Forgetes Cespedes

November 30, 2016
 
The Mets just announced that they’ve resigned slugging outfielder Yoenis Cespedes to a four-year deal worth $110 million bucks.

When the news broke, I thought it was a good move. Cespedes is coming off two years of excellent production at the plate, and he’s a reliable middle-of-the-order hitter who play corner outfield to a draw, and doesn’t totally embarrass himself in center. He’s thirty, but he’s been reasonably healthy, and four years isn’t a long contract. He’s hit well in the park, and he seems to like the team.
 
And I liked it for the Mets. If they didn’t resign Cespedes, they were looking to enter the 2017 season with a middle order of Asdrubal Cabrera, Neil Walker, and the Ghost of Curtis Granderson. Call me biased, but that doesn’t seem like a lineup that will strike fear in the hearts of opposing pitchers.
 
And the Mets should be looking to win games. They’ve made the playoffs two years in a row, and they certainly should be looking to compete in 2017 and 2018. Cespedes was their best offensive player last year, so it’s reasonable for the team to keep him.
 
I said all of this in our discussion board. Did you know we have a discussion board? It’s fun. Mostly we talk about politics and Waffle House. At least that’s my general sense of how things are going.
 
One reader, BarryBondsFan25, posted the following response:
 
True, the Mets did need a hitter but Dexter Fowler in my humble opinion was the better fit. Fowler in CF with Bruce and Granderson or Conforto on the corners. Both Bruce and Granderson come off the books after 2017 which would allow the Mets to take part in a bigger free agent class which includes JD Martinez, Lorenzo Cain, Brantley and Cargo. 
 
My initial response….well, you can probably predict my initial response. I like Dexter Fowler a lot, but he’s not Yoenis Cespedes. If I’m a team looking to win baseball games, I want the big Cuban.
 
What about the money question?
 
T’aint my dough. Get the better player.
 
I thought, too, that the Mets already had players like Dexter Fowler.
 
Dexter Fowler is a prototypical leadoff hitter. He gets on-base, and can run decently, though he’s not a burner anymore. He can play a tough defensive position, but he’s on the wrong side of 30, and probably can’t be expected to play that position well. He walks ‘nuff, and strikes out too much.
 
Dexter Fowler is a lot like Curtis Granderson, the guy the Mets slotted in the leadoff spot for the first half of 2016. And Dexter Fowler is a lot like Jose Reyes, who was the Mets leadoff hitter over the second half. The Mets don’t need three leadoff hitters…they need a real hitter. The Mets needed a big name to slot into the #3 spot on the lineup card.
 
*             *             *
 
That’s bad sckeyence.
 
It’s not even science….it’s assumption. I assumed that Fowler was like Jose Reyes and Curtis Granderson because a) all three of them have similar skill sets, and b) all three players have hit leadoff for most of their careers.
 
I was looking at the Mets offense mostly through the lens of traditional lineup structures: speed and on-base guys at the top, good sluggers in the 3rd and 4th spots, and fill out the rest as you can. The Mets had their leadoff hitters in Grandy and Reyes. Asdrubal Cabrera was a logical #2. What the Mets really needed was a #3 hitter. No player on the club could match the production of an absent Cespedes, and there weren’t any better bats on the market. The Mets needed Cespedes, and they got him. Kudos. High-fives all around.
 
And then I remembered RER.
 
RER stands for Run Element Ratio. It’s a stat that Bill invented in one of his annual Abstracts, and one that I think about every two or three years, before forgetting all about it again. 
 
Run Element Ratio attempts to give numerical weight to whether a player is better at starting trouble or finishing it. Phrased more simply, it tells you if you’re a better at starting off an inning, or wrapping it up. 
 
The formula is simple:
 
(Stolen Bases + Walks)  /  (Total Bases – Hits)
 
The dividing line is 1.00. If a hitter has an RER above 1.00, he’s better at starting trouble than he is at finishing it. He’s good at getting himself on base, and he’s good at getting around the bases.
 
If a hitter is below 1.00, he’s better at finishing trouble: he’s better at getting other guys around the bases.
 
Let’s look at two examples:
 
Name
H
SB
BB
TB
RER
Rickey Henderson
3055
1406
2190
4588
2.35
Joe Carter
2184
231
527
3910
0.44
 
Rickey Henderson, the best leadoff hitter of all-time, posted an extremely high RER. Joe Carter, a low on-base power hitter who drove in an insane amount of runs, posted a rather low RER.
 
The metric doesn’t tell us who the better player is. Ron LeFlore has a higher RER than Harmon Killebrew, but that doesn’t mean that we should put the Tigers speedster in the Hall of Fame. It’s just a way to see what dimensions of offense a player contributes to for his team. Are they better at starting trouble, or finishing it?
 
If you look at the Mets as a team, they have a low RER. Their RER ranked 13th in the NL last year, ahead of just the Cardinals and the Rockies:
 
Tm
H
SB
BB
TB
Run Elem. Ratio
MIL
1299
181
599
2168
0.90
PIT
1426
110
561
2226
0.84
SFG
1437
79
572
2215
0.84
ATL
1404
75
502
2119
0.81
CHC
1409
66
656
2359
0.76
MIA
1460
71
447
2187
0.71
CIN
1403
139
452
2238
0.71
WSN
1403
121
536
2338
0.70
SDP
1275
125
449
2115
0.68
PHI
1305
96
424
2089
0.66
LAD
1376
45
525
2257
0.65
ARI
1479
137
463
2446
0.62
NYM
1342
42
517
2274
0.60
COL
1544
66
494
2568
0.55
STL
1415
35
526
2453
0.54
 
We shouldn't read too much into this. The Rockies scored more runs than any team in the NL last year, and the Cardinals finished 3rd in runs scored. Having a low team RER doesn’t mean that the offense is bad…it just means that the offense is crowded with hitters who finish trouble.
 
But I also don’t want to make nothing of this. I watched a fair number of Mets games last year, and one of the most distinct characteristics of their team was the inability of their hitters to just get rolling. This is anecdotal, but it seemed like the Mets lineup had a lot of trouble just stringing together baserunners. Even when they scored runs early, the offense never seemed able press. You’d see the opposing starter give up two or three runs in the first, and then follow that up with a string of easy innings. Suddenly we'd be the seventh, and the opposing pitcher who looked like garbage in the first would be notching a Quality Start.
 
You can see the problem by just looking at individual players: the Mets just didn’t have anyone to start trouble on a regular basis. Here are the guys with 150+ plate appearances last year:
 
Name
H
SB
BB
TB
RER
Yoenis Cespedes
134
3
51
254
0.45
Curtis Granderson
129
4
74
253
0.63
Asdrubal Cabrera
146
5
38
247
0.43
Neil Walker
116
3
42
196
0.56
Wilmer Flores
82
1
23
144
0.39
James Loney
91
0
16
136
0.36
Michael Conforto
67
2
36
126
0.64
Jose Reyes
68
9
23
113
0.71
Kelly Johnson
49
3
15
84
0.51
Travis d'Arnaud
62
0
19
81
1.00
Alejandro De Aza
48
4
26
75
1.11
Jay Bruce
37
0
17
66
0.59
Rene Rivera
41
0
16
63
0.73
Lucas Duda
35
0
15
63
0.54
David Wright
31
3
26
60
1.00
Juan Lagares
34
4
11
54
0.75
Kevin Plawecki
26
0
17
35
1.89
 
Ranked by total bases, you have to go all the way down to Travis d’Arnaud to find a player with an RER over 1.00.  Among Mets regulars, only Jose Reyes (0.71) had an RER that was higher than the NL average (0.70).
 
*             *             *
 
I assumed that Dexter Fowler was a similar player to Reyes and Granderson. He’s not. Or, he wasn’t last year:
 
Name
H
SB
BB
TB
RER
Dexter Fowler
126
13
79
204
1.18
Jose Reyes
68
9
23
113
0.71
Curtis Granderson
129
4
74
253
0.63
 
Dexter Fowler was very good at starting trouble last year, as he has been throughout his career.
 
But Granderson and Reyes weren’t particularly good at starting trouble: they both posted RER’s that were comfortably below the 1.00 dividing line, and pretty close to the league average. The Mets had two players who looked like leadoff hitters, but weren’t all that great at getting the offense started.
 
This shows up in the Mets performance in the first innings of games: the Mets ranked 12th in first inning runs last year. Their optimal lineup just couldn’t get things started.  
 
I want to be clear: I am not suggestiong that Dexter Fowler is a better player than Yoenis Cespedes: I do not expect Dexter Fowler to, say, out-WAR Yoenis Cespedes next year, or for the durations of their contracts.
 
But I do think we’ve fallen into a tendency to see things in one narrow way. Certainly, that’s what I did when I initially answered BBFan25’s comment about Dexter Fowler. I knew that Cespedes was a better player than Fowler, so I assumed that he would be a better player for the Mets.
 
I’m no longer certain of that. Cespedes is a better hitter, but his hitting gives the Mets more skills that they already have. He’s another hitter whose skills are firmly in the ‘finishing trouble’ side of the spectrum.
 
Dexter Fowler would give the Mets a player capable of starting trouble:
 
Name
H
SB
BB
TB
RER
Fowler
126
13
79
204
1.18
Cespedes
134
3
51
254
0.45
 
And in case you think that we’re relying too heavily on Fowler’s strong 2016 season, the three-year splits actually show a bigger gap between these two players:
 
Name
Years
H
SB
BB
TB
RER
Fowler
'14-16
395
44
229
622
1.20
Cespedes
'14-16
474
17
119
867
0.35
 
Again, I don’t want to imply that Dexter Fowler is a better player than Cespedes…that isn’t what Run Element Ratio measures. But Fowler does provide something that the Mets offense really lacks: a player capable of getting on and around the bases at a healthy clip. A player who can jumpstart an offense.
 
And I wonder if that matters more than having the best player.
 
We’ve been talking a bit about the way that statistics can blind us, and we’ve been talking about the difficulty of understanding a complex system (the ebbs and flows of a baseball team) through numbers. I mean, I think that’s what some of us are talking about. At the very least, those are subjects have been central to a lot of my thinking recently, and I think that Bill’s latest entries are at least near this terrain.
 
BBFan25’s comment about Fowler and Cespedes at least touch on these subjects. A lot of us think that Cespedes is the better player for the Mets because he is the better player, full-stop. What I would suggest is that those are actually very different questions, and I think that I’ve been mixing them up for a while.  
 
Is Yoenis Cespedes a better baseball player than Dexter Fowler? Sure. Absolutely. I'm fine with that. 
 
Is Yoenis Cespedes a better baseball player for the Mets than Dexter Fowler?
 
Eight hours ago, I would have said that the answer is still ‘yes.’ I was really quick to dismiss the possibility that Dexter Fowler might have more to contribute to the Mets than Yoenis Cespedes. Eight hours ago, I looked at the Mets lineup and thoughtthat their real weakness was a big bat to anchor the lineup.
 
Now I’m not sure of that. Now I think that the big absence in their lineup is a hitter who can reliably start the offense. And for all his talent, Yoenis Cespedes isn’t nearly as good at starting trouble as Dexter Fowler. Purely as hitters, I've come around to the idea that Fowler's abilities are more useful to the Mets team than the skills that Cespedes offers. Factor in secondary details like defense and cost, and I've come around to BBFan25's opinion.
 
Dexter Fowler was a better fit for the Mets team than Yoenis Cespedes. 
 
 
Dave Fleming is a writer living in Wellington, New Zealand. He welcomes comments, questions, and suggestions here and at dfleming1986@yahoo.com.  
 
 
 

COMMENTS (71 Comments, most recent shown first)

flyingfish
Poor David Wright, so sad to have that ailment, especially in someone so young. Do you think he'll ever play again? I'm not optimistic.
3:40 PM Dec 9th
 
OldBackstop
Alderson was talking up Lagares...Gold Glove a few years ago, signed for a few years yet. He also said Conforto and Granderson.

Boy they are fire saling Bruce, Too many lefties. Duda, Conforto, Bruce, Grandy. Walker, Cabrera, Reyes are switchies.

I still like four 30 hr guys in the lineup. Wright even had 30 back in his day.Just need some table setters.
11:16 PM Dec 8th
 
flyingfish
The Mets need a center fielder. A second baseman wouldn't hurt, either, but let's focus on CF. Seems to me that good center fielders tend to be fast and--I haven't done this study but it should be easy--I'll guess they probably have higher RERs than say first basemen or even corner outfielders.

So, Mets, get yourself a good center fielder and solve two problems at once.
10:41 PM Dec 8th
 
steve161
"...the Mets probably need a high-RER player more than they need another low-RER contributor."

The keyword is 'another'. Cespedes isn't 'another' low-RER contributor. He is THE low-RER contributor. Emphasis on the word contributor.

Nobody here is denying that they need more baserunners. It seems possible that the current roster can supply them if used properly. But lose Cespedes and add Fowler and the result will be an increase in Left-on-Base, not Runs Scored.
10:29 AM Dec 2nd
 
OldBackstop
Yes, the title gave me a chuckle. Once. I'm not mad.
4:28 AM Dec 2nd
 
MarisFan61
P.S. Just thought I'd add, I think the title is brilliant, and I don't stop getting the same laugh from it every time I click on the site.
3:19 AM Dec 2nd
 
MarisFan61
Dave: I doubt that anybody.....well there's always somebody :-) so I'll say, I doubt that more than 1 person in 100 would read the article as not saying the Mets should have gone for Fowler rather than Cespedes. I understand what you mean about the "fit" thing -- but heck, it sure looks like you also mean the other. And moreover, "fit" includes thing besides a player's "RER," such as the various things that several of us have mentioned, but OK.

If you're saying you never meant anything other than that from a theoretical RER standpoint, Fowler is a better fit, I never would have said boo, and I suspect that nobody else would have either.
2:17 AM Dec 2nd
 
OldBackstop
This is one of the more thought-provoking articles I've read here, Dave. Very educational.

I think the Mets offense is very much a team in transit. The only people after 2017 that have contracts are Cabrera...one year option for 2018, Lagares...paid until 2019 for some reason, Wright, and now Cespedes. M any guys under control, of course, but not a bunch of long term plans in place outside the starting rotation.

In fact...given Wright's injury and Lagares' fade...they have no set pieces in place in the field to build around except Conforto, who better start hitting lefties, and now Cespedes. They can add other pieces and shop for talent that fits going forward, but to get Cespedes without giving up kids in trade or losing a draft pick is a great guy to build an offense around for 2018.

Fowler, on the other hand...is he a guy you want to have as a contracted piece three four years out? He only plays center, and he doesn't bring any eye-popping stats to demonstrate his RER pedigree to count on as he ages.

Note: I like Jeff Keppinger. Can't Rivera turn out to be Jeff Kemp or even Gregg Jefferies at least? You know, Kepp left the Mets in 2005, but he played until 2013, and batted .325 in 400 something at bats with the Rays in 2012 with an .800 OPS.

I had him a few times as a bench guy....great position eligibility.

2:07 AM Dec 2nd
 
DaveFleming
The thesis of this article wasn't that the Mets should have signed...the 'thesis', if you absolutely need to have one, was to wonder which player better 'fit' the specific needs of the Mets offense. That's why I didn't mention the defense differences between Cespedes and Fowler, or the likely cost differences in their contracts.

But...I don't write with a 'thesis.' I was just thinking about someone else's comment, and I wrote an article about how my thoughts evolved. My thoughts went from 'Cespedes is obviously better' to 'Christ, the Mets offense is super duper weird', to 'Jeez, maybe Fowler is a better fit.' That's it.
1:15 AM Dec 2nd
 
MarisFan61
(very sorry - typo.
Should be "Don't you need to add....")
12:05 AM Dec 2nd
 
MarisFan61
You need to add some factor of how good the players are, before concluding that a team should have gone after the other player?

Your thesis would (or at least could) hold true, if it's an "all else being equal" situation.
But not if it isn't.
12:04 AM Dec 2nd
 
DaveFleming
I don't know if this helps, Don, but I want to reiterate that we're not really using RER to determine pure value. We're just using it to pinpoint where a player lies on a spectrum. We can call it an offensive spectrum, if that helps.

On one side of the spectrum you have the slappy singles hitters who can draw a walk and get around the bases....your Billy Hamilton's, essentially. On the other side you have the pure power hitters who can drive the ball out of the yard, but don't really get a lot of hits or draw too many walks. In the middle you get the really good players: the guys who contribute to EVERY dimension of the offense....guys who get hits, hit homers, and move around the bases. The Goldys and Trouts.

I wasn't using RER to communicate value: I hammered this point at least once or twice about Cespedes and Fowler, and once about RER as it relates to a team's ability to score runs. I also used two examples (Rickey and Joe C., LeFlore and Killebrew) to illustrate this. But I'll say it again: RER doesn't communicate value. All it communicates is where a player lies on a spectrum of ways to contribute to offense.

What it can tell us, if you look at a team or a league, is where the balance of an offense is. Baseball in general is trending towards lower and lower RER's. The Mets, as a team, have a particularly low RER. They hit a lot of homers, but they don't score a lot of runs.

We can use that broad information to look at what a team's offense is doing, and what it might need. Cespedes and Fowler are at different places on this spectrum: and while we'd all probably wouldn't call them equal players in a pure vacuum, we can cut some of the gap in their anticipated contribution to the Mets offense by noting that the Mets probably need a high-RER player more than they need another low-RER contributor.

It doesn't matter that Cespedes has an RER of 0.40, or that Fowler is at 1.18. That tells us nothing. What tells us something is that the Mets have a bunch of guys with RER's that are close to Cespedes' RER number, and no one who is anywhere in the ballpark of Fowler. When we look at that and then look at the big gap between the team's (impressive) HR totals, and their lackluster count in overall runs scored, we start to understand - maybe - a little bit more about the complex structure of the Mets offense.
11:35 PM Dec 1st
 
MarisFan61
It's not better or worse, just different -- sort of totally different.
For looking at setting the table, the simpler stat of on-base-average absolutely is not only as good as RER, it's better -- except to the extent that a leadoff home run doesn't leave a guy on base (which gets to the thing Dave just mentioned).

On-base tells about setting the table. "RER" tells about what type of hitter the guy is, and how it might be best to use him in the batting order.
8:33 PM Dec 1st
 
doncoffin
I keep struggling with the RER concept...again:

RER = (W+SB)/(TB-H)

(W+SB) is supposed to measure "setting the table." How is it an improvement on just using OBA? I ask because I'm not at all sure what it actually measures.

TB-H is supposed to measure "finishing things off." It assigns (implicitly) a zero value to singles, right? How can that be a useful measure?

But...The (simple) correlation of (W+SB) with runs scored (using all players with 300 + PA in 2016) is 0.72--there is a strong correlation between this component of RER with scoring runs--because the player with more of these is getting on base a lot and/or advancing a lot. (The simple correlation between times on base--H+W--and runs scored at the level of the individual player is 0.91).

But...the simple correlation of TB-H with runs scored is 0.79...because players who hit for more power tend to drive in more runs. (The simple correlation between total bases and RBIs at the level of the individual player is 0.90)

And the correlation of RER with runs scored is -0.08--negative and insignificant. Combining two useful measures into a third measure that means...what, exactly?

Didn't we already know that getting on base is valuable? That hitting for power is valuable? What does RER--which combines these two things in a way that I am struggling to comprehend--add? Why don't we just say: "The Mets seem to have decent power, but their ability to get people on base is an issue." We can do that simply by looking at the team's OPA and SA, can't we?

What does dividing a measure of getting on base by a measure of power even mean?
8:20 PM Dec 1st
 
OldBackstop
Was there a BJOL author meeting this week on developing controversial sticky content?


8:17 PM Dec 1st
 
MarisFan61
I agree with Dave's point about HR's. (!)

It doesn't mean we're saying HR's are bad, or that something else is better, but IMO there's a point to it. I likewise sometimes have the same feeling he expressed about a pitcher giving up a HR in such an inning.

Some years ago, a related thought led me to do a little study of a Yankee season, looking at the chance of having a "big inning" after a leadoff HR vs. a leadoff walk or single. Wish I had the results right here (I posted them on www.forums.mlb.com/discussions/New_York_Yankees/ml-yankees); what I came up with was that at least in that season, the Yanks did have more big innings (3 or more runs) after just getting the leadoff guy on 1st than after a leadoff HR.

BTW there was a slightly different result if we looked at it in terms of "average runs scored in the inning" than "chance of a big inning," because the leadoff HR of course guarantees at least 1 run, while the other thing resulted in 0 runs most of the time.
8:17 PM Dec 1st
 
OldBackstop
re: Fowler and career years....how many guys go eight seasons, 13.9 WARs (Warses?), never over 2.8....have a bust out year at 30 with 4.2 WARies, and then perform there for four years? 1.7 average WAR

Well....how about Daniel Murphy? Seven years, 12.5 WAR, never higher than 3.0, 1.7 average WAR (c'mon, that's pretty damn good for the first guy I tried! :-)) He was a playoff hero, but his regular season Age 30 was only 1.4 WAR. Then he had a great Age 31....4.6 ERA, 25 homers after a high of 14 in his twenties.

How about Joey Bats? 8 years 15.0 1.7 average WAR (just kidding, 2.1) that includes his insane breakout Age 31 with a 8.1 WARsy. Anyone would take the first four years of his 30s.

I think those scenarios might give optimism for people looking at Fowler, except for one thing...power. Those guys' renaissance was built on power, Joey Bats becoming a .540 slugging guy from a sub .400, Murphy having the greatest postseason power run in history at Age 30, followed by a jump of .150 points in his slugging this year.

Bats and Murph hyperspaced into stardom at similar periods most obviously based on power surges.

Now....one thing that might argue that Fowler could do this is his size....he is 6'5", which surprised me. But I don't see it for Fowler because the baseline for his power has been at Coors and Wrigley and he still only averaged 10 home runs and a .219 slugging avg. I hit 16 home runs and slugged .445 there in 2009, or, well, I bet I woulda.

I look at Fowler and I'm reminded of another 6'5" guy, Jason Heyward....Fowler being an older, slower, less powerful, poorer fielding Jason Heyward. I suspect Heyward will rebound, he's only entering Age 27, did win a Gold Glove, but, damn.....

The more I look at his numbers, the more I'm starting to hate Dexter Fowler :-) With that wonderful atmosphere and the lineup at Wrigley you couldn't write a better script for a one off career year. (Note Murphy played for another NL division winner in his breakout year.) If Fowler moves to a bigger park and has to run the 6'"5 body around and try to keep up his SBs into his 30s.....I dunno. I think every facet of his game is poised to decline to sub-replacement level. He better have a sparkling personality.

I'll say it is more likely that his WAR is more like 2.2 than 4.2 in 2017.





8:13 PM Dec 1st
 
DaveFleming
Well, no...I'm not kidding.

We talk a lot about high-stress and low-stress scenarios all the time, right? Every team is studying this sort of thing: what stresses a pitcher out and what doesn't. We know that throwing your 32nd pitch in an inning is more stressful than pitch #5, and we know that throwing pitching #129 in a game is more stressful than pitch #17.

We can probably speculate that pitching with runners on is more stressful than pitching with no one one, right?

I mean, that seems like it would be evident to me. I'd rather pitch with no one on base than someone on base. I play outfield, and I know I feel a lot less stress when there's no one on base...I imagine the pitcher feels the same way.

So doesn't it follow that a pitcher is less stressed in a no-out situation, even if he has allowed a homer, than a situation where he allows a sequence of hits and walks, but no homers?

What is a more stressful inning?

1B, walk, walk, HR, groundout, pop fly, strikeout

or

1B, walk, walk, pop out, strikeout, single, groundout

The first inning is worse on the scoreboard....it's four runs allowed versus two. But the second inning is a helluva lot tougher on the pitcher: he's had six stressful at-bats, whereas the first guy only suffered three stressful at-bats.

Again, it's just speculation. It's possible I'm crazy.
7:42 PM Dec 1st
 
OldBackstop
I dunno, Dave....kinda fond of the 6-4-3 myself....

I get it maybe with a guy on third in the first inning with 3-4-5 on deck and number 2 hits a 2-run home run....nowadays an early grand slam usually means you are reading vicious tweets in the clubhouse in a minute.
7:20 PM Dec 1st
 
studes
Dave...you're kidding, right?
7:14 PM Dec 1st
 
studes
BBF, I don't think Fowler will be as potent a bat as Cespedes for the next four years. I don't use OWP. I tend to like situational stats like RE24. Over the last two years, Cespedes has been solidly better than Fowler in less situational stats. Over the last two years, his wRC+ has been around 135. For Fowler, it was 110 and 129.

I guess the key is whether you think Cespedes' performance over the last two years represent a legitimate improvement as a hitter and/or whether you think Fowler just had a career year.

You make good points in dollars saved and a better fit with the current Mets outfielders. Also I guess the Mets wouldn't lose a draft pick given the new CBA if they sign Fowler, which is something else in the back of my mind.
7:12 PM Dec 1st
 
DaveFleming
This is out from left-field, but do you ever think about the ways that giving up a homer benefits the opposing pitcher?

Bear with me...I'm just spit-balling here.

Sometimes, I'll be watching a game where the pitcher I'm pulling for has a crummy inning. Let's say he gives up a double and then sort of loses the plate for a few batters...can't quite get the slider to hit the spot. He walks a couple guys and then gives up a home run. Suddenly he's down four runs.

Whenever that happens to a pitcher I'm rooting for, I confess that I feel a little relief. Yeah, his ERA is going to take a battering, but he's out of trouble for the moment. He gets to hit the reset button on the inning. There's no one on, so there's nothing to worry about. The bad thing has already happened. Now he gets a clean slate. He can take a breath.

I think this is one of the difficult-to-calculate problems with having an offense that is too reliant on the long-ball: while clearing the deck is nice for the team, it's also provides some benefit for the opposing pitcher who was struggling through the inning.

I tried to look for the number of Quality Starts logged against each team, but I couldn't find the number. I'd be curious where the Mets ranked in terms of Quality Starts by opposing pitcher...


7:03 PM Dec 1st
 
MarisFan61
(that was to OldBackstop....BBF25 beat me by a minute in getting his post in between)
6:23 PM Dec 1st
 
MarisFan61
Dave is gonna wanna kick my arse for making this tiny point, but you flip-flopped Granderson's and Cespedes's HR's. :-)
6:23 PM Dec 1st
 
BarryBondsFan25
Is offensive winning percentage an obsolete statistic? I use it occasionally. In terms of offensive winning percentage here are both players' numbers:

Fowler:

2016- .678
2015- .586
2014- .630

Cespedes:

2016- .646
2015- .711
2014- .587

Fowler averages out a little better. In terms of more frequently used current stats, Cespedes edges out fowler in wOBA and wRC+ but not by much.

When I originally said fowler was the better fit for the mets, I was looking at position, dollars and flexibility for the Mets. If we are looking purely at offense, I still believe fowler is the better fit. The mets hit home runs but it did not translate necessarily to many runs relatively speaking.
6:22 PM Dec 1st
 
OldBackstop
Okay, I'm better.

So here is the thing. We all agree that the Mets were not an efficient offense last year, wasting the run production they should have got from 218 home runs, the second largest number We all agree, group hug.

Fowler might be a better overall offensive boost replacing Yo-C in center, or he might be a better leadoff hitter than Grandy....but let's say we found Fowler and a Fowler clone to replace both them, and he hit as many home runs at Citi as he did at Wrigley (13).

The reason we are talking about the Mets HR frustration is all Yo-C (30 home runs) and Grandy (31 home runs). Replace those two with Fowleries and we net out 35 home runs less....from 2nd in the NL to the median at 8th.

That feels like it would be a bad thing.

6:08 PM Dec 1st
 
studes
Great point about WPA, Dave. So do you choose to give more valuable opportunities to a worse cleanup hitter? Cause that's what you're recommending, right?

Also while knowingly passing on someone who thrives in the rough NYC environment and is also worth millions in extra revenue due to his marketing appeal? In exchange for someone who is a risk in both categories?
6:00 PM Dec 1st
 
OldBackstop

This is like when mom and dad fought. I don't feel safe, and my therapy dog is still in Berkeley. :-(
5:49 PM Dec 1st
 
DaveFleming
I think, actually, that the WPA/LI stuff is pretty revealing about Cespedes AND the Mets.

Well...sort of backing up, so we don't lose anyone. Leverage Index is the measure of how important someone's at-bats were, over the course of a season. If you came up a lot in big situations, you'd have a high Leverage Index. If you were always coming up to bat with no one on in blowout games, your LI is low.

Do you know who had the worst Leverage Index among eligible players in baseball last year?

It was Cespedes. No batter in baseball came up to the plate in lower leverage situations last year than the Mets outfielder.

You know who ranks behind Cespedes? It Eugenio Suarez, a Reds infielder who hit mostly 6th or 7th for the year. Then it's Blackmon (leadoff), and Cesar Hernandez, who hit either leadoff or 8th for the Phillies. Then DJ LeMahieu. Then Alcides Escobar. Then Calhoun....

Cespedes was the extreme last year, but he was also an outlier: he was a legitimately good hitter whose leverage index was what you'd expect to see for a leadoff/late order hitter. He's a guy on a team that made the playoffs, surrounded by players whose teams were mathematically eliminated in mid-April. He doesn't belong on this list, but he's there.

That's not a knock on Cespedes, of course....it's a knock on what the Mets have put around him. If they really want to get the most bang for their buck out of their flashy OF, they should find someone to get on base ahead of him.

***

And no Maris...your feedback is not constructive. It's hectoring. It's being rude.

We're all having a conversation. But because you don't know how to participate in the conversation the rest of us are having, you jump in with these little nits to pick. You tell me that I should change my 'take' because of one iota of information that you presented, and when I ignored your demand, you upped the criticism, calling me an 'embarrassment to the field,' and informing me that I should be more 'modest' in how I present my articles.

That's not constructive feedback. That's being a self-important jackass.

All you are doing is derailing that other conversation, the one the rest of us want to have. And, frankly, ALL of your comments are aligned towards that exact end: you are never able to just participate in the ebb and flow of a conversation.....you have to jump in pointing out every single error anyone gets wrong. And then you act like some righteous crusader when someone tell you to knock it off.

It's gross. You're the guy at the dinner party who checks his phone to find out exactly who played the sidekick to John Wayne in that movie no one quite remembers...by the time you've looked it up, the rest of us are on to some other topic, but you have to tell us who it was anyway.

We've all moved on. That's what happens in a conversation. Go with it.
4:57 PM Dec 1st
 
MarisFan61
Studes: What you said might have been close (although still not right) if not for what Lidsky pointed out:

Isn't the use of a ratio stat like RER a problem when comparing two players? RER says for a given production that a batter is better early or later in an inning. To say that somebody with a higher RER is better early in the inning compared to another batter with a lower RER without considering anything else isn't accurate.
In other words, Hitter A can be be better than Hitter B in both early and late inning skills and have a lower RER compared to Hitter B.


We've been pretty gentle in the criticism, but since Dave is sort of doubling down....."RER" is simply not an apt metric to be using for this question. It's dangerous to be trying to address a complex question (which this is) with any single metric, but if you must, at least make it a good one.

Since, as Lidsky noted, RER only assesses a given player's own relative offensive qualities, it's a very poor choice for assessing the relative abilities of different players. I can think of sort-of-related things that RER would be good for assessing, like, whether a given player is or isn't being used in an apt spot in the batting order, or how much his theoretical value is being wasted by having him in a wrong spot. But, for comparing one player to another?? It's the wrong hammer for the wrong nail. Lidsky noted this, and it's important; nobody really followed up on it. I would agree that it does approach the question of whether theoretically the Mets may have been going after a wrong type of player altogether -- but that gets back to all the specific things about Cespedes in relation the Mets, which is a very large part of this question and trumps much of the rest, and which the article completely bypassed.
4:43 PM Dec 1st
 
studes
A few comments:

I think this article is fine, Dave, and in the best spirit of sabermetrics: questioning consensus thinking. Nice job. I have a feeling MarisFan is reacting to your strong last sentence, not so much the rest of the article.

I wouldn't get too excited about T.J. Rivera. He's Jeff Keppinger.

I don't think your first point about WPA holds, Dave. LI doesn't vary much between hitters over a full season. In fact, Cespedes' LI was LOWER than Fowler's last year.

Cespedes also performed better in RE24 and WPA/LI. You're right that these stats aren't as predictive as something like wOBA, but my point is that if you give up Cespedes in order to balance out the Mets' strengths, you're giving up the best run producer on the team.
3:39 PM Dec 1st
 
OldBackstop
The heck with it, let's just hit more home runs. Everybody uppercut. Let Grandy lead off with 30 home runs.

One thing the amazed me last year was the guys that played positions they hadn't (or almost had never) played before.

Reyes goes to third and came within a whisker of left field.

Wilmer plays first.

Conforto to right. Conforto to center.

TJ Rivera to first base.

Ty Kelly first base.

OldBackstop first base.

Matt Reynolds, left field (he hit a home run, Collins bragged about it post-game)

Eric Campbell 2nd base

==========================

Let's pull in the Citifield fence play to be top 3 in home runs and not care if we are 12th in hits. It won a pennant two years ago.

3:08 PM Dec 1st
 
MarisFan61
Dave: I think indeed I am doing you exactly that, if you could think about it. When someone tends to overplay his hand in such extreme and evident ways and to overlook certain key kinds of things (I mean, don't just look at my comments, look at others here), I think it's safe to say he benefits from having his attention drawn to it and to the kinds of things he has tended to overlook -- about which some of us have been specific, both here and on the Pineda article. That's what's called constructive criticism.
3:05 PM Dec 1st
 
DaveFleming
On Studes' point about WPA (Win Probability Added)...I think that metric generally favors middle-of-the-order hitters over lead-off batters, especially in the NL. Fowler, hitting leadoff or just after the pitcher's spot, is likely to get fewer high-leverage at-bats than, say, David Ortiz. Fowler had a Leverage Index of 0.94, which is among the lowest in the top-50 players.

And (I'm asking this as a question)....is WPA generally consistent among players, or does it vacillate? Can we expect Cespedes to continue to produce in high leverage situations, just because he did last year?

***

No one's under siege here, MarisFan...we're all just having a conversation. It's an interesting conversation, and we're all enjoying it.

If you want to chime in about the article you think I should have written, why don't you stop harassing me about it, and just write it yourself. And if you think I'm 'embarrassing the field' somehow, well: there are plenty of other fields for you to go play on. But don't come and sh*t on myfield, and pretend like you're doing me a service.
2:44 PM Dec 1st
 
steve161
Well of course the Mets win more when Cespedes plays. Their offense depends on the long ball. Take away the long ball and they lose.

Now clearly the Mets do not, at the moment, appear to be a well-assembled team. They lack a center fielder and outfield defense in general. They lack table setters. The infield is in flux and catcher is an open question. The young starting staff broke down. Despite all of this they contended. Who would have thought that Terry Collins is a genius?

I wonder if the issue is not who they have but how they use them. As a franchise it feels like they have always preferred offense to defense, going all the way back to Howard Johnson and the like. They play d'Arnaud over Plawecki. Their infield has featured the likes of Murphy, Flores and Duda. They sit Juan Lagares and put Michael Conforto in center. They play Granderson in right and Cespedes in left, possibly the most mystifying decision of all. (I'd play Cespedes-Lagares-Bruce with Granderson as fourth OF. I'd trade Conforto while he has value based on his potential. What I'd really like to do is play Conforto-Lagares-Cespedes and trade Bruce, but for whatever reason Cespedes appears not to want to play RF.)

A wild card in this calculation may be TJ Rivera. He doesn't appear to be the second coming of Brooks Robinson or Bill Mazeroski, either, but he can hit, to the tune of .333/.345/.476/.821 in 133 PA in 2016. He and Reyes might well be a nice top of the lineup.

Even setting aside the very real possibility that Dexter Fowler is being overrated on the strength of his 2016, I don't see him as filling a need that can't be filled internally, at least on a short term basis (and of course you weren't going to get Fowler for only the short term). But setting the table isn't enough if you don't also clear it. They had to keep Cespedes.
7:37 AM Dec 1st
 
studes
I track something a little different. I think that offense can be divided into three things: get men in scoring position, hit with men in scoring position, and hit home runs. The Mets were just about last in the league in both of the first two elements. They have to do better in at least one of those two if they want to have a decent offense next year.

Hitting with RISP may be kind of random (they improved in the last month of the season) but getting into scoring position seems less so. So I agree with Dave's basic point and agree Fowler would fit well in the Mets lineup.

But not at the expense of losing Cespedes. Cespedes hit for a decent average, walked, hit home runs, hit with RISP, played fine left field defense. Plus he actually wants to play in NY, seems to thrive on the big stage and is worth millions in star power.

If you want a stat to underline the point, try WPA. Cespedes was eighth in the league, Fowler was 29th.
6:51 AM Dec 1st
 
OldBackstop
@BBF I don't think Dave or anyone had been focusing on money, or defense for that matter, it is more the academic argument of fitting into what the offense needs.

The defensive shortfallings of Fowler in a head-to-head surprise me....I just think their reputations would be that Cespedes was a big negative in center compared to Fowler.

Fowler (who by the way had never played a corner outfield position in a game, something to consider) has a career of 36 assists and 35 errors in center, with a fielding percentage in center of .984 and a dWAR since 2012 of -3.5.

Cespedes has 57 assists and 25 errors, with 12 assists and four errors in center with a fielding percentage of .990 and a dWAR since 2012 of -0.9.

I think that an economic argument might come out Fowler's way, but we won't know that until he signs. If, for instance, we lose Harvey in 2019 because we are paying Cespedes X million more, than....the point is made.

As a Mets fan, though, I can't get past that the solution to our 25th ranked offense is losing the guy who has had 48 home runs and knocked in 130 ribbies over the past 189 games at a slash if .282/.348/.554/.903

To frame what would be a more pragmatic discussion, what if the finances presented the choice of either signing Yo-C or signing Fowler AND upgrading in the middle infield to more of a rabbit run producer.




5:44 AM Dec 1st
 
BarryBondsFan25
Marisfan- you have yet to produce one statistic that suggests Cespedes at $27 million AAV was the better route for the mets to take rather than signing fowler for a lot less that fills a mets need.

This reminds me of your Derek jeter posts on the message board this past summer- no numbers to support your opinion.

Just to review, the 2017 projections for fowler and Cespedes are very similar. Secondly, fowlers three year average offensive winning percentage is greater than cespedes'. Third, the mets have proven they can't score runs with a lineup that just hits home runs. Fourth, they are much weaker defensively in the outfield with granderson in centre. Last, the mets w/l with Cespedes in the lineup is a mirage and secondly not a fair assessment because the mets do not have an adequate replacement for Cespedes during the year.

Feel free to provide some statistics that prove your point otherwise.
4:02 AM Dec 1st
 
MarisFan61
Dave, I think you're under a little bit of siege.... :-)

But seriously, it does look like you ought to back off from how you expressed it in the article. What you posted is interesting, but not close to dispositive and I think hardly even indicative about what the Mets did or should have done; too much was omitted from your consideration. I think it would have been fine if you had expressed it simply as factoids of possible interest, to be thought about in our assessment of what the Mets did -- but in view of all that has been noted here, it's hard to see that it's more than that.

To that extent, it reminds me very much of your article of last year about Michael Pineda, how a particular metric seemed to show him as the best pitcher in baseball (or some such). Just like in this current thing, it was an eye-popping thing, perhaps unexpected, but not nearly as strongly indicative as you stated, because of key things that were ignored.

Just a little more soapbox: As I've said about various sabermetric work (but tellingly, never about Bill's work, and not at all because this site belongs to him but because I've never seen anything of his that I felt came close to such a thing), I think such pieces embarrass the field, in varying degrees. They lend support to the remaining skepticism and resistance to the field, because of how obvious is the focus on narrow aspects, with neglect of relevant broader aspects. It's truly an Emperor's New Clothes thing, in which all one has to do is.....look. There's no problem when such things are presented more modestly -- like, "Here's an interesting tidbit which maybe tells us a little something about this." The problem comes from the overreaching. Perhaps the basic problem is that many in the field have become too comfortable approaching issues narrowly, and there haven't been enough of us hammering it home to them that this isn't good, or that we haven't done a good enough job of that.
3:07 AM Dec 1st
 
OldBackstop
I just realized that Fowler has a negative career dWAR. I guess I am missing his charm :-) Up to this year, playing at Coors and Ebbetts, he was a low teen/low teen (not 20/20) guy who slashed .267/.363/.418/.781 with five consecutive years of negative dWAR.

Grandy had 16 HRs leading off innings this year, with a .947 OPS.

In fact, leading off innings in 2016:

Fowler slashed .272 .373 .453 .825
Grandy slashed .270 .355 .582 .937.
Reyes slashed .276 .366 .551 .917

I didn't really see the high buzz on Heyward last, but he had the glove and he was 26. Fowler look likes somebody who is going to be a marginal center fielder before the end of this contract without the pop you would want from him sliding to a corner.


1:12 AM Dec 1st
 
MarisFan61
Dave: In view of what you say in that last post, does that mean your mind is changed about the basic thesis of the article, i.e. that the Mets would have been better off taking Fowler?
(to the extent that it was a 'binary' choice as OldBackstop put it -- which it wasn't, but that's how you were approaching it)

I would think that your realization of how much better the Mets seem to have played 'with' Cespedes has to have changed your basic take.
11:34 PM Nov 30th
 
DaveFleming
Doncoffin: my table didn't include the 2016 Blue Jays...it included the b]2010[/b] Blue Jays.

I think the Mets W-L record with Cespedes is an important point: they really do play better when he's around. And while I realize that I made this into an either/or thing between Fowler and Cespedes, my bigger point is that the Mets still haven't solved a big problem with their offense: they have a lot of finishers, and no real offense starters.

I'm rooting for Reyes to bounce back, and it seems like the Mets are going to try him at the top of the order. I just wish that they had a second option for a table setter, on the (good) chance that age or injuries catch up to him.
10:27 PM Nov 30th
 
OldBackstop
Hey Maris...looking at Cespedes missed games, there were only maybe two in 2015? We were riding him like a horse. In 2016 it looks like the standout date in the period was the two weeks he missed in August, where the Mets went 1-6 for a stretch. That was a bad period...Cabrera went down for almost the same period...I forget what was going on with the rotation, but Logan Verrett got blown out twice early in August. When Cabrera and Cespedes came back so did the Metsies winning ways.

I do think Cespedes is the straw that stirs the drink and he was the best signing we could make....I wouldn't have looked at this but.....omigod you wouldn't stop!:-)
10:22 PM Nov 30th
 
doncoffin
RER might be useful in understanding differences between players, but I'm not seeing anything (yet) that tells me it matters at the level of the team.

Where I'm having trouble with this discussion is that the evidence I keep seeing suggests that, at the team level, a higher RER is associated with fewer runs scored. Take Dave's 3 team example of teams with high Guillen numbers:

2016 Blue Jays - 53.11%
2016 Orioles - 51.88%
2016 Mets - 51.12%

Runs per game?
Jays--4.79
Orioles--4.51
Mets--4.20
MLB average: 4.48
AL average--4.50--or, essentially the same.

RER?
Jays--0.707
Orioles--0.470
Mets--0.656
Major league average--0.660

So the Jays have a high RER and scored more runs than the average.
The Orioles have a very low RER and scored an average number of runs.
The Mets have an average RER and scored a below-average number of runs.

Overall, the correlation (as I have pointed out before) is negative--a higher RER is, on average, associated with scoring fewer runs.

And I close by repeating myself: RER might be useful in understanding differences between players, but I'm not seeing anything (yet) that tells me it matters at the level of the team.
10:15 PM Nov 30th
 
OldBackstop
Dave...I think the issue is, if taking a look from team needs, this wasn't a binary decision between Cespedes and Fowler, it was a trinary assessment between potential leadoff guys Reyes, Fowler, Granderson.

In wrestling with my worksheet skills like Professor Moriarty on a cliff, it looks to me like Granderson is well below the other two in RER, but Reyes was usually over 1.00. The three are pretty consistent in their annual ranges, which I guess is a test of the validity of a metric.

It appears the Mets will try Reyes at leadoff, and it looks like Reyes best RER years, in his first stint with the Mets, were higher than any Fowler has produced. To cherry pick a number somewhat in the conversation, in their careers Reyes has batted 19 points higher than Fowler when leading of an inning, .291 to .272. And last year....well, I don't think it is fair to grab that as a comparison. Reyes' year reads like one of those stress measures of major life events -- domestic charges, reviled and dumped by Colorado, moving houses, changing jobs, marital issues, health problems, time in the minors, facing the NY media about the abuse, learning a new position in a pennant race....and...uh....probably both his parents and dog died, I'm just guessing.

So...if the Mets want to think a full year of Reyes will spark their RER number, then the numbers that are stark are the salaries....Reyes is making near the minimum, I believe. They could also look at that team list and hope that some of the guys with higher numbers -- D'Arnaud jumps out --- will get more at bats in 2017.

Great article and a cool metric.


9:34 PM Nov 30th
 
doncoffin
I looked at the team data for 2016, which is obviously noisier than MLB data. The correlation between RER and runs per game is, again, negative--a higher team RER score is associated with lower scoring (the correlation coefficient is -.21, which is not quite significant.

I'm beginning to think that we might want to argue, at best, that very high or very low team RERs are inefficient when it comes to scoring runs. Which would suggest a parabola-shaped relationship. But that does not fit the 2016 data at all. I'm going to need a lot more data to check that.
8:18 PM Nov 30th
 
evanecurb
This article and ensuing discussion has been very enlightening. Thanks guys.
8:15 PM Nov 30th
 
MarisFan61
Cliff's Notes: The team had a worse record in the 41 games without Cespedes than they went in any of the one hundred and forty nine 41-game sequences with him in the game.

That's a helluva random stat.
8:03 PM Nov 30th
 
MarisFan61
Here's another one for you guys to chew on. :-)

I mentioned the difference between the Mets' record "with" Cespedes and "without" him in this last year and a half.

I recognized that you could say (and of course it would be true) that a team with that overall record could easily go just 18-23 by random, regardless of who is or isn't in the given games. I also said I didn't think this difference was just chance.

I did a small extra exercise.
This is a test, folks. :-)

If you say this doesn't mean anything, I'll then want to know if it gives you any tension in your gut to say so. :-)

NUMBER OF 41-GAME SEQUENCES DURING THIS YEAR AND A HALF WITH CESPEDES IN THE GAME
(counting each and every unique 41-game sequence, i.e. each next game with Cespedes in it creates a new set, and taking the 2 seasons as though they're a single unit, to give us a fuller sample -- i.e. it counts as a sequence even if it overlaps the 2 years):
149 (one hundred and forty nine)

NUMBER OF THOSE SEQUENCES IN WHICH THE TEAM WENT 18-23 OR WORSE
zero
8:00 PM Nov 30th
 
DaveFleming
The Guillen Number! I was actually looking for that exact stat (% of team runs from homer), and I couldn't find it. Thanks for pointing it out, BBFan.

As BBFan25 noted, the Orioles and Mets ranked 1st and 2nd in the percentage of their runs that came on homers, both crossing the 50% threshold. That's actually surprisingly rare: over the last twenty-five years, only three teams have gotten MOST of their runs from homers:

2010 Blue Jays - 53.11%
2016 Orioles - 51.88%
2016 Mets - 51.12%

It's worth mentioning that despite hitting tons of homers, none of these teams were particularly good at generating runs scored: the Mets ranked 12th in runs scored in the NL, while the Orioles ranked 7th in the AL this year. The 2010 Blue Jays, despite 255 team homers, ranked just 6th in the AL that year.

We have at least SOME evidence that teams that are too reliant on the long ball have trouble building efficient offenses.
7:20 PM Nov 30th
 
doncoffin
Five basic charts for thinking about Run Element Ratio, covering MLB from 1901 to 2016. This is for all MLB players combined, using annual data.

isolatedpower.blogspot.com/2016/11/run-element-ratio-charts.html#!/2016/11/run-element-ratio-charts.htm​l
7:07 PM Nov 30th
 
BarryBondsFan25
To build on dave's point regarding the mets hitting the long ball, baseball prospectus tracks this sort of thing with a stat called the Guillen Number. It is the percentage of runs a team scores via the long ball. The mets were 2nd in 2016 at 51.12%. The mets scored 671 runs. The Orioles who were first at 51.18%, scored 744 runs.

The Braves who were dead last in guillen number at 28.51% scored 649 runs. Finishing just ahead of them was the Giants at 28.53%. They scored 715 runs.

If there was ever a perfect fit for Cespedes, it was San Fran. However, would the Giants really pay a player like Cespedes more than buster Posey?
6:20 PM Nov 30th
 
DaveFleming
Another way to understand the Mets offense last year....

The Mets ranked 2nd in the NL in home runs:

1. STL - 225
2. NYM - 218
3. COL - 204
4. WSN - 199
5. CHC - 193

You would expect that teams who hit a lot of home runs also scored a lot of runs. This was generally true. Here are the top teams in the NL by runs scored:

1. COL - 845
2. CHC - 808
3. STL - 779
4. WSN - 763

For the Rockies, Cubs, Cardinals, and Nationals, there was a clear correlation between hitting homers and scoring runs. The team that's missing is the Mets:

10. SDP - 686
11. NYM - 671
11. MIL - 671
13. MIA - 655
14. ATL - 649
15. PHI - 610

Most teams who hit a lot of home runs also score a lot of runs. The Mets didn't: homers were ALL of their offense last year. I don't think an offense that can only score on homers can be effective.
5:18 PM Nov 30th
 
MarisFan61
(P.S. That wasn't a typo, putting Walker and Reyes in that order. I assume most people would do the opposite -- not necessarily here, but out there.)
5:04 PM Nov 30th
 
MarisFan61
I don't think the Mets' batting order is hurting nearly enough for them to have been desperate to sign a slightly-aged oft-injured leadoff guy, especially if it would have been seen as a "him-or-Cespedes" thing, which, I have to tell you, seems like an absurdity from a hometown standpoint. (You don't have to tell me that hometown views aren't necessarily accurate.) Re-signing Cespedes seemed not only important and worthwhile, but crucial. Signing Fowler? Real nice, but uncertain and possibly fraught. BTW notice that the Cubs aren't rushing to open their pocketbooks for him.

BTW this is how it looks to me like the Met batting order could work, and I don't see it terribly hurting, although yes, if you put a Rickey Henderson or even a Johnny Damon at the top, it looks better. BTW I'm not saying this is what they should do. I assume that whatever Terry Collins actually does will be better, not just because he knows better but because he knows things we can't possibly know. But on paper, I'd be very, very comfortable with this (assuming they don't add any other position players, which they might):

1. Neil Walker
2. Jose Reyes
3. Asdrubal Cabrera
4. Cespedes
5. Granderson
6. Duda or Flores
7. Conforto
8. D'Arnaud​
4:59 PM Nov 30th
 
BarryBondsFan25
The Mets would be losing a big gun yes but they would also be adding a gun in return that strenghtens their defense at the same time.

3 year average offensive win percentage for Cespedes and Fowler:

Cespedes .607

Fowler .630


No matter how I look at it, to me fowler was the better fit. Now if the mets acquire a centre fielder and deal one of their corner outfielders, the Cespedes signing makes sense but what are the chances of that? Granderson gets them a return of almost nothing. Bruce gets them perhaps a bullpen arm. Conforto would probably yield the mets the greatest return but do they really want to give up six years of a good controllable bat?

I hear mccutcheon is on the market. Do the mets have what the pirates would be looking for? If the pirates traded mccutcheon, I would assume their outfield would be Marte in centre, polanco in right and either conforto or Austin meadows in left one of which would be blocked.
4:38 PM Nov 30th
 
lidsky
Good article. Thanks.


Isn't the use of a ratio stat like RER a problem when comparing two players? RER says for a given production that a batter is better early or later in an inning. To say that somebody with a higher RER is better early in the inning compared to another batter with a lower RER without considering anything else isn't accurate.

In other words, Hitter A can be be better than Hitter B in both early and late inning skills and have a lower RER compared to Hitter B.
4:29 PM Nov 30th
 
OldBackstop
Great article!

This RER stat is very cool, and exactly spot on for the Mets, who had a table setter problem. They had the second highest percentage of their runs generated from home runs last year, and it would have been nice to see Fowler trotting in ahead of the sluggers once in a while.

If I were the Mets looking at Fowler, though, I might have thought “yeah, nice Age 30 contract year with the World Champs, congrats on your first All Star Game. But that 4.4 WAR was your first year above 2.8, and maybe we have a 27-year-old guy here under contract through 2019 who looks a lot like you in Juan Lagares.”

In 2014 Lagares won a Gold Glove in centerfield, and batted .281 with a 5.5 WAR.... in his rookie year in 2013 Lagares had a 3.7 WAR. He was making circus catches out in center and had a gun, but he broke his thumb at the circus and banged up his elbow. If you look at the last 4 years, Lagares averaged 5.1 WAR per 650 PAs to Fowler’s 3.5 at Coors and Wrigley. And Lagares is batting .666 in the Dominican Winter League right now…well, he was after Opening Day last week :-).

Yo-C….well, Yo-C was just an imperative for fan morale and season tix…it is hard to lose the biggest gun from a team that was 25th in runs scored. Right now, we are standing flat-footed at 25th in runs scored, back to status quo, and they are shopping Bruce and Grandy. So maybe they will come up with a better answer on the leather and speed than we are showing right now.

4:12 PM Nov 30th
 
doncoffin
Well, stimulated by Dave's addition to the data dump...

I calculated RERs back to 1901, using data in Baseball Reference. Over that period, the RER has generally declined, largely (I suspect) because hitting for power has largely increased--so the denominator in the RER formula (TB-H) has gotten larger. Of more importance, the correlation between RS per game and RER is -0.586--a lower RER is associated with scoring MORE, not fewer, runs, over the 1901-2016 period. When I get the chance, I will post these results with the charts in case anyone is interested, and post a link here (probably later tonight).
3:41 PM Nov 30th
 
DaveFleming
Just adding to the Don's data drop....

The RER across all of baseball last year was .657, the lowest mark in history. The ten lowest RER years:

2016 - .657
2015 - .667
2006 - .684
2005 - .692
2004 - .695
2001 - .699
2003 - .702
2012 - .720
1930 - .727
2013 - .730

The seasons with the highest RER totals, perhaps predictably, were the years between 1910 and 1919, and then the WWII seasons. It's likely that seasons earlier than 1910 had even higher RER's, but we don't have accurate walk data for them.
2:35 PM Nov 30th
 
doncoffin
Incidentally, and because Jonathan Schoop had one of the lowest RER...and he's a second baseman. In 647 PA, he had 1 SB and 21 walks--and 25 home runs. That seems unique for a second baseman.
2:26 PM Nov 30th
 
doncoffin
Because I'm who I am, I was fascinated by this, and so I looked at RER for 2016. Interesting results. I restricted myself to players with 300+ PA. There were 279 of them. The player with the highest RER in 2016 was...Francisco Cervelli. 3.263. Second was Travis Jankowski. (3.130). Then Billy Hamilton (2.765), Dee Gordon (2.175). 45 players (about 1/6 of the total) had RER's in excess of 1. Let's do a table (>.9 means greater than .9 and less than 1; >.8 is greater than .8 and less than .9, and so on).

>1.......45
>.9......17
>.8......26
>.7......27
>.6......44
>.5......34
>.4......42
>.3......32
>0.......15

The mean is 0.707 (roughly, Xander Bogaerts, with the 112th highest RER).
The median is 0.640 (exactly Max Kepler 140th))
The standard deviation is 0.389.

This is not a normal distribution.

At the level of the individual player, there is no correlation between RER and Runs scored (-0.087); the correlation between RER and OPS+ is (probably unsurprisingly) -0.263, significant, but not large.

Bryce Harper, Paul Goldschmidt, and Mike Trout are probably the three best players with RER > 1 (1.290,1.279. 1.132, respectively).

Among the players with the lowest RER (the bottom 45), the three best players are probably Manny Machado, Robinson Cano, and Carlos Beltran (.314, .305, .253).

In general, players with higher RER appear to be better players, probably because they have a broader range of skills.

Spreadsheet on request (or tell me how to upload it, because every time I try to do that I fail).
2:12 PM Nov 30th
 
BarryBondsFan25
Fowler is connected because Fowler fills a need that the Mets have- CF. It is a very obvious need. He also fills the need the Mets have in terms of a leadoff hitter. That is the pro-Fowler camp's point.

Define breaking the bank? Fowler would not have cost the Mets what it cost them to sign Cespedes. The Mets have limited resources to allocate for players. Had the Mets gone the Fowler route, they would have more money left over to get bullpen/catcher help. They still may do that but obviously they have less resources to do so.

Cespedes' 2017 forecast wOBA is .338. Fowler's is .334. Cespedes is forecast WAR is 2.9. Fowler's is 2.1.

Of course, this is all assuming Fowler even wants to play in NY which is not a given. However, given Fowler's difficulty in signing last year, I do not see why he would not want to sign with the Mets especially since he would be certain to play CF.
1:13 PM Nov 30th
 
MarisFan61
I don't see that there's "a Fowler/Cespedes problem" here. I don't think there's any connection between what the Mets should have done or should do on these two players. They're different and separate questions.

I don't mean that I think it was an option for the Mets to sign both guys. Since it's not that long since they were pleading poverty, I'll assume it wasn't. But, Cespedes fills these bills, or at least is seen as filling them, and I think he really does:
-- Power hitter
-- Clutch (probably better I shouldn't have said that because it risks getting into how stupid it is to believe in such a thing, but....FWIW he's definitely seen that way by most in the Met world)
-- Makes the whole lineup work (ditto)
-- Can handle New York (perhaps triple ditto), and loves it
-- Immensely popular; puts fannies in the seats and keeps the town buzzing about the Mets.

It's not known that Fowler is any of those things, and again FWIW, it is felt in the Met world that Cespedes encompasses an unusually high mix of those things, in a way that very few other players would.

Yes, the Mets definitely are hurting in putting men on base, they especially could use a real leadoff guy, and Fowler would seem to be a good candidate for that. But..... (let's now look at the separate "Fowler problem"): Is he worth breaking any bank for it?

It looks very debatable to me. If we knew he was going to have at least a couple of years like 2016, he might be -- but we can hardly count on it, and it looks to me like the percentages are against it:
-- 2016 was his best year; his average year has been less good.
-- He's going to be 31 next spring.
-- He has missed significant playing time in most years. On paper, that's likely to get worse in the next few years, not better.

So tell me, folks: How is Fowler connected to Cespedes in all this?
12:16 PM Nov 30th
 
BarryBondsFan25
Thanks for the article Dave. I'll put Run Element Ratio in my stats tool box.

We'll see what happens with the Mets. I assume they will now try to deal Bruce. I'm not sure there is that much of a demand for Jay Bruce. Conforto can always be sent down to assure he gets enough ABs to continue to develop.

One other thing is that we are looking at the Mets only from an offensive standpoint. The Cespedes signing also makes the Mets weaker defensively. Granderson cannot play CF.

We'll see where Fowler ends up and at what price.
11:49 AM Nov 30th
 
bjames
Enjoyed the article and, more to the point, came out of it with a better understanding of the Mets and of the Fowler/Cespedes problem. Thanks.
10:12 AM Nov 30th
 
nettles9
I still think the Mets need to get more men on base.
10:01 AM Nov 30th
 
evanecurb
Come to think of it, maybe the Orioles could use a guy like Nick Markakis. Hmmmm....
9:54 AM Nov 30th
 
evanecurb
Interesting article, Dave. And I think you make a point that is often forgotten. The construction of the elements of an offense matter. It's better to have some starters and some finishers. It's also true that the Mets need a run starter. But the Mets' offense isn't that great to begin with. So while they do indeed need Fowler, they can't afford to lose their best offensive player.

The case for Fowler might make more sense for a team like the Orioles, who have a bunch of home run hitters but only one guy (Machado) with a high on base percentage (he's also a home run hitter, so he's awesome). If the Orioles could swap Mark Trumbo or Chris Davis for Dexter Fowler, that would make a lot of sense. They'd be adding defense, speed, and OBP to a lineup that needs all of those elements, while losing some home runs, which they already have in abundance.

Baseball is indeed a complex system. There are many things we don't know, including:
1. How many teams are interested in Dexter Fowler.
2. How much those teams are willing to pay Dexter Fowler.
3. Whether Dexter Fowler is interested in playing for the Mets.

What we do know is that Cespedes signed with the Mets, and that the Nationals were trying to sign Cespedes. The Mets made the right move.
9:53 AM Nov 30th
 
steve161
Another way of saying what Maris is saying: yes, Fowler would be better than Cespedes at starting trouble. But without Cespedes, who's going to finish it?
7:47 AM Nov 30th
 
MarisFan61
The thrust of this article reminds me a lot of Bill's current article on how breaking things down, in the way that sabermetrics often does, can "dissolve the bonds of the game so completely that the game no longer exists." It's not exactly like that, because Bill wasn't particularly talking about misjudging a given player when we break him down into his elements; he was talking about larger aspects of the game, and of things that occur just by unpredictable chance. But I'm feeling the same principle going on here.

You can say and prove to high heaven that the Mets "already have" enough of the exact skills that Cespedes brings, and you could be right; the numbers here do seem to show that you're right.

But in the overall, you're not right. This thing that the numbers are showing doesn't tell the full picture, and the parts of the picture that they miss are key. Cespedes brings a kind of overall package that none of the other individual players have, not nearly -- and it's a very important kind of package, and by all accounts (except perhaps some sabermetric ones) :-) .....OK, I'll say by many accounts, it's a package that makes the rest of the batting order "work," in a way that it doesn't otherwise.

BTW, granting that this doesn't tell a full story either, chew on this one for a second:
(data gotten from current articles; I didn't fact-check them)

Mets Record With Cespedes (past year and a half)
107-74

Without Cespedes (during same period)
18-23

I know, I know, a 107-74 team (or a 125-97 team, which is the total) can easily go 18-23 just by chance....
Do I think it was just chance? NO.
4:11 AM Nov 30th
 
 
©2019 Be Jolly, Inc. All Rights Reserved.|Web site design and development by Americaneagle.com|Terms & Conditions|Privacy Policy