Lately, projecting future performance has become a trendy pastime; it seems like everyone and his uncle has a crazily acronymed projection system (PECOTA, CHONE, MARCEL, ZiPS, CAIRO, and of course BILL JAMES--I wonder what that stands for?). These systems are useful tools, but when it comes right down to it, they are just algorithms based on certain assumptions, and thus not really worth discussing or analyzing.
The system that I find most interesting is FanGraphs' Fan Projections, which isn't really a system so much as a poll. Basically, anyone with a (free) FanGraphs membership can vote on how they think a player will do in 2010. These votes are then combined to give a projection for that player. This is the first year FanGraphs has done this, so there are a few kinks in the system. The most obvious is that some starting pitchers--particularly those also predicted to pitch in relief--have an inflated number of projected starts. For instance, the last time I checked, Joba Chamblerlain was projected to make 39 starts AND 13 relief appearances. Still, the numbers for the most part seem reasonable, and provide a good sense of the consensus for how players will do in the coming season.
In this post, I am going to look at the Fan Projections for all the starting pitchers in the NL East. Here at TC, most of us probably assume that the Braves have the best top-to-bottom rotation in the division. Are we being overly optimistic, or do the FanGraphs voters agree? The answers (and lots of fun graphs) are below.
The Best Rotation in the NL East Is...
First, a note about the process. I compiled a list of SPs in the division who had at least 8 votes (most had far more) and were projected to make at least 10 starts; this ended up being 6 or 7 SPs per team. I then compiled a 5-man "rotation" for each team based on the highest projected WAR* values from among these pitchers. The only exception to this rule was Oliver Perez who, despite projecting as the 6th-most-valuable SP on the Mets, will surely be in their rotation (you don't put a guy making $12M in middle relief).
* For those of you who don't know about WAR, it is basically a measure of how many "wins" a player is worth compared to the average AAA replacement player. Since each win is worth about 10 runs, you can just multiply any WAR value by 10 to get the number of runs a player is worth. A star gets about 6 WAR, a good player gets about 4 WAR, and an average player gets about 2 WAR. Garrett Anderson was worth -1 WAR in 2009.
So, who has the best rotation in the division? I summed up all the WAR values from each team's 5-man rotation, and these are the results:
The Braves' total WAR of 18.3 is a full 2 wins better than the Phillies' total, almost 3 wins better than the Marlins', and more than 4 wins better than the Mets'. The Phillies do have the best projected pitcher in the division (Halladay), but they also have a black hole in their 5th-starter spot. The Braves have only the 3rd-best #1 Starter, but they have the best #3, the best #4, and the best #5. In fact, our 5th starter (Kawakami) projects to be better than the Marlins' 3rd starter (Volstad) or the Nationals' 1st starter (Lannan). That's pretty solid.
Overall, I'd say these projections match up almost exactly with my perceptions of the relative strengths of the NL East rotations. I agree that the Braves clearly have the best staff, followed by the Phillies, Marlins, and Mets in that order, with the Nats a distant 5th.
Of course, a lot can change during the season, between injuries, trades, surprise performances, and call-ups. This is not a prediction; it is only a projection based on how things look right now. For instance, it's certainly possible (though unlikely) that Strasburg comes up early, pitches dominantly, and makes the Nats' rotation the best in the division. Based on the way things stand at the moment, though, I think the fans did a very good job.
Below, I examine the projections for each team in more detail.
For each SP in the sample, I list the # of votes cast for that player, plus his projected starts, innings pitched (IP), ERA, and WAR. I also included 4 of the "per 9 innings" rate stats: hits (H/9), walks (BB/9), home runs (HR/9), and strikeouts (K/9).
A few caveats:
- Though each starter is ranked by WAR, I wouldn't worry too much about the order. There's a lot of wiggle room in these numbers, since they aren't, y'know, real.
- I would also not worry too much about the projected number of starts, particularly for the figures that are italicized (italics indicates that the player is also projected to make at least a few relief appearances, thus skewing his Starts total a bit).
- Finally, you should keep in mind that the projections with fewer votes are more likely to be outliers than those with more votes, so if a player has fewer than 15 votes, take his projections with a grain of salt.
Let's start with the hometown team.
All of these projections seem pretty reasonable to me. Hanson should do well, though he could of course do even better than his projection--the sky's the limit for Tommy. Then again, you can never tell with young pitchers; he could just as easily underperform. Jurrjens' projection seems dead-on, assuming he's able to stay healthy all year (knock on wood). The fans seem to think Lowe will have a nice bounce-back year, which makes sense to me, though at his age, this projection may be a tad high. Hudson's numbers all seem reasonable; the key for him will be staying healthy; if he makes all his starts, he should be in the 4-WAR range as well. I'm optimistic on Kawakami and Medlen; I think they'll both beat their projections, though neither WAR value is far off from what I'd go with.
I combined the projections from the Braves' top 5 pitchers and compared those numbers to the projected league-average numbers. The graph below gives you a sense of how the Braves compare to the averages in 5 rate categories: Innings Pitched per Start, H/9, BB/9, HR/9, and K/9.
To read the graph, keep in mind that closer to the center is always worse and closer to the edges is always better. In other words, the bigger the shaded area, the better. The percentages given are relative to the league average. The Braves' rotation rates as "+8%" in the H/9 category--this means that they project to be 8% better than the league average at preventing hits, not that they will give up 8% more hits.
The one stat that stands out is that the Braves project to be 23% better than the league average at preventing home runs--a phenomenal number. In fact, every Braves starter projects to be below the league average of 1.0 HR/9, and the top 4 project to be well below that number. Ground-ball pitchers + pitcher's park = great home run prevention. Given our lack of strikeout pitchers (aside from Hanson), the ability to prevent HRs will be critical to the success of our staff. The Braves also rate as above average in preventing hits and preventing walks, and about average in IP per start.
(I'm going in alphabetical order by location, not in order of projected finish, just FYI):
The Marlins have a great 1-2 punch in Johnson and Nolasco, but there is just a huge drop-off after that. All the talk about all the great young Marlins hurlers has mostly been about potential, and most of those guys have yet to really harness their potential. Clearly, the fans doing these projections would have just as hard a time filling out the Marlins' 3-5 slots as the Marlins themselves appear to be having. There's really just not that much difference between Volstad, Miller, West, and Sanchez (another pitcher, Rick Vandenhurk, will likely get some starts, but only has 5 votes, so I didn't include him). All of them are talented, but none have had much success in the majors. Any of those 4 guys could potentially blow away their projection--but then, any of them could also easily end up performing at or below replacement level. If the Marlins are to contend, they likely need at least 1 guy from that group to step up and put up a 4-WAR type season.
Here's the Marlins' graph:
Basically, this is a completely average staff. A bit better than average in HR prevention (thanks largely to whatever that stadium is called now) and strikeout rate, but a bit worse than average in walk rate. This fits with the perception of the Marlins' staff--a bunch of good arms who have trouble with their control.
What's interesting here is how badly Maine and Perez project; clearly, fans have little confidence in these two, despite all their talent. Maine is basically the epitome of an average pitcher, while Perez, well, he's a disaster area. The fans actually project Kelvim Escobar to have more WAR than Perez in about half the innings, though I think there is virtually no chance that the Mets give Escobar any starts, due to concerns about his shoulder. Pelfrey and Niese project fairly well mainly because of their decent walk rates and excellent HR rates (thanks, Citi Field!). As for Santana, he is certainly capable of surpassing these projections, but given his health, I think it's fair to project him to have a very good, but just short of dominant, year.
Here is the Mets' graph:
The Mets' main strength is their HR prevention, though A) they aren't nearly as good as the Braves in this department, and B) much of that strength is due to Citi Field. The Mets also have the worst projected walk rate in the division--a terrible 12% below league average. I guess that's what happens when you let Oliver Perez make 30 starts. Classic Minaya.
The Phils have an excellent top 4, headlined by the amazing Mr. Halladay (check out that projected walk rate--very Madduxian). I think the projections for all 4 are spot-on. Things get a bit dicey, however, in the "battle" for the 5th spot. Basically, they have a bunch of terrible options. Seriously, any time 2010 Jose Contreras is the best of 3 available options, you're in bad shape. They'll probably end up going with Moyer until they can't take it any more, and then they'll probably call up some middling prospect from AAA. Kendrick is nothing special (I actually think his projection is *optimistic*), and Contreras belongs in the pen. The Phillies have to be considered the favorites in the division, but they do have some glaring holes (the bullpen is another).
Here's the Phillies' graph:
Despite the lack of an adequate 5th starter, the Phillies have a good staff overall; they project to give up some HRs, but part of that is due to their ballpark. They're about average in K/9 and IP/S, better than average in H/9, and project to be a fantastic 17% better than average at preventing walks, thanks mainly to Halladay and Hamels. Considering their HR-friendly ballpark, preventing baserunners is very important, so Phillies fans should feel good about these projections.
Finally, for the sake of completeness, here are the Nationals' projections:
Let's see... The Nats have 5 guys who project as #4 or #5 starters. Well, at least they have good depth! The sad thing is that they are likely to give Livan and Olsen a bunch of starts; that's a terrible idea, though, as both those guys are just terrible, and don't project to be better in 2010. The one source of upside in this list is Wang, who could certainly have a big comeback. The fans don't seem to think that will happen (neither do I), though his projection is decent. If he gets 30 starts, he's probably their best pitcher in terms of WAR. Lannan is what he is: a decent pitcher, but not really more than a #3 at best. I find it interesting that Mock and Marquis project to be so similar, especially since Mock might not even make the rotation and Marquis just got $15M to be their "ace." Barring a trade, Marquis' playoff streak will be ending this year.
Here is the Nats' graph:
Wow, that is ugly. Well below-average in 4 of the 5 categories, including an abysmal 23% below average in K rate. Aside from Garret Mock, none of their pitchers projects to strike out even 6 batters per 9 innings (the league average is projected to be 7.2 K/9). That puts a lot of pressure on a defense, which is not something you really want to do when your first baseman is Adam Dunn. The one bright spot here is the HR prevention number, though again park effects influence that somewhat. (We sure do have a lot of extreme pitchers' parks in the NL East--I wonder how many HRs the Braves would hit in another division.)
According to the fans at FanGraphs, the Braves have the best rotation in the NL East, at least on paper. Do you agree? How would you rank the rotations in the division? Which players do you think will outperform their projections? Which will underperform? Are these fan projections better or worse than the other major projection systems, like CHONE or Bill James? I'd be interested to hear what the TC community has to say.