Jays From the Couch examines how the Toronto Blue Jays could surpass their projections in 2017…it is very doable.
I’m an optimist –but more than that, I’m a realist. The optimist in me is happy to know that we’re more likely to succeed than at least 21 other teams, amongst the Mets and Giants, in a group right behind the six division favourites. But the realist in me accepts that odds makers give the Jays a roughly 5-8% chance of winning the 2017 World Series, while Fangraphs projects them to have an even lower 3.4% chance. Since we (optimistic) Jays fans are hoping that that 3-8% chance pays off, wouldn’t it be fun to examine how a (realistic) “good case” version of the Blue Jays season might play out? Yes? What if math is involved? Still yes? Alright!
First, we’ll look at Fangraphs’ baseline projections for each position. Then, I’ll highlight areas where certain Jays could over-perform and see how that could affect the team’s overall performance. But again –I’m a realist and my intention is to remain rooted in reality throughout. I promise I will not get carried away and suggest that Devon Travis is due for a .350 AVG season, or that Jose Bautista will anger-hit his way to 73 HRs (although…).
A few disclaimers are probably useful for an article like this:
- I prefer using weighted runs created plus (wRC+) when examining hitters, as it controls for different parks. But since the projections use wOBA, I will too.
- To keep things simple, projection-beating hitters saw their wOBA increase, as it is essential to recalculating their WAR. I left their AVG, OBP and SLG as projected by Fangraphs.
- Similarly, I only altered a pitcher’s innings pitched (IP) and WAR, leaving all other stats as they were.
- The math behind my adjustments is correct to the best of my knowledge. That said, math is hard and mistakes happen.
- Any or all of this can be made irrelevant as soon as the season starts and players get hurt. Nevertheless, we’ve got to do something to pass the time until baseball starts and the last week is always the hardest.
- Due to rounding, sum totals may not add up.
- I’m happy to respond to any questions or comments about my methodology (or anything else) left in the comments section.
First, an overview of Fangraphs’ Depth Charts projections (as of March 29):
The projections are a combination of the Steamer and ZiPS projection systems, with Fangraphs’ employees taking care of playing time distributions. The Blue Jays are projected to have the eighth best team in the majors, according to wins above replacement (WAR). Position players are key to this production, with the Jays getting Top 5 production from the group as a whole. The Jays don’t appear too top-heavy either, as they are expected to get Top 10 production from six of the nine positions. On the pitching side of things, the Jays project to get league average production (15th highest WAR) from both their starters and their bullpen. As you can see, there’s a lot to be excited about as a Jays fan.
Now let’s look at the positions that I will leave as currently projected by Fangraphs:
At the C, SS, 3B and RF positions, established veterans will take nearly all of the plate appearances. Fangraphs’ projections put their 2017 performance in line with their performances in recent years.
- Russell Martin is projected to repeat his 0.322 wOBA in 2016, with his defence rebounding to a happy medium between his rough 2016 and his league-best 2013-15.
- Troy Tulowitzki is projected to roughly repeat his solid 2016 performance. Last year, he was the 10th best SS according to WAR, this year he’s projected to finish 9th.
- Josh Donaldson is projected to have the highest WAR among all position players (non-Mike Trout division). Sure, I could boost his PA up to the 700 level he has hovered around for four years, which could add about half a WAR. But the reasonable thing to do is take the 6.1 WAR and be very happy with it.
- Jose Bautista is projected to replicate his 2015 ½ season, by which I mean a level between his 2015 and 2016 performance. Plus, it would appear that the projections give him credit for an arm strength rebound as well. All told, I have no quibbles with Jose putting up a 3.2 WAR this year.
Shifting gears to the bullpen, I have to agree with the conventional wisdom that relief pitching is an almost un-projectable position. There are issues with small sample sizes and the year-to-year variation to contend with. Plus, WAR might not even be the best way to evaluate relief pitchers, with win probability added (WPA) often considered a good statistic for the purpose. All in all, the Jays’ bullpen is projected to be worth 3.3 WAR, tied for 15th with the White Sox, very similar to the 3.9 WAR (tied for 15th with the Mariners) the bullpen put up in 2016.
Now, onto the potential projection beaters. Any Jays fan reading this post has likely spent some part of their winter pondering the Jays’ 1B/LF situation. Fangraphs projects the Jays to get a grand total of 1.7 WAR from these two spots, so things can only get better from here. (Right? Please say I’m right.)
The two positions are connected by an exciting gentleman named Steve Pearce. He may not seem like much but, health permitting, Pearce is my pick for signing of the season. Fangraphs projects Pearce to contribute 1.6 WAR in 371 PAs for the Jays this year, mostly playing 1B and LF with a little RF, DH and 2B sprinkled in. The key to salvaging 1B/LF without any outside additions will be for Pearce to stay healthy enough to push 600 PAs. I’m optimistic that he will. A big part of this optimism is Pearce’s experience thus far with the Jays’ High Performance team. He seems very confident that his new focus on flexibility and building stability over raw strength will help him decrease his likelihood of injury. Another reason for optimism (that can apply to other Jays as well) is that “injury prone” may not be a thing (behind a pay wall). Obviously, there are chronic injuries. But the data seems to suggest that a player who keeps injuring different muscles is usually just unlucky.
Back in my first post with Jays from the Couch, I discussed the double platoon the Jays seemed intent on using in 1B and LF. While my preference is to use Pearce at 1B full-time and come up with something better at LF (my first choice is a healthy Dalton Pompey), this double platoon has its merits. Pearce would play 1B against LHP, Justin Smoak would play against AL RHP and Kendrys Morales would mainly play in away interleague games. I’ll assume, as Fangraphs did, that Rowdy Tellez will get in a handful games, perhaps as an injury replacement for a couple of weeks. In LF, Pearce would play against some RHP, Ezequiel Carrera would play against the other RHP and Melvin Upton Jr. would play against LHP. In order to keep things a little less complicated, I assumed that Dalton Pompey and Harold Ramirez would not feature this year.
Organizing the players in this way should help the Jays beat their 1B and LF projections:
- Pearce should outperform his projected 1.3 UZR/150 defence at 1B. Over the last three seasons, he has maintained a 11.6 UZR/150 in the position.
- Pearce should also outperform his projected 0.352 wOBA. Over the last three seasons, he has a 0.383 wOBA vs. lefties and a 0.354 wOBA vs. righties.
- Smoak should outperform his 0.314 wOBA projection if he is used exclusively against RHP (0.322 wOBA vs. RHP in 2016).
- Upton Jr. is worth a negative number of batting runs vs. RHP, so less is more for him. He can beat projections if he is used exclusively against LHP (0.333 wOBA vs. LHP with the Jays in 2016).
All told, these adjustments suggest that the Jays could squeeze an extra 2.4 WAR from 1B/LF. The Jays would now have “solid starters” (Fangraphs’ description of players worth 2-3 WAR) at both first base (2.0 WAR, 13th) and left field (2.1 WAR, 7th). As a bonus, it would come at the cost of $13 million in salary, a rate of $3.2 million per WAR (less than half the market rate of $8.5M/WAR). This is fundamentally why the double platoon is the Jays’ Plan A: its value potential. These “value plays” help reduce the payroll, leaving space for eventual investments in our young core. If the plan doesn’t work out and LF is the only thing standing in the way of success, I’m cautiously optimistic that the front office will invest in an improvement there.
As an aside, my ideal solution to 1B/LF (without making trades) would involve full-time roles for Steve Pearce and Dalton Pompey. Over the last three seasons, Steve Pearce has a claim as the best 1B defender in the majors (his 11.6 UZR/150 is highest among 1Bs with 900+ innings played). Over the same time, Pearce produced the 24th highest wOBA (0.363) among all qualified batters, ahead of former Oriole teammates Manny Machado (.361) and Chris Davis (.349). He’s a lefty masher, but he can do plenty of damage against righties too.
Dalton Pompey, on the other hand, is an enigma at the MLB level. He’s had a grand total of 148 MLB PAs, so projecting him is a tricky task. I’ll keep things simple. His defence and base running appear to be his key sources of value. He has a career 4.1 UZR/150 in the outfield and has generated 8.1 base running runs above average (BsR) per 600 PAs. On the other hand, Fangraphs projects him to have a .296 wOBA (in line with his MLB career). Using these numbers, Pompey would produce 0.6 WAR in 400 PA out of left field. Upton would still bat against LHP, while Carrera would get his at-bats as the fourth outfielder.
Combined, the two positions would be worth 5 WAR, an improvement to the aforementioned double platoon. Additionally, this setup has intangible long-term value, as the Jays can finally find out what Pompey can do in the majors. We all hope (especially for his own sake) that Pompey’s concussion issues of 2016 won’t repeat themselves in 2017.
Devon Travis stands to benefit from the possibility that being injury-prone isn’t a thing. Cracking 500 PAs for the first time in his career would help add to his projected WAR by way of quantity. Repeating his career defensive performance (1.8 UZR/150) would represent a relative increase in quality, as he is projected to be league-average defensively (0 UZR/150). His hitting is projected to regress from his .349 career wOBA, mainly due to a projected regression in his batting average for balls in play (BABIP).
The theory is that hitters can’t really control whether a ball in play will become a hit or an out, so the expectation is that all hitters should have a BABIP close to .300. In practice, players have run consistently high or low BABIPs (.250-.350). If Travis hits for his career wOBA this season, he’d project to be worth 3.3 WAR and produce the 3rd highest WAR among 2B. I will err on the side of BABIP caution and leave his wOBA as projected by Fangraphs.
Fangraphs projects the Jays to have the 10th best production from CF, driven mainly by Kevin Pillar‘s defence. That said, Fangraphs projects a bit of a regression for him (11 UZR/150). Pillar’s first two full seasons in the majors have seen him maintain an impressive 21 UZR/150, the 4th best mark in the majors among qualified position players. I split the difference and estimated that he puts up a 16 UZR/150, equal to his 2015 performance. I’ve also adjusted his walk rate, incorporating the idea from the post I wrote last week, which was corroborated this week at BP Toronto.
To avoid getting carried away, I’ve estimated the effect of a modest 1% increase in his walk rate (5.3% from 4.3%). The two changes each contribute about 0.4 extra WAR, increasing the Jays’ production from CF to 3.5 WAR (5th best in the MLB). One projection component that I didn’t change is his base running value. Over the last two seasons, he has very quietly amassed the 6th highest BsR in the MLB (10 BsR in 1212 PAs). A repeat of his 2015-16 base running production would be roughly worth another 0.4 WAR.
At DH, the Jays are anchored by Kendrys Morales. Fangraphs projects Morales to make small improvements across the board relative to his 2016 performance. The big question is whether the “real Kendrys” is closer to the 2015, .364 wOBA (25th of 141 qualified batters) version or the 2016, .339 wOBA (77th of 146 qualified batters) version. Tony Blengino of Fangraphs might argue that both versions are basically the same. According to Blengino’s 2015 hitter contact quality report, 33.6% of Morales’ batted balls were fly balls, 20.4% were line drives and the average exit velocity of these flies and liners was 95.6mph.
In 2016, Morales hit 33.6% of batted balls for fly balls, 20.1% for line drives and the average exit velocity of these flies and liners was 96.1mph. Morales’s 2016 batted ball performance earned him the 9th highest “Adjusted Production” from Blengino among all qualified batters. As per Blengino, “Adjusted Production expresses…what a hitter “should have” produced based on the exit speed/launch angle of each ball put in play.” Just to show that this statistic isn’t complete BS, the eight guys in front of him were Miguel Cabrera, Mike Trout, Josh Donaldson, David Ortiz, Joey Votto, Freddie Freeman, J.D. Martinez and Daniel Murphy. Good hitters clearly score well with this metric.
Blengino went so far as to predict that “a healthy Morales outperforms both Jose Bautista and Edwin Encarnacion, for far less money, in 2017.” With all this in mind, I’ll just split the difference and guess that Morales produces a wOBA of .352, half way between his 2015 and 2016 levels. Doing so would make Morales worth an extra half a WAR, bumping up the Jays’ DH position from 4th to 3rd best in the majors.
Finally, we arrive at the starting rotation. This has been another contentious area for Jays fans. While the pitchers themselves have been locked in all winter, their 2017 projections have not made Blue Jays fans happy. Jays from the Couch has discussed this in some detail but, in a nutshell, Fangraphs projects that the production of the starting rotation will fall off a bit. The group is projected to be worth 13.4 WAR, good for 15th in the MLB. In 2016, a very similar group of starters was worth 15.3 WAR, good for 5th in the MLB.
There are essentially two reasons for the drop off. Reason one is easy: lower innings counts for everyone but Aaron Sanchez. In 2016, our Top 5.5 starters (both R.A. Dickey and post-trade Francisco Liriano included) pitched 983 innings, tops in the majors. In 2017, the starting rotation is projected to pitch 871 innings, with the difference projected to be made up by near-replacement level depth starters and relievers. Is there any wiggle room in these innings projections? Yes (see earlier comments on the Jays High Performance team), but since pitcher health is a dangerous game, I won’t get carried away.
- Sanchez is projected to see his innings pitched rise to 205 in 2017. I’m happy with that.
- J.A. Happ is projected to see his innings fall from 195 to 181. 181 IP would still be his second highest IP in a season. I’d be happy with 181 IP from Happ in 2017.
- Marcus Stroman‘s IP are projected to plummet from 204 in 2016 to 169. This fall is likely because Stro only pitched 157 in 2014-15. Let’s split the difference and project 184 IP from him.
- A modest fall in IP is projected for Marco Estrada (167 from 176 in 2016). Given his back issues, this seems reasonable.
- Liriano is projected to see his lowest IP total since 2011, with 149. Given his consistent ability to crack 160 IP, I’ll go with that nice, round number.
Reason two for the drop in WAR is more complicated. At this point, it’s best to pause and make clear how Fangraphs calculates pitcher WAR. While there is no published equation, the gist is that they use fielding independent pitching (FIP) to calculate WAR rather than good, old earned run average (ERA). There is a good reason for this: the data shows that pitchers generally cannot control what a non-home run batted ball becomes. It might become a double, it might be caught for an out, it might roll slowly to an infielder. FIP focuses only on a pitcher’s home runs, walks and strikeouts, the three things we can assume the pitcher has some control over. Is this system perfect? No. Is it founded on sound logic? Yes.
The problem with this system is pitchers like Marco Estrada. He genuinely seems capable of generating weak contact by opposing batters on a consistent basis. Over the last five seasons, overlapping with Estrada’s career as a regular starter, 57 starters have accumulated 700+ IP. Estrada has a BABIP of .251, comfortably the lowest of the bunch and much lower than the .300 BABIP we expect pitchers to hover around.
Since coming to the Blue Jays in 2015, his BABIP has fallen to .227, also comfortably lowest among 71 pitchers with 300+ IP. His killer changeup and fastball command makes it very tough for hitters to make dangerous contact. Ditto for Sanchez’s and Happ’s sinking two-seam fastballs. Their hard to hit stuff led them to produce BABIPs of .267 (14th lowest) and .268 (15th lowest) in 2016. These low BABIPs resulted in large gaps between a pitcher’s ERA and FIP, as ERA counts a pitcher’s actual outcomes on batted balls (which is favourable for the low BABIP guys) while FIP assumes league average outcomes on a pitcher’s batted balls against (which is unfavourable for the low BABIP guys). Estrada, Happ and Sanchez produced solid ERAs of 3.48, 3.18 and 3.00, but could only muster FIPs of 4.15, 3.96 and 3.55. As a result, they only produced 3.0, 3.2 and 3.9 WAR, ranking 31st, 22nd and 15th out of 71 qualified starters in 2016. Very good, but not as valuable as they seemed to Jays fans.
So how do we resolve this? By following Fangraphs own advice, pretty much: “When evaluating knuckleball pitchers, it’s better to use RA9-Wins, as those pitchers have been shown to consistently generate weaker contact than typical pitchers”. RA9-WAR is the same as regular WAR but it uses runs allowed per nine innings (similar to ERA) rather than FIP. While Estrada, Happ and Sanchez are not knuckleballers, their extreme pitches are knuckleball-like, in the effect they have on batters. The three Jays were worth 4.0, 5.4 and 5.6 RA9-WAR, ranking 25th, 14th and 12th out of 71 qualified starters in 2016. That’s more like it.
Looking ahead to 2017, I think there is a good chance that a healthy rotation can crush projections:
- In his career, Aaron Sanchez has generated 5.7 RA9-WAR per 200 IP. Split the difference between that and his projection and you get 6 WAR over his 205 IP.
- Similarly, J.A Happ has generated 4.8 RA9-WAR per 200 IP over the last two seasons. Focusing on his last two years may seem arbitrary, but reflects the huge difference in approach he has adopted, in particular just throwing a steady stream of fastballs for strikes. Splitting the difference with his projection produces 3.4 WAR over his 181 IP.
- Marco Estrada has also generated 4.8 RA9-WAR per 200 IP over the last two seasons. Splitting the difference with his projection produces 2.9 WAR over his 167 IP.
- Francisco Liriano is not a FIP breaker like the previous three. However, his projected WAR for 2017 is hurt by his terrible first half of 2016. Back in 2013-14, the last time he pitched to Russell Martin, he produced 3.2 WAR per 200 IP. Splitting the difference with his 2017 projection produces 2.2 WAR over 160 IP.
- Unlike his fellow starters, Marcus Stroman has a pretty normal career BABIP (.303). As a result, Fangraphs projects him to produce a FIP of 3.68 in 2017, only slightly higher than his career mark and slightly lower than in 2016 (3.71). Fangraphs also expects his value (3.8 WAR per 200 IP) to be in line with his career performance (3.9 WAR per 200 IP).
With these modifications, Jays starters are projected to produce 17.4 WAR in 2017, 6th best in the majors and much more in line with the stellar group performance we saw last season. While the Fangraphs projection system says what it does, there has been an increasing acceptance among the site’s writers regarding the FIP-beating talents of Estrada (here, here, here, here and my personal favourite here), Sanchez and Happ. Specifically referencing the Jays starting rotation for 2017 in a recent chat, Fangraphs writer Travis Sawchik mentioned that he thinks they’ll finish the season with the 8th to 12th highest WAR in the majors.
All told, my adjustments increased the Jays’ team WAR from 40.2 (8th highest in the MLB) to 49.4 (3rd highest in the MLB). The group of Jays’ position players would now finish tops in the majors with a 28.7 WAR. Seems a bit high? If I split the difference (I promise that’s the last time I use that phrase in this article) between the Jays’ position player WAR in 2015 and 2016 I get…29 WAR. Jays’ pitchers would now be worth 20.7 WAR, 7th highest in the majors. That represents a small improvement over last year’s 19.2 WAR, fair given that Estrada, Happ, Stroman and Sanchez are coming back and Liriano > Dickey.
In each case, I generally erred on the side of caution when adjusting a player’s projections. As a result, I feel confident saying that a 49.4 team WAR does not require absolutely everything to go the Jays’ way in 2017. Maybe Devon Travis misses time with a brand new injury, but maybe Upton, Smoak and Pillar hit like they did in 2015. Maybe Estrada’s back flares up and he misses 30 of his 167 projected IP, but maybe Stroman cracks 200 IP like he did last year.
Do these projections mean that the Jays are the 2nd best team in the majors? Not necessarily. My main point is simply that it doesn’t take much to put the Jays in that top group with the Dodgers, Cubs, Indians, Astros, Red Sox and Nationals. They have the hitting, defence and pitching to compete with these teams. And, ultimately, if they beat their current projections by only 4-6 WAR rather than 9.2 WAR, they’d still find themselves in the middle of that top pack of contenders. Plus, the front office likely has some salary left to spend on in-season upgrades and, while I really want to see all of the Jays’ prospects flourish in Toronto, it is comforting that there is once again enough prospect depth to make a modest trade or two. I have a good feeling about 2017.
*Featured Image Credit: C Stem
THANK YOU FOR VISITING JAYS FROM THE COUCH! CHECK US OUT ON TWITTER @JAYSFROMCOUCH AND INSTAGRAM. LIKE US FACEBOOK. BE SURE TO CATCH THE LATEST FROM JAYS FROM THE COUCH RADIO