From the Handbook: Bill James leaders

Over his long career as a pioneer of sabermetrics, Bill James has concocted a number of analytical stats. Some. such as runs created, have become foundation pieces for further metrics; some never caught on in the sabermetric community.

Win shares is one of the latter. I took note earlier this offseason of the philosophical debate on social media over WAR -- wins above replacement -- which has the same goal as win shares (to put one all-encompassing number on a given player's contributions). James' criticism of the various formulas for WAR is that, despite its name, the stat is disconnected from wins. Win shares is not.

In the leaderboard section of the 2018 Bill James Handbook, following the traditional stats (common and obscure) comes the Bill James Leaders. And I see that Joe Mauer is nestled between Robinson Cano and Chase Utley in career win shares, 290, seventh among active players. The top 10 is, really, made of guys I expect to get to the big stage at Cooperstown some induction Sunday down the road -- Albert Pujols, Miguel Cabrera, Adrian Beltre, Ichiro Suzuki, Joey Votto ... and, yes, Mauer. 

From the Handbook: 2017 Leaderboards

Another longtime staple of the annual Bill James Handbooks: top ten lists in virtually every statistical category. A double of easy Brian Dozier nuggets:

Dozier was third in the AL in plate appearances (705), but he was "only" eighth in at-bats (617). That's because his 78 walks were ninth most in the league, and most of the regular top-of-the- order guys didn't show on that list..

Sticking with Dozier: his 306 total bases ranked seventh in the league. But three, maybe four, American League middle infielders had more; it depends on what you consider Jose Ramirez, who split 2017 between third base and second base. The others are Francisco Lindor, Jose Altuve and Jonathan Schoop. Another shortstop, Elvis Andrus, is ninth. (Andrus? Yep. The once light-hitting shortstop not only hit .297 but spiced it with 44 doubles, four triples and 20 homers.)

Moral of that list: Teams are demanding, and getting, some serious thump in their middle infields. And since most of these guys are hitting at or near the top of the order, they're getting more opportunities than the outfielders and first basemen hitting in the middle.

From the Handbook: Lefty-righty statistics

Back in the earliest days of the annual Bill James Handbook -- before Baseball Reference -- this was the section that made the book a worthy investment for a Strat-O-Matic fanatic like me. Today it's a bit redundant.

But it's still part of the book. A handful of Twins of note:

Joe Mauer in 2017 slashed .308/.377/.377 against lefties, .304/.386/.430 against righties. One doesn't often see a player with equal on-base percentage and slugging percentage, even in a low-figure split such as this (Mauer had 130 AB versus lefties).

Eddie Rosario, who has had platoon issues in the past, hit .279 against lefties last year. That sounds acceptable, but his OBP against them was a mere .293 and his slugging just .380. These are good when compared to Max Kepler, but that's an exceedingly low bar.

Ervin Santana was, oddly, more effective against lefties last year: .215/.260/.386 vs. lefties, .234/.308/.397 vs. righties. I wouldn't care to wager on that repeating.

Why Taylor Rogers didn't hang on to the full-time eighth inning role last summer: Righties hit .287/.362/.404 against him.

From the Handbook: Ballparks and Park Indices

First impressions are lasting impressions. Target Field has, at least superficially, the same dimensions as the late Metrodome, but it got a reputation in its debut season as a difficult home run park, and I'll wager it continues to be seen that way.

The numbers don't back that up. From the 2018 Bill James Handbook:

In 2017, the Twins and their opponents combined to hit 232 homers in Target Field, 198 when the Twins were on the road. BIS puts the park's home run index at 114, meaning homers were 14 percent higher there than in the "average" MLB yard. That's not the highest (Yankee Stadium's was 132, for one), but it's pretty steep.

The hitters' advantage is a muted a bit over a long time frame. In the three-year period 2015-17, Target Field's HR index is 106, a more reasonable figure -- but  still higher than that of, say, Houston's stadium (101), which is routinely referred to as a bandbox.

From the Handbook: Manufactured Runs, Productive Outs and Unproductive Outs

It's Dan Gladden's favorite play, or so I deduce from how often he praises hitters for it and complains when it doesn't happen: the groundball out to the right side of the infield with a runner on second and nobody out.

It is a longstanding staple of sabermetrics that the "productive out" is a fallacy. No out is productive. Production comes from avoiding outs.

Baseball Info Systems has been tracking the actual numbers for a few years now.  BIS has a six-part definition of a manufactured run, but it's rather commonsencial. You know one if you see one.

The Twins last season manufactured 164 runs. Only three teams had more. One of them was the World Series champs (Houston), and another was the Boston Red Sox, who won 93 games. The third was the San Francisco Giants, with the worst record in the National League.

Brian Dozier was involved in 26 manufactured runs, most on the team and ninth most in the majors. Byron Buxton was involved in 21, and Eddie Rosario in 18.

The Twins had 243 productive outs -- outs that advanced one or more runners. Seven teams had more (and another had the same number). The Twins with the most productive outs were left-handed hitters who pull a lot of ground balls: Rosario (31), Joe Mauer (30) and Jorge Polanco (29) -- Polanco, of course, being a switch-hitter.

Polanco and Rosario also led the team in unproductive outs, 88 and 77 respectively. (With relatively low on-base percentages, they make a lot of outs, period.)

From the Handbook: Pinch-hitting

Darn near a half-century ago I was fascinated by a piece about the history of pinch-hitting. I can't remember who wrote it or when it was published, but I know it's hopelessly outdated today as a source of information about pinch-hitting accomplishments.

But the topic of pinch-hitting -- the art of coming off the bench for one at-bat -- is fascinating. It's a difficult task. The average MLB hitter last season hit .255 with a .750 OPS (On-base Plus Slugging). The average pinch-hitter hit .223 with a .648 OPS.

Four members of the 2017 Twins had at least 10 pinch-hit appearances.

Ehrie Adrianaza went 1-for-13 with a walk as a pinch-hitter. Eduardo Escobar when 2-for-9 with five walks (!), two of them intentional, so he had a really good on-base percentage. Neither had an extra base hit. Robbie Grossman was 2-for-10 with three walks and a double -- good OBP, but again, not much power.

And then there was Kennys Vargas. Vargas went 4-for-11 with a walk and a homer. He drove in four runs. Batting average .364, OBP .417, SLG 636. Not bad at all, Mr. Vargas.

From the Handbook: Average fastball velocity by age

This is a new one to the Bill James Handbook -- a year-by-year listing of the average fastball velocity of each pitcher.

This is probably most interesting for veterans. For example, Yu Darvish, the free-agent right-hander the Twins have made a display of coveting. Darvish just finished his fifth season of actual pitching in the majors (he missed 2015 with to Tommy John surgery) and had his highest FB velocity, 94 mph.

Ervin Santana, over the eight seasons shown in the chart, has been rather remarkably consistent -- 93 or 92 every season, with 93 for 2017.

One other Twins pitcher I'll take note of: Tyler Duffey was at 90 in 2015 and 2016, when he was a starter. Out of the bullpen in 2017, 92. I guess I would have expected a bit larger uptick.

From the Handbook: Pitcher repertoires

This is one of my favorite and most used portions of the annual Bill James Handbook. I have generally written something annually out of this section, and almost always about Phil Hughes -- in large part because his repertoire changes so much from year to year.

Tracking it backwards:

2017: 46 percent fastballs, 21 percent cutters, 20 percent curves, 13 percent change ups.

2016: 49 percent fastballs, 26 percent cutters, 23 percent curves, 3 percent change ups.

2015: 59 percent fastballs, 20 percent cutters, 16 percent curves, 5 percent change ups.

2014: 65 percent fastballs, 21 percent cutters, 14 percent curves, less than 1 percent changeups. (This was his first year with the Twins and the best season of his career.)

2013: 62 percent fastballs, 0 percent cutters, 17 percent curves, 16 percent sliders, 5 percent change ups, less than 1 percent splitters. (With the Yankees; note the use of the slider and the absence of the cutter.)

Obviously the thoratic outlet issues -- with surgery after the 2016 season and now again -- has had something to do with this, but the fastball useage keeps declining, and the percentage of changeups reached a new high last year.

I don't know what to expect from Hughes in 2018, other than that if he's sound he's going to be on the staff. The man's signed through 2019. He might be in the bullpen, he might be in the rotation. I don't think he's going to be effective if his velocity remains below 90 mph, as it was last season. 

And I don't think he's going to be effective throwing a lot of changes, either. I always thought that Neil Allen and his change up obsession was a poor fit for Hughes as pitching coach. It will be interesting to see if Hughes backs away from the change with Allen gone -- although if the velocity is gone for good, it probably doesn't matter.

On "short" home runs

A reader given to emailing comments to me suggested (regarding Monday's post) that Brian Dozier might lead the league in "short" home runs.

Well, let's look at the data from Baseball Info Systems:

Dozier in 2017 had 22 flyball outs that traveled 330 to 349 feet, no homers of that distance.

Edwin Encarnacion had 20 flyball outs in that range -- and three homers.

Aaron Judge had just 11 outs in that range -- and two homers. Scooter Gennett, 10 and 3. Mookie Betts, 20 and 3. Carlos Santana (who, like Encarnacion, played for Cleveland), 18 and 2.

There are a LOT of players with more flyball outs in that range than Dozier. Whit Merrifield of Kansas City: 31 outs of 330 to 349 feet, no homers.

Obviously, the ballpark has a lot to do with that. Kansas City's yard is unforgiving of fly balls, Yankee Stadium the opposite. Pull a popup down either line in Boston, you got a chance. Something in Cleveland seems condusive to particularly short homers. I don't think Target Field has any particularly soft touches in that regard.

Dozier hit 33 homers in 2017, and 22 of them went at least 390 feet. Five were between 350 and 369, six between 370 and 389. Encarnacion (38 total) had 15 homers shorter than 390.

Mike Moustakas -- big home park -- also had 38 total. Nine of them were between 370 and 389. Only three were shorter than 370, but with about the same number of fly ball outs shorter than 370 than Encarnacion.

And here's an oddity that caught my eye: Gary Sanchez of the Yankees had a total of 37 flyball outs of 330 feet or more, and 33 homers. That's really strong percentage. He has a favorable park, but that's a lot of homers per flyball.

Anyway, if I had to crown somebody as the king of the cheap homer in 2017, it would be Edwin Encarnacion, not Brian Dozier.

From the Handbook: Pitcher analysis

Let's take a dive into the Twins' newly signed closer, Fernando Rodney.

According to the Bill James Handbook, in 2017 Rodney

• threw a total of 978 pitches to 231 hitters
• struck out 65 men and walked 26
• 70 ground balls, 21 line drives and 43 fly balls
• threw strikes on 61 percent of his pitches and got swinging strikes on 21 percent
• opened 114 at-bats with a no-ball, one-strike count, 94 one-ball, no-strike
• had 37 full counts
• went to two strikes on 133 hitters, three balls on 60

All of which tells us ... probably not very much. We already knew that command is sometimes a problem for him. We're going to see a lot of deep counts. (I didn't know, however, how many grounders Rodney got last season.)

Brandon Kintzler last year threw 64 percent of his pitches for strikes. The difference between Rodney and Kintzler on that doesn't sound like much ... just three percentage points. It amounts to about one less strike every three outings. But that difference adds up. It's why Kintzler walks less than two men per nine innings and Rodney more than four.

From the Handbook: Long outs and home runs

It's been long enough since I delved into the Bill James Handbook for a topic that you might have hoped that I've given up trying to mine that vein. No such luck.

This section deals with long fly balls -- those that went out of the yard and those that were outs. Brian Dozier, for example, hit 10 fair balls at least 400 feet (according to this accounting); all 10 were homers. Obvious, you say? Well, no: Miguel Cabrera of the Tigers also hit 10 400-footers, and four of them were outs. (The Detroit park has some space to its centerfield.)

Most 400 foot (or more) flies do leave the yard. Joe Mauer was 3-for-3, Jorge Polanco 5-for-5, Byron Buxton 5-for-5, Eddie Rosario 8-for-8, Miguel Sano 14-for-14, Max Kepler and Eduardo Escobar 9-for-9.

Things start getting dicier just a little shorter, however. Dozier hit 20 balls between 390 and 400 feet; 12 were homers, eight were outs.  Rosario hit 17 balls that distance; nine were homers.  Kepler hit seven; four homers, three outs. Mauer's four were evenly split. Just eyeballing it, that seems the "usual" percentage; a bit over 50 percent of balls hit 390 to 400 feet are homers.

And then there's Sano. Sano hit nine balls that distance; eight were homers. I'll assume it's a fluke until I see it repeated.

(There's an obvious flaw in all this -- it doesn't account for balls neither caught or a homer, say a double off the wall.)

From the Handbook: Holding Runners

This is actually from a section entitled "Pitcher hitting, fielding and holding runners, and hitters pitching." And since we all already know Chris Gimenez pitched in six games last season, we'll narrow it down to something else a bit off beat -- holding runners.

Kyle Gibson pitched 158 innings for the Twins in 2017. Six runners attempted to steal off him; only one was successful.

There there's Jose Berrios -- 145 innings, nine attempts, six steals. Ervin Santana: 211 innings, 13 attempts. 11 successes.

To be sure, there isn't as much basestealing today as there was when I was young and artificial turf was in vogue. Managers wisely are less interested in one base in an era when there are twice as many homers hit and higher strikeout rates make "manufactured runs" less likely. And more precise data on "pop times" -- the time between the pitch hitting the catcher's glove and the catcher's throw to second hitting the infielder's glove -- and base runner speed gives managers a better handle on the real chance of a specific runner stealing a base successfully.

The result is that while basestealing attempts are down, success rates are up.

Unless you're running on Gibson. In which case, not only are attempts rather infrequent, success is even more rare.

From the Handbook: Relief pitcher use

One of favorite segements of the annual Bill James Handbook in recent years has been the relief pitching breakdown, detailing how different relievers were used.

One of the running themes in the blog last season was Paul Molitor's handling of his always thin bullpen. He never seemed to have more than three guys he trusted -- and the identities of those three changed frequently.

The Handbook breaks down the usage of 12 Twins relievers -- the 12 with at least 10 relief appearances.

A few items:

The Twins reliever who inherited the most runners was Ryan Pressley, 38 -- one more than Tyler Duffey. Eleven of Pressley's inherited runners scored (29 percent). But 17 of Duffey's scored (46 percent, worst of the 12 listed). The best strand rate: Michael Tonkin, who has been sold to the Nippon Ham Fighters in Japan (10 inherited runners, one scored).

The reliever with the highest leverage index -- a measurement of the cruciality of the at-bats -- was Brandon Kintzler, the closer who was traded at the end of July and is now a free agent. His leverage index was 1.7. For context, 16 of the 30 major league teams had at least one qualified reliever with a higher leverage index.

Kinztler didn't get a lot of chances to pick up saves with Washington; Dusty Baker generally gave the ninth inning to fellow mid-season pickup Sean Doolittle. Kintzler's leverage index with the Nats was 1.8.

No real surprise here: The Twins reliever with the most consecutive days pitching was Taylor Rogers, 19. What is a surprise, to me at least: the next highest was Matt Belisle  with 13. I'll wager that the majority of them came after he ascended to the glory job after Kintzler was traded.

And the Twins reliever with the most "long" outings --  more than one inning -- was Duffey with 18, more than twice that of any other Twins.

From the Handbook: Stolen base attempt times

The Handbook lists the average time recorded by every player with at least six attempts to steal second base. The fastest is Mallex Smith of Tampa Bay, 3.44 seconds. Byron Buxton is fourth at 3.52 seconds, one one-hundreth of a second slower than Bradley Zimmer of Cleveland. (Billy Hamilton of Cincinnati is wedged between Smith and Zimmer.)

Three other Twins are in the chart; Jorge Polanco is 63rd on the list at 3.69 seconds, Brian Dozier is No.72 (3.71) and Eddie Rosario is 74th, one of seven listed with 3.72. (Max Kepler, who had six steal and one caught stealing, doesn't appear to be on the list.)

For me, at least, there are a few surprises on this list. One is the notion that Polanco is faster than Dozier and Rosario. 

Then there's #OldFriend Aaron Hicks, darn near the caboose on this train. His name is 104th on the list of 106, tied with two others at 3.85, a third of a second slower than Buxton and more than 0.10 behind Polanco, Dozier and Rosario.

From the Handbook: Team baserunning

Strictly speaking, this is in the same section as the previous post, but I thought it worth its own post.

Baseball Info Systems' formula rates the 2017 Twins as the third best baserunning team in the majors, and just one base shy of second-place Cleveland. (Arizona was tops at + 103, then the Tribe with 83 and Minnesota at 82.)

This is a bit up from 2016, when BIS' data had the Twins a +61 and the sixth best baserunning team, and sharply up from 2015, when they lost 19 bases and were solidly in the bottom half.

Only one team (San Diego) grounded into fewer double plays than the Twins, and the Padres had more than 100 fewer chances to do so. As I said in the previous post, I don't know that I agree with working GIDP -- that's grounded into double plays, for yesterday's commenter -- into the baserunning formula, but it's there.

Twenty-two of the 30 teams were in positive territory as baserunners in 2017. And while the absolute worst baserunning teams (Detroit and Philadelphia) were bottom-dwellers in the standings, the Houston Astros -- who won 101 games and the World Series -- were -14 for the season, sixth worst.

From the Handbook: Baserunning

We all know Byron Buxton is fast. We also know that speed alone does not equal outstanding baserunning.

But as Baseball Info Systems figures such things, Buxton was the best baserunner in baseball in 2017, at +55 bases.

The essay that preceeds the numbers notes that Buxton gained at least 27 bases from both basestealing and from "bases gained" -- for example, going first to third on a single or advancing on a wild pitch -- making him "the most balanced of the leaders."

Plus he grounded into just one double play. I'm not completely sold on BIS' inclusion of GIDP in the baserunning metrics, but I don't make the rules for them.

Joe Mauer has long been viewed -- accurately for the most part -- as an outstanding baserunner without outstanding speed. He has become visibly slower in the past few years (he is 34, after all) and in 2017 he scored a -7 in BIS's metric. For his career -- the Handbook contains career baserunning totals for players with 1,000 career games started in 2002) Mauer is +86, which isn't close to being the active leader but is certainly no embarrassment.

From the Handbook: Defensive Statistics

This section of the Bill James Handbook has changed considerably over the years as defensive metrics have been devised, tested and improved.

The Twins player I was most interested in checking out here is Jorge Polanco, the regular shortstop who I believe is stretched at the position.

What these stats say is: He's not great, but he gets the job done.

He's at zero in "bases saved" and -1 in "runs saved," which adds up, essentially, to: He's an average shortstop. Which, if he's capable of being a productive middle-of-the-order bat -- and he hit third regularly down the stretch -- is quite acceptable.

It's worth noting, however, that 14 of the 24 shortstops listed as "regulars" at the position were better in the metrics. He's average compared to the entire set of shortstop innings, but less adept than most the guys who play most of the innings. As one might imagine for a key defensive position, the scrubs bring the average down -- and Polanco is good enough to be better than them.

Eduardo Escobar, in about an eighth of Polancos innings at short, is -2 in runs saved, and Ehrie Adrianza is +1 in about a quarter of Polanco's innings at short. That Adrianza is the best of the trio with the glove sounds right. I'm not convinced that Polanco's better than Escobar, but I'm not sure the difference matters much in either direction.

From the Handbook: Shifting

Combining two sections here, complilations of how often teams shifted and how successful those shifts were.

By Baseball Info Systems' accounting, the Twins shifted 906 times in 2017, an increase of 71 shifts from 2016. This includes the at-bats in which the hitter gets to two strikes and THEN Miguel Sano trots over to the first-base side of second base. It might be a one-pitch shift, but it counts because the AB ended with the defense shifted.

This puts the Twins pretty much dead center in the AL in terms of deploying shifts. Seven shifted more often, seven shifted less often.

Brian Dozier is spotlighted in the short introductory essay on hits lost and gained to shifts. Dozier faced 225 shifts in 2017, and BIS says he lost 21 hits to the shift -- and gained 24. Few right-handed hitters face as many shifts as Dozier (Albert Pujols faces more, but that's about it), and it doesn't appear to do much for the opposition.

From the Handbook: Win Shares

I'm going to skip ahead again today because last week, in conjunction with the AL MVP announcement, Bill James posted this essay dissecting what he sees as a fundamental problem with WAR -- Wins Above Replacement -- and then Joe Posnanski posted this piece to amplify James' points, and ... well, here I am.

WAR was developed about the same time -- maybe a little later -- that James concocted Win Shares. He made Win Shares the basis of his major revision of The Bill James Historical Abstract near the close of the 20th century, followed that with a book to explain Win Shares in detail ("Win Shares" -- an obvious title) and then pretty much dropped out of public sabermetrics. Whatever current cutting-edge analysis he does, he does for the Boston Red Sox. He's far more likely to publish books about crime than baseball these days.

He didn't do a lot to promote Win Shares in the battle for acceptance. WAR won. You can find competing versions of WAR on Baseball Reference and Fangraphs; it's not easy to find Win Shares online.

But I've always preferred Win Shares. I'm sure part of that is that James writes so well. I've never encountered an explaination of WAR that didn't leave me gasping for air as I drown in concepts I can't grasp. James I can follow. I'm more confident I understand Win Shares than WAR.

Anyway ... back to the Handbook. The Handbook annually publishes Win Shares. And Win Shares has a notably different take on who was important for the Twins than WAR does, or at least the Baseball Reference version.

bWAR has Byron Buxton as the most valuable Twin (5.1 WAR). Some other position players: Brian Dozier is at 4.4, Joe Mayer 3.4, Jason Castro and Miguel Sano 2.5 each, Jorge Polanco 2.1, Eduardo Escobar 1.3.

Win Shares has Dozier well ahead of the rest with 26. Mauer has 18, Buxton, Polanco, Sano and Escobar 14 each, Castro 12.

My sense is something of a mixture, and I suspect it of being heavily biased in Buxton's favor because I have maintained for so long that the Twins needed to prioritize outfield defense. I think Buxton's the most important player on this team. I think bWar has that right.

But I think Win Shares is more correct on the rest of them, and in particular Escobar. He's underappreciated by bWAR (in my opinion).

From the Handbook: Defensive Runs Saved

On Monday I took note of the "runs saved" by position for the Twins. As a team, the Twins had 28 runs saved -- and 27 of them came from centerfield.

A few pages deeper into the book we get to the individual leaders. Byron Buxton had 24 runs saved in center -- most of any individual center fielder in baseball, but not the most of any outfielder. Mookie Betts, who plays right field for the Red Sox, had 31.

So there were another three runs saved by the others who played center for Minnesota -- mostly Zack Granite and Eddie Rosario.

The worst center fielder by this metric? #OldFriend Denard Span, who cost the San Francisco Giants 27 runs last year. A lot of things went wrong for the Jints in 2017. An aging Span in the middle of that big outfield was one of them.

The only other Twins to appear among the leaders in runs saved in 2017 are Joe Mauer, seventh at first base (7 runs saved) and Jason Castro, tied for ninth at catcher (10). But no Twins were among the trailers -- not even Aldaberto Mejia, who struck me as perhaps the worst fielding pitcher I've ever seen in a Twins uniform.