Rockies Boost Bullpen with Bryan Shaw and Jake McGee by Craig Edwards December 13, 2017 Last season, the Colorado Rockies made the playoffs with a below-average offense, great defense, decent starting pitching, and a very good bullpen. It remains to be seen if the Rockies will completely double-down on that strategy for next season, but they certainly appear motivated to remain solid at the end of games. Last night, it looked like the team was close to a three-year deal for around $27 million with former Cleveland pitcher Bryan Shaw. It looks like they are also in the process of retaining Jake McGee to a three-year deal with a fourth-year option. Bryan Shaw might not have gotten much notoriety in Cleveland with Andrew Miller and Cody Allen closing out games, but Shaw is a very good pitcher. He uses a mid-90s cutter on a vast majority of pitches, and it gets swings and misses and a tons of ground balls. For a rather cherry-picked analysis, here is a comparison of Shaw with the potentially departing closer, Greg Holland. Bryan Shaw and Greg Holland in 2017 Name SV IP K% BB% GB% ERA ERA- FIP FIP- WAR Bryan Shaw 3 76.2 23.4 % 7.1 % 55.9 % 3.52 79 2.96 67 1.6 Greg Holland 41 57.1 29.8 % 11.1 % 41.6 % 3.61 72 3.72 81 1.1 Shaw pitched more innings and they were higher quality innings than Holland provided last year. We can see that Holland’s ERA was considerably lower once we factor for Coors Field based on the ERA-, and Holland does strike out more batters, but every other factor favors Shaw. The walk rate favors Shaw and a great ground-ball rate keeps balls in the park, something of particular importance in moving to Colorado. Now we don’t want to just waive our magic wands and say that Holland was just a lot luckier than Shaw last year to explain away the difference in ERA. A chart should suffice, though. Bryan Shaw and Greg Holland in 2017 Name LOB% BABIP xwOBA Bryan Shaw 66.3 % .311 .260 Greg Holland 75.2 % .252 .270 xwOBA from Baseaball Savant Holland ended up with a BABIP well below reasonable expectations, Shaw’s LOB% is well-below league average and his career norms, and Shaw’s xwOBA based on launch angle and exit velocity was lower than Holland. This is not to say Holland was bad last year. He was worth more than a win above replacement, which is quite good for a reliever. Shaw was just better, and if he had been a bit more luck neutral, he would have had a sub-three ERA. The projections put both players equally in terms of FIP and ERA, right around the 3.8 or 3.9 mark, which would be worth roughly a win in Coors Field. Whether that is worth a three-year deal for $27 million might be debatable, but that is certainly what teams are paying for relief pitchers of Shaw’s caliber. If the alternative to Shaw is double the money for Greg Holland or Wade Davis, then Shaw is the prudent choice. In terms of production, McGee mirrors Shaw. He had the 2.93 FIP to go along with a 3.52 ERA, 23% strikeouts, 7% walks, and he also uses one mid-90s pitch about 90% of the time. For McGee, it is the four-seam fastball. He doesn’t have the sweeping breaking ball of a traditional LOOGY, but he also doesn’t have much in the way of platoon splits. It’s why Colorado tried to install him as closer in 2016 after the trade with the Tampa Bay Rays. The trade didn’t work out quite as expected for the Rockies in 2016 as McGee’s lost velocity at the end of 2015 carried over. Here’s a chart with McGee’s velocity by month from 2013, though last season from Brooks Baseball. It’s not difficult to see why McGee’s 2016 season was bad, and looking back it shouldn’t have been hard to see coming given the loss in velocity the previous season. McGee’s velocity was back up last season, and he put together a good year. However, McGee’s performance in the early part of the season was significantly better than the stretch run. We are dealing with relatively small sample sizes, but in the first half, McGee faced 145 batters and struck out 30% of them. In the second half, McGee faced 84 batters and struck out just 17% of the hitters he faced. He did induce a few more infield fly balls in the second half, and his FIP remained a solid 3.56 after the break. However, much of that success had to do with allowing just a single home run on 4% HR/FB rate. That doesn’t seem repeatable and McGee could struggle if he doesn’t go back to missing bats like he did in the first half of last season. In terms of the bullpen overall, Shaw can pretty easily take Holland’s role in terms of production. The question will be whether McGee can repeat his overall numbers from last season. The projections for McGee are right in line Shaw and Holland as roughly one-win relievers. McGee pitched at a rate twice that for the first half last season and his home run prevention had him pitching like his projections the rest of the way. Right now, it looks like the Rockies formula is pretty much the status quo. Signing McGee and Shaw combined is what it would have cost to bring Holland back on either his vested option or the qualifying offer, though committing to future years doesn’t always work out. It’s possible the Rockies could still add another piece to the bullpen, which can be especially helpful in Colorado with a rotation that can’t be expected to shoulder that many innings in the mountain air. They do have some other holes that they would be wise to fill. Colorado should have some money left to get another bat or another starter to bolster a team with reasonable hope for a wild card. They have room, either at first base or in the outfield to make a significant upgrade. It’s possible that’s where the Rockies should have been focused all along given the unpredictability of relievers and their poor offensive season last year. The Rockies are going back to the formula that worked last season, though whether they can repeat a 21-14 record in one-run games is tough to predict. If they are to compete again next year, the bullpen is going to be a big part of their success.