three per team. Stop and think about that for a second. If all was distributed equally, each team would only have two, maybe three pitchers who pitched enough innings to qualify. That’s insane to me. What’s even more insane is that we know that the pitchers are not distributed equally, which means there is the potential that a team could have no pitchers who meet the requirements. There simply isn’t enough talent there.
The other thing I noticed was that there really doesn’t appear to be any sort of trend with this data. It’s pretty random across each year as to what you’ll get. Some years we see over 90 guys and other years we see just below 80, but things seem to hover in the mid-80’s.
Now that we have the knowledge that there aren’t enough quality starters to go around, we have to figure out a good way to then classify pitchers. Seeing that we don’t have enough guys to simply take 1-30 and call them aces and downward through 150, I decided that I would simply divide the average number of qualifiers by five, the number of spots in a rotation, and group guys into one of those five tiers. What I got in terms of results was actually quite nice.
For sorting purposes, I decided to rank players by WAR, which summarizes nicely into a single number that tends to lend itself perfectly to Cy Young award / MVP voting. Generally the higher WAR guys tend to be right in the mix for those awards, so it seemed the most logical stat to use. After taking all the results, I organized it into a summary that showed the average WAR by spot in the rotation, assuming a 17 slot tier for each spot in the rotation. For the 5th spot, I used 17+ players as some years there were leftovers. The result was fascinating.
There were two main things that stood out to me about this data that I thought were worth pointing out.
First, we need some perspective as to what the scale is generally considered when it comes to WAR. Fan Graphs does a nice job explaining the stat and providing some context.
Aces, should you be able to get one, are huge difference makers. – A true # 1 starter is a guy who is capable of winning the Cy Young award when all is said and done. These are guys that are losing streak stoppers. They are guys that the other team comes into the game essentially expecting to punt and try again the next day. However, how difficult is it to get a guy who pitches so well that he compiles a WAR at 6+?
Seeing that I write a Cubs blog, I decided to look back at the rotations we’ve seen in the past and see what we find. I looked only at the expansion era (1961-2013) and came up with a total of just 15 instances where the Cubs had a player who did this. Here is the list:
It’s not as easy as you might think, so when it happens in Chicago, we need to pay attention. That’s what made 2003 so special. Watching guys like Mark Prior and Kerry Wood pitch was something I’ll never forget. That year was wonderful. Most teams don’t have a single guy with 6+ WAR and we had two of them that season. However, what’s even crazier is that, aside from Fergie Jenkins, we haven’t had a guy actually repeat this for more than a season.
The 5th starter spot is essentially garbage. – Looking at the chart and the legend, we see that the average WAR of 0.2 is essentially replacement level. Again Fan Graphs comes in handy to define this term:
We can define a replacement level player as one who costs no marginal resources to acquire. This is the type of player who would fill in for the starter in case of injuries, slumps, alien abductions, etc. If we use replacement level as a baseline for our ‘runs above <x>’ statistics, we introduce durability into our statistic, and we’re therefore measuring marginal productivity, which is what we want.
Basically this is the equivalent to calling up that organizational filler type player and putting him into the starter role. That got me wondering if it was even worth it to go with the five man rotation. Would it be better to find a way to pitch on a four man rotation in an effort to get more out of the quality guys and simply use them less innings or is it better to go with the five man rotation? I’m not sure what the answer is there, but I’d like to know. Ultimately, it shows me that if you can have even a replacement level type guy in the 5th spot, you’re probably ahead of the game. Essentially just minimize the damage in that spot.
The other thing I found interesting was that over the last five seasons, there is a clear downward trend when it comes to the quality produced by the fifth spot in the rotation, and I don’t know a clear reason to account for that. It could just be small sample size and in the grand scheme we’d see it balance itself out, but I’m not sure. In the 14 year period, 2012 & 2013 were the only years we saw a negative production from the guys who were the back end guys. I feel like there is something there, I’m just not sure how important it is.
After collecting all the data and summarizing it into the chart that shows us what we can expect from each slot, I decided to go back to look to see how the Cubs grade out since 2000 on each of these spots in the rotation based only on guys who qualified for the ERA title with their WAR in parenthesis.
2000 – (65-97 record) – Jon Lieber (3.7) & Kevin Tapani (1.2)
2001 - (88-74 record) – Lieber (3.9), Wood (3.3), Jason Bere (1.7), Tapani (1.0) & Julian Tavarez (-0.1). (NOTE: Tavarez missed the 162 inning mark by 2 outs, so I included him anyway.)
2002 – (67-95 record) – Matt Clement (4.4) & Wood (4.3)
2003 – (88-74 record) – Prior (7.4), Wood (6.2), Carlos Zambrano (5.5) & Clement (2.8)
2004 – (89-73 record) – Zambrano (6.7), Clement (3.7) & Greg Maddux (3.2)
2005 – (79-83 record) – Zambrano (5.6), Prior (3.6) & Maddux (2.8)
2006 – (66-96 record) – Zambrano (5.2)
2007 – (85-77 record) – Ted Lilly (4.1), Zambrano (3.4), Rich Hill (3.4) & Jason Marquis (0.8)
2008 – (97-64 record) – Ryan Dempster (7.0), Zambrano (4.3), Lilly (4.1) & Marquis (2.5)
2009 – (83-78 record) – Lilly (5.0), Randy Wells (4.2), Dempster (3.5) & Zambrano (3.0)
2010 – (75-87 record) – Wells (3.2) & Dempster (3.0)
2011 – (71-91 record) – Matt Garza (2.8) & Dempster (0.8)
2012 – (61-101 record) – Jeff Samardzija (1.8)
2013 – (66-96 record) – Travis Wood (4.4), Samardzija (1.0) & Edwin Jackson (-1.3)
Finally, there were some interesting trivial things that I noticed doing my research that make for fun nuggets.
- In 2000, Pedro Martinez was the top WAR pitcher with an astounding 11.7, which was 3.6 higher than the 2nd highest qualifier, Randy Johnson. It was also 7.7 above the 2nd place finisher, Tim Hudson, in the Cy Young race in the AL.
- In 2011, Tim Lincecum finished the season ranked 17th, which put him in the top tier, “Ace” category. Just the next year he finished, not only in the bottom team, but with the lowest WAR of all qualified starters in baseball.
- In 2010, James Shields did the opposite and finished with the worst WAR only to rebound the following year and be ranked in the “Ace” tier.
- 2007 was the only year in my data that did not yield a pitcher with a 7+ WAR.
2013 yielded the lowest WAR of all the seasons examined and was turned in by Edinson Volquez, who posted a dreadful -2.4 WAR. What does that look like? Try a 5.71 ERA over 32 starts.
This article was originally published on the ESPN SweetSpot Blog.
Like what you see here? Never miss new content. Follow Us: