It’s that time of year again! Pitchers and catchers are due to report to camp very soon despite the threat of impending white doom here in the tristate. The former makes the latter bearable, at least for me as I tell myself, "it won’t be long before Old Man Winter’s icy grip melts away to the time honored sounds of Louisville sluggers striking cushioned cork encased in white Rawlings leather!" Even while shoveling the last acts of defiance of a season that can’t know the joys of hearing “Play Ball!” bellowed into the air or the delight of sharing the season’s first ballpark dog with your wide-eyed, awestruck kids, my mind drifts to the age old question oft asked both by 8 year old boys and the 8 year old boys existing within the psyche of grown men, “Is this the year for my Redlegs?”.
Anyway, that was my mindset last week while sitting on I75 for several hours during the snowstorm. In order to see what the Reds chances in ’07 might reasonably be expected to be, I turned to the five projection systems that are currently available to the public: 1) Bill James, 2) CHONE, 3) Marcel, 4) ZiPS, and 5) Pecota. There are several advantages of using these systems. First, modern projection systems are really quite advanced these days though they predict offense better than they currently predict pitching. Second, and importantly, the projection systems are devoid of fan-related biases that may unknowingly exist when examining our favorite team. Third, by taking an average of all of the projection systems (which is how I approached the question), individual biases of any single system can reasonably be mitigated. These projections then could be used to estimate RS and RA in order to determine Pathagorean wins. One last word, this was a lot of work and I tried to thoughtfully be as unbiased as possible when making assumptions so please play nice when offering critiques (but please feel free to critique!). Here’s the summary (caveats and methodology below):
[IMG][/IMG]
Wow! A winning season! The projections indicate Krivsky has assembled a collection of guys that could theoretically be above .500. This suggests that, if things break right and some key guys beat their projections, the Reds could be in the thick of a pennant race again this year. However, I'm a bit skeptical about the playoffs. First, that's a lot of things that need to go right while assuming nothing goes wrong. The Reds have little margin of error as a significant injury to a starter or two could doom the Reds to a losing season. Second, projections for Griffey/Dunn don’t consider the shift. Both are predicted to rebound from last season but is there any reason to believe that the shift won’t mess with their BABIP’s similarly this season? Third this assumes Homer spends at least half a year in the bigs and other decisions dependent upon Narron. For instance, will he play Denorfia enough to meet his projections? Will the roster have Moeller on it?
All that being said, there is reason to hope. If the Reds are in contention in July, the Reds might be able to get over the hump with the addition of a rent a player or perhaps small tweaks to sure up a weakness. My take on the 2007 Reds? If Krivsky was seeking to build a roster that hits the league mean with the hopes that the chips fall well and they catch lightning in a bottle, then he's achieved that. If 90 wins are required to win the divsion, then reds fans should expect to be in the division race until deep into September. Wohoo! I'm looking forward to some more discounted tickets and dollar hotdogs.This could be a good season or a season to forget. Most likely it will just be an average season with reason to hope for more. There will be plenty to cheer about.
So a few words about my methodology and some caveats:
1. Obviously, it is very difficult to predict what the roster will look like. I tried to make the best estimate of how the Reds will emerge from spring training. Also, there are more than 25 players listed above. Some are guys that will likely take the ride on the AAA shuttle during the course of the season so the projected 25 man roster reflects some fluidity.
2. I’m assuming we’ll see Homer for roughly half of the season and I’m giving the nod to Saarloos as the 5th guy/swing man given his salary and the comments during recent press conferences that Lohse and Milton are penciled in at #3 and 4 in the rotation.
3. A few words concerning the parsing of AB and innings. I first determined the average projected performances (averaging all five projection systems) for players on the projected roster. Using regression analysis, total team AB’s and IP by the starting rotation were estimated from national league totals from 2002 through 2006 by using the '07 Reds projected performance levels to find similar teams during the past. Once IP by the starters were determined, subtracting them from 1450 determined innings by the bullpen. I can post the equations but suffice it to say that I could predict a given team’s AB within .5% for any given year for instance. So once, these targets were determined, I examined the projected averages to see if individual players AB or innings needed to be adjusted. Fortunately, the projections were almost dead on with a few notable exceptions. For hitting, I decreased Hamilton’s at bats in order to give Valentine an addition 77 Ab. This was necessary in order to raise the total projected AB for Reds catchers to a reasonable expected number. Also, Elizardo’s projected innings were reduced from slightly over 100 to 61. Given these limited tweaks, I’m confident that personal bias didn’t dramatically influence the RS and RA estimates (other than decisions concerning who would be in the roster of course).
4. In my original analysis, defensive estimates using PRM to runs (using ’06 data) were also included. However, for the projected roster, the net was -1 run. Given the uncertainty in the defense of the roster (i.e Griffey in RF, actual value of Denorfia and Freel etc), I felt pretty comfortable considering the roster neutral defensively and dropping defense from the equation for simplicity.
5. No projection system is destiny-all have built in biases and flaws. If there is sufficient interest, I'd like to initiate a redszone fan-based estimate drive to compile a similar set of projections during spring training to see how fan estimates stack up against the projection systems. Hopefully this will be fun.
GO REDS!
Originally Posted by dougdirt
