2019 Atlanta Braves player projections: relief pitchers

I say this every year, but everything about projecting relievers is tough. Projecting reliever playing time is tough. Projecting reliever performance is tough. Hell, even projecting which relievers are going to don which uniform for the bulk of the season is tough. For all we know, a bunch of 2019 relief innings for the Atlanta Braves will be hurled by someone (or multiple someones) not yet in the organization at the time you read these words. Because of that, I’m not inclined to spend a lot of my time and energy worrying about reliever projections, nor giving you things along those same lines to spend your time reading. You could be doing something better, like developing a framework to appropriately adjust wRC+ for each park in which it’s accrued, or something.

For quick reference, if you’re coming to this post with no background, I recommend quickly looking at the following:

• Discussion of reliever projections from last year — https://www.talkingchop.com/2018/3/9/17083396/2018-atlanta-braves-player-projections-relief-pitchers-steamer-zips-iwag
• The position player projection post in this series, which contains a bunch of introductory errata about projections — https://www.talkingchop.com/2019/3/8/18247367/2019-atlanta-braves-player-projections-position-players-steamer-zips-iwag
• The starting pitcher projection post in this series, which describes the specific metrics examined in this post — {FILL IN}

So, where are the Braves in terms of their relief corps? Well, they’re in a weird place. It’s not so much because they’re still doing what they did from 2015 through 2017, where the bullpen was basically the equivalent of a local community theater production of “We Be Pitchers.” (This isn’t much of an exaggeration either, as the Braves barely managed 1 WAR/season from their entire bullpen in those three years — they were below-replacement in 2015 and bailed out only by the fact that 2016 featured a really strong quartet of Jim Johnson, Ian Krol, Mauricio Cabrera, and Chaz Roe, all of whom were then never seen again.) Rather, it’s because at this point, it’s not actually clear whether their bullpen will be composed of “usual relievers” or basically filled with a bunch of guys heretofore considered “starting pitchers.” There are a bunch of pitchers discussed in this post, but the reality is that the 2019 bullpen may lean heavily on pitchers discussed in the starting pitcher post instead.

But, we’re talking about “usual relievers” here, and the below table shows how they stack up according to three projection systems.

In addition, rather than quantifying uncertainty directly via probability distributions, which are way too erratic for relievers to have much meaning, below is a table that shows the bounds of the interquartile range (that is, the 25th to 75th percentile) for IWAG for these guys.

Basically, here’s 14 bullets about these 14 pitchers.

• Arodys Vizcaino’s usage has always been askance relative to his pitching quality. Since 2015, he has 168 innings, which is nowhere near “a lot” for relievers over a long span, but he has the 22nd-highest game entry leverage among all relievers with 160 or more innings in that span, as well as the 18th-highest overall average leverage. Yet, his FIP- among this set of relievers is only 55th, and his xFIP is 85th. Since we’re only talking about 124 relievers here, those aren’t really great numbers. It’s not that Vizcaino is bad or anything, he’s just more of a generic relief guy than a prized high-leverage arm. A notable platoon split makes him more like a situation-specific middle reliever, and his numbers bear that out, but the Braves have kept using him in high leverage and will probably do so again in 2019, health permitting.
• A.J. Minter is the premier relief arm on the roster, and the three projection systems agree on that, with IWAG thinking he’s even better at run prevention than Steamer or ZiPS. That isn’t to say that any of the systems see him as putting up Josh Hader numbers or anything, but he is projected as the pitcher with the 38th-lowest FIP in 2019 by Steamer, so that’s something. The big question for Minter, as always, is injury — IWAG only sees about 45 innings as a reasonable expectation, and with Minter likely starting the season on the shelf, the Braves just have to hope he won’t struggle with other injury or injury-caused ineffectiveness problems as the season goes on.
• Darren O’Day is similar to Minter in more ways than one, despite handendess and age differences. The projection systems don’t quite expect him to be as good as Minter, but they still see him as one of the better relievers on the team. Once again, health is the main question, as IWAG’s reasonable innings total for him is 36 or so, and it sees 20 innings as considerably more like than 60 innings from the veteran. He probably won’t be great while healthy, but he should be solid-to-good.
• ZiPS loathes Jonny Venters’ 2019 forecast for some reason, while IWAG is pretty high on him representing another solid relief arm. Venters’ unique injury situation makes him a beast to project, though, so the fact that his WAR/65 point estimates feature the biggest spread among the 14 pitchers listed here isn’t at all surprising.
• On the other hand, ZiPS joins IWAG in thinking that Dan Winkler is for real as a high-quality reliever, while Steamer is pretty meh on him as a generic relief human. IWAG does see a range of outcomes for Winkler, but most aren’t quite as pessimistic as Steamer. Still, getting “good Winkler” could provide a fair bit of value for the team vis-a-vis “okay Winkler.”
• Jesse Biddle is the first of a troika of relief holdovers from the 2018 squad that IWAG sees as definitely generic relief humans. The interesting thing about Biddle is that ZiPS is crazy-low on him (below replacement!) while Steamer has a 90 FIP- as in the cards. IWAG’s spread of outcomes for Biddle is actually very wide, so ZiPS’ low mark makes sense in that context, as Biddle’s had a winding road to the majors and a pretty erratic rookie season (e.g., compare his lights-out August to his awful September).
• The much-maligned Sam Freeman is pretty similar to Biddle in the generic reliever outlook, with ZiPS having him as replacement level instead. People like to designate Freeman as their whipping boy as his -1.21 WPA in 2018 was the worst among the team’s 30 relievers, but he had a better FIP- than Vizcaino and Biddle last year, and a better xFIP than Venters, Vizcaino, and Shane Carle, yet he’s really the only guy that gets targeted by jettison demands.
• Shane Carle’s central estimate of a generic reliever, but he too seems to be able to run the gamut anywhere from unplayable to replacement level (see Steamer and ZiPS) to decent. Carle took the world by storm in April but wasn’t very good after that, and was awful after returning from injury for his last few innings of the year. There’s a chance that his even without allowing zero homers as he did in April, he has a decent K/BB ratio in him to serve as a decent relief option. But, his peripherals were an abomination the rest of the way, and he had twice as many walks as strikeouts in the second half, so there’s a pretty sizable chance of a prolonged implosion as well.
• I always check the IWAG outputs for certain players to make sure that things aren’t ever getting too crazy due to some syntax error or something else I missed, and when Luke Jackson’s central estimate ends up at 0.0 WAR, I know that things are aligning as expected with respect to the modules responsible for calculating his production. The real question is why ZiPS estimates his outlook as 0.6 WAR/65, because... huh? Jackson does have some potential for good production given his 2018, but even that was driven by one month in which he had a 4.0 K/BB ratio, and he was his usual self in the other months, so I’m hopeful that the Braves choose essentially any other option in an attempt to get some positive bullpen production.
• The projection systems don’t like Chad Sobotka, and that’s pretty much that. His walk rate is off the charts (in a bad way), so unless he can sustain a 35-percent-plus strikeout rate like he did in his short major league stint last year, the FIP isn’t going to look good. It was weird when he made the playoff roster, and it’ll be weird if he makes substantial contributions to the Braves’ 2019 hopes without substantially cutting his free pass rate.
• Grant Dayton could be lights-out, as he was once before. Or, he could be someone completely different at this point, given the fact that he hasn’t pitched at all for a year, and hasn’t been healthy for two. Steamer and ZiPS are sort of splitting the difference; IWAG is way more skeptical but sees a wide range of outcomes, with a lot of upside hindered by the potential for sub-replacement performance.
• Jacob Webb is a name that comes up when discussing potential relief reinforcements for the 2019 bullpen, but he doesn’t really have interesting projections at all. Maybe if he spends the beginning of the year improving on his 2018 Triple-A numbers the expectations will rise accordingly, but he’s not looking like a hidden gem just quite yet.
• Corbin Clouse, on the other hand, has some eye-widening minor league numbers, and his projections are accordingly more positive. He seems to have less downside and a lot more upside than other relatively unknown options.
• Lastly, Thomas Burrows is kind of like “somewhat better Jacob Webb” as far as projections go. The weird thing about Burrows is that he was 24 last year when he was dominating Double-A, and actually spent the bulk of the year a level below that. He didn’t allow a single homer last year — if that’s a fluke, then meh; if he has some kind of anti-homer-magnet-gun, he gets more interesting.

Summarizing this with playing time considered, using the Fangraphs Depth Charts, yields the following:

A 3 WAR bullpen is pretty generic and is exactly what the Braves managed last year. Pushing it up to four wins generally makes it an above-average but not quite top 10 unit; 4.2 WAR would have placed 11th, 14th, and 13th over the last three years on a team basis. But, I’m not sure IWAG is actually as optimistic as shown above, because the above seems to ignore the ability of some players to stay healthy enough to provide all those innings. Instead, IWAG sees the breakdown of innings more like the following:

The above is probably more in line with a health-dependent reality. The good news is that what it doesn’t include is any would-be starter that works out of the bullpen — those guys would probably take innings from the likes of Shane Carle and Luke Jackson in the form of longer relief outings, and thereby add value. So this may be a bit of a pessimistic take on overall bullpen production. But, the bad news is that if the Braves can only rely on Vizcaino, Minter, O’Day, Venters, and Winkler for 180 total innings, i.e., only about a third of what they’ll need to get through the year, they may end up two to four replacement-level-or-worse pitchers before the year is out.

In sum, the projections mostly see this relief unit as an average group. There’s upside in the ability to give high-leverage innings and/or just a plain big ol’ innings total to starters whose stuff should play up in the bullpen and generate additional value for the team that way; there’s downside associated with injury risk as many of the actual relievers, and especially the good ones, are somewhat brittle, and the corresponding reliance on generic-to-worse arms to backfill the difference. The Braves’ position players are definitely the team’s strength, the rotation carries a ton of uncertainty but could very well be a weakness. The bullpen is just kind of there. So it goes.

We’ll wrap up next.