Evaluating the Mets’ Arbitration Projections

In October, MLB Trade Rumors released their projected arbitration salaries, so I have decided to write about which of the thirteen Mets I would sign or non-tender. Due to the shortened season, it is difficult to determine how arbitration salaries will work and MLB Trade Rumors provided three different projections:

• “Method 1: Applies model directly with actual statistics from this 60-game season
• Method 2: Extrapolates all counting stats to would-be 162-game totals.  One home run becomes 2.7 home runs.
• Method 3: For non-first-time eligibles, finds the raise they’d get in a 162 game season, then gives them 37% of that raise”

The teams will be arguing for method 1 while the players will be arguing for method 2, which creates a lot of uncertainty. Because of the economic effects of the coronavirus on teams, owners are likely to attempt to slash payroll. This should lead to a lot of non-tenders; however, that doesn’t seem to apply to the Mets’ new owner Steve Cohen.

Obvious:    I don’t think these really require an explanation because these players have shown that they are worth much more than all three of the projections.

Michael Conforto: $9.0MM /$13.6MM / $10.1MM                                                                  

Seth Lugo: $2.2MM / $3.1MM / $2.4MM                                                              

  Brandon Nimmo: $3.0MM / $5.2MM / $3.3MM    

Dominic Smith: $1.9MM / $3.6MM / $1.9MM

J.D. Davis: $1.7MM / $2.9MM / $1.7MM    

Easily Yes:     These are players who I think are easily worth all three of their projections, but still need an explanation.

Edwin Diaz: $5.1MM / $6.5MM / $5.6MM:

Diaz angered many Mets fans in 2019 when he posted a 5.59 ERA and a 4.51 FIP in 58 innings. However, It was clear that Diaz was still very talented as he posted a 2.63 SIERA, 2.80 CRA, 2.95 DRA, and a 3.25 xERA in 2019. In 2020, he recorded much better results with a 1.75 ERA and a 2.18 FIP in 25.2 innings. His advanced statistics were still very good as he posted a 2.52 SIERA, 2.19 DRA, 2.56 CRA, and a 2.46 xERA. Diaz is an elite reliever and he is well worth 6.5 million which is the highest projection.    

Amed Rosario: $1.8MM / $2.6MM / $1.8MM: 

Rosario has been awful this year, posting a 76 wRC+ and a 0.1 fWAR in 46 games. He also overperformed his xwOBA which was .246 (4th percentile). Despite how bad Rosario was this year it was a small sample size and the highest estimate puts him at 2.6 million, which is going to make him an easy yes because of his potential and the fact that he only played 46 games this year. 

Miguel Castro: $1.3MM / $1.8MM / $1.3MM:

The Mets traded for Castro at the deadline this year, a move I liked at the time and wrote about here. Castro didn’t impress with the Mets as he had a 4 ERA and a 4.19 FIP in 9 innings. Most concerning is his 17.8 BB% because control is his biggest weakness and the key to his success. Nine mediocre innings upon joining a new team doesn’t change my mind, and he still has the same potential that I previously alluded to. Even for the Mets, it would be foolish for them to trade for Castro and then non-tender him, so he is an easy yes.

Noah Syndergaard: $9.7MM / $9.7MM / $9.7MM: Under normal circumstances, this would be one of the easiest decisions the Mets would make this offseason, however, Syndergaard’s injury complicates this. Syndergaard is a great pitcher and is worth much more than 9.7 million if he were healthy. Even if he is only going to pitch half the year it is an easy decision and the Mets get the ability to give him a qualifying offer in free agency. The only situation where the Mets non-tender Syndergaard is if they doubt that he will pitch at all in 2020, a situation that seems highly unlikely.

Debatable: These are players whom I would consider non-tendering. I will give much longer explanations for these players since they are not obvious one way or the other.

Steven Matz: $5.0MM / $5.3MM / $5.1MM:

Matz had an awful 2020 season posting a 9.68 ERA and a 7.76 FIP in 30.2 innings giving him a -0.7 fWAR. Matz’s 7.81 DRA and 6.57 xERA further prove that 2020 was an awful season. On the positive side, his predictive metrics look a lot better. In 2020, Matz posted a 4.05 SIERA and a 4.13 pCRA, which is very encouraging. Even more encouraging is the fact that he increased his K% to 25.4, a career-high. Additionally, his walk rate decreased to 7.0%, his best since 2017. Matz was awful whenever batters made contact. He had a 13.5% barrel rate (average is 6.4%), a 91.5 mph average exit velocity against (average is 88.3), a 41.7% line drive rate (average is 25.7%), and a .478 xwOBACON (average is .376). This is very concerning especially since there isn’t much to suggest that his career-high K-BB% of 18.3% is a result of some improvement on behalf of Matz. His whiff rate, chase rate, and zone rate didn’t significantly change, and his edge rate went down. The only change that likely contributed to his better K-BB% is an increase in first-pitch strikes. It is also likely that that may have contributed to his troubles on batted balls because when the batter put the first pitch in play they averaged a 99.4 mph exit velocity and a .677 xwOBA. For Matz’s career, he has hovered around an average pitcher, with a 109 career ERA-, 108 career FIP-, and an 88 career DRA-. Even a below-average starter is worth 5 million dollars in free agency during a normal offseason and there is a chance that Matz can be better than that. Obviously, this year’s market is a lot different and I wouldn’t be surprised if Matz is non-tendered. Hopefully, the Mets will spend enough on starting pitchers that Matz is not one of the top 5 however I think the Mets should keep him as a long man in the bullpen and as an injury replacement.

Robert Gsellman: $1.2MM / $1.4MM / $1.3MM:    

Gsellman has a similar career trajectory to Matz as he impressed in his first year, hovered around slightly below average for a while, and then had a disastrous 2020. In 2020, Gsellman started 4 games and came out of the bullpen twice giving him a 9.64 ERA, 7.55 FIP, 8.56 xERA, and a 7.63 DRA in 14 innings. Gsellman’s predictive stats won’t save him either as he recorded a 5.74 pCRA and a 5.99 SIERA. Gsellman has been haunted by injuries recently. A hamstring strain prematurely ended his 2019 season, and a triceps strain delayed his 2020 season. His 2020 season ended with a fractured rib during a relief appearance where he gave up 6 runs. In his final appearance, he had a 5% whiff rate and opponents averaged a 91 mph exit velocity against him. He also walked 2 and his pitch chart looked like this:

If we remove that game he has a 7.84 ERA, 6.19 FIP, and a 5.75 SIERA in 10.1 innings which is less bad but still not anything close to average. Removing that game he has a 7.9% barrel rate which is also bad but is a lot better than his 11.3% barrel rate on the season. His command was awful the whole season, especially on his breaking balls and he lost some movement and velocity on his fastballs. Should the Mets non-tender Gsellman just because of one poor short season? I would say no. Gsellman posted a DRA- below 100 and an xERA under 4 in 2018, and 2019. In 143.2 innings from 2018-2019, Gsellman has a 4.45 ERA, 4.03 FIP, and a 4.09 SIERA. If he returns healthy, he will likely be an average reliever who can go long or make a spot start if needed. His potential is still there despite the 2020 disaster. He still has a high spin curveball, a 90 mph slider, and has shown willingness to use his 4 seam fastball more and his sinker less, a decision that should benefit him. The estimates for his arbitration costs fall somewhere between 1.2 million and 1.4 million dollars which I think is a good deal for a pitcher like Gsellman.

Chasen Shreve: $900K / $1.1MM / $800K 

Shreve had a good year with the Mets in 2020 as he posted a 3.96 ERA, 3.99 FIP, 3.74 DRA, and a 3.46 ERA in 25 innings. His predictive stats were similarly optimistic as he recorded a 3.53 pCRA and a 3.45 SIERA. His control was still subpar which is responsible for his 11.8 BB% however his K% was an impressive 33.3% giving him a 21.6 K-BB%. The high K% is in large part a result of a high 37.8% whiff rate (league average is 24.5%) which puts him in the 96th percentile for that statistic. His best pitch in that department is his split-finger which has a stunning 55.1 whiff rate per baseball savant. To emphasize how impressive that whiff rate is, it ranks 12th out of 963 types of pitches thrown by pitchers with at least 25 plate appearances in 2020. However, Baseball Savant gives the pitch a positive run value which means the pitch is worse than the average pitch. This is because the pitch is only in the strike zone 23.3% of the time. While Shreve’s control on the pitch may have been awful his command was a little better. Check out this heat map of his split fingers from baseball savant.

As you can see Shreve keeps his split-finger low and away to a righty and while his command is erratic it almost never leaks into the middle of the zone where he could get badly hurt. While he might miss out on some called strikes this means that when hitters swing at his split-finger they miss a lot as evidenced by his whiff rate. I think that the correct metric to judge Shreve’s split-finger on is CSW, which measures the called strikes and whiffs per total pitches. You can read more about it here. According to Alex Chamberlain’s pitch leaderboards, Shreve’s split-finger has a 32% CSW indicating that the pitch isn’t as good as his whiff rate makes it seem but is still a good pitch. His split-finger isn’t his only pitch that can be evaluated by CSW.

Shreve’s slider had a CSW of 30.6% in 2017 and 30.3% in 2018 but his CSW increased 35.3% in 2020. After a quick Google search, I found an article from March of 2019 that says that Shreve adopted a new slider grip from Adam Wainwright. Baseball savant doesn’t classify any of Wainwright’s pitches as sliders however his curveball has gained horizontal movement over time and has become more sliderish. The theory that Shreve picked up Wainwright’s curveball grip seems to be backed up by the fact that his vertical movement on the pitch increased a lot. 

Shreve’s slider never had a lot of vertical movement until the 2020 season when his slider gained 6 inches of drop. The spin axis on Shreve’s slider also changed a lot from 2018 to 2020. Spin axis measures where the ball spins from, to show the movement on the pitch. A pure topspin curveball will have a spin axis of 360, while a four-seam fastball with backspin will have a spin axis around 180. A lefty slider will have a spin axis around 270, while a changeup that moves in the opposite direction might have a spin axis around 90. Shreve’s slider had a spin axis below 270 for his whole MLB career until 2020 when the spin axis was 324.9 which seems more like a curveball than a slider. Shreve only threw his slider 8% of the time in 2020 and it has never been a main pitch for him however, I think he can increase the usage in 2021 with his new slider. Without any free agent signings, Shreve and Matz are the only lefties who the Mets can turn to out of the bullpen in 2021 and Shreve provides release point variance that the Mets need. I think that the Mets should bring back Shreve for the 2021 season for 1.1 million or less.

Guillermo Heredia: $1.4MM / $1.5MM /$1.3MM

It is hard to judge Heredia’s 2020 season since he only had 36 plate appearances and hit for an 83 wRC+ which is in line with his career 84 wRC+. Although small sample defensive metrics don’t mean a lot Heredia played great defense this year according to UZR, DRS, and OAA although he has always been a good defender. He had a .227 BABIP this year but that doesn’t mean that he got unlucky since he averaged an 85.9 mph exit velocity, had a 50% fly-ball rate, and 16.7% of those fly balls were hit in the infield. Heredia is a fast base runner and a good defender however I doubt he will hit enough to justify paying him 1.3 million or more in such a deflated market.