But things happening after the fact have a way of making any insightful analysis look off. The Mets were forced to fill even more rotation holes after losing Matt Harvey for the season, and they went back to the Jon Niese well. Wheeler won’t see a major-league mound this season. In hindsight, bringing back Colón was a great move, and it’s not just because he’s adequately filled a hole in the rotation for much longer than originally anticipated. His spot is entrenched, and he’s put up his best season as a member of the Mets that demand questions about 2017.

Colin’s surface results have far surpassed anything else he’s done in Mets colors. In his first two seasons with New York, Colón posted a 4.13 ERA in 397 total innings pitched. He didn’t strike out a lot of batters–6.5 per nine for those two seasons–but he also didn’t walk many, just 1.2 per nine innings. Still, every adjusted metric, from Baseball Prospectus’ DRA-, to Baseball Reference’s ERA+, to FanGraphs’ ERA-, suggested that Colón was performing below the league average. He was, indeed, someone who could be expected to be adequate. He’s been more than that. In 2016, Colón’s ERA has fallen almost a full run, to 3.14. While Noah Syndergaard has been at another level, Colón has proven to be just as valuable as Jacob deGrom and Steven Matz.

But Colón is walking a tightrope. The results have been different, but he’s been essentially the same pitcher in 2016 as he was during his first two seasons with the Mets. He’s experienced small changes. Colón has struck out 5.88 batters per nine innings and walked 1.62. Those figures are slightly worse than his first two seasons with the Mets, but they both contribute to a higher FIP. Still, that FIP is just 3.96. Colón’s improved ERA might be explained by his strand rate. In 2016, it’s 78.2 percent compared to the 70.2 and 71.9 percent strand rates he had in 2014 and 2015, respectively. Colón’s been better on the surface, and he’s done far more than fill a gap in the rotation. Still, with the emergence of Robert Gsellman, he’ll probably still end up the third starter if the Mets make it to the Division Series.

The fun part in all of this, of course, is that Colón’s doing this at the age of 43. He’s a better, right-handed, version of the extremely-close-to-washed-up and near the end of his career 35-year-old Jorge De La Rosa. Colón is having one of the 16 best age-43 seasons for a starter in the history of baseball—there have only been 16 recorded, ranging from Cy Young’s in 1910 to Colón’s this season. In five of those 16 seasons, the pitcher in question threw fewer than 100 innings. Colón has thrown 177.2. As noted, his results have been better this season, even though his underlying statistics aren’t trending in the right direction.

It’s strange to think of where numbers are “trending” for a 43-year-old, unless it’s about retirement savings, but that leads to the final question. Colón 2017 would probably look more like the 2014-2015 version than 2016. The underlying numbers suggest his 2016 ERA would be difficult to repeat. The 2014-2015 version wasn’t bad though. If Colón is ready and willing, would it be a good idea for the Mets to bring him back?

Even now, the Mets’ rotation going into 2017 looks like it will be more crowded than it was in 2016. Syndergaard, Matz, and deGrom, and Gsellman should all be locks for the rotation. Harvey should return, and after sitting out all of 2016 recovering from Tommy John surgery, Wheeler really should come back. Colón could find himself seventh on the depth chart. But, as happens, things tend to turn out differently than one expects. What if those bone spurs continue to bark or lead to something worse (yes, I knocked on wood)? What if Harvey returns, but the Mets get the bad version that we saw in 2016? What if Gsellman isn’t quite what he’s seemed to be? If any of those things happen, the Mets might be lucky to have Colón at the ready to throw fastballs and consume innings.

But that’s not to assume that Colón doesn’t come with risk as well,. He is, after all, (relatively) old. His downward aging curve could come all at once. There’s also a romantic reason to be skeptical of a return. Colón has been so much fun to watch. The way he plays catch with himself during dead time says “I’m having fun out here.” He also hit a home run! It would be a little dispiriting to have Colón’s sendoff be a June DFA due to ineffectiveness. Ultimately, what he’s done this year just might be Bartolo Colón’s perfect sunset season.

Photo credit: Mark J. Rebilas-USA TODAY Sports

Even when it benefits a team you want to win, it can be hard to muster the energy to cheer for someone else. Luckily for Mets fans, neither of the teams the Mets need to lose are in the NL East, so there’s no need to go through the indignity of cheering for the Phillies. Heavy intra-division September scheduling means that support will go to NL Central and West teams.

Let’s start with the Giants. Nineteen of the Giants’ final 26 games will come against the Dodgers, Padres, and Rockies. They have a home and away series against each of those teams.The Giants play the Dodgers six times, in Los Angeles September 19-21 and in San Francisco September 30-October 2. The Giants are three games behind the Dodgers in the NL West, so these should be competitive games. It’s also notable that Clayton Kershaw will be making his return to the team soon, so there’s a chance the Giants will have to face the best pitcher alive in two of those six games. Even if the Dodgers wrap up the division by the final series of the year, they might have a reason to pitch Kershaw as he readjusts to a normal workload.

The last-place Padres should give the Giants a breather. They play each other in San Francisco from September 12-14 and in San Diego September 23-25. So: go Dodgers; go Padres.

The Cardinals play just two teams multiple times, the Pirates and the Cubs. They wrap up a three game set against Pittsburgh tomorrow, and they will finish the season at home against the Pirates for a three-game series. These shouldn’t be easy series. One of the hidden caveats of the article you’re reading right now is that the Pirates are just a couple games behind the Mets in the Wild Card race. As long as the Mets stay ahead of the Pirates, and they will as long as the Mets continue winning against teams under .500, then cheering for the Pirates is well and good.

The Cards also play three at home (September 12-14) and three on the road (September 23-25) against the major-league best Chicago Cubs. Unfortunately, the Cubs won’t have much to play for at this point, other than pursuing the best record in the National League. But a mild-effort Cubs team is still pretty good. Go Pirates; go Cubs.

There are two more wrinkles to pay attention to. First, the Cardinals play the Giants in San Francisco for a four game series from September 15 through September 18. The first reaction is to wish that they can both lose, but some quick research reveals that it’s not possible for them to both lose. One way to approach this series is to hope for a split. If they can’t both lose, cheer for neither to win the series. The Wild Card standings might also look different in ten days. By that time, it might be prudent to hope for one or the other team to land a deathblow for the Mets.

One final thing: I didn’t mention the Giants’ two series against the Rockies. They’re currently playing three against them in Denver, and they have a three-game set from September 27-29 in San Francisco. Notably, the Cardinals also play the Rockies. They’ll visit Coors Field from September 19-21. They’re the only team that plays both the Giants and Cardinals from now until the end of the season.

The Rockies aren’t contenders, but they’re also not as bad as you might think. The Rockies are 65-71, but Baseball Prospectus’ third order winning percentage is .518. Not only that, but they are 40-41 with a plus-31 run differential against teams above .500 in 2016. The Rockies play one or the other eight more times, and five of them are at Coors Field. So, go Rockies.

Photo credit: Ed Szczepanski-USA TODAY Sports

In particular, the Mets have been doing some strange* things on offense. As early as May, Adam Rubin observed that the Mets were producing an unusually high percentage of their runs by way of the home run. And a few weeks ago, we looked at the squad’s nigh-impossible batting average with runners in scoring position. Today, let’s look at the chance for Curtis Granderson to yield a season sui generis.

* – Strange is a possible stand-in for “maddening,” “aggravating,” “soul crushing,” and the like.

Granderson has hit 20 home runs this season, but he has driven just 34 batters in. So few RBI with that many home runs just doesn’t happen. The lack of RBI with a lot of home runs is odd because it’s contrary to a perfunctory matter—hit homers, drive in runs. In general, there will be dudes enough on base to push the RBI total higher. According to Baseball Reference’s Play Index, there has only been one season since 1901 where a batter hit at least 20 home runs and drove in 40 or fewer batters.

That season belongs to Chris Hoiles of the Baltimore Orioles. In 371 plate appearances over 96 games in 1992, Hoiles hit exactly 20 homers and drove in 40 runs. That was an incredible feat for a couple of reasons. First, Hoiles hit .274, so he wasn’t exactly a boom-o- bust hitter. And second, 233 of his plate appearances came when he was batting either fifth, sixth, or seventh, which should have given him some opportunity to hit with runners on base. It’s the single season with the fewest RBI for a hitter who had at least 20 home runs. If Granderson manages to only get six or fewer RBI for the remainder of the 2016 season, he’ll join Hoiles as the only other batter to hit 20 home runs and drive in 40 or fewer runs.

If we expand the search to identify players who have hit at least 20 home runs and driven in fewer than 50, we get more players to accompany Granderson. The Play Index search yields 50 results, including Granderson. We can safely remove three of them. Mitch Moreland and Miguel Sano each have 50 RBIs on the season, and each one should post at least one more in the next six weeks. Michael Saunders has 20 home runs and 49 RBI, so he only needs two in order to be excluded from this list.

Half of the 26 players on this list have just 20 home runs—they barely qualify for the parameters set. The other half have anywhere from 21 to 23 dingers, with 10 of those 13 having 21 home runs. Just one player has the 23: Rubén Rivera. In 1999, Rivera hit 23 home runs and drove in 48 in his 475 plate appearances. He hit in every spot of the lineup that season, but most of his plate appearances came in the seventh and eight spots. But Rivera didn’t put the ball in play enough, and he didn’t get enough hits when he did. His totals were so low because of a .195 batting average. Rivera holds the distinction of having the most home runs with fewer than 50 RBI in a season.

Now we have something to root for. Let’s hope for Curtis Granderson to hit four more home runs and get no more than RBI from now until the end of the season. If Granderson can manage to do that, he’ll finish up with 24 dingers and 39 RBI. Sticking in the leadoff spot and keeping pace with his .224 batting average are good starts for this. If this takes place, Granderson will best Chris Hoiles by having more than 20 home runs and one fewer RBI than Hoiles did in 1992. This conclusion to Granderson’s season, if it were to happen, would also mean he’d replace Rubén Rivera as the player with the most home runs while producing fewer than 50 RBI—and he’d do it by driving in fewer than 40.

Without much to cheer for in terms of the postseason, we can at least turn to that fun combination of obscurity and futility. Let’s do it, Grandy!

Photo Credit: Mark J. Rebilas-USA TODAY Sports

It took a long time for this to change. Over the first half of the season, the Mets’ playoff odds plateaued at 44.5 percent, when they were still 2.5 games behind the Nationals. The Nationals were underachieving the lofty expectations placed upon them, but they were still good enough to keep the projection system on which the odds were based faithful. The Mets’ low point throughout the season appeared on July 30, when their odds stood at 15.1 percent. At that point, the Nationals’ odds were an even 92 percent, just a touch lower than they began the season. The Mets were three games out of first place and had the Pirates, Giants, and Cubs ahead of them in the Wild Card race.

Because the race was so tight, however, the Mets were able to close the gap rather quickly. After July 30, the team notched seven consecutive wins, during which the Nationals lost five of seven. A three game deficit in the division turned into a 2.5 game advantage. In playoff odds terms, the Mets climbed from their season low point to a then season high 59 percent chance; the Nationals’ odds dipped below the Mets’ odds for the first time of the season, 57.4 percent. The two teams traded places a couple times over the next few days, but by August 12 the Mets had a playoff odds advantage that would remain until it was 100 percent. The promised graph:

Prior to the 2016 season, the Mets had the playoff odds advantage. They entered as favorites, with an 80.4 percent chance to make the playoffs. The Baseball Prospectus staff, informed but not determined by similar projections, collectively guessed that the Mets would win the National League East. The Nationals started at 72.3 percent playoff odds. The gap to start 2016 wasn’t nearly as large as the one in 2015. It was essentially a toss-up. The two teams flipped and flopped one another in the playoff odds advantage through May, even though the Mets only held a division lead for four days during that time. Mired together at roughly 80 percent playoff odds on May 28, the Nationals then began to separate. As the Nationals have risen, the Mets have fallen. Through Sunday—August 7, the day they pulled ahead in 2015—the Mets were at a season low 26.5 percent playoff odds.

Another difference between the arcs of 2015 and 2016 is that in 2015 the Mets had a better chance at winning the division than earning a Wild Card berth. Ultimately, the two Wild Card winners had seven and eight more wins than the NL East division-winning Mets. It’s tougher in 2016. If the 2015 chart is called joy and the second disappointment, this one might be an admission. It is, indeed, tougher this year:

And yet, the Mets are still very much alive in the Wild Card race. They’re 1.5 games out of first place. The season has felt like a disappointment only because the expectations were higher, for both the casual viewer as well as the consumer of playoff odds. There’s still time to give the season a new name.

Photo credit: Anthony Gruppuso-USA TODAY Sports

Through Saturday’s games, the team as a whole is hitting an inconceivably bad .204 in such situations. In the expansion era (since 1961), that’s tied for the second worst mark, the worst being the 1969 Padres, who hit .201 with RISP. They are tied with the 1968 Mets. Though to be fair to the Padres and 68ers, the league only hit .248 in 1969 and .237 in 1968, The Year of the Pitcher. League-wide batting average in 2015 is .255. In sum, it both feels like the Mets turn into Brendan Ryan with one eye closed with the bases loaded, and they actually have been.

That’s not the end of the story though. It also feels like the Mets aren’t lacking for good hitters having fine seasons. And, indeed, they are not. Judged by the park adjusted and context neutral wRC+, the Mets have the seventh best offense in the National League. It approximates the Nationals, Marlins, and Dodgers offenses.

It also feels like the Mets are suffering from a serious lack of depth on offense and have given far too many plate appearances to unproductive players. But this feeling is only partially true. To get a sense of this, we can turn to a stat that Matt Gross of SB Nation’s Colorado Rockies blog Purple Row developed a few years ago. The stat is called Drag Factor. As Gross has put it, Drag Factor “is designed to measure how much ‘bad offense’ . . . drag[s] down the positive production of the good hitters.” It offers a glimpse into a team’s depth and balance.

Drag Factor is determined with the following equation:

Drag Factor=((100-wRC+)*PA)/Team total non-pitcher PA

The formula uses wRC+ and will punish a player more for being bad and sucking up a lot of plate appearances—just like that player tends to punish the team for which he plays. Players with an above average wRC+ (which is 100) aren’t included because, ostensibly, they are not dragging down the squad’s offense at all. An excellent Drag Factor will be around 5, an average mark is around 10, and a Drag Factor above 15 is considered poor. Here’s how the Mets have done so far this season (stats through Friday):

Mets Drag Factor

The Mets total Drag Factor, which is between good and average, tells us that the worst hitters haven’t been getting enough plate appearances to really bring the offense down. Of the eight hitters who are not accounted for here because they have been above average according to wRC+, José Reyes, Kelly Johnson and James Loney might be reasonably expected to tumble below 100 and begin dragging the offense down. The chart also indicates that players with a wRC+ close to 100 don’t burden the offense.

To provide additional context, let’s look at the Drag Factor held by a couple division rivals, the Nationals and Marlins, as well as the Cubs, who are considered a great team with a lot of depth and few weak spots.

Nationals Drag Factor

Marlins Drag Factor

Cubs Drag Factor

The Cubs, fortifying the feeling that they have excellent depth and few weak spots, come out spit-shined here, besting the Mets by a decent margin. Their chart is also a reminder that Drag Factor is only about offense. Jason Heyward might not be contributing on offense, but he adds value in other ways. The Marlins and Nationals, however, have more offensive hindrances, on the whole, than the Mets. Each team has given more plate appearances to equally or comparably poor hitters than the Mets have.

One way of reading these charts as a mark against the Mets’ quality of depth is that they have simply had to rely on too many different players. While Ty Kelly and Eric Campbell haven’t accrued enough plate appearances to really hurt the team, the Mets have still had to call upon both players at different times. The Mets have 12 position players represented here; the Nationals and Cubs have eight and the Marlins seven. The Mets’ Drag Factor is spread out thin, whereas more than half of the Nationals’ Drag Factor is concentrated in Ben Revere and Ryan Zimmerman. But just by looking at the Mets’ Drag Factor (like the Mets’ overall wRC+ vis-à-vis the rest of the National League), nothing appears amiss with the offense.

This is where context-dependent stats conflict with context-neutral ones. There shouldn’t be anything wrong with the Mets’ offense. Their best hitters have been producing, and even the bad ones haven’t been disastrous. With the added context of runners in scoring position, however, they’ve been historically bad.

The Mets-inclined observer is left with the hope that this will all normalize. On the one hand, the Mets offense has been extremely poor in just the way it feels like they’ve been. On the other hand, they’ve been quietly good in a way that is not immediately apparent. If the context-neutral outcomes continue apace, the Mets should see their offensive fortunes turn soon. It seems impossible that, in light of everything else, that the team can’t manage a collective batting average with runners in scoring position upwards of 50 points higher than it is right now. After all, context-dependent statistics are more useful to describe rather than to predict. But a .204(!) average with runners in scoring position in four months of baseball should also have been impossible.

That’s baseball, I guess.

Photo credit: Adam Hunger-USA TODAY Sports

The Mets need some heroes to lift them to the postseason, which, as the days pass, appears like it will have to come by way of the Wild Card rather than a division title. Those heroes, to use a baseball idiom, are unlikely to come out of left field. To get a sense of what a player can do for a team over the course of roughly a half-season of baseball, I sought out the best second halves in Mets’ history. The sample for a second half (usually a bit more than 81 games) is relatively small, but it’s large enough to demonstrate that the Wilmer Floreses of the world might be single game heroes, but over months of play, the stars stand out.

According to Baseball Reference, the Mets players with the highest sOPS+ for the second half of a single season yield a list of great players in the midst of great seasons:

There aren’t many surprises here. Perhaps Bernard Gilkey’s run of excellence is just as forgettable now as it was when it was actually happening, but the man posted an eight win season in 1996. Cleon Jones doesn’t stand out, but the “1971” next to his name elaborates why. But Jones’s second half that year was just a slightly better continuation of his first half. He finished up the season with a 144 OPS+. Perhaps Lucas Duda’s second half run in 2011 can be seen as surprising because it was the first time he’d gotten that much playing time, but in hindsight it’s not unexpected. Otherwise, this list is a who’s who of great Mets hitters, from Keith Hernandez to David Wright.

The best second half starting pitching performances, this time measured by sOPS+ against, are slightly more varied with regard to stars and others, but it is still dominated by players we’d expect:

Of these, Steve Trachsel’s 2001 is the only second half that was not also part of an excellent full season. He posted a 2.74 ERA over 14 second half starts and 98.2 innings. For the season, however, he produced a 4.46 ERA and an ERA+ of 93. Otherwise, there aren’t surprises on this list either.

The second half of the 2016 season is just a week old, so the split OPS+ figures for hitters range from Neil Walker’s -8 to James Loney’s 158. Bartolo Colón’s 150 represents the highest (worst) among the starters, Jacob deGrom’s 69 is the lowest. Right now, the samples are too small to draw any second half conclusions—the heroes, if there will be any, are yet hidden. If they do emerge, however, they’ll be from the places we expect. The Mets need more than great second halves from players like Yoenis Céspedes, Michael Conforto, and Noah Syndergaard, but they can’t get there without them.

Photo credit: Brad Penner: USA-TODAY Sports.

The outcome has been in the middle. As Mike Vorkunov recently wrote, the Mets’ rotation this season has been better than it was at the same point in 2015. He also notes, however, that the rotation has seemed to teeter on the precipice of disaster. Indeed, one day after that article went up, the Mets announced that Matt Harvey would undergo season-ending surgery. But even that news generated responses that would have seemed out of place before the season: perhaps it was a merciful end to a peculiarly poor season for Harvey. Additionally, Syndergaard and Matz seem to have weekly injury scares on account of the bone spurs they live with, and this all culminates in the threat of too many Logan Verrett starts. As a result, it’s starting to look more and more likely that the Mets could pursue external help to shore up the rotation. One pitcher that might fit the bill for the 2016 Mets is Colorado Rockies left-hander Jorge De La Rosa.

De La Rosa’s season stats thus far don’t recommend him: a 5.47 ERA, a 5.40 FIP, and a handful of adjusted metrics that peg him as a below average pitcher. De La Rosa even had a stretch this season where the Rockies demoted him to relief, although the woeful lack of major-league pitching depth brought him back to the rotation after a short while. However, De La Rosa’s extensive run of reliability and underlying performance indicate that he’s much better than his ERA and FIP suggest.

First, let’s look at the long view. De La Rosa is now 35 years old, but he’s been consistent since 2013, his first year back after undergoing Tommy John surgery in 2011. In 563.7 innings pitched in those seasons, De La Rosa has posted a 4.12 ERA and 4.25 FIP while pitching half of his games in the unfriendly confines of Coors Field. In the same time span, he’s struck out 7.1 batters per nine innings and walked 3.6. These figures represent the best and worst tendencies of De La Rosa.

Over the past two seasons, De La Rosa has maintained strikeout rates similar to his pre-TJS seasons, when he was on the other side of 30, and better than his 2013 and 2014 seasons. In 2015, he maintained a K/9 rate of 8.1, and this season he’s maintained it at 8.3 through 62.7 innings. But while his strikeouts have gone up over the last two seasons, his walks have as well—3.93 and 4.16 per nine in 2015 and 2016. De La Rosa’s greatest weakness is that he’ll have starts where his command completely abandons him, he ends up throwing a lot of pitches early, and he has to exit without recording 15 outs. For instance, in an April 9 start against the Padres, it took De La Rosa 97 pitches to make it through four innings, during which he walked five batters. Here’s the start, in picture form:

Still, comfort can be found in De La Rosa’s velocity and the sustained quality of his secondary pitches, which are the primary reasons why he’s been relatively effective and able to get a lot of strikeouts. De La Rosa’s fastball velocity has declined in recent years, as happens to pitchers, but it has not been a dramatic fall as he’s entered his mid-30s. According to Brooks Baseball, De La Rosa’s fastball velocity has only declined 1.13 mph since his return from Tommy John surgery in 2013. This year, it has sat at 91.25 mph. In addition, he can still dial it up to the mid-90s when he has to.

De La Rosa’s best secondary pitch is his splitter, which is more of a split-change. He’s averaged 82.71 mph with the pitch so far this season, which gives it enough velocity difference to be effective. When he has a feel for it, the pitch generates a lot of whiffs due to its downward vertical movement. In 2016, batters have swung and missed the pitch 22.04 percent of the time, which is the best whiff rate the pitch has had in his post-surgery career. While De La Rosa’s surface stats won’t inspire a lot of confidence, his underlying performance suggest that he could be a valuable role player in this Mets’ rotation.

The Mets’ rotation does not need a savior; it needs insurance and support. De La Rosa could provide both. Significantly, he’s also attainable. De La Rosa is in the final year of a contract set to pay him $12.5 million, so the Mets would only have to cover less than half of that for the remainder of the season. And gone are the days when a team would offer a prospect like Eduardo Rodríguez in exchange for De La Rosa, as the Orioles reportedly did in 2013. The Mets wouldn’t have to give up a top-tier prospect. De La Rosa does have 10-and-five rights, so he would have to approve any trade the Rockies make, but it’s almost certain that he would do so if it meant going to a contender.

De La Rosa’s not someone who should take the ball for the Mets if they do end up competing in the single elimination Wild Card game, which is increasingly looking like the best-case scenario. Instead, he’s someone who can help make sure the Mets get there in the first place.

Photo Credit: Richard Mackson-USA TODAY Sports

Reed’s path to the Mets is a lesson in the fuzziness of assigned bullpen roles as well as the small sample life relief pitchers lead. Reed began his first full season in the White Sox’ bullpen as a late inning reliever, but not the closer. In 10 of his first 11 appearances, he entered the game in either the seventh or the eighth inning. His single ninth inning appearance was in a tie game. Reed logged four holds for then-closer Hector Santiago. Santiago, however, was not terribly effective as the closer. He allowed 12 hits and six runs in 7.3 innings across eight games. He saved four games but blew two. This set the stage for the White Sox to flip Santiago and Reed. Santiago began entering games late and ended the year as a starter, while Reed held on to the closers role for the remainder of 2012 and 2013. In those two seasons, however, he posted numbers that spoke more to Internet memes than ninth inning dominance: 69 saves and a 4.20 ERA. Additionally, his walk and strikeout rates were good but not great, as Reed struck out 23.6 percent of the 533 batters he faced in those two seasons and walked 7.6 percent of them.

Non-dominant numbers notwithstanding, Reed was a closer. And because of that he demanded a lot of the trade market. Indeed, the Diamondbacks traded Matt Davidson—a top 10 prospect in the organization and a player who ranked as the 93rd-best prospect in baseball after the 2013 season, according to Baseball Prospectus—for Reed in December 2013. As the Diamondbacks’ closer in 2014, Reed saved 32 games. He did so on the back of an ERA that fit his career mark, 4.25, and he even upped is strikeout rate a few percentage points while dropping his walk rate. In 2015, however, he got off to a bad start and soon lost the “closer” designation.

In his first 11 games and 10 innings pitched, he gave up 15 hits, posted an ERA of 7.20, and saved as many games as he blew: two. From the periphery, Reed’s strikeout rate dipped below 20 percent, and his walk rate rose about one percentage point. Through August, Reed got season ERA back down to the familiar weed number it was then that the Mets traded for him. Still, the Mets didn’t have to give up a top 10 prospect to get him. It just took two non-prospects named Matt Koch and Miller Diaz. Reed was excellent for the Mets in late inning relief during September, though he struggled in the postseason.

From Reed’s first season in 2012 to the end of 2015, his role and trade value fluctuated much more than his performance. With Familia firmly in place as the Mets’ closer to start 2016, Reed found himself in the same place he was in his breakout 2012 season: a late inning reliever, just not the closer. Except this time, his performance has changed, and it’s changed for the better. In his first 38 innings of 2016—a small but not insignificant sample size—Reed has posted a 2.37 ERA, and his strikeout and walk rates are higher than they’ve ever been, 29.7 and 6.1 percent, respectively. And Reed hasn’t just been an excellent member of the Mets’ bullpen. He’s been one of the better relievers in the NL. He has the fourth best DRA, 3.02, among NL relievers with at least 35 innings pitched. And Reed’s 82 cFIP suggests believability for his 85 DRA-.

At this point, one might wonder whether or not the Mets’ closer is one of the relievers above Reed on these rankings. He is not. In fact, Reed, former closer, has been more effective than Familia, current closer. Familia’s DRA is a run higher at 4.02. His adjusted DRA- is a close to league average 98, and his 91 cFIP is less optimistic looking forward than Reed’s. This isn’t to suggest that Familia is having a poor season, neither is it to say that Reed should be the Mets closer. Indeed, Reed might prove to be even more valuable in pre-ninth inning high leverage situations, as he was on Saturday when he struck out four batters in a strong Cubs lineup before handing the ball to Familia. It does show that the Mets’ acquisition of Reed for practically nothing in 2015 has proven savvy. The Mets wanted a former close to bridge the gap in the late innings. They got that—except that he’s turned out better than he was when he was racking up saves.

A dive into Reed’s Brooks Baseball page doesn’t reveal any dramatic underlying changes. His fastball and slider usage aren’t career anomalies, and his fastball has remained a steady 93 mph, which is where it’s been since 2013. The same applies to the number of whiffs his fastball and slider have generated. His BABIP against is below his career average .302, but a .280 BABIP is far from extreme.

In other words, the results have improved, but the process by which they have done so is not clear. He’s not been overly lucky, but at the same time, he hasn’t exhibited any underlying changes to which his improvements might be ascribed. Once Reed doubles the 38 innings he’s thrown so far this season, the story might change. For now, however, it’s enough to note that Reed has been one of the best relievers in the NL, and perhaps the most important part of a strong Mets’ bullpen.

Photo Credit: Adam Hunger-USA TODAY Sports

These weren’t misguided opinions. In fact, all of the analysis of Cabrera as a player was pretty accurate. He’s a decent hitter who can hold his own at shortstop, but won’t be mistaken for a particularly good shortstop. Additionally, the consensus was that nobody should expect Cabrera to replicate his peak years, when he was a three-win player for Cleveland.

In fact, Cabrera has been an utterly unsurprising player. He’s currently hitting .268/.327/.418. Cabrera’s career batting like is .267/.328/.412. He’s walked in six percent of his plate appearances, which is down from his career 7.4 percent rate, but is almost exactly in line with the 6.5 percent he posted in 2015. The same can be said about his strikeout rate. In 2015, it was 19.4 percent; it’s been 19 percent in 2016, and his career rate is 17.4 percent. According to FRAA, his defense has stayed about the same as it’s been for the past several years: below average, but not dramatically so.

Cabrera has shown different tendencies in his batted ball profile, but the messages there are mixed. First, his hard hit rate, as measured by Baseball Info Solutions, is higher than it’s ever been in his career: 35.5 percent. His career hard hit rate is 29.8 percent, and it was 26.4 percent in 2015. His balls in play measured as softly hit sits at 16.6 percent, which is in line what he did last season but is higher than his career 13.4 percent rate. That means that balls he’s hit classified as medium have gone down.

It looks like a positive re-distribution of batted balls, but Cabrera, on the whole, doesn’t appear to be hitting the ball harder. According to exit velocity data collected by Statcast, Cabrera’s average batted ball velocity is essentially unchanged. In 2015, Cabrera’s average exit velocity was 88.6 mph. It is 88.4 mph in 2016. Additionally, his batted ball type distribution is almost exactly in line with his career averages. Cabrera’s 2016 line drive rate is 22.6 percent, while his career rate is 20.6 percent; his groundball rate in 2016 is 42.8 percent, against a 42.3 percent career rate; and Cabrera has a 34.6 percent fly ball rate in 2016, while he has a career rate of 37.1 percent.

Cabrera is neither defying expectations, nor is he proving the doubters wrong. And yet, he’s provided immense value to the team. That’s because while Cabrera is, essentially, what we thought he’d be, the Mets are not. Injuries have thrown the corner infield spots into disarray, whereas Cabrera (along with Neil Walker) have proven to be steady forces in the middle. It is true that Cabrera is miscast in the batting lineup, though it’s out of necessity. As noted before, Maggie Wiggin suggested that Cabrera would fit best as an eight-hole hitter. Instead, 51 percent of Cabrera’s plate appearances have come from the two-spot. It’s still true that he probably shouldn’t have anything to do with that part of the daily lineup, but it’s also true that the daily lineup has been so discombobulated as to make his appearances there justifiable.

A better situation would be Cabrera and the two-win season he’ll likely produce being overshadowed by the rest of the team. But that’s simply not how things have shaken out. He hasn’t been the shining star that Céspedes has been, but he’s provided what the team has lacked in other areas: the precise fulfillment of expectations. It has turned out to be an excellent off-season signing.

Photo Credit: Jason Getz-USA TODAY Sports

Baseball Prospectus’ PECOTA projection system, like others, focuses on one set of numbers for each player. But there’s much more. The single set is the median of a range of projected outcomes, from a 90th-percentile peak to a 10th-percentile valley. That’s what we’ll focus on here: Where have 2016 Mets’ players landed within their range of projections, and what can we conclude about where they’re going based on those? To simplify things, we’ll use True Average (TAv) for hitters with at least 100 plate appearances and Deserved Run Average (DRA) for the five pitchers who have mostly composed the starting rotation. TAv and DRA serve as all-in-one stats representative of the player in questions season so far. Let’s start with the bats. (All stats are through Saturday’s game.)

The Mets have two position players hanging around their 90th-percentile projections: Yoenis Céspedes and Neil Walker. Céspedes, in particular, is pushing the limits of “90th-percentile.” His .334 TAv is far ahead of his 90th-percentile projection, which was .312. Walker’s performance fits the confines of his pre-season range more neatly, although he’s still hanging around in rarefied air. The pre-season trade acquisition has posted a .303 TAv thus far in 2016. Coming into the season, he’s been surpassing his .296 90th-percentile projection.

One should expect most players to land anywhere from their 30th- to 70th-percentile projections. Players are more likely to be in this bulky middle than they are to be at the top or bottom, based on past performance and trends. This is where most Mets’ hitters land.

Three Mets players have performed roughly within their 70th-percentile projections: Asdrubal Cabrera, Wilmer Flores, and Curtis Granderson. Cabrera’s .272 TAv fits squarely in his 70th-percentile projection, which was also .272. Wilmer Flores, who is being featured in the Mets’ lineup more often than most expected, has also landed in his 70th-percentile projection with a .282 TAv.

Curtis Granderson and Michael Conforto are studies of contrast. The 35-year-old Granderson’s 70th-percentile .287 TAv feels like receiving too much change back from the soda you bought at Target. Just look ahead, don’t say anything, do not ask whether or not you deserve it, and maybe you’ll get to keep it. On the other hand, it feels like Conforto has achieved his 60th-percentile .284 TAv despite himself. All of that feeling can be attributed to his six-week slump.

Kevin Plawecki is the only Mets’ position player with substantial playing time to sit under his median projection. Getting playing time while Travis d’Arnaud is on the mend, Plawecki has posted a .239 TAv, which fits his 4oth-percentile projection from prior to the season.

The elephant in the room here is that names like Lucas Duda and David Wright are absent in this analysis. That’s due to their absence on the field. It also demonstrates the concession of an analysis like this. The players are up against themselves. To wit, James Loney’s .279 TAv in 118 plate appearances is closest to his 90th-percentile projection. It’s not much worse than the .285 Duda posted prior to his injury. The difference is that Duda’s performance fit into his 50th-percentile projection. He was capable of much more. Similarly, that’s why the Mets can have so many players within their 70th- to 90th-percentile projection while simultaneously offering frustration for Mets fans pleading for runs. Mets’ position players are almost uniformly playing above their pre-season median projections, at least at the plate. But that doesn’t mean the picture is rosy, regardless of whether the players start trending back toward their median projections.

The pitchers tell a different story:

Prior to the season, season previews for the Mets frequently asked: “who is the best pitcher on staff? Is it Jacob deGrom or Matt Harvey? Could it possibly be Noah Syndergaard?” It has been Syndergaard, and he’s pitched in such a way as to eliminate that question from 2017 previews. Syndergaard has posted an excellent 2.39 DRA that somehow is not even his 90th-percentile projection. This is simultaneously a compliment to Syndergaard’s season and PECOTA’s ability to recognize that his Valhallian peak.

Syndergaard is the only starter to exceed his pre-season median projection. Matz has faltered a bit lately after a hot start to the season, and his cumulative DRA thus far has only matched his 40th-percentile projection. That deGrom’s 3.50 DRA sits at his 20th-percentile projection is at once disappointing and heartening. As Jarret Seidler recently noted in these pages, deGrom’s “down year” is the exceptional season that proves deGrom is an ace. His current season would be the best on a lot of marginally competitive teams and among the best on competitive ones.

Matt Harvey is to Mets’ pitchers as Yoenis Céspedes is to Mets’ hitters—except the exact opposite. Harvey’s 4.55 DRA is well below his pre-season 10th-percentile projection of 3.49. For a few starts, Harvey looked like he fixed what was ailing him, probably due to one of the suggestions we made about how to fix him. His most recent outing, in which he allowed four runs in six innings to the Braves, will keep the question mark present. Harvey’s pre-season projections suggest that it can’t get worse, but it’s already been worse than PECOTA could imagine. Bartolo Colón’s 4.96 is keeping Harvey company, which could be a sign that the 43-year-old is mortal after all.

Ultimately, the portion of the team that was supposed to carry the team, starting pitching, has been mostly playing below its preseason median projections. The supposed weaker side, the bats, have been playing above theirs. Still, the result is that the starting pitching is playing the largest role in the team’s early success, even if it hasn’t been doing so as effectively as it could.

What each player has done so far this season also informs PECOTA’s rest of season (RoS) projection. New information leads to new projections. Here are the RoS median projections for the position player’s looked at here, alongside their pre-season median:

There is not a lot of change for position players. Despite performances above and below median projections, PECOTA sees most of them staying the same. Their ranges are consistent. Just like before, the story is different for the pitchers:

The RoS median projections for the five pitchers looked at here are worse than their pre-season projections—even Syndergaard. To be honest, I don’t know how to interpret this information. The revised median DRAs for Colón and Harvey appear too harsh, but reasonable given their seasons so far. deGrom’s RoS projection suggests that his center is what he’s done so far this season. That’s not the case for Matz, whose median is worse than he’s performed so far. Syndergaard’s revised 90th-percentile DRA is roughly where he’s been so far in 2016. But where he’s been is his pre-season 80th-percentile DRA.

This view accounts for the most prominent Mets players to log substantial playing time this season. The season could still hinge on how role players and returning players complement those referred to here. Having Travis d’Arnaud and Zack Wheeler return to compete with the rest of the team and against their PECOTA projections should strengthen the whole. And there’s still the question of who will get the most playing time at third base, and where that person is playing right now. Projections are oriented to the future, but these are some of the questions that have to be answered first before we can begin considering what PECOTA can tell us.

Note: At the time of his injury, David Wright’s .304 TAv fit into his 80th-percentile projection. His RoS projection is the not yet invented emoji that simultaneously conveys sadness and acceptance.\

Photo Credit: Adam Hunger-USA TODAY Sports