Sunday, April 20, 2008

NBA Win Shares, 2007-08 season

I came up with this system shortly after Bill James published win shares for baseball. It could probably use some updating but I don't have the time or motivation to do so. I think it does a reasonable if not perfect job of rating players, and in his excellent book Basketball on Paper, Dean Oliver briefly reviews the system and basically says it doesn't suck. So I'll put the numbers for another year on my site:

Latest season

The latest 3 seasons can be found on this page.

I don't watch the NBA much anymore. I don't think I've ever seen Chris Paul play, but he had quite a season. He'd be a fine MVP choice as would Lebron James (who I've seen play quite a bit). Kevin Garnett would have probably run away with the award for the way he turned the Celtics around but it's hard to say what the voters will think of him since he played less than 2400 minutes. For all I know Garnett might still win it.

The other numbers on the page include rating, my version of the Tendex rating, but adjusted not for game pace but for opponent points scored. Rebound percentage is rebounds as a percentage of total rebounds when the player was on the floor - 10% is obviously an average figure. Adjusted shooting percentage is points divided by (FGA*2 + FTA). If you take 6 shots and score 6 points, you will rank the same regardless of whether you hit 3 2-pointers or 2 3-pointers.

Saturday, April 12, 2008

Pitchers and velocity loss, part one

The Hardball Times has a new pitching mechanics expert, Paul Nyman. He comes highly recommended from the previous expert, Carlos Gomez. Carlos, also known as ChadBradfordWannabee, is not working for the Arizona Diamondbacks.

The discussion on BTF was interesting. Poster JPWF13 says:

A LOT of pitchers see their raw velocity peak in their early 20s and then decline
Avery only appears different became came up so young and was effective at an early age, most guys whose fastballs peak when they are 22/23 we don't lose sleep over, because we never see them with a good fastball if/when they make the majors.

You see it all the time, prospect supposedly throws 95, 4 years later, in the show, he ocassionally reaches 92, and you think, "eh so much for the scouting reports".
Paul Responds, here is a part of it.

There is no physiological reason for someone to peak at 22 or 23. How do you rationalize the Nolan Ryan's, Randy Johnson's, or any other player who is under 30 throwing 95 mph. And let's not muddy the water by dragging steroid or HTH use into the discussion.
You really don't need to go into steroid/HGH use to explain them. Ryan and Johnson are freaks of nature, the outliers. The vast majority of hard throwing young pitchers do not keep their velocity past age 40 or anywhere near it. What I wonder is how likely is it that a pitcher loses velocity?

For the first study, I looked at Baseball America's 2003 Prospect Handbook. The usually give a range for a pitcher's fastball velocity, though in some cases I inferred one. For example, if all they say is "mid 90's" I'll count it as 93-96 or so. I compared the ranges to the velocity data from Fangraphs for 2007. I've got 54 pitchers that I wrote down from the book who pitched in 2007. For 39 of them, they are within 1 MPH of their 2002 range. Seems like Baseball America did a pretty accurate job, and prospects losing velocity is not as commonplace as I may have thought. 12 were under the 2002 range, and 3 were over. This is of course selective sampling, if we were able to put a radar gun on the prospects who did not pitch in the majors in 2007, we may very well come across many cases of injuries and reduced velocity.

The biggest drops were Brandon Webb, who threw 88 last year but 92-95 as a prospect. He seems to be doing OK though. Jose Contreras threw 90 last year, 94-96 in 2002. But then again, he was already over 60 year old back then. Noah Lowry was down to 87 from 91-92, and Joe Blanton was 89 after 93-97. I've used Blanton as a poster boy for not trusting velocity in scouting reports, but it turns out he's an exception and not the norm.

Three pitchers gained velocity in this time: JJ Putz, 95 after 88-91; Heath Bell, 95 after 89-92; and Andrew Brown, 95 after 89-93. All three were able to max out their throwing in short relief last year, Brown and Putz had been starters in the minors, though Bell was always a reliever.

Next project: Look at Fangraphs data to estimate aging curves for fastball velocity.

Monday, April 07, 2008


Not one, but two game winning homeruns. One to go ahead in the 8th, and one to walk off in the 9th. Torii bailed out a stinker of an outing by K-Rod. Matthews continued to get on base with his third walk of the game, Vlad finally gets a single. Garret Anderson refused to swing outside the strike zone, walking on for pitches. That must have really confused JoeBo.

While watching the game I was putting together my wins over replacement database. I've got batting baseruns, baserunning, double play avoidance, reaching on errors, and totalzone fielding.

Hunter's runs over average, and over replacement, since he established himself in 2001:

2001 20, 38
2002 13, 32
2003 10, 30
2004 2, 19
2005 10, 23
2006 19, 37
2007 16, 36

Since a win over replacement goes for close to 5 million these days, Hunter is worth 15-18 million in his best years (5 of the last 7) and around 10 million the other two (including injury shortened 2005).

But tonight was priceless.

Saturday, April 05, 2008

Where do bullpens come from?

I have a theory, which is hardly a unique theory. It states: Minor league relievers are usually irrelevant to major league success. Most key major league relievers are former starters. I'm not sure the extent to which this is true. Obviously there are exceptions, like Troy Percival and Bryan Harvey who pitched in the bullpen from the start of their careers.

On the current team we have:

K-Rod - began as a starter, made 42 minor league starts, switched to the pen and made the big leagues in a year (pretty good finish too).

Shields - Made 68 starts in the minors, did a good deal of relieving too. Did a fine job towards the end of 2003 in the rotation, but the Angels thought he was more valuable in relief.

Oliver - 228 big league starts

Bootcheck - 128 minor league starts

Moseley - 141 minor league games, all starting

Speier - The exception, he's been a reliever his entire time in the majors or minors. He does have 7 minor league starts, but those aren't real starts, they look more like minor league rehab appearances where he just got his 1-2 innings in at the beginning of games.

Then there's Bulger, Thompson, and O'Day. These guys make up the back end of the bullpen.

If anyone who reads this follows another team than the Angels, can you give me the breakdown of your team's key relievers?

Thursday, April 03, 2008

Nice Pitching

Joe Saunders was awesome last night, throwing strikes and keeping the ball on the ground. The Twins didn't threaten very often, but when they did Joe got another double play ball. Aybar looks really good on turning the DP, showing off his acrobatics on the turn a few times with the runner bearing down on him.

Until I found out that Nick Blackburn was starting game 3, I had no idea who he was. Baseball America rated him the #1 prospect in the Twins system (though a weak system before the Santana trade was completed). His ERA was over 7 at the end of last year and his minor league numbers show a guy who barely strikes out 4.5 per 9, though his control is good. The numbers look like a batting practice pitcher. The reports say that due to injuries he had pitched without good stuff for most of his time in the minors. The stuff looked pretty good last night, with a 94 MPH fastball, sharp command of a breaking ball, and a slider that Vlad Guerrero absolutely could not distinguish from his fastball until it was too late. This is rare, Vlad may swing at anything, but he generally picks up on what type of pitch is thrown. If Blackburn can keep the stuff he had last night and stay healthy, he's got a good future ahead of him.

That is of course a big if, but last night his stuff looked like the stuff of vintage John Lackey.

Jon Garland pitched a fine game himself. He didn't strike anybody out, so sooner or later he's going to give up a ton of hits, but I can't complain about a guy getting through 8 innings on a low pitch count, or the Angel offense from Tuesday night.

2-1. Good start.