The pain of the M’s season-ending loss to Oakland is still fresh; it’s been less than a week before that crazy game officially eliminated the M’s from the wild card chase. In a season in which their late-season push came up *just* short, you can’t help but wonder if this or that game, or this or that at-bat could’ve swung something. These are all counterfactuals, and so by definition there’s nothing to really learn here. That said, as the Chicago Cubs play their first playoff game tonight, I find myself thinking about an important piece of their (juggernaut) club, a piece they picked up from Seattle in late July: lefty Mike Montgomery.
Montgomery opened some eyes with the Cubs, pitching well in relief and as a spot starter (he made 5 starts for the Cubs in August/September) and stabilizing a bullpen that’s now one of the league’s best. Corinne Landrey’s article at Fangraphs goes over what he’s doing differently (throwing a ton of curves) and what he’s maintained (high velocity) since the trade, and given the plaudits Monty’s racked up and some of the crushing bullpen collapses the M’s suffered after the trade, well…would the M’s have won a wild card berth if they’d kept him?
Obviously, it’s impossible to know, but if you think he was the missing piece, I’d think you’d need to show a clear pattern in the 2nd half losses: 1) that the M’s bullpen had fewer/worse left-handed options, and thus lacked the platoon advantage more often; 2) that this led to lefties enjoying more success against the M’s pen, and 3) the way Montgomery was typically used would’ve made a difference. The third is important, because we don’t just want to take the M’s worst 2nd half reliever and swap him out for Montgomery. We can’t just plop this hypothetical Monty in any situation that went poorly and say the M’s would’ve won the game. I’m going to be up front here, the data for this is a little spotty. I can’t get platoon splits vs. relievers for a certain date range. I’m using first half/second half splits because they’re easy to get, but you and I will make a mental note that Monty was traded before the AS break. We’re going to have to do the best we can with limited data. Ok? Ok.
Let’s start with a bit of context. The M’s bullpen in the first half of 2016 was a very high-K, high-HR club, and the plusses and minuses even out and produce the 14th-most valuable bullpen by fWAR. Their ERA was 3.44 and their FIP was 3.97; the ERA was aided by an impressive .278 BABIP-against. The club’s most-utilized reliever in the first half? Mike Montgomery, with just over 50 innings, over 10IP more than 2nd place Steve Cishek. In the second half, the bullpen’s K rate came down substantially, but this was balanced by an improved walk rate and a slight improvement in HR rate. All of this and a good-but-not-great BABIP pushed their ERA up to 3.68, and sent their FIP soaring from 3.97 to 3.98. All told, they were, again, the 14th most valuable unit. Nothing changed.
That’s not true, of course. They were led by Edwin Diaz, who logged the most innings in the second half, again with a 10IP margin over 2nd place Cishek/Nick Vincent. Diaz’s emergence was a critical factor in the M’s push; he finished with a FIP *under 2* and struck out nearly everyone. With Evan Scribner’s return and Steve Cishek’s improvement, it’s kind of amazing that the bullpen didn’t really change; the M’s added the best reliever they’ve had in years and yet the bullpen’s overall numbers were unchanged. It’d seem that regression came for some of the lesser lights of the ‘pen.
In the first half of the season, the M’s bullpen logged a total of 280 1/3 IP. Of these, left-handers pitched 86 1/3, or 31%. In the second half, the Monty-less bullpen tossed another 242 IP, but lefties pitched just 44, or 18%. The M’s pen was clearly less left-handed, and the lefties that filled in (Nuno, David Rollins, and sometimes-lefty Pat Venditte) weren’t exactly world-beaters. It’s not clear that these guys pitched the innings Monty used to get, though. Despite his success, Montgomery wasn’t given particularly high-leverage innings in the first half; his leverage index was a bit under 1. That would’ve probably gone up, but not as high as Edwin Diaz’s. By WPA, the guys who “got” Monty’s innings were Arquimedes Caminero, Drew Storen and Nick Vincent, with Vidal Nuno thrown in as well as the team’s primary lefty. Caminero, Vincent and Nuno combined to put up a -1.24 Win Probability Added, with Vincent and Nuno finishing 2nd-to-last and last on the club in reliever WPA. This is circumstantial evidence, but you could make a case that Montgomery would’ve led to the M’s using less of their most unhelpful relievers, but the picture’s still mixed: Storen was oddly effective, putting up a plus-1 WPA all by himself.
Since I don’t have platoon splits that’d shed some light on if lefties suddenly started destroying the M’s pen, we’re going to have to do this the old-fashioned way. Let’s take a look at the M’s second half bullpen losses and see where we think Monty may have been used. Of course, these situations may have been different if Monty had been there, and the M’s may have suffered different bullpen losses if they hadn’t made the trade, but this is what we can do without time machines and alternate universes. If you’d like to dive into some very masochistic qualitative data, follow on:
1: The first one was one of the most painful, not just because the M’s held a win expectancy of 96% in the 9th, but because Mike Montgomery actually pitched in this game…against the M’s. This was the extra-inning loss to the Cubs, spoiling a great Felix start, on July 31st. The major damage came in the bottom of the 9th, when Steve Cishek yielded 3 runs. Yes, he faced some lefties, but there’s really no way to argue Montgomery would’ve come in for his first save opportunity of the year in this game. This was on Cishek. Cody Martin then pitched the 10th, 11th and part of the 12th before losing it on a sacrifice bunt. Not sure that Montgomery would’ve come in for extra innings, but Martin was pretty good all told. I don’t think we can say that losing Montgomery made a difference here, though if you WERE to make such a case, you’d probably go all the way back to the 7th, when Tom Wilhelmsen faced Dexter Fowler, Kris Bryant, Anthony Rizzo and Ben Zobrist. Zobrist ended up tripling in Fowler, who walked, cutting the lead from 6-1 to 6-2. Montgomery could’ve seen action there, though Zobrist and Fowler are switch hitters, so he only really would’ve had the platoon advantage against Rizzo. Verdict: Montgomery probably wouldn’t have made a huge difference.
2: Literally the next day, the M’s lost to Boston 2-1, spoiling a gem from starter James Paxton. Again the culprit was Cishek, who gave up a go-ahead HR to Mookie Betts in the 9th inning. This one seems even more clear-cut. If Servais hadn’t yet given up on Cishek as his closer/high-leverage-innings guy, there’s just no way to argue that Montgomery would’ve been in there, and even if you do, that Monty was a clear-cut better choice to face a righty like Betts. You might have more luck saying that Montgomery’s presence would’ve given Paxton less of a leash, meaning he may not have pitched the 8th, where he gave up a game-tying HR to Aaron Hill. That’s equally tough, as Paxton was pitching a shut-out and the Sox had their right-handed line-up in against the lefty Paxton; Monty would’ve been a decidedly odd choice there instead of forcing the Sox to pinch hit. Verdict: Nope.
3: A few days after that, the M’s bullpen suffered another loss to Boston, this one in 11 innings. Ariel Miranda started and pitched well, with Drew Storen, then Vidal Nuno, Tom Wilhelmsen, Blake Parker and Cody Martin pitching in relief. With another lefty starter on the mound, Monty probably wouldn’t have come in when Storen did, but there was a perfect opportunity in the 8th, when the Sox led off with David Ortiz, Dustin Pedroia, and Jackie Bradley jr. 1-2-3 in the inning. Vidal Nuno faced these three, getting the lefties, but giving up a single to Pedroia, who was stranded after Wilhelmsen pitched the final 1/3 IP. This was a high leverage spot that screamed out for a lefty. And Nuno essentially did his job. With Monty on the club, Nuno’s available for the mop-up duty that fell to Cody Martin, but I’m not sure it would’ve made much of a difference. I’d much rather have Martin in there vs. Betts, for example. Verdict: Nope.
4: After that, the M’s went 9-1 in their next ten, including a walk-off win against Detroit. They had a hiccup on August 16th, though, against Anaheim. To that point, Caminero had seemed like a revelation, giving up no runs and showing oddly good command. That stopped here. Miranda started this one, too, and he was relieved by Nick Vincent, Vidal Nuno and then Caminero. The M’s were up 4-1 in the 6th, when Miranda left after yielding walks to the first two batters of the inning. Nick Vincent promptly allowed a game-tying 3R HR to Albert Pujols, then after retiring a batter, gave up a go-ahead dinger to Jett Bandy. Again, with a lefty starting and facing a righty-heavy line-up starting with Pujols, there’s just no possible way Monty would’ve been pitching here. The M’s retook the lead, but ended up losing it when Caminero suddenly blew up with 2 outs in the 8th. Jefry Marte hit a tying HR, and then Cliff Pennington tripled in Bandy for the final run. Marte and Bandy are RHBs and Pennington’s switch – this was Caminero’s inning. Nuno pitched the 7th with Kole Calhoun and Nick Buss up, and Montgomery may have tossed this one if he were still on the team. Didn’t much matter, as Nuno got the Angels 1-2-3. Verdict: Nope.
5: This one… this one still hurts. The M’s took a 6-3 lead into the 9th at home against the rebuilding Milwaukee Brewers, and Tom Wilhelmsen coughed up 4 in the 9th to lose it. The starter was, yet again, Ariel Miranda. The Bartender was the 4rd righty reliever that day, following Cishek, Vincent and Caminero. It made sense to stick with the righty, as the heart of the Brewers lineup (2-3-4 were due up) is all right-handed. You know what happens next – Broxton HRs, Braun Ks, Perez doubles and then M’s killer Chris Carter ties it on a HR, followed by a single by Santana. At this point, the M’s went to Vidal Nuno, who gave up two straight singles to score Santana. If Montgomery were on the team, he may have come in right when Nuno did, to face either Manny Pina or Scooter Gennett. Would he have gone hit-hit-walk? Maybe not, but given the situation, it’s not clear that Montgomery would’ve been the first lefty to pick, but it’s certainly possible. Verdict: ehhhh, maybe? The damage was already done, so maybe Monty could’ve sent it to extras.
6: A few days after that heart-wrenching loss, the M’s turned a 6-3 lead into a 7-6 walk-off in Chicago. By this point, Caminero was in free fall, and he never really recovered. With the M’s win expectancy at 90%, up 3 in the bottom of the 7th, Caminero came in to face JB Shuck. An error and a walk brought up three big righties, Tim Anderson, Jose Abreu and Todd Frazier. Anderson chased Caminero with a ground rule double, and then Steve Cishek gave up a 2R single to Frazier. The beginning of this inning feels like a natural spot for Montgomery, with the first and third batters, Shuck and Eaton, batting lefty. Remember though that Shuck grounded to SS and Ketel Marte booted it, and in any case, Montgomery’s walk rate wasn’t much better than Caminero’s. With the three righties in a row due up after Eaton, I think it’s pretty likely that Servais would’ve gone to Cishek anyway. This one comes down to the likelihood that Montgomery would’ve also walked Eaton, which…we’re getting down in the weeds. Verdict: Maybe.
7: On August 30th, the M’s lost a lead late, then regained it, then watched as lefty Rougned Odor walked it off against Edwin Diaz in the bottom of the 9th. This one seems fairly clear-cut: there’s no way Servais goes to anyone but Diaz in that situation. Up 1 in the 9th with Beltre/Odor/Lucroy – that’s a tailor-made situation for a RH closer like Diaz. With a lefty in Paxton starting, Montgomery probably wouldn’t have come in much earlier, either. Nuno pitched to one batter, Odor, in the 7th, and while Odor singled, he was thrown out at 2nd. The three batters Cishek faced to start the inning were two righties and a switch hitter, so it doesn’t seem likely that Montgomery would’ve been brought in there. The bottom of the 8th makes the most sense, with Mitch Moreland and Nomar Mazara batting, but Diaz held the Rangers scoreless there. Verdict: Nope.
8: The final bullpen loss of the year still feels raw, but it was that awful game against Oakland on October 1st. Given the way the game went, Montgomery could’ve been used any number of places, including following Iwakuma in the 5th, where the M’s used Vidal Nuno. Nuno pitched a 1-2-3 inning in the 5th, but gave up a run in the 6th on a single by LHB Joey Wendle. Would the M’s have used Mongtomery so early when facing a 3-run deficit? Maybe, and it’d be a much better spot for Monty given his platoon splits weren’t anywhere near as bad as Nuno’s; having Nuno pitch multiple innings was asking a lot. Wendle singled and went to 3rd on an error in the 8th, and you can again say that if Monty hadn’t pitched earlier, he could’ve faced a few batters here, but the big hit in the inning was Marcus Semien’s ground rule double, and if you hadn’t used Cishek to face Wendle, it would’ve been awfully tempting for Servais to go righty-righty there. If they didn’t change there, they almost certainly would’ve when Khris Davis’ spot came up. This is a tough one; I think Monty might’ve mattered here, but it’s tough to make a definitive case. Verdict: Maybe?
What have we learned? The M’s suffered some tough losses in the second half, but I’m shocked at how many of the key, WPA-swinging hits came off the bat of right handers facing right handed relievers. We’ve talked about 7 big home runs, and *6* of them were by righties, in every case facing a righty. Six of eight games were started by lefties, making it slightly harder to use Montgomery in the middle innings. The “maybes” like the game at Wrigley or the game that knocked them out of the playoffs feel a bit forced; you’d have had to use Montgomery creatively for it to matter, and he’d have to have been perfect. I’m actually a bit surprised by this; given the huge shift in the percentage of bullpen innings that went to lefties, I’d have imagined it would’ve been easier to pinpoint places where the M’s obviously would’ve preferred Montgomery.
The specifics of the M’s most painful losses highlight that the problem with the M’s bullpen wasn’t its lack of lefties, it was that the back of the bullpen was running on fumes and started to break down. Caminero, Wilhelmsen and Vincent scuffled, while Nuno was either a LOOGY or used in mop-up duty. Meanwhile, Edwin Diaz posted reverse splits, and while Cishek was generally dominant against righties, it seems like his every failing is detailed above.
A more interesting line of attack would be to argue that keeping Montgomery would’ve allowed the M’s to use him in the rotation, pushing Miranda to the pen. That’s an even tougher argument to get a handle on, as committing to that likely changes the return they get for Wade Miley; seems unlikely they would’ve picked up another lefty swingman if they still had Montgomery.
I gave the Montgomery-for-Vogelbach trade some qualified praise on the blog, but the context was critical: the M’s wild card odds were under 10% on that day, with the M’s sitting with the 9th best record in the AL, and with the surging Astros 3 up on the M’s in the West. In that context, the trade makes perfect sense. I still can’t quite believe that the M’s surge muddied the waters a bit, leaving us to look back on a trade that’s firmly in the ‘rebuild’ and not ‘win now’ camp with a bit of regret. All of that said, and as bad as these losses were, I just can’t see how to justify the belief that Montgomery would’ve made THE difference. I laid out three criteria to justify that belief, and we’ve found evidence for one of them. Montgomery would certainly be nice to have, with several more years of club control and an emerging curveball and sustained plus-velocity, but the trade wasn’t the difference in the wild card chase.