Wednesday, February 27, 2008

Rox Talk - WPA Revisited

To put a final nail in the 2007 season, I wanted to compare the weekly WPA numbers I captured last year and compare this to the expected wins and see who really made a difference for the team last year. If you recall WPA is simply I way to look at a player's contribution to a team's success. As a team moves through a 162 game season, each day a player has the opportunity to help or hurt a team. Obviously if you continue to contribute positively you will have a better WPA over the season. So let's start with the starting pitching:

Okay that is a headache of a graph! The blue dots show the expected wins over the season after each day. Obviously winning a bunch of games at the end helps you get to 90 wins! WPA and pitcher's are at somewhat a disadvantage as they only get to improve on their WPAs every 5th day. So one bad start can certainly tank your WPA. Francis was definitely the Roxs pitching MVP especially down the stretch with some help from Jimenez and Morales. Fogg had some good starts at the end but his early pitching woes certainly hurt him over the entire season. Cook was the real disappointment (let's hope the contract extension wasn't too big). Now lets look at the relievers:

Obviously down the stretch it was the Corpas/Herges show. Interesting to see that WPA does in fact mean something in the real world as Fuentes collapsed so did the Roxs right before the All Star Break. And finally the hitters:

I think the big surprise to me was how valuable Helton and Hawpe were down the stretch. Helton especially since he has been beaten down due to his lack of power but he really got the job down along with Holliday. Tulo was a nice additional and although Matsui certainly led the team early in the year he leveled out. Glad we got Yorvit for the handling of pitchers because offensively he ain't it.

Sunday, February 24, 2008

Spring Training

"Today the Times reported the arrival of the first pitchers and catchers at the spring training camps, and the morning was abruptly brightened, as if by the delivery of a seed catalogue."

You know baseball is just plain fun...
Well it has been a while since the ol' Roxs had any expectations. Is this going to be a good thing? Will they hit the ground running or will they start slow and have the doubters start. Ughhh to have a target on your back and teams suddenly gunning for the NL Champions. Heck of a ring to it. To be truthful I haven't been too happy with Hurdle and O'Dowd. They have certainly tried our patience with these last 3 years. But I have to eat some crow and say they have put a pretty good product together. These guys have won together through the minors, they are a "team", not just a collection of free agents trying to win this year. The team has the feel of O'Dowds Indians a few years back. Definitely a potential to run the table the next 2 - 3 years before the dollars come calling. It would be nice to have a hard time getting tickets once in a while. So anyway let's see how it rolls, if Nix and the starting pitchers can get it done then it will be truly magical year and maybe we can get another shot at the mean ol' Sox. Hey maybe Hurdle can win home field this year for the NL and we can get it done!

Thursday, February 21, 2008

Rox Talk

From the news that just won't go away...more steroids in baseball. Thought the Mitchell Report would end it all but it ultimately just opened a can of worms. At the NY Times Opinion Section featuring our friends, the Freakonomics, an article (including the graph to the left) tries to show that Clemens stats over the last 8 years of his career show something fishy. This was somewhat in response to Clemens legal staff that created a statistic treatise of Clemens careers showing how he didn't get a boost from any PEDs. All of this brings to light the world of sabermetrician's which I guess is good but also it brings to light that numbers and statistics can be a slippery slope. Stats can often show what you want by how you present data. In fact other analysis (here and here) show some other thoughts. Even reviewing the comments about this article seem to doubt the curve fitting that was done on the graphs to the left. Anyway interesting article and a decent follow up graph below using ERA+ rather than age vs WHIP.


Friday, February 1, 2008

Division Hot Stove

Back to the Hot Stove. With my Game Score+ meanderings I sort of dropped the Roxs. Anyway I think Garett Atkins said it best (when asked about the D-backs signing of Haren),

"Yeah, we needed more good pitchers in the NL West"

Anyway the NL West seemed the most active of any of the divisions in the Senior Circuit. Major signings included Haren to D-backs; Iguchi, Hairston, Edmonds, Wolf, and Prior to San Diego; Jones and Kuroda (not to mention Schmidt's return) to LA; and Rowand to San Fran. Obviously something seems to be missing from this list....hmmmm, yup the Roxs signed no one of significance. Is this good or bad? Well basically what I did was I took last year's Win Shares and see what the addition of these players meant.

A Win Share is statistic developed by James that assigns a single number to each player for his contributions for the year. All pitching, hitting and defensive contributions by the player are taken into account. Statistics are adjusted for park, league and era. A Win Share represents one-third of a team win, by definition. If a team wins 80 games in a season, then its players will share 240 win shares. So for last year based on the team's final standings

W

L

RS

RA

Ex - W

WS Factor

Arizona

90

72

712

732

79

0.88

Colorado

90

73

860

758

90

1.00

San Diego

89

74

741

666

89

1.00

LA

82

80

735

727

82

1.00

SF

71

91

683

720

77

1.09

So assuming that all the teams current players earn similar Win Shares in 2008 (probably not, but I will assume that some players will go up and some down...this is just a back of the envelope analysis!). So basically looking at the 2008 roster versus the 2007 roster and looking at only starters and players with at least 100 at bats, I determined the Win Shares (divided this number by 3 for wins) returning and then I also adjusted these Win Shares with the WS Factor above. I did this because Arizona's win total last year was way out of whack with the pythagorous win total based on their runs scored and runs against. I doubt very much that they will be as "lucky" as last year especially with them trading Valverde this off-season.

WS Return

WS Correction

Arizona

82

72

Colorado

81

81

San Diego

60

60

LA

72

72

SF

55

60

So Colorado, which kept its team pretty much intact and only losing Matsui's 14 Win Shares, Carroll's 4 Win Shares, Fogg's 6 Win Shares. What about the other divisional rivals and their additions?

+ WS Wins

Total Wins

Arizona

6

78

Colorado

0

81

San Diego

8

68

LA

11

83

SF

11

71

Well based on this little exercise it looks like LA has the early club house lead. They had a bunch of good acquistions in Andrew Jones and possibly a full season of Schmidt. The unknown factor is the Japanese pitcher Kuroda. San Diego has a lot of offensive holes to fill with losing Cameron, Giles, and Bradley and trying to make this up iwht Iguchi, Hairston, and Edmonds (who is old!). Haren doesn't solve Arizona's offensive woes (although their young players should start to hit being a year older). What about our Roxs? Well with full seasons from Jimenez and possibly Morales along with a year older Hirsh the pitching should stay fine. The only real hole is 2B. Can Nix get 8 - 12 Win Shares this year (to make up for Matsui's 14)? I think the Roxs will be fine even with the rivals additions.