clock menu more-arrow no yes mobile

Filed under:

Thursday Rockpile: Rox Girl is slightly afraid of Snakes, you should be too.

Okay, I spent all morning writing a drawn out response to Andrew Fisher's Diamondbacks post yesterday, so for links you just get bullet points from the Denver Post from me. If you've noticed some more yourself, feel free to post them in the comments.

 

 

I'm feeling a little sassy today as certain opinions I've expressed on the blog the last few months are getting assailed from different quarters including via friendly fire here at Purple Row. Time to step up and defend myself, but where to start? I think I'll go with what to me is the easy one, which is to actually defend the MLB team I dislike above all others, the Arizona Diamondbacks. Troy Renck's already written them off, and in the post just below this one, Andrew Fisher does the same. While I agree with Jeff Aberle that the Rockies are going to win the West, and I really think it's going to be a wider margin than people outside of Purple Row are expecting, overlooking the Snakes is a big mistake. Let me start with an obvious flaw in Andrew's conclusion yesterday:

 


Okay, I know this is doing open heart surgery with a butter knife if ever there was such a thing. There is certainly room for improvement above projections from some, but I think these predictions are certainly on the optimistic side of things. It assumes no injures, no personnel flux, at times includes more ABs than are possible for a given position, and assumes Kennedy and Webb stick to their workload potential.

Assuming marginal improvement from the bullpen, this roster looks to be about 7.0 wins better than they were in 2009. That equates to 77 wins. Or - if you utilize their 2009 Pythagorean, an 82-80 record. So - are you scared of the Diamondbacks?

 

Put those seven wins back into the scale they came from, rather than the scales that Andrew switched them to at the very end of his piece. In 2009, the Diamondbacks had 34.0 WAR overall according to FanGraphs. 34.0 WAR plus 7.0 WAR equals 41.0 WAR. In 2009, the Dodgers and Rockies both had 43.9 WAR in finishing with over 90 wins. If we give teams a plus or minus of five wins from their projections, which I hope you realize as you read the rest of this is a very stingy figure and putting a lot more faith in projections than you probably should, the Diamondbacks are certainly capable of winning the division. Also, they are certainly capable of finishing fourth or even fifth again.

That said, I still think they're going to finish second unless I'm underestimating the Giants. Let me spell out the reasons that I'm right and you guys are all wrong, ;)

 

  • Projections show the thinnest possible strip in the middle of what can be a large band of possible outcomes. At the end of the season, when we say teams are "surprising" or "disappointing" what we're usually really saying is that they were either above or below that thin strip in the middle by a large margin, as such, if you're looking for teams to surprise or disappoint in any given season, just looking at that mid-range projection won't be as useful as looking at the upper or lower ranges of those bands. The D-backs have a surprisingly high upper range (around 90 wins) for a team coming off a 70 win season.
  • Somewhat related to that first point, projection systems will miss certain events completely. Obviously they can't foresee season ending shoulder surgeries for ace pitchers or drug suspensions for star outfielders, but beyond this, even some on-field events go outside what they typically convey to us.That is to say, some players' performances will be in the less than 1% tails at either end of their projection bands, and since most systems only concern themselves with 25th to 75th percentile range at best, seasons such as the 2009 versions of Pittsburgh's Garret Jones and Arizona's Chris Young (to use two examples at opposite ends of the spectrum) will seem to come at us out of nowhere.
  • Because baseball is a meritocracy with a large player pool fighting for a small amount of jobs, these individual outliers and other unexpected events (the shoulder surgeries and drug suspensions) will tend to have an exaggerated impact on the standings. Player events like Jones' are likely to give teams an unseen boost as they muscle their way into previously weak lineup spots, player events like Young will leave teams vulnerable if they don't have adequate bench depth to replace them.
  • Even though projection systems by their nature can't foretell these events, does not necessarily make them unpredictable. Specifically identifying the Garret Jones' of the baseball world may be a next to impossible task, but with knowledge of teams' systems one can identify likely sources for players like this to emerge. A team with a weak MLB starting roster but a lot of useful 25 to 28 year old AAAA types backing them up will likely see somebody emerge from the latter category to take the place of the former at an average or above MLB level of play. Conversely, teams which are overly reliant on players that are outside the seasons where we'd expect near peak production (let's just say from ages 26 to 32, you can quibble with the range, but the point stands) are far more likely to suffer disappointments relative to projections than those with cores in their prime.
  • Many people find WPA as a somewhat useless stat when it comes to predicting stuff, which can make the few of us that know otherwise seem smarter than we really are. At the team level, it's a marvelous device for identifying teams that are being misjudged by either projections or the mainstream. Start in 2001 and just go through these charts year by year and look for some consistency. What you'll likely note is that good teams do tend to have positive WPA's and bad teams negative, but the range fluctuates and leaders in the category one year will often fall back to the middle of the pack or lower the next. In 2006, the Rockies had a deeply negative WPA that flipped to one of MLB's best in 2007. In 2008, the team fell back again before rebounding to the middle of the pack in 2009. Why do I feel Los Angeles is overrated and Arizona underrated? Partly because in 2009, the teams were at opposite ends of the WPA spectrum. Los Angeles was both lucky and good in 2009, Arizona was certainly unlucky, but I really don't think they were that bad talent wise, mediocre might be more like it. This will balance out, and it's likely to this season.
  • The unbalanced schedule skews our perception of teams. After 2007, the Dodgers were underrated going into 2008 in large part because of their 34-48 record against the rest of the NL West despite having a more talented group than that. After 2008, the Rockies were underrated going into 2009 because of their 31-41 record against the West. After 2009? I'm guessing it's the Diamondbacks turn after a 30-42 record and that the Dodgers are most certainly being overrated after going 46-26. Like Tommy Lasorda's Slim fast commercials, a small type disclaimer needs to be made that these results are not typical given the talent levels of either club.

 

In fact, I believe the 2009 results are fudging the data, effectively making Los Angeles appear stronger talent-wise than they really are in 2010 projections, and Arizona weaker, similarly to how some projections for the Rockies heading into 2008 may have led us to be overly optimistic. Thankfully for our team in 2010, we're passing most of these stress tests (one notable warning flag is in the category of percentage of players that had the best season of their careers) I use to try and gauge teams as "likely to disappoint", whereas I see the Dodgers as really weak in several of the categories. The Giants likewise seem like they played above their heads in 2009. The Diamondbacks, on the other hand, keep on looking pretty good in the "likely to surprise" categories, which are pretty much the opposite of the above.

Since baseball's a zero sum game, if I'm predicting those two California teams (not to mention San Diego, who I also think will be losing more than they did in 2009) to drop back significantly this season (by more than five games), some other team has to be the beneficiary. To me, Arizona is easily the most likely candidate given that the Rockies probably don't have a lot of upper growth for this season from their 92 win campaign last year (I do think we're going to be right around 95 wins, but that's only a plus three).

So next, I'm going to tell you all why you're wrong about Brad Hawpe, but that one's admittedly going to be far trickier to prove.