Twins Defense

by SBG on Nov 19, 2005 9:30 AM EST Comment 9 comments

This article first appeared at my site, SBG, concerning the Twins defense.

Before the season ended, I wrote that the Twins' defense, contrary to popular belief, was improved in 2005. I got virtually no reaction to this assertion, partially because I talked about Barry Bonds in the same post and Surprise! that is what people wanted to talk about. Over at The Bleacher Bums, a very nice joint blog from MPR, at least one of the authors was complaining about the dropoff in the Twins defense. Being the acerbic fella that I sometimes am, I openly challenged the author in his comments, which promptred the one of the authors to respond, "SBG, you are full of [fecal matter]." Well, he didn't actually say that. What he said was that was an example of statistics run amok.

Keep reading for more on the 2005 Twins Defense.

Baseball Prospectus has listed the 2005 Twins as the fourth best defensive team in the majors in terms of PADE (park adjusted defensive efficiency -- seriously, these guys work HARD to come up with their stats).

How can this be? Didn't our eyes tell us something completely different?????

I will start by saying this. Defensive numbers, especially with regard to individual players are not necessarily reliable. Certainly, errors, long regarded as the primary indicator of defensive excellence is not really a good indicator. A player with concrete feet who doesn't make errors isn't a better defensive player who gets to every ball, but makes a few errors.

The number that I like is defensive efficiency. It's a team defensive number that calculates the percentage of balls put in play that result in outs. The league average was 69.47%. That is, 69.47% of balls put in play were converted into outs. The Twins' percentage was 70.33%. Above average, yes, but that doesn't seem like a whole lot above average. So, the 2005 Twins are a little above average. The 2004 team must have been way above average, right? Guess what. That's not the way it was. In 2004, the Twins had a defensive efficiency of .688, good for 10th in the 14 team AL.

Counter to everything you've heard anyone else say, it appears that the 2005 Twins defense was actually one of the best in the league. That is, if you value making a high percentage of outs on balls put into play.

According to BP, the Twins' PADE was 1.54, meaning that they turned 1.54% more balls into outs, given the parks that they played in. In other words, if the league average is 69.47%, the Twins turned 71.01% of balls put into play into outs, adjusting for park effects. That's probably one fewer base runner for every two games. BP describes a 1% increase as "not insignificant."

BP provides a statistic called RAA, which is the number of runs saved (defensively) above average. As a team, the Twins saved 35 runs above the average defensively. Collectively, then, the Twins defense meant four wins for the Twins over the average defense (to come up with this, I added 35 runs to the total allowed and calculated the Pythagorean Winning Percentage).

So who saved the most runs defensively over the average? Torii Hunter? Well, maybe not, he only play 93 games. Well, then, who?

Two players each saved 12 runs over the average: Juan Castro and Jason Bartlett. And Bartlett played 64.8 games. (Castro played 80.0 games.)

Next was Jacque Jones. He saved 11 runs over the average player at his position. Hmmm. How about Torii? How'd he do? One. Gold Glover Torii Hunter saved one run over the average. One. How does that grab you? I know what you are thinking. He better hit. How about this? Jones hit .249/.319/.438/.757 this year. The average AL Right Fielder hit .270/.332/.451/.783. The average AL Center Fielder hit .268/.322/.407/.729. So, he's a below average right fielder offensively, but his an above average center fielder, offensively. With his defense, he might be a decent option in center field. Now, he's not a great option in center, but he's probably just as good as Hunter. Actually, neither one earned their money last year.

Next was Joe Mauer with 10. Man, we really miss Henry Blanco.

So where did the Twins suffer? Well, Justin Morneau was 14 runs below average. Michael Cuddyer was 8 runs below average. Note on Cuddyer (and everyone else): since Michael played multiple positions, I added the results at each position to get a total.

I'm not sure exactly how these numbers are calculated. I'm not sure if they are an accurate reflection of what players are the best and worst. Like I said, I do like the defensive efficiency numbers, but as a team, four games seems about right.

The manager and general manager have both indicated that the Twins 2005 defense was below par and needs to be upgraded. Obviously, what needs to be upgraded is the offense. But, hey, the defense was pretty darned good. In fact, it was one of the best in the league.

0 recs | Comment 9 comments |

Story-email Email Printer Print

Comments

Degree of difficulty

Why did they seem so bad? I think it's because the Twins made themselves into an extreme G/F team, thanks to a lot of help from Silva and some help from Lohse and Mays. People whine about groundballs that get through the infield that seem like they should get caught. They whine about it a lot. But this is a difficult way to judge a defense, because no matter how good you are, there will always be plays at the edge of your range that you can't make.

Outfielders, on the other hand, get a free pass on this. If they run hard and dive for a ball that they got a bad jump on, all of the blame shifts to the pitcher, because, well, the batter hit the ball to the outfield, which looks worse for the pitcher, and it really looked at the end like the fielder tried hard, so he must've made a good play, right?

To me, the difficulty in judging individual defense is that the bulk of defensive value is contained in the plays that you make look easy. It's tough to say, after Jacque Jones camps under a ball for a while, that's a ball other players might not have made because they wouldn't have gotten a good jump. You generally have to go look back at video to see if someone got a good jump, too, because while the outfielder is getting his jump, you're watching the ball, just like everyone else in the stadium.

Also, I know Morneau was no great shakes, but there is no first base defensive rating I trust at all. The importance of defense at first base is generally underrated, because the first baseman handles the ball more than anyone else on the field, save the catcher. Most of those chances are easy, but there are so many that even small differences in rates can accumulate to big differences over the course of the season. Morneau's height at first could be a real advantage, so hopefully Gardy can work with him to improve his consistency on handling throws in the dirt and fielding hit balls up the line. He did win some defensive awards in AAA, if I recall correctly.

by ubelmann on Nov 19, 2005 7:26 PM EST reply actions

About first base defense...

I agree that there are no good tools to measure the defensive contribution of first baseman. I think part of the problem is that the importance of the position depends a lot on how good the other infielders are. Good infielders make the first baseman's job much easier. I've always contended that it was Mientkewicz that made Guzman and Rivas serviceable defenders. Guzman, in particular, had a maddening tendency to simply 'let it fly' and put the onus on Doug to make the play (which he often did). A great defensive first baseman can allow a team to get away with having infielders with poor throwing accuracy.

by hornbakr on Nov 22, 2005 10:07 AM EST up reply actions

BP defensive stats

These guys still haven't figured out a way to rule out the influence of pitchers. I understand "putting the ball into play" stats, but balls put in play do not have the same velocity or trajectory for every pitching staff, and the variability in the staffs is bound to be too great to be able to analyze defenses.

The Twins sucked, though Castro (and even Bartlett) was good.

Sometime your low self-esteem is just good common sense.

by statman on Nov 22, 2005 1:04 PM EST reply actions

Huh?

"The Twins sucked, though Castro (and even Bartlett) was good."

Who exactly do you have a problem with on defense, then?

Cuddyer wasn't good, but he wasn't that much worse than average.

Morneau wasn't great, but compared to your league average first baseman, he's not hurting the Twins much.

Rivas/Punto each have their weaknesses but are at least league average 2B.

Joe Mauer was possibly the most valuable defensive player in the league (at any position) last year.

Castro/Bartlett, by your admission, were good.

Hunter is still a great defensive CF, and Lew made a competent stand-in.

Stewart makes up in range what he lacks in arm strength.

Jones is one of the most underrated fielders in the game, and is a real asset in the outfield.

Even looking at the lineup player-by-player, it seems like the most reasonable conclusion is that the Twins had at least a league average defense, if not above average. The stats presented by SBG are not so far out of line with reason that we should just summarily dismiss them.

by ubelmann on Nov 22, 2005 2:19 PM EST up reply actions

Huh (back)?

Neither Castro nor Bartlett played much, and Bartlett wasn't that good at the beginning (his first stint, if you remember). Mauer is outstanding when he's not giving up a passed ball. Hunter was very average again this year. He finally got payback for trying to make another catch tougher than it really was. Jones makes bad decision after bad decision and can't hit a cut-off man. Bad routes and still can't play the lights. Stewart was not good at all (injury took away what range he had), and the arm is the worst in the majors.

Cuddyer reeks. No range, bad hands, bad decisions, and bad throws. Improving, though. Rivas is the worst 2B defensively in the league if you remove his adeptness at pivoting the DP. Morneau was bad, though better than I thought he'd be. Zero range, and he doesn't field one hop throws well yet. All you had to do was watch the keystone cops in action every day to see this.

Fielding stats are horrible, horrible indicators of how good a fielder is. They're not even on par with the old use of BA to determine a hitter's worth. And a team defensive rating is a ridiculous measure when it can't account for how the balls coming off the bats.

If you think that was a good defensive team, well, then have at it. We who watched know better.

Sometime your low self-esteem is just good common sense.

by statman on Nov 22, 2005 4:40 PM EST up reply actions

Right...

"If you think that was a good defensive team, well, then have at it. We who watched know better."

Hey, I watched, too. How does that strike you for a counter-argument? I watched, and I listened to outsiders' opinions, too. Almost every team the Twins visited, their broadcasters lauded them for having the best defensive outfield in baseball. If you thought this was a terrible defensive team you were an awfully lazy observer.

Saying Mauer was outstanding when he wasn't allowing passed balls is like saying that Barry Bonds is outstanding when he's not striking out. Amongst catchers with 900+ innings played last year, Mauer was 9th of 20 in passed balls allowed, on a staff where plenty of guys throw pitches in the dirt regularly, a side effect of having a groundball pitching staff. That, and watching him, I certainly remember no problems with passed balls.

And don't worry, I was watching when the Twins' "keystone cops" were turning double play after double play for Carlos Silva. There is no "Carlos Silva, quality pitcher" without "Minnesota Twins, quality defense."

Just saying you watched, and your observations are superior to someone else's observations is an amazingly weak and meaningless position, though.

by ubelmann on Nov 22, 2005 5:35 PM EST up reply actions

Okay...

You saw a good defense. I saw a horrible one relative to where the defense was in the past. My real point is that, if these defensive stats have any merit, then they should be independent of teammates. That is not anywhere close to the case and, unfortunately, very difficult to test. Mostly, you will have to take players who move to new teams and predict their defense (predict their stats), if in fact these stats have merit.

The overall team defense could probably only be tested if the pitching staff is mostly new.

Sometime your low self-esteem is just good common sense.

by statman on Nov 23, 2005 2:31 PM EST reply actions

Huh?

"My real point is that, if these defensive stats have any merit, then they should be independent of teammates."

Why? The collective object of the defense is to get batters out. If you have Ozzie Smith at SS and Stan Laurel at 2B are you a better defensive team than if you had two major league quality players?

Why can't you measure team defense? It's probably MORE valid than individual defense.

SBG -- Your source for half-baked crap

by SBG on Nov 23, 2005 6:36 PM EST reply actions