The Comeback Mets

            Historically, the New York Mets have a penchant for disappointing performances and this season, much of that has come from their bullpen. While led by an elite closer in Jeurys Familia and a set-up man who used to be a closer in Addison Reed, they have found countless ways to blow leads in the late innings. Last night was no different as Reed took a blown save and later Familia put himself in position to take a loss. Yet despite those performances, the Mets came back to win the game. While the Mets have blown countless leads in a fashion similar to last night, they’ve also managed to provide some of the most exciting moments in baseball this season because of them.

            The Mets win last night was their sixth walk-off win of the season, just one shy of their seven last year with a few games still left to play. Remarkably, five of those six walk-off wins have come after what FanGraphs terms a “meltdown”, or MD. A meltdown is a statistic for relievers that credits a pitcher for their impact on their team’s chance to win. A reliever earns a meltdown if their win probability added, WPA, is less than -0.06, or -6%, for the game. In this way, a meltdown matches up closely with a blown save or hold, though it can be applied more widely in other innings or situations outside of the strict definitions of saves and holds. So while Mets relievers have seemed to blow things in big moments all too frequently, the offense has often managed to come back to provide exhilarating wins shortly thereafter. I’ve decided to run through a couple of those specific wins in terms of win probability (WP) below.

May 27^th: Mets 6, Dodgers 5

            After a sparkling start of 7 IP, 1ER by Jacob DeGrom and a combined scoreless 8^th by Jerry Blevins and Reed, the Mets entered the top of the 9^th up 5-1 with a 97% win probability. Over the first five batters of the inning, Familia recorded a strikeout, three singles, and a walk to make the game 5-2 with two outs and the bases loaded. Another strikeout brought the WP to 92% before Chase Utley unleashed a three-run double to tie the game and earn Familia the MD. After another strikeout, the Mets went into the bottom of the 9^th with a win probability of 63%. In the first AB of the bottom of the 9^th, Curtis Granderson ripped an 0-1 pitch over the right field wall for a walk-off homer, a WP swing of 37% in one swing.

September 17^th: Mets 3, Twins 2

            In a closely-contested game throughout, Yoenis Céspedes’ clutch RBI single in the bottom of the 8^th tied the game 2-2 and gave the Mets a win probability of 61%. Starting off the top of the 11^th, with both teams having a WP of 50%, Byron Buxton stepped up to the plate and capped an eight-pitch at-bat with a leadoff home run to left, dropping the Mets WP to 16% and giving Hansel Robles the MD. As we’d seen before, Curtis Granderson led off the bottom of the inning and cranked an 0-1 pitch the other way and over the left-center field wall to tie the game. That hit was worth a WP swing of 44%, and despite continuing to put a runner on third with one out which peaked the Mets WP for the inning at 70%, they couldn’t capitalize. A scoreless 12^th by Josh Edgin got the Mets up to a 63% WP, but two quick outs to start the bottom of the inning brought it back down to 53%. Once again Granderson stepped to the plate and this time worked a full count. On the sixth pitch of the AB, Grandy pulled one down the right field line for his second home run in two innings, creating a WPA of 47%. Grandy was clearly the star of the game, accruing a massive 91% WPA in his last two ABs alone.

September 22^nd: Mets 9, Phillies 8

            In a very back-and-forth game, the Mets entered the 8^th inning leading 4-3 with a win probability of 75%. Trusting in Reed to shut the Phillies down, the Mets and their fans were disappointed as he quickly served up a three-run home run to Maikel Franco, dropping the Mets WP to a lowly 14% and giving Reed the Mets first meltdown of the game. Fast forward to the bottom of the 9^th and a Jay Bruce strikeout brought up Jose Reyes, down 6-4 with one on and one out, translating to a 10% WP. Reyes took Jeanmar Gomez’s offering deep to right for a game-tying home run, a WP jump of 48% in one swing. Moving ahead to the top of the 11^th, the Mets entered the inning with a WP of 50% even. Despite recording 2 outs in the first three batters of the inning, Jeurys Familia served up an RBI single to AJ Ellis and exited the game leaving runners on 1^st and 3^rd after earning the Mets second MD of the game. Blevins came in and promptly hit and walked the first two batters to score another run. Just like that, entering the bottom of the inning the Mets WP was down to just 9%. After a Brandon Nimmo groundout to start the inning brought it even lower to 4%, a walk and a hit set up Asdrubal Cabrera for the biggest hit of the game. He launched an Edubray Ramos pitch way over the right field wall, accounting for a walk-off win, a massive 81% WPA, and the best bat flip of the season.

Watson's Secret Weapon

            When it comes to evaluating Deshaun Watson as a pro prospect, there are plenty of mixed opinions. While he is an electrifying playmaker with both his legs and his arm, many question the role of running quarterbacks in the NFL today. Some wonder if his arm is good enough to overcome NFL defenses that will partially neutralize his rushing impact. While bigger, faster, and more athletic NFL defenses may hamper his ability to break long runs, his new team will rely on him to continue to use his legs for what he does best- avoid sacks.

            In his breakout 2015 campaign, Watson had an FBS-best 107.3 EPA, or expected points added, according to ESPN. Total EPA is a combination of four parts: passing, rushing, sacks, and penalties. Watson ranked 5^th in the FBS in rushing EPA, 10^th in passing EPA, and 11^th in penalty EPA. Those three numbers were bolstered by him taking part in a 3^rd-best 746 action plays. Where taking part in so many plays might hurt him, sacks, is where he really stands out. While he ranked a modest 28^th of 128 players in sack EPA with -8.5, nobody who ranked ahead of him took part in even 650 action plays. No player ranked ahead of him except Navy’s rush-first Keenan Reynolds, registered more than 23 rush EPA, a far cry from his 33.2. This unique combination of running and avoiding sacks while still being an elite passing QB are what make Watson special and bode well for his NFL future.

            Quarterbacks in the NFL that pose a threat to run can add a new dynamic to their team-- the top three teams in rush yards during the 2015 season (Buffalo, Carolina, Seattle) were led by the top three quarterbacks in rushing yards. Being a rushing threat, though, often leads to increased attention by the defense and thereby more pressure on them. The signal-callers leading two of those leading rushing teams, Seattle’s Russell Wilson and Buffalo’s Tyrod Taylor, felt this pressure often as both teams were in the bottom 10 in the NFL in sacks allowed. Newton’s Panthers ranked a middling 14^th, and they lost only one regular season game on their way to Super Bowl 50. All three of those players were threats with their legs in college, each ranking in the top 15 in rush EPA their final seasons, but none of them were very adept at avoiding pressure. Newton was the best of the bunch with -14.2 sack EPA, ranking 72^nd of 117, Taylor had -19.6, good for 98^th of 117, and Wilson had -15.1, 87^th of 122. Of the three only Newton had more rush EPA than Watson did in 2015.

If Watson can continue to use his legs and avoid sacks as well as he has, he could go on to lead one of the premier rushing attacks in the NFL while simultaneously staying upright when he drops back to pass. That’s a balance that many NFL QBs have struggled to strike- the one who did it best in 2015 won the NFL MVP and came one win from a championship. While pundits will continue to worry how his ability to run will translate to the NFL, it’s really his ability to get out of pressure that will make him stand out.

The Next Kentucky?

             Duke enters the 2016-17 season as the presumptive favorites to win the National Championship, according to many experts and amateurs alike. Despite losing Brandon Ingram, the #2 overall pick in the NBA Draft, they return a host of talent. Additionally, Coach Mike Krzyzewski has put together one of the best recruiting classes in recent memory. The last team to be hyped this much in the preseason was the 2014-15 Kentucky Wildcats, who ripped off 38 straight wins to start the season before stumbling in the Final Four. While Duke’s rigorous ACC schedule makes an undefeated regular season unlikely, it is reasonable to ask the question of which squad is better. Without Duke’s incoming players having seen a single minute of NCAA game action, it’s a difficult determination to make.

            To compare the returning talent of the two teams, I’m going to use Basketball Reference’s Win Shares from the year before, an estimate of how many of a team’s wins a player contributed, as well as the rate statistic Win Shares per 40 Minutes. Here’s how the returners stack up:

                  

A quick glance at the tables shows that while Kentucky brings back more total Win Shares, Duke brings back the best overall player in Grayson Allen and a higher quality of player on average. Both teams return two players that can essentially be labeled non-contributors, posting a WS/40 of under 0.1. Neither Dominique Hawkins nor Derek Willis saw meaningful minutes for Kentucky in the 2014-15 season, and with both playing in a frontcourt crowded by more talented players, it appears safe to say that neither Chase Jeter nor Sean Obi will this year for Duke. Removing those two players from each side gives total WS and WS/40 of 15.3 and 0.682 for Duke and 17.8 and .915 for Kentucky. While removing those small contributors increases Kentucky’s advantage in total, it adds to their deficit on average. Duke’s new averages come in at 3.825 and 0.171 while Kentucky’s are 2.967 and 0.153.

            While Duke is bringing back better players on average than Kentucky, how those players are used could come a long way in determining team success. That Kentucky team blitzed opponents with a seemingly never-ending line of reinforcements- they had 10 players average double-digit minutes per game. It seems unlikely that Coach K uses a similar strategy- he returns less contributing players than Coach Calipari did at Kentucky (4 players with WS/40 over 0.1 compared to 6). Additionally, in the past 3 seasons, only one Duke player outside of the team’s top 7 played more than 10 minutes per game. That player was Semi Ojeleye in 2014-15 who averaged 10.8 MPG over 6 games played, all coming in the easier nonconference portion of their schedule. While Kentucky relied on a long line of players ready to come in and compete, Duke will likely try to use their massive talent advantage over a shorter rotation to win games.

            To evaluate the recruiting classes of the two teams, I am going to use the Recruiting Services Consensus Index (RSCI), a service that combines a number of recruiting rankings from different expert sources to get a more normalized ranking. Duke’s class is composed of five players on this list while Kentucky, returning more players from the year before, had just four:

                  

Despite getting one more prospect, Duke has the better average ranking. Because of Coach K’s shorter rotations and the aforementioned crowded frontcourt, it is possible that Javin Delaurier doesn’t see meaningful minutes unless Duke’s young bigs encounter serious foul trouble. Removing him brings their average ranking down to 7.25, half of Kentucky’s. These rankings are by no means infallible- Towns went on to be one of the best players in the nation and the #1 overall pick in the NBA Draft. Because of Kentucky’s larger rotation, however, the impact of those freshman were limited- the four played a combined 3348 minutes, or an average of 837. In comparison, the three marquee freshman on Duke’s 2015-16 squad (Brandon Ingram, Luke Kennard, and Derryck Thornton) played a combined 3142 minutes, or an average of just over 1047 each. Because of Coach K’s shorter bench, his best freshman should have the opportunity to play more than Kentucky’s did. By all indications, Duke’s incoming class is set up to have a bigger overall impact than Kentucky’s did two years ago.

            While they may not tear through the regular season like Kentucky did, this year’s Duke team stacks up favorably. It seems likely that a deep March run is in their future, especially since they’re headed by one of the top coaches in the NCAA, and one with a history of winning at that. If things go well and they click as a team, they could be the better overall team and capture the championship Kentucky couldn’t.