Small sample size theater

"Small sample size theater." That's a phrase I heard Monday afternoon on MLB Network in discussing Shohei Ohtani. I like it, amd I'm stealing it.

Scene I

A regular reader of these posts dislikes the Twins decision to play Eduardo Escobar at shortstop during the suspension of Jorge Polanco. Shortstop, he opines, should prioritze defense, and Escobar is not as good with the the glove as Ehire Adrianza.

I agree with the second point, but think the first is something of a sliding scale. You can sacrifice some defense at any position to get more offense. (See the second half career of Jeter, Derek.)

And then there's this:

Which defenses are doing the best job of turning ground balls & bunts into outs?

1. Indians 84%
2. Twins 78%
3. Mets 78%
4. Giants 77%
5. Rockies 77%

Worst
30. Rangers 69%
29. Yankees 70%
28 Pirates 70%
27. Marlins 70%
26 Tigers 70%

— Sports Info Solutions (@SportsInfo_SIS) April 16, 2018

Small sample size theater, yes. But so far, Escobar isn't killing the Twins infield defense.

Scene II

I watched Justin Verlander and Bartolo Colon duel on "Sunday Night Baseball" the other night. Wow. Verlander allowed one hit -- a homer -- in seven innings, and Colon threw seven PERFECT innings before cracking in the eighth. 

The game went 10 innings, and the Texas Rangers eventually won, 3-1 in 10, so neither starter got a decision. (I thought this during the recent Astros-Twins series and tweeted it Sunday night: If the Astros have a weakness, it's their bullpen.)

Colon, who finished 2017 filling out the Twins rotation, has started twice and relieved twice for the Rangers this year. He has a 1.45 ERA in 18.2 innings. Still throwing almost nothing but fastballs, still surviving on location.

Scene III

The Twins have played 11 games so far this season, the fewest in the majors, and will have another off day Thursday after the Puerto Rico series. They're going to pay, eventually, for such sporadic play.

Love the glove

Two runs a game.

Six games into the 2017 season, that's what the Twins opponents are averaging. Two measly runs per game.

Six games isn't much; it's two series, and one of those series was against a team that went into teardown mode this winter (the Chicago White Sox). Minnesota's 5-1 record may prove a mirage. But what we've seen so far is a far cry from the pitching-and-defense ineptitude of 2016.

That the outfield is impressing afield shouldn't be a surprise. Nothing falls but raindrops is supposedly their mantra, and they are living up to it. (Those calling for Byron Buxton to hit the bench over his struggles at the plate should remember: Were Danny Santana playing center field on Friday, Phil Hughes probably doesn't get out of the first inning.)

But Jorge Polanco has been far better at shortstop than I expected. Miguel Sano had a rough game Saturday at third -- two errors, one on a botched rundown that directly resulted in a run -- but other than that, the infield defense has also been superb.

If this pattern of stellar defense continues, 5-1 may not be a mirage.

From the Handbook: Team defense

Only two player on the  2016 Minnesota Twins started as many as 100 games at a given position: Brian Dozier, 151 at second base, and Max Kepler, an even 100 in right field. One hundred is less than two-thirds of the games, so it's stretching things to assert that Kepler was the everyday right fielder. He had a majority of the playing time, certainly, but hardly an everyday guy.

So in evaluating and measuring the defensive skills (or lack of same) on this team, individual metrics are a bit lacking. There's hardly enough playing time for Eduardo Nunez, Eduardo Escobar and Jorge Polanco to be accurately assessed as shortstops.

But the Bill James Handbook does contain an intriguing table -- new to the book -- that assesses the defensive runs saved by each position for each team. And that is rather illuminating with the Twins.

The Twins had some ugly positions. No team's catchers cost more runs than the Twins (-18), No team's left fielders cost more runs than the Twins (-22). Shortstop cost the Twins 16 runs, third worst in the majors. Right field was -9, third base -8; oh, hello there, Miguel Sano.

Only three positions were in the positive zone for the Twins. First base was +8, the pitchers were +6 and centerfield was barely positive at +1. Second base was -1.

All told, BIS estimates that the Twins defense gave away 50 runs compared to the league average. That's not the worst in baseball -- Oakland was -71, and the Detroit Tigers were also -50 -- but it is certainly a problem.

The hidden problem

Jorge Polanco fails to handle a "base hit"
off the bat of Lorenzo Cain with two outs
in the sixth inning Sunday. The next batter
doubled Cain home for what proved to be
the winning run.

I voiced optimism here about the Twins direction last week. Then they went to Kansas City and got swept. Two blowouts, one extra-inning rain-interrupted marathon, one pitchers duel -- all with the same "L" for the standings.

One of my pet theories about "intelligent fandom" is that the broadcasters reflect the thinking of the organization. Dick Bremer, Bert Blyleven, Cory Provus, Dan Gladden -- these guys are not only around the team (as are the beat writers), they hold their jobs at the mercy of the team (unlike the beat writers). When Billy Beane was in the process of imposing the use of advanced stats on the Oakland A's almost 20 years ago, their TV broadcasts rather quickly ditched the traditional "triple crown stats" on their graphics to show on-base percentage and slugging percentage.

So when Bremer spends much of his airtime bewailing the Twins starting pitching while barely acknowledging the flawed fielding, the presumption here is that he's reflecting the thinking of the manager and coaches.

There's really not much good to say about the Twins starting pitching in three of those four games in Kansas City. Tyler Duffey, Jose Berrios and Hector Santiago all had short starts, and that strains the already strained bullpen.

But defense is the hidden problem. The Kansas City Star on Friday described a pair of plays by Minnesota outfielders Eddie Rosario and Robbie Grossman on Thursday as "two of the worst defensive plays of the year." The video-laden breakdown linked to above concluded:

According to FanGraphs, the Twins’ defense has a negative-25.3 runs saved, the second-worst in the American League.

Yeah.

Sunday's final, and decisive, run was handed to Kansas City by the Twins fielding, specifically the shortstop play of Jorge Polanco. As has been noted here, the heavy use of Polanco at short in recent weeks directly contradicts the obvious conclusion drawn by the farm system, which devoted more than two seasons to playing him at shortstop and decided he's not a shortstop.

Polanco, as depicted above, failed to come up with a grounder in the sixth inning. Blyleven and Bremer were still questioning the scoring call when Eric Hosmer doubled over Rosario's head (might Byron Buxton have reached that ball?) to score Cain from first base -- and suddenly the TV talk went from a missed out to a pair of two-out hits.

I say this a lot: The easiest way to improve the pitching staff is to improve the defense. That we don't hear that from the broadcasters suggests that they aren't hearing that from the organization.

R

eIad more here: http://www.kansascity.com/sports/spt-columns-blogs/for-petes-sake/article96623782.html#storylink=cpy

From the Handbook: Defensive shifts

There are, I suspect, two types of sophisticated defensive metrics: the publicly available ones, and the proprietary ones designed by individual teams' analytics departments.

Baseball Info Systems, which generates the numbers in the annual Bill James Handbook, is prominent in the world of public defensive metrics. Its work has, among other things, laid out for public viewing the rationale behind the current trend toward infield shifts,

In his essay accompanying the shift stats in the 2016 Handbook, John Dewan says that major league teams shifted 13,298 times in 2014 and 17,733 times in 2015. (Those numbers represent balls put in play when the defense is in a shift, so at-bats in which the defense is in a shift early and out of it later aren't counted, and neither shifts in which the batter strikes out, walks, etc.) That's a 33 percent increase last year.

The shifts, by BIS' calculations, prevented 266 runs over all. That's a 36 percent increase over 2014's 196, and suggests that teams are becoming more efficient at shifting.

One other general point: BIS believes that the optimal number of shifts is more than double last year's total, that the upper limit is probably around 40,000 shifts. MLB isn't there yet, but if shifting continues to increase, 40,000 is possible in a few years.

OK, that;s the general stuff. On to the Twins specific shift data:

In 2014 -- with Ron Gardenhire as manager and Paul Molitor assigned to handle shifts (an assignment widely viewed as being an override of Gardenhire's preferences) -- the Twins deployed 478 shifts. Only three teams in the American League used fewer shifts.

In 2015 -- with Molitor now the manager -- the Twins used 724 shifts, sixth most in the AL. That's an increase of 246, That's a 34 percent increase.

The Handbook does not provide a team-by-team breakdown of runs saved by shifts, so I don't have a precise number for the Twins. Across the majors, teams saved 1.5 runs per 100 shifts, so if the Twins shifted at average efficiency they saved roughly 10 runs with the shifts. Ten runs sabermetrically equates to a win, so that's hardly an insignificant matter.

Evaluating defense in the shift era

My last post of 2013 was about the closing off of new-age evaluative stats to the general public. For much of my adult life, savvy fans had access to statistical insights of which the teams themselves were (for the most part) ignorant.

Today the teams — or at least some of them — are developing new metrics for their own use and keeping them out of the public domain.

Some of that is visible on a daily basis as teams rely ever more heavily on exaggerated defensive positioning. These shifts are based on detailed charts of hit patterns and pitch sequences.

And it occurred to me Sunday that these shifts are going to raise havoc with the best of the public defensive metrics.

The play that crystallized this thought for me wasn't a defensive play at all. It was Brian Dozier stealing third — not off the pitcher, not off the catcher, but off the third baseman, who was playing Jason Kubel essentially at the traditional shortstop position. Manny Machado had no chance of getting to third far enough ahead of Dozier to take a throw.

My favored defensive metric, the Baseball Info Systems plus-minus, is a video-review system that splits the field into zones and measures how often balls hit into a specific slice of the field are turned into outs. But that system makes more sense when infielders are in their traditional spaces, not congregated on the first base side of second base as they are for Kubel.

Aesthetically, I'd prefer to see the shifts backfire enough to discourage such heavy use. I don't see that happening, because such hitters as Kubel can't/won't/don't take advantage of the shifts. (The Orioles kept the shift on even after Dozier took third; they knew Kubel wouldn't push a bunt to third and take the RBI single.)

Teams will keep shifting until the hitter make them stop. It's just incidental that they will, in the process, make it more difficult to get a objective handle on the quality of an individual's defense.

What's wrong with the Tigers?

Miguel Cabrera and Prince Fielder expected to do a lot
more celebrating this year than they have.

The Detroit Tigers last season won 94 games and pulled away from the rest of the AL Central, then won their first round of the playoffs and came close to knocking off the Texas Rangers in the ALCS.

Then they signed Prince Fielder, and everybody -- well, pretty much everybody -- figured them for 100-plus wins, an easy divisional title, a strong contender for the World Series.

We're now halfway through the season, and things aren't going that way for Detroit.

What's wrong?

Roy Smalley, doing the commentary on the Twins telecast Thursday, at one point said something along the lines of: The difference between the Twins and the Tigers is that the Twins play better defense.

Which is an odd thing to say, because if that were true — that the difference is defense, and the Twins are better — the Tigers shouldn't be 5.5 games ahead of the Twins in the standings.

But defense was the obvious flaw in the Tigers roster before the season, and it has been bad, and it's tempting to blame their disappointing first half on that flaw.

I'm not so sure that narrative is supported by reality.

Consider this: The average American League team entered play Thursday having scored 365 runs. The Tigers have scored 361.

The Tigers are fourth in on-base percentage (.331), and closer to second than to fifth. They're just barely above average in slugging percentage. That combination ought to provide more than average runs scored.

The problem, I think, is that the offense is basically three men: Austin Jackson, Miguel Cabrera and Prince Fielder. Nobody else with 200 plate appearances is above average in either OPB or SLG.

The Tigers have allowed 368 runs, a bit worse than average — and given the defensive profile of their lineup, that's not too surprising. The "runs saved" defensive metric compiled by Baseball Info Systems suggests the defense has cost Detroit 30 runs compared to the average team; take 30 runs off their runs allowed, and they go from 10th most allowed in the AL to fifth. Good pitching negated by lousy defense.

If they were scoring runs the way they anticipated, the defensive problems wouldn't matter. That's what hasn't happened — yet — for them.

For love of the glove: Ben Revere, right fielder

Ben Revere: His .750
OPS is is higher than
those of Ryan Doumit
and Denard Span.

Ben Revere made three notable plays in the field Sunday, including one of his "how-did-he-outrun-that-ball" efforts on the warning track.

Defense -- specifically range -- is a major part of his value to the team. This is true even in right field, the outfield spot at which Revere's subpar throwing arm is most noticable.

Still, having two legit center fielders out there is a plus for the pitching staff, and I thought it worthwhile to take an early read on Revere's defensive metrics in right field.

Emphasis on early: Revere didn't open the season in the lineup and spent some time in the minors before being recalled. He has 218 innings in right field; that's less than half the innings of the season.

Still, by Baseball Info Systems' plus-minus and runs saved metrics, Revere's raw numbers compare quite well to even the best of the American League's full-time right fielders.

Revere, as of the figures I consulted online Monday, is +8 in plus-minus, +3 in runs saved. These are good figures, especially considering the limited innings, but neither is close to leading the AL's right fielders. Ichiro Suzuki of Seattle, who has played 595 innings, is +15 and +13. Tampa Bay's Ben Zobrist, who has just 329 innings in right (he plays multiple positions), is +11, +8.

Worth noting: Trevor Plouffe and Darin Mastroanni have a combined 187 innings in right field (less than Revere), and they are a combined +5 and +3. Ryan Doumit, meanwhile, is -4, -2 in 49 innings, which I'm sure didn't escape Ron Gardenhire's attention. Doumit hasn't played right field since May 4.

Twins killings: Turning the double play

Jamey Carroll, playing second base on Tuesday,
turns a first inning double play.

The Twins turned double plays in the first and second innings Tuesday night — and none after that, breaking a four-game streak with at least three double plays.

Those two DPs gave the Twins 70 double plays on the season — by far the most in baseball. (Toronto has the next highest total at 58, a dozen behind Minnesota.)

This is a mixed blessing — a multitude of double plays is characteristic of a team that allows too many baserunners, and this is certainly the case with the Twins.

But it also has links to some strengths. The Twins' opponents entered Tuesday hitting ground balls in 35 percent of their plate appearances —a high figure, just slightly behind Cleveland for highest in the AL. (This is fueled in large part by Scott Diamond; according to Baseball Reference, Diamond in his five starts has gotten 65 ground balls and 38 air balls.)

And there can be no serious question that getting Tsuyoshi Nishioka and Trevor Plouffe out of the middle of the infield (and Jamey Carroll and Brian Dozier into the middle infield) has helped tremendously.

Plouffe was a disaster at shortstop last year,  but he has been pretty decent in his first dozen or so games at third base — considerably better than I expected when he started playing there with Danny Valencia's demotion. He has been charged with two errors at third and has had a hand in seven double plays. It's a rare third baseman who has even as many DPs as errors — to have that kind of ratio is almost unheard of.

Valencia, in more than twice the playing time at third, had two errors and eight double plays — last year, 18 and 20. The contrast in around-the-horn double plays from 2011 to 2012 suggests that the improvement may have less to do with the third baseman than with the second baseman.

On the defensive

A run-scoring single gets past a diving Jamey Carroll
during the series in Anaheim.

One of the running themes here this season has been the idea that part of the Twins' pitching problems is with the other players on the field — that the defense is not up to snuff.

But I'm being forced by facts to reconsider that argument.

The defensive metrics developed by Baseball Info Systems — plus-minus and runs saved — actually looked pretty good on a position-by-position basis when I broke them down earlier in the week.

And then came into my email inbox a BIS "stat-of-the-week" message on team level defense.

The Twins, as a team, were plus nine in runs saved at the end of April— fifth best in the majors, but only fourth best in the American League. Toronto was first with 31 runs saved, followed by Tampa Bay at 23, then Arizona and Texas at 14 and 13 respectively.

(On the other end of the spectrum were Colorado, -26; Detroit, -19; Washington, -15; the Mets, -13; and Milwaukee, -12. Four NL teams and the Tigers, with all the DHs they can stuff into a lineup.)

My view is shifting slightly.

I'm still sure that the decisions made late in spring training — specifically the de-emphasis of Ben Revere, the demotion of Brian Dozier and a rather heavy use of Ryan Doumit at catcher — all served to enhance the team at the plate while eroding the defense.

Despite that, the defense is much better than the 2011 atrocity and is a bit better than average.

But "a bit better than average" isn't enough to save this starting rotation. Quite possibly even a top-of-the-line defense wouldn't be.

A quick look at the defensive metrics

Jamey Carroll has been an iron man
at shortstop so far for the Twins.

My Monday print column repeats an assertion that I've made before: That the Twins defense has declined steadily over the Ron Gardenhire era.

My eyeball sense is that, while Jamey Carroll is an improvement at shortstop over last year's regular-less disaster, there are still plays left unmade; that the corner outfielders don't get to enough balls; that Danny Valencia's range leaves much to be desired.

But that's an eyeball sense. What do the metrics say?

As I did last year, I will periodically check the "plus-minus" and "runs saved" stats (explanation here) compiled by Baseball Info Systems and which I'm gleaning through an iPad app.

A quick disclaimer: It's just 21 games, and considerably less for several individuals. This is a snapshot of what has happened and not to be relied on as talent evaluation.

First base: Chris Parmelee has started 11 of the 21 games, with Joe Mauer starting five, Justin Morneau 4 and the departed Luke Hughes the other. They have combined for -3 in plus-minus (plays made), -1 in runs saved.

Second base: Alexi Casilla has 17 starts, with two each for Trevor Plouffe and Hughes. Casilla is +6 and 3. Hughes, suprisingly, is +2 in plus-minus despite a lousy .923 fielding percentage. The three are a combined 8 and 3 at second; a running total of 5 in plus minus, 2 in runs saved.

Third base: Danny Valencia has 18 starts, Sean Burroughs three. Burroughs has no score; Valencia is -1, -2 (better than I had expected). Running total: 4, 0.

Shortstop: Carroll has played all but one inning so far and is +1, 2. Running total: 5, 2.

Left field: Josh Willingham has 18 starts, Ben Revere two, Parmelee one. Willingham is -1, -2, which is better than I expected; Revere is +2, 1, which is pretty impressive for 18 innnings; Parmelee is -1, 0. Left field totals 0, -1; running total 5, 1.

Denard Span, by the metrics, has been the Twins
best defensive player so far. Ryan Doumit, not so much.

Center field: Denard Span has all 21 starts, with Clete Thomas picking up six innings. Span is +10, 6; Thomas +1, 1 (interesting). Center field totals 11and 7, lifting the running total to 16 and 8.

Right field: A mishmash. Plouffe has seven starts, Thomas and Ryan Doumit five apiece, Revere three and Parmelee one. Plouffe is +1 and 1 (surprise); Thomas 0 and 1; Doumit -2, -1; Revere 1 and 1; and Parmelee -2, -1. Right field totals -2, +1, making the running total 14 and 9.

Catcher: Mauer had 12 starts, Doumit nine. Mauer is 0 and 1; Doumit is 0 and -1, making catcher 0 and 0 and the running total unchanged.

Pitcher: I'll spare the individual breakdowns, but will note that Liam Hendriks isn't doing himself any favors (-1, -2 in 16 innings). As a staff, the Twins are -1, -4, making the running total +13 in plus-minus and 5 runs saved.

Which is better than I had expected to find, and far better, I'm sure, than the Twins were in these metrics at any point in 2011.

Coming up short

Baseball Info Systems's
advanced metrics suggest
Trevor Plouffe is the worst
defensive middle infielder
in the American League.

I have this season periodically dug into the advanced fielding metrics designed by Baseball Info Systems, and after seeing Ron Gardenhire promise "a lot of research and a lot of digging on what's going to make us better in the infield," I returned to my iPad app this morning to see what it tells me.

And the top take-away: The best way to "make us better in the infield" might be to stop playing Trevor Plouffe.

By BIS's "plus-minus" and "runs saved" measurements, no team in the American League is getting worse shortstop play than the Twins.

Derek Jeter has the worst individual totals; the Yankee captain is at -22 in plus-minus, -18 in runs saved. (I write in present tense, but the app appears to be about four games behind on the Twins, so the numbers here are not up-to-date.) But Jeter complied those numbers over 969 innings.

Tsuyoshi Nishioka is -10 and -11 in 508 innings; Plouffe is -13 and -9 in 247 innings (and, since that certainly doesn't include his performance in Kansas City this week, that understates how poorly he's played). That makes the Twins duo -23 and -20, worse than Jeter in almost 200 fewer innings.

And Plouffe isn't merely a dismal shortstop. He's even worse, on a per-inning basis, at second base (-9 and -8 in just 127 innings there.) In fact, he has the worst totals at second base in the AL, in about one-tenth the playing time of a regular second baseman. The Twins' second base play, by these metrics, have been about average if you take Plouffe's innings out.

Now the disclaimer: Even Baseball Info Systems is wary of putting a lot of weight on one year's numbers, much less one year of part-time play. But these figures jibe with what we think we see. If Trevor Plouffe isn't the worst defensive middle infielder in the AL, he's in the running.

And Nishioka isn't much better. The more we see of this defensive duo, the better Brian Dozier looks — and we haven't seen him yet.

Knocking on the cellar door

Poetry in motion: Danny Valencia and Tsuyoshi Nishioka
turn this soft grounder into an infield single — and allow
a runner to score from second base.

Decades ago, one of my buddies was in a play. I don't remember the name or the plot. All I remember was one of Ron's lines, which he delivered with gusto: "Down in the cellar with the rest of the rats!"

Which, it becomes increasingly clear, is the Twins ultimate destination this season. Monday night's sloppy loss reduced the tragic number to 12. It also left Minnesota a mere half-game behind the Kansas City Royals in their desperate battle for last place in baseball's worst division.

When Dick Bremer is repeatedly conceding that the Twins won't win the division — and he seemed to alert the audience to this fact Monday night with the frequency of CNN flashing "breaking news" on two-day-old stories — things have gotten so bad that even the professional optimists of the hometeam broadcast booths can't ignore it.

Monday morning's post cited some sophisticated defensive numbers that put a spotlight on Danny Valencia's shortcomings defensively. Joe Christensen gets specific on what he's doing wrong, particularly on plays to his glove side.

Note there the dissonance between the player and the manager. The manager is focused on the runs the Twins gave away Monday; the third baseman is focused on the runs the Twins hitters didn't create.

When everything is broken, it's not easy to identify what to fix first. The weekend announcement that the Twins will add three days to spring training — not for games, but for defensive drills — suggested that the first priority is going to be defense. Gardenhire's postgame talk about the defensive flaws suggests the same.

Measuring the defense

The Twins have had far too much of this in 2011.

Monday's print column deals with the question of who should play center field for the Twins in 2012, Ben Revere or Denard Span, and in the process of writing that I gathered up this information on the Twins overall defense that won't fit.

Using the two sophisticated defensive metrics available from Baseball Info Systems on my iPad app, the Twins entered the weekend:

• Two runs below average at first base (-2 total); 
• Minus 8 at second base (-10 total); 
• Minus 22 at third base (-32 total); 
• Minus 15 at shortstop (-47); 
• Plus four in left field (-43); 
• Plus 17 in center (-26); 
• Minus 5 in right (-31) 
• Dead even at catcher (-31); 
• Plus 10 at pitcher (-21 total)

In plus-minus, which measures plays made or not made compared to MLB average:

• Zero difference at first base;
• Minus 11 at second (-11);
• Minus 25 at third (-36)
• Minus 18 at short (-54)
• Plus three in left (-51)
• Plus 34 in center (-17)
• Minus 19 in right (-36)
• Zero at catcher (-36)
• Plus five at pitcher (-31)

Comments: 

1) The Twins entered the weekend having allowed 77 more runs than the average American League team. The BIS numbers suggest the defense is to blame for 21 of those runs. I’m inclined to believe that they understate the defensive problems.

2) There are voices on the Internet complaining that Ron Gardenhire is overly critical of Danny Valencia. These numbers suggest that Valencia's defense deserves heavy criticism.

3) BIS itself is wary of putting a lot of weight on a player's single-year numbers, and playing time at every position other than third base is badly fractured, meaning that sample size make individual conclusions even more suspect. 

But with that caveat, I'll note that Alexi Casilla and Matt Tolbert accounted for 517 innings at shortstop, while Trevor Plouffe and Tsuyoshi Nishioka combined for 637. Casilla and Tolbert (as shortstops) combined are -3 in plus-minus, zero in runs saved — in other words, almost exactly average defense. Plouffe and Nishioka are a combined -15 and -15. Similarly, the second base hole is almost entirely Plouffe (-8 and -7 in just 86 innings).

4) The obvious problem areas defensively: Third base, shortstop, right field, second base.

Problem spot: Infield defense

This dropped fly ball in left field by Ben Revere
was one of three errors charged against the Twins
Sunday. There were other unmade plays, not
scored as errors, that hurt their chances.

A basic rule of thumb: Shuffling middle infielders in and out of the lineup is bad for the defense. Double play combos thrive on familiarity.

So it shouldn't be overly surprising that that the Twins have had poor middle infield defense this season. They've played five different men at second base (Tsuyoshi Nishioka, Luke Hughes, Michael Cuddyer, Matt Tolbert and Alexi Casilla) and three different men at shortstop (Tolbert, Casilla and Trevor Plouffe).

Plouffe is supposed to be the regular at short for now -- at least until Nishioka is ready to play, and maybe beyond. But his play afield Sunday was somewhat less than airtight.

Not that he was alone. Consider the third inning, in which the Red Sox scored four runs and took a 5-3 lead. All four runs were earned, but the Twins had numerous opportunities to get out of the inning. (I should note here that  I saw none of the disaster; I was, at that point, following the game on the radio and hence a hostage to the observations of Gladden and Gordon; this is less than optimal.)

Carl Crawford triples. The radio boys suggested that Denard Span might have been able to catch the ball but instead settled for trying to play the carom instead. Said carom got away from him, allowing Crawford to reach third. So Span passed on the risky play of trying for the out, then failed in the fallback position of holding Crawford to a double.

Jason Varitek grounds to first, scoring Crawford. In an alternative universe in which Span catches Crawford's ball, there are now two outs; in the one in which Span holds Crawford to two bases, Crawford advances to third.

Jacoby Ellsbury singles past Plouffe.The radio boys and Joe Christensen of the Strib agree: This could have been ruled an error on Plouffe. (The ruling of single extended Ellbury's hitting streak.) Hsad this play bene made ... it could be the third out in one scenario, or it could have scored Crawford in another, or it could have left him at third. I'll assume it would have scored Crawford. But at least there'd be two outs and nobody on.

Ellsbury steals second. Of course he did; Pavano is notoriously poor at holding runners. In a cleanly played inning, of course, there's no steal because he's not on base.

Dustin Pedroia walks. No defense for the base on balls. Adrian Gonzalez singles. Ellsbury scores, Pedroia to third.

Kevin Youkilis grounds into a force out. Gordon's initial call anticipated an inning-ending double play, but Plouffe and Casilla couldn't turn it. Pedroia scores.

At this point there have been multiple ways out of the inning. Maybe Span catches Crawford's triple. Maybe Plouffe snares Ellsbury's grounder. Maybe the double play is turned. Maybe the Twins yield one run, maybe two. Instead, they've yielded three, and the inning is still alive.

And it goes on to include two more singles and a throwing error by Drew Butera trying to pick Youkilis off. The wonder isn't that they gave up four runs in the inning, but that it was only four runs.

Yankees eat some Humber pie

Two former Twins: catcher A.J. Pierzynski and
pitcher Phil Humber.

Monday offered one of those confluences that tend to plague me: A night off work for me fell on a night off for the Twins.

So I watched the ESPN broadcast of the White Sox-Yankees, secure in the knowledge that one of them had to lose (and tormented by the fact that one had to win).

It's  been well over 12 hours since the final out, and I still can't quite get my brain around the notion that Phil Humber — sour fruit of the Johan Santana trade, he of a 6.10 ERA in 20 scant innings with the Twins — no-hit the Yankees into the seventh inning. Nor can I credit the fact that after four starts with the Sox that he is 2-2 with a 3.20 ERA.

Can he keep this up? I have my doubts, but I'm the guy who greeted the spring training news that he would step into the Sox rotation for the recovering Jake Peavy by referring to him as "throwing batting practice."

As unlikely as I find the notion of Phil Humber getting outs, his performance may not have been the biggest surprise of the game for me. The Yankees made two uncharacteristic defensive mistakes, each of which led to Chicago runs. (And since the game ended 2-0, those miscues, neither of them an error, lost the game.)

One: Center fielder Curtis Granderson made an overly aggressive dive attempt on a leadoff single and turned it into a double. Two ground balls later, the Sox had their first run in something like 28 innings.

Two: In the ninth inning, Alexi Ramirez hits a little pop-up over the pitcher's mound. Rafael Soriano does what he's trained to do: Point to it and get out of the infielders' way. (Major league teams don't want their pitchers trying to backpedal up and down the mound; that's a good route to ankle injuries.) Only Derek Jeter never got there. Base hit.

The ESPN boys immediately blamed Soriano, I assume because the broadcast rights contract requires that nobody ever criticize Jeter for anything. (Rick Sutcliffe spent much of an inning earlier denouncing a new book that is apparently critical of Jeter's "leadership" on the basis that Jeter had nothing to do with the book, as if that has any relevance to the truth or falsity of the content.)

It's just one game and two plays, but it sure didn't look like the Yankees who won the World Series in 2009 by playing it safe on defense and letting the other team make the mistakes — and knowing that the thunder in their lineup was sure to generate runs.

And the Twins best defensive player is ...

Let us begin this post with the standard disclaimer: Nine games is too few from which to make reliable judgments.

It's too few with hitters -- anybody really think that hitters like Delmon Young and Michael Cuddyer are going to remain below .200 all season? -- and it's too few with fielders as well.

And, as regular readers of this corner of the Internet know, I bear a dose of skepticism for the audit-proof defensive metrics that have emerged in recent years. I want to know the numbers, but I decline to base any conclusions solely on them.

So the numbers to be cited in this post are illustrative of what's happened so far, but not predictive of the future.

End disclaimer. On to the meat.

A nice Delmon Young catch from last
season, when there weren't many of them.

Through the wonders of my iPad and a new app, I now have access to Baseball Info Systems' two sophisticated defensive metrics, plus-minus and runs saved. The figures for the Twins aren't pretty.

Tsuyoshi Nishioka is already -3 in plus-minus. Cuddyer, with one game played at second base, has (says BIS) already cost the Twins a run there (and another run at first base, where he has two games played.)

Indeed, the only infielder who isn't in the negatives in runs saved or plus-minus is Matt Tolbert, who has just 17 innings in the field.

But the outfield — Cuddyer and Jason Repko are both +1 in plus-minus in right field, and each (and Jason Kubel as well) is credite with one run saved. Denard Span is 2 to the good in both categories (he's played every inning in center field so far).

But the best defensive player so far: Young. He's +4 in plus-minus with three runs saved.

Those are incredible figures considering how incompetent these metrics have held him to be in the past. He's -57 over the past three years in plus-minus, -27 runs saved in the same period.

And now he's saving the Twins a run every three games? Really?

We'll see how long this lasts.

Flawed defense, flawed numbers

From a comment to my Friday post:

So the Twins had the best second baseman, a very good shortstop and a good centerfielder and defense hurt them? Am I missing something?

Perhaps. Like this game. And this game. And this utter travesty. And ...

A slimmer Delmon Young looked better
on defense in 2010 to most fans,
but the defensive metrics say he was as
bad as before. So do you believe your
lying eyes or the newfangled stats?

Let me back up. I did a lousy job on the Friday post. It was supposed to be about Nick Punto's rather remarkable plus-minus and runs saved numbers, and it wound up being about almost everybody else's first — and the underlying problem with the whole idea is that I don't completely trust those stats to begin with.

I like the idea behind the stats, but just try to describe how Baseball Info Systems decided that Punto saved eight runs over an average third baseman in just 344 innings, or seven runs over the average shortstop in just 258 innings. You can't. I have the books, and they can't describe it well.

What's more, they don't fully trust those numbers either. That's why they supplement the 2010 leaders with a rolling three-year figure, and why they left Punto out of the leaders rankings.

Several years ago, when BIS rolled out plus-minus, I wrote a Monday print column on how we're being asked to take these fielding metrics on faith. Unlike the hitting stats, they can't be readily audited. If you doubt that Joe Mauer hit .327 last season, you can go through the box scores and check it yourself. If you question his 91 runs created, you can work though the formula. If you doubt that Orlando Hudson was +23 in plus-minus, you literally can't figure it yourself.

I'd rather know the plus-minus and runs saved stats than not. But I'll continue to doubt their validity as long as I'm watching players who look good in those numbers give away outs and runs and games.

Back to SoCalTwinsFan's comment:

So, the Twins are weak defensively in three of eight positions on the field and those three are the least important on the field, RF, LF and 1B.

Are they the least important? No; they are the least demanding, which is a different concept altogether. Let's ditch for the moment my reservations and work with runs saved. Over the past three years, Hudson has saved his teams 18 runs. Delmon Young, by the same measurement, has cost his team 27. If a run is a run, Young is costing his team more runs than Hudson saves.

It takes more ability to play second base at a major league level of skill than to play left field. That is undeniable. That doesn't mean the runs Hudson saves are more important than the runs Young allows.

The magic glove of Nick Punto

Measuring defense has not been perfected by any means, but it's still better to know what the metrics are saying than to ignore them. (Imperfect knowledge -- coupled with an understanding of the limitations of that knowledge -- is better than ignorance.) The Bill James Handbook, of which I wrote a few days ago, publishes two yardsticks devised by Baseball Info Systems: plus-minus and runs saved.

Glovework is a big part
of what Nick Punto offers.

Notable Twins in these stats:

Orlando Hudson, now a free agent, had 17 "runs saved" at second base in 2010, tied for the best in baseball with Tampa Bay's Sean Rodriguez, and led all second basemen in plus-minus with a +23. Over the past three years, the O-Dog is credited with 18 runs saved and a +27 -- most of both coming last season.

Alexi Casilla's three-year total in runs saved at second base is -13, tied for fifth worse.

J.J. Hardy has a three year total of 18 runs saved at short, at 2010 total of 4. In plus-minus, Hardy was a +6, tied for ninth in BIS' rankings, and a solid +29 over three seasons.

Delmon Young ranks last among left fielders in the three-year totals, -27 in runs saved; in 2010, he was -8, still among the league's worst. In plus-minus in 2010, he was -20, worst in all the majors; over three years, he's -57.

Michael Cuddyer, despite not playing at all in the outfield after Justin Morneau got hurt in early July, was a -6 in runs saved in right in 2010, tied for fourth worse in baseball. In plus-minus, he was -17, which is ghastly.

Denard Span was top 10 among center fielders in runs saved (nine) and in plus-minus (+14).

And then there's Nick Punto.

Punto was (in plus minus) a +1 at second base, a +10 at third base, a +8 at shortstop. In runs saved, he was 1, 8, 7. These are very impressive numbers for relatively few innings.

BIS didn't include him in their rankings; one of their people told me it was because he didn't play enough innings. But he'd have easily made the top ten at both third base and shortstop in both runs saved and plus-minus.

The question is: How much weight can one put on those numbers? Answer: Not a lot. BIS publishes three-year numbers in part because the single season figures, even for a regular, can fluctuate so much (see Hudson). For a part-timer, the fluctuation problem figures to be worse.

I continue to believe that defense, both the portion we can measure and that which is still hidden from statistical analysis, is one aspect of the game that kept the 2010 Twins from being a truly outstanding team.

Error, hit or misplay (Alex Rios edition)

Eighth inning Tuesday night, bases loaded, and Denard Span drives a shot to deep left-center. Alex Rios, the White Sox center fielder, appears to have it in his glove, but the ball shoots up and out of the webbing. He grabs for it with his bare hand, but it again eludes his grasp and falls to the ground.  Three Twins run home.

Alex Rios' juggling act in center

Scoring ruling: Double.

And let the argument  begin.

My take on that play and that ruling:

1) I expect a major league center fielder to make that catch. Perhaps I've been spoiled by watching Kirby Puckett, Torii Hunter and Carlos Gomez patrolling center field for the Twins, perhaps my standards are too high, but I expect it to be caught.

2) There's not an official scorer who'd rule that play an error (perhaps if it's late in a no-hitter, then they might, but they'd feel cheap and guilty about it).

The two thoughts are compatible in this way: Imagine that the White Sox had in center field a player who isn't a bona fide center fielder. Carlos Quentin, let us say. He might give good effort, but he's not going to lay a glove on that ball. And the scoring ruling would be a double. (Or triple.) Rios is fast enough to lay a glove on it; should he really be penalized for being a better outfielder than Quentin?

I don't know what the advanced metrics say about Rios as a center fielder. I have to say that I'm  not impressed. He overran a single earlier Tuesday to allow a run to score. Back in July, his bizarre throw to nowhere handed the Twins one of the signature games of the 2010 season. It's possible that he's a good center fielder who just happens to have made his worst plays when facing the Twins. I think it's more likely that he ought not be a center fielder, just as I'm increasingly convinced Span isn't living up to the the Puckett-Hunter-Gomez standard.

It doesn't have to be an error to be a bad defensive play. Rios had a couple of bad plays; Span had one.