Monday, December 21, 2009

Rox Talk - Links

A simple link dump before the Holidays...

Surprising article in how blunt it is about Atkins performance...I mean testing began in 2004 and Atkins got his first big shot in 2005. Drug testing is supposed to be more intense in the minors so I guess I would have a hard time thinking he was on the juice and then had to go cold turkey...

Todd Helton was ranked as one of the top 3 hitters in baseball history according to this study (link to the link here). Wired declared it a baseball version of six degrees of Kevin Bacon...

Monday, December 14, 2009

Rox Talk - WAR Part IV

WAR - What is it good for? - Part 4

Last week we talked batting, this week we will close out with pitching. Again I must mention that WAR is only suppose to show the previous year...no one claims that it is predictive. Below is a graph plotting the Rox FIP vs WAR since 2002 (min of 20 innings pitched). These graphs might be a bit more correlated than the batting ones because the fielding isn't part of the Pitching WAR calculation. I did find it interesting that starters (empty squares, red line) and relievers (black squares, blue line) broke into two distinct groups.

Lower FIPs lead to bigger WARs. Again I fitted the points and used the expression to estimate possible outcomes for next year. The spreadsheet below shows Bill James FIP projections for the 2010 season. Using my handy dandy expression I get the following WARs (see sheet below, click on pitchers). The Bill James FIP can be found at Fangraphs and the website is currently taking fan guesses to what the player's WAR will be. As mentioned last week the fans WAR total for our 2010 team is 108 wins. My roster below is missing another arm. I would expect Rox to carry 13 pitchers(?)


Based on this WAG I have the Rox winning about 82 games. Just my gut says that we probably have about 6 more wins in there...accounting for the any missing players and Tulo and Jimenez being a bit below averaged based on historical numbers. As the season gets closer I will revisit this but 88 wins doesn't sound too bad.

Monday, December 7, 2009

Rox Talk - WAR Part III

WAR - What is it good for? - Part 3


OK so we are all smarter when it comes to WAR but what does it all mean? Well it is not a predictive tool, so it simply shows the value of the player for the previous year and shows whether a player was truly worth what the team forked over for him. Back in May on the Purple Row website a pretty good explanation is shown using WAR and Rox payrolls to show whether the team was worth its wins.




The Rox management team should be commended. They earned alot more wins then what they are paying. Essentially this is because they haven't delved into the free agent market and they have a lot of WAR coming from players in their pre-arbitration days. Tulo and Jimenez are outright steals right now! It will be interesting to see during the next union bargaining whether players try to change how they can bargain when they first make the big leagues. Small market clubs are simply jettisoning players once they become arbitration eligible. Look at Atkins, he is 28 and should be coming into his attractive payroll period and yet the Rox just threw him aside for younger cheaper talent. It is a strong possibility that Atkins is done and probably will never see $7 million a year again. Now of course he could have signed a long term deal too so a player does have that option...stability over free agent big bucks...you better not swan dive during that pre free agency year!

Back to WAR...I wonder if there is any predictive value to it? WARNING: WAR is only suppose to show the previous year...no one claims that it is predictive. Below is a graph plotting the Rox wOBA vs WAR since 2002 (min of 60 plate appearances). As I mentioned this really shouldn't tell me anything because position played and defense ability does factor into WAR in the transition from wOBA but what the heck...with excel and a new tool I just found I can do whatever I want! (it is my web page).

Higher wOBAs lead to bigger WARs. My new tool helped me create a nice line through the data. After some simple ;-) math I got this expression...whoopee!

f(x)= 2.38733*exp(15*x- 5.16638)-0.44

So back to estimating. The spreadsheet below shows Bill James wOBA projections for the 2010 season. Using my handy dandy expression I get the following WARs (see sheet below). The Bill James wOBA can be found at Fangraphs and the website is currently taking fan guesses to what the player's WAR will be. Currently the fans WAR total for our 2010 team is 38.9 for our players and if you add the pitching data in (next week post) the Rox would end up with 108 wins (fans might be thinking a bit too highly)! My 2010 team below also is missing another two or three bats (a catcher , infielder, and pitch hitter). Also I don't think Hawpe is going to get 642 bats!

Of the numbers using my estimating tool I would think Tulo and CarGo might have better seasons then what their wOBA says. Rox do have some talent, the big question is 2010 going to be like 2009 or 2008? Will expectations once again bite?