Monday, November 30, 2009

Rox Talk - WAR Part II

WAR - What is it good for? - Part 2

So last time we delved into batting wins above replacement. A convoluted system starting with a batter's wOBA and ending with runs above replacement which subsequently is reduced to wins. In the NY Times article, it stated that Zack Greinke's desire was to have a WAR value of 10. If he had done it (he achieved a WAR of 9.4), he would have been the first to do so. To put his achievement in perspective, the highest offensive WAR ever achieved by a Rox is 7.9. An even more stunning comparison is to note that the 2004 Rox team's pitching staff eked out a whooping WAR of 6.5 (that team had a record of just 68 - 94 and this year's KC record was 65 - 97 so if every team should win approximately 48.5 games based on just AAAA replacement player's performance then Greinke single handily had almost 9.5 of KC's other 16.5 wins above replacement!).

So where does the madness begin? With another esoteric statistic called Fielding Independent Pitching or FIP. The FIP calculates a pitcher’s responsibility for the runs he allows based on the three factors that a pitcher has demonstrable control over which is walks, strikeouts, and homeruns. The FIP is used because it provides a context neutral formula to a pitcher's performance similar to that of the batter's. In this way the WAR for a pitcher can be compared directly to the batter. The FIP also takes out the contribution of the fielders behind the pitcher both good or bad. Using FIP has some detractors. Another stat called tRA has some believers but it is suggested that at this point, using FIP is as good as anything else out there (or at least comparable). Once FIP is established, some hocus pocus entails using differences in replacement level for each league and role (AL is harder to pitch in then the NL), run environments (a pitcher due to his ability limits the numbers of runs scored per game), the dynamic runs-to-wins conversion, and park factors (Coors Field vs. Petco). Finally a runs above replacement is generated and once again converted to WAR.

So the spreadsheet below shows the top FIPs in Rox history (min 50 innings pitched) and the top WARs.

FIP leaders tend to be your dominant relief pitchers. Ubaldo 2009 campaign is the first starter but it is also the highest WAR in Rox pitching history. 2009 was a monster year for Rox pitchers. Four of highest WARs since 2002 were from this year's staff. Rox won 92 games, there WAR total for a team was 90.3 and so the Rox had approximately 42 wins above a replacement level team. Almost 45% of the team's wins above replacement could be attributed to last year's starters. With Ubaldo, Cook, Hammel, de la Rosa (?), and Francis (?) back next year I can say with some certainty that as the pitching goes so goes the Rox!

Next I will check out what all this WAR means from a team perspective...

Monday, November 23, 2009

Rox Talk - WAR Part I

WAR - What is it good for? - Part 1

There has been a big seismic shift in the world of baseball this week. With Greinke and Lincecum's Cy Young award wins this past week, the world of statistics has firmly kicked the old baseball school purist in the groin. To think that a 15 game winner would lose out to a 19 game winner back in the day would be pure heresy. As the NY Times put it, the pool of voters are changing which is causing a shift away from traditional voters to a different and more statistical attuned group. This is probably similar to the front offices in baseball that are getting away from the scout centric view and moving more into a numbers only view of players. First there was Moneyball and Billy Beane and then the Red Sox hiring of Bill James a number of years ago with Epstein bringing with him a new look front office which may (or not) have been the reason for the Red Sox burying the curse in 2004. However you cut it baseball is moving in a new direction.

So what about this WAR? It stands for Wins Above Replacement. It is a complex formula (to say the least) that attempts to quantify a player's worth in comparison to that cheap triple A call up. In addition the WAR is calculated different for offensive players and pitchers (see Part 2). For offensive players it is calculated by providing a "number" for batting (using wOBA, wRAA, and then a park adjustment factor), fielding (using UZR with no factor involving outfield arm skills and all catchers are rated at 0), a positional adjustment (good shortstop is harder to find then a good left fielder), and a replacement factor (that deals with how many plate appearances and awards players who play everyday). These four factors are then added which provides a runs above replacement value. This number basically gives the runs a player would provide over a replacement player. Remember a replacement player is that AAAA player who is playing for the league minimum. Once this number is determined it is then divided by 10 (again a determined number which indicates a 10 run change in a team's runs scored/runs against suggests that one win is 10 runs). Thus the player's runs above replacement level is divided by 10 to provide the almighty WAR.

Now if you want to get even more advance you can factor in how much teams payed for free agents last year and by determining their WAR over the years you can determine a going rate for a win. Therefore one can valve what a player with a 5 WAR is in comparison to a 2 WAR player. Again this ain't your father's statistics. These numbers are becoming so esoteric that one wonders if we lose sight of the diamond? One can view this page to see 2009 wOBA for our Rox as well as WAR.

Basically Tulo led the team with 0.393 (over 0.400 is considered a superstar and over 0.370 is all star caliber). He was followed by Helton, Hawpe, CarGo, and Smith (and why wasn't Smith an everyday starter?). The spreadsheet below provides the top 50 wOBA in Rox history. Larry Walker was superman...wonder if he will even get a look see with the Hall of Fame?

Also in the spreadsheet above is the top 50 WAR in Rox history (note: data only exists back to 2002). As you may recall I am not a huge fan of Matt Holliday but boy for two years he was worth every penny and more for what he did on the field for the Rox. It will be interesting what he does this off season.

Monday, November 16, 2009

Rox Talk - Link Dump

Link dump kind of week...hey this is the off season...simpler than doing my own calcs ;-)

Brad Hawpe...who knew he was such a terrible right fielder? I always thought he had at least a good arm (guess 2006 really stuck in my head). Time for him to move on. Mr DH

On to some positive things...the Rox were the best base running team in the league last year. The statistic is based on base running metrics developed by former Colorado resident Dan Fox. Dan has since moved on to the Pirates helping in their IT department and player development (as an aside we published this together back in 2007). Fowler's development and a full year of Cargo and EY should only increase possible mayhem on the bases next year.

OK here is another infamous Coors debacle...pre-humidor. Since 1954 there have been two games with 5 blown saves...guess where?

Saturday, November 14, 2009

Rox Talk - Minor League Winning Percentage

Minor League Effects (?)

OK for the most part the Rox have decided to become the NL version of the Minnesota Twins. That is, develop your minor league system, field the majority of your big league squad with that talent, and when they become free agents trade them away for more minor league stock. A decent model to follow since there will always be big money teams eager to throw big dollars at proven commodities while sending along unproven minor leaguers. So if this is the case and you have a loaded (or competitive?) minor league system, then shouldn't eventually your major league team see spikes in their winning percentage? If your loaded AA wins the Texas League championship, then shouldn't that same talent be fielded and face similar players in the big leagues, eventually? Of course if your AA team stinks perhaps you have no big league talent and thus they never reach the big leagues but then you would still have an age gap for a few years against those teams...Then of course there are those teams that develop talent but not enough at the right time and end of trading it away (think of the Pirates trades over the last few years, had they developed together who knows...)

So figure each year there are 8 playoff spots. Pencil in the big spenders...the Yanks, Red Sox, Angels, Mets, Cardinals, Cubs, Giants (?), and Dodgers as perennial playoff teams that take five or six of those spots. That leaves just two or three teams from the non-big spenders teams that have a chance to get to the dance. Now let's assume that those remaining 22 teams, about 8 are just plain poorly managed by the front office (i.e., Nats, Baltimore, Royals, and Reds for instance) and don't have a clear plan on how to win now or in the future so that leaves about 14 teams with some clue. Now if those 14 teams follow the model of raising good talent from within and having a competitive minor league system then don't you think that looking at the minor league records and tracking the success among those other teams would suggest that as this talent matures you should be competitive among those non big spenders and thus get that hand me down playoff spot? The graph above attempts to show some sort of correlation to a minor league record and major league output. Does minor league winning percentage in past lead to major league success in the future? All minor league percentage includes rookie ball and short season records and the A, A+, AA, AAA is the records of just the big 4 minor league clubs.

There does seem to be about a 3 - 4 year difference between a peak in the minors to a peak in the big leagues. The biggest example would be 2003 peak in the minors to the 2007 big league record. The 2003 had Holliday and Hawpe and a young Jimenez. If we assume that the 2007 big league squad overachieved then the next peak was in 2005 which again had Holliday, Hawpe, Jimenez, and added Tulo and then you have 2009 winning record.

Since 2003 - 2004, the Rox minor league system has consistently faired well. Depth in the minors can't be bad. Obvious alot of talent moves in and out of the minors...some of this talent succeeds, some fails, some remain in limbo, some are fast movers, others take time to develop. I think another problem is you need talent to mature together (think Pittsburgh) and for this success to continue then talent must come in waves. If you think Holliday, Hawpe, Atkins were Group 1, then this was followed by Tulo, Iannetta, Smith and then followed by Fowler and Young. As older talent moves on then good talent exists to take there place. I doubt whether one can truly quantify a correlation but for non big spenders it feels good to think that young talent will eventually mature and sneak into the playoff picture once in a while (and even win!).

Wednesday, November 4, 2009

World Series - Game 6

2009 in the books. Phils tried but just couldn't overcome a 3 - 1 deficit. I thought tonight they might sneak one with a Pettitte on the mound and on 3 days rest. Pedro couldn't tame the demons and A-Fraud finally gets a ring. There are some decent guys on the team...Jeter, Mariano, and Matsui...and I am glad for them. Some of the high dollar free agents...well glad you bought one.

Here is to our Rox connection in Girardi...he deserves some credit. He did a great job with tremendous pressure. Three man rotation got it done. Kind of showed how the Phillies really had a severe weakness on the mound. Too bad they couldn't clone Lee. Still think the Rox were a better team...ha!

Godzilla chomps Philadelphia and wins the MVP...nice series Matsui.

Monday, November 2, 2009

World Series - Game 5

Not dead yet...Phillies made the hometown fans sweat a little but they salvaged a game at home and send it back to NY on Wednesday. Hey, hey more November baseball! So what kind of hair gel does Utley use and I bet if they win he has a major sponsor deal...and what is up with Nike. I don't recall seeing so many swooshes at the "V' of uniform ever.

So Yanks win we have Mariano or Jeter as MVP candidates and if the Phillies make it back then it is either Lee or Utley. So does Manual go with Pedro again? and then Hamel for the deciding game? Gotta like the Yankees with Petite and C.C. I am just glad I can watch some more baseball!

Sunday, November 1, 2009

World Series - Game 4

As soon as Damon got a second life (swinging third strike in the dirt, guess he got a piece of it), I knew the Yanks were going to win. You just don't get to give second chances to the team of improbability without getting thumped. Then you go and let Damon steal with a shifted infield (way would you do this with the winning run on?)and then gets to jog on to an empty base (and what were you doing Lidge or Ruiz). Don't you think the catcher, pitcher, or even left fielder have a clue that if you are shifted that someone needs to cover third? I don't think it was that big of an impossibility that Damon was going to try and steal (again hand it to Yanks, they are forcing the issue). Poor Lidge, once again he will be the goat. Just makes Mariano look all that more impressive...unbelievable how he comes in and just shuts the door. Ice in the veins for sure.

Broadcast often does a no out walk end up scoring, anyone?