Marco Estrada Isn’t Just a One-Year Fluke
Toronto has made the first move toward retooling its starting rotation, reportedly resigning Marco Estrada to a two-year deal worth $26 million.
Estrada is 32 years old and coming off a career-best season, but also had just ~$10 million in combined career earnings before this offseason, and would have entered the market with draft pick compensation tied to him in a rich free agent class for starting pitching.
The move feels like a win for both sides. Estrada takes something of a middle ground between the risk of accepting the qualifying offer in lieu of guaranteed years and testing the market in hopes of cashing in on his 2015 with a long-term deal. In making the decision, Estrada likely considered the recent situations of similar pitchers like Kyle Lohse and Ervin Santana who went unsigned until March after being extended a qualifying offer and ultimately chose to avoid that possibility by staying with a team that should contend for both years of his contract, while getting to throw to Russell Martin, one of the game’s best catchers and one with whom he’s already familiar.
From the Blue Jays’ perspective, they return their most consistent pitcher from 2015 to a mostly depleted rotation, and fill one of potentially three open spots with a short-term deal at a completely reasonable price, leaving room for a higher-profile pitcher to slot above Estrada.
Zooming in just on Estrada, there seems to be a perception among some that, had any team signed him to a multi-year deal, they’d be taking a risk. After all, he’s still just one year removed from a replacement-level season in Milwaukee, and for a 32-year-old, he doesn’t have much of a track record to stand on. To the Estrada naysayers, his 2015 season was a fluke, propped up by a historically low BABIP and a career-low HR/FB% that helped hide his ever-declining strikeout rate.
However, I’m not so sure that’s the case.
I mean, yeah, it wouldn’t be wise to bet on Estrada running a .216 BABIP again, but since becoming a regular in 2011, he’s suppressed production on balls in play better than any pitcher in baseball, and his four-year BABIP of .254 makes this year’s outlier figure look a bit less like an outlier. Since 2013, that number is .242. You don’t ever want to bet on someone repeating a .216 BABIP, but if you had to bet on anyone, Marco Estrada might be your guy.
Then there’s the track record. You’ve heard folks refer to him as a one-year fluke, which probably isn’t incorrect. In fact, Estrada has had a fluky season, where his year-end production stuck out from the rest. Only thing is, that year wasn’t this year. It was 2014. Consider Estrada’s yearly ERA’s as a starter since 2011:
- 2011: 3.70
- 2012: 3.76
- 2013: 3.87
- 2014: 4.96
- 2015: 3.28
It’s easy to spot the outlier, the “fluke season” if you will, and it wasn’t this most recent one. Four out of five years, Estrada’s turned in an ERA that begins with a three, and he’s done so each time pitching in extreme hitter’s parks.
This being FanGraphs and all, you might be inclined to scoff at ERA and compare it to Estrada’s FIP, finding that these last two years it’s been at an all-time high and among the worst in baseball. However, Estrada’s unique profile lends himself well to being one of those pesky guys who makes a living beating their peripherals, and while FIP is a great tool for evaluating most pitchers, no one around here is going to pretend like it’s perfect. For some guys, FIP just misses, and Estrada and his extreme pop-up inducing, high-rise fastball appears to be joining ranks of the Chris Youngs and Jered Weavers where you can essentially throw the fielding independent numbers out the window, because contact management is their game and they’re sticking to it.
And this is why the four-year decline in strikeout rate isn’t as concerning as others might be. In those first couple seasons, Estrada was striking out nearly a quarter of the batters he’s faced. The next two, that rate dropped to the low-20s. This year, it was just 18%. Would Estrada be more appealing if he were still striking out a quarter of all batters? Of course he would. But the reason we usually grow concerned with declining strikeout totals is because many pitchers live and die with the strikeout. Estrada is not one of those guys. For Estrada, the strikeouts are just a bonus.
If the declining strikeout rate gives you any pause, consider the fact that the biggest part of Estrada’s game comes from contact management, and in that area, the trend is headed in the opposite direction. The direction that’s a big plus for Estrada:
Hard contact rate allowed, by year
- 2012: 37%
- 2013: 36%
- 2014: 34%
- 2015: 27%
It’s probably not a coincidence that, as Estrada’s career progressed and he missed fewer and fewer bats, he’s also caught fewer and fewer barrels. As time’s gone on, Estrada has reinvented himself as a pitch-to-contact guy, rather than a swing-and-miss guy, and this season, he finished the year as a top-10 pitcher in generating soft contact.
It isn’t the sexiest signing in the world, but it comes with relatively low risk, and could provide great value if Estrada continues to prove that it was actually 2014 and not 2015 that was the fluke, posting an ERA in the mid-to-high 3s like he has throughout his career as a starter. The Blue Jays were going to have to go get someone like this to fill out the middle part of their rotation anyway, so why not go with the guy with whom you’ve already had success in the past? Estrada comes with one of the game’s best changeups, a penchant for soft contact and low BABIPs that are unmatched by nearly any other pitcher today, and a track record that’s better than you might expect.
August used to cover the Indians for MLB and ohio.com, but now he's here and thinks writing these in the third person is weird. So you can reach me on Twitter @AugustFG_ or e-mail at august.fagerstrom@fangraphs.com.
Early favorite for bargain of the winter? Estrada is a perfect fit with his batted ball profile and that defense behind him.
This post is absolutely insane and 100% in the opposite direction of everything fangraphs has told us to believe in the past.
A guy with declining K rates, who’s BABIP is only low due to having multiple gold glovers behind him, who’s FIPs have gone in reverse from underperforming to being exactly the same, to outperforming them isn’t the trend any pitcher wants to see.
Is he terrible? No. He’s a fine 5th SP. If he throws 180 IP you are happy, but in the back of your mind you have to think it could all fall apart at any moment.
Apart from Pillar, Bautista was terrible in right, revere is barely passable in left and I’d rather never speak of Colabello or Valencia in the OF.
Ben Revere didn’t play for the Jays last year, huh?
Are you trolling or waking up from a coma?
Anyone who watches the Jays knows that Bautista is far from terrible and Valencia was traded in August.
Bautista’s routes ARE terrible. His ability to react to the ball off the bat and read where it’s headed has gotten pretty bad.
HE JUST TOLD YOU TO IGNORE HIS FIP AND EXPLAINED WHY!!!!!!!!!!!
Yeah and the explanation was a crock and in the opposite direction of everything this site has told me about how to gage pitchers.
No worries, I’ll be more than happy to revisit this post in 11 months.
This other article on Fangraphs goes into greater detail about why FIP may not provide the most accurate representation of Estrada’s skill: http://www.fangraphs.com/blogs/marco-estrada-al-contact-manager-of-the-year/
Slow your roll there Forrest, wouldn’t to bust your leg braces or anything like that! Many a great sabermetric mind has criticized FIP for being incomplete as it completely fails on certain types of pitchers as which is what August is getting at here. Fan graphs actually teaches you to look at all the variables when assessing performance not to be rigidly adhere to once statistic. After all the field is in continually developing!
i think fangraphs has been increasingly clear that it’s far from impossible to consistently beat your fip.
it’s very hard to evaluate and predict, of course.
that being said, estrada’s career babip of .260ish is likely a indicator of a true talent in suppressing contact quality.
I’m not sure which gold glovers you’re referring to. Pillar was nominated in CF, and probably would have won if not for the magical Kiermaier. And maybe Ryan Goins will one day win the award, but he wasn’t even a finalist. Martin was nominated but didn’t win, and, as is discussed above, Estrada threw to Navarro for a lot of his starts, not getting the benefit of throwing to Martin.
Also, Pillar was in LF for the first month or two, Goins really only played regularly in the second half, and while he benefited from the fielding prowess of Tulo for two months of the year, that was more than offset by four months of the utterly terrible Jose Reyes.
Lastly, and most importantly, if you want to attribute his success this year to having pitched in front of a ‘gold glove defense’, the question is: so what? This is (probably) the same defense he’ll be pitching in front of for next year, too. His ability to repeat may be better than you think.
Toronto did have the best AL defence in the 2nd-half
@TBJESE: Where do you get fielding splits? I’ve tried Baseball Reference, the MLB sites, etc. but I can’t get fielding splits for 1st and 2nd half for the team? thanks
…actually he’ll be pitching with a better defense behind him because, as other have pointed out, he had a pretty sketchy defense behind him for the first half. Next year will be Tulo all year instead of Reyes, with Goins the guy who steps in when Tulo is out of the lineup, and Travis/Goins at 2B… and not a trace of Collabello in LF unless Satan possesses the Blue Jays decision makers.
Goins benefited from the fielding prowess of Tulo? That might be true if Goins wasn’t already a better fielder than Tulo. Goins would have won a Gold Glove if he had played the entire year at second base. Martin, Tulo, Goins, Pillar, Donaldson & Smoak are all premium defenders that are all Gold Glove worthy candidates.
Smoak is horribly overrated at first, mostly because our broadcasters slobber over him. Also Tulo is declining, that much is obvious.
While I dont think its insane I do agree that this article kind of does run contrary to alot of things people have said in the past.
Essentially we are being told to ignore a larger set of data that says Estrada’s 2015 season was a bit of a fluke and instead to trust a smaller set of data that says it wasnt.
Estrada spent most of his time throwing to Navarro in 2015, not sure how much Russell plays into his decision.
14 games thrown to Martin, 20 to Navarro. It certainly doesn’t hurt to have Martin back there, is the point.
Oops! Thought there was much more of a distinction between starts caught. My bad.
Toward the end of the season the Jays switched to using Navarro for Estrada based on (SSS) results and also for a bunch of Buehrle starts. This was because they didn’t want to keep Thole up for Dickey and wanted the young pitchers to get the benefit of pitching to Martin. That didn’t leave a lot of room for days off for Martin so when Buehrle and Estrada showed good results with Navarro they just ran with that. Shockingly Navarro started showing real improvement behind the plate – especially at controlling the running game.
Yes but he didnt start pitching good until they switched him to Navarro in the second half of the year. Marco was quoted as saying something along the lines of “no disrespect to Martin, but Navarro is a better catcher for me”. I doubt Martin had anything to do with is decision.
While Martin is certainly not a bad catcher to throw to, the 14/20 split is a bit misleading given Estrada threw almost twice as many innings to Navarro:
61.1 IP to Martin / .710 OPS allowed
119.2 IP to Navarro / .589 OPS allowed
If you look at the #s, Estrada had better K:BB and HR/9 numbers with Martin. But why were his #’s better with Navarro? Because he caught all of Estrada’s starts with Tulo/Revere, instead of Reyes/Colabello playing SS/LF.
In case it’s not clear the defence was significantly better in the second half which is why Estrada’s #’s look better with Navarro even though Estrada was better in terms of K/BB/HR with Martin.
That’s a good point about the defense. A taco stand would’ve been more effective in LF than Colabello. Valencia out there was not much better.
And Jose Bautista was playing through an injury in much of the 1st half which forced guys like Carrera into RF. Not a good outfield situation at all for an extreme flyball pitcher like Estrada.
So I guess it might not be a coincidence that his 1st half .252 BABIP fell to .180 in the 2nd half when Jose was healthy and the LF taco stand was removed.
Taco stand removed and replaced with a hyperactive Chihuahua who chases the ball around randomly at incredible speed. Man, the moment Revere slows down even a half-step he is going to be brutal out there because that guy cannot read a hit.
In 2011 he only had 7 starts, but still a 3.70 ERA, 3.87 FIP, and 3.55 xFIP.
In 2012 he had a 3.64 ERA, 3.35 FIP, and 3.48 xFIP.
In 2013 he had a 3.87 ERA, 3.86 FIP, and 3.63 xFIP.
I’m not sure how he can be considered to be someone who always beats his peripherals.
Yeah, I suppose I didn’t do a great job of explaining that. It has more to do with his current approach lending itself to being a peripheral-beating guy. He isn’t the same pitcher he was in his Milwaukee days. I updated that graph to do a little better job explaining myself.
Why did you go back to 2011 with some stats and 2012 with others? He’s less of a FIP-beater if you include 2011. Estrada’s previous good ERA seasons were largely build on good FIPs (he was a great K:BB guy). He’s really only significantly beat his FIP in one season: 2015.
I don’t buy this narrative that Estrada can beat his FIP. As a SP from 2012-2013, Estrada had a 3.81 ERA and 3.59 FIP. 2014 he certainly beat his FIP (4.96 vs 5.73). He beat his FIP in 2014, and if you look at the 3 year sample entering 2015 (2012-2014) it was 4.15 ERA, 4.22 FIP. Even Estrada’s career numbers are just a 0.20 difference between ERA/FIP.
I’ll buy he can induce weak contact, sure. What I don’t accept is that he can beat his FIP to the point where he’s 1 run better by ERA than FIP. Since 1980 among SPs with 1000 innings, there are only 4 pitchers out of 334 that have had a ERA-FIP difference greater than 0.50.
I’ll accept that Estrada is excellent at managing contact and can have a 0.50 difference between ERA/FIP, which based on last year would be around a 3.90 ERA.
If his peripherals decline anymore than they already have, then it’ll be more like high 3, low 4 ERA than your claim of mid to high 3 ERA.
Excellent points. I think Estrada can still beat his FIP if he continues to induce weak contact and has an above average fielding team behind him. But certainly not to the point he did in 2015.
At $13M per year I’m sure the Jays will be happy if he throws 180 innings again with an ERA around 4.
I’m expecting 160-200 innings with a 4.00 ERA, so in essence what I’d expect from Dickey but with fewer innings.
I read it, and your research was very poor in terms of saying stuff like “he makes a living beating his peripherals” when he’s only done the last 2 years, when his previous 3 tell a vastly different story. Fangraphs taught me to always ignore ERA and focus on FIP – why did that suddenly change?
Your entire argument is “he has done well in lowering his hard contact rate, and has a nice changeup” but instead you focus on things like ERA and ignore how he led the NL in HR allowed last year.
His FIP / xFIP / siera were as much outliers this year as his era, in the opposite direction.
His career line as SP is probably the best guideline here – 100era-, 107fip-, 106xfip-.
That’s about what i’d expect, with about 6ip/gs…..but with the acknowledgement that he could vary wildly from this line in either direction, and he’s not the safest of bets.
As a jays fan this seems reasonable, but i’d much rather be using our payroll space to get more top of the rotation type talent.
I dunno, kind of like it. Annual salary less than what they gave Dickey or were paying Buehrle.
And I think they may already have a decent #1 starter who’s not going to cost them much at all over the next few years….
well, now we’re paying $25m for guys I’d be most comfortable calling #4sp.
market value, sure, but not my preferred method of team building. I’d rather pay more money for top guys, and then let kids (hutch/sanchez/osuna) and cheapo vets battle it out for the bottom spots.
Well at least $25m gets you two years on those guys. Beats $22m per one year of Porcello lol.
Yes but that money went to building an offense that is simply on a different level than any other team in baseball, so there is some logic to how they allocated their limited resources.
A caveat to the above however is that next season I actually expect the Red Sox to have an offense that challenges the Jays – in contrast to this season when I just didn’t understand how anyone thought they would.
you think? I don’t see it tbh.
their best hitter would be what? 4th or 5th best on the Jays? jays are a bit deeper too i think.
Steamer wRC+
Bautista 137 —— Betts 125
En’cion 136 ——- Ortiz 123
Donaldson 131 —– Hanley 120
Tulowitzki 110 —- Bogaerts 110
Colabello 107 —– Pedroia 105
Martin 105 ——– Sandoval 104
Travis 104 ——– Bradley 97
Saunders 100 —— Castillo 90
Pillar 97 ——— Vazquez 88
Smoak 98 ———- Shaw 101
Pompey 93 ——— Brentz 93
Goins 67 ———- Holt 88
Thole 70 ———- Swihart 86
Hague 105 ——— Cecchini 83
Revere 88 ——— Rutledge 82
Kawasaki 66 ——— Marrero 71
Jimenez 66 ——– Hanigan 77
Bear in mind that Saunders’ 100 wRC+ will be applied to about 22 AB.
Easier said than done. What top FA SP will sign with Toronto? Going by MLBTR rankings, we can dream about Price but his price for one rotation spot makes no sense with this payroll. Greinke isn’t coming here at any price. Zimmerman may eat innings but is he top of the rotation material? Cueto? Highly variable at a bigger number and I don’t see him coming here. Is he a 1-2? More like a 3. After Cueto Samardzija and Chen are left.
Back to Price as the only top rotation talent that we may have a shot at. Are you good with $25-$30 million over 7? He’ll be good but there will be no trade deadline rabbits in 2016
I think Price at $30m is an ideal fit, yes. Even the likely dead money the last few years of the deal will come after the other contracts expire and when we’re likely rebuilding anyways.
and yeah i wouldn’t want to pay for the other guys you mentioned either, even if they did agree to come here.
After price, the next guy down the list that intrigues me is Iwakuma.
News is that he just signed back with Toronto for 2-years, $26 M. If he was willing to accept that, why not just take the QO and hope you can get $10M next year (which even an outsider would think he has a good shot of doing, and I’m sure he believes in himself more than even we do)
Injury risk.
I would say regression candidate more so.
If injury risk is what he’s worried about, why not seek out something like 4/$34M with $6-10M more in incentives? If he gets hurt in any significant way either of the next two years, he’s going to have trouble making $10M on a new deal at age 34. If he doesn’t get hurt, he’s still not likely to get much over $10M per year, right? I dunno, maybe I’m doing the theoretical math wrong here, but it feels like he’s given the Jays a hometown discount without the years you usually get for those.
Remember, it takes 2 to tango. I’m sure he would have loved to sign for 3 or 4 years, but with a ton of good SP on the market this year, and a draft pick compensation attached to his name, it’s unlikely a team would have gone those extra years.
It’s also possible that he likes pitching in Toronto.
Because you’d rather have 26M guaranteed as opposed to 16.
Navarro actually almost exclusively teamed with Dioner Navarro this season. in fact Navarro even started in all of Estrada’s playoff starts. a growing list of pitchers really love throwing to that man, despite him appearing mediocre in most areas as a defensive catcher.
See one of August’s earlier replies.
yes i was mistaken. however, it was probably a more exclusive pairing in the second half of the season and obviously in the post-season.
I actually exclusively teamed with Dioner Navarro my entire career. I also slugged around .470 from 2012 to 2013, same as Yoenis Cespedes
did you really? now I’m really impressed
i like this deal. Estrada’s declining K/BB numbers appear to have everything to do with a change in approach rather than a decline in stuff. he throws as hard as ever, changeup is as deadly as ever (check out it’s whiff rate inside the strikezone). his heat maps from the last two seasons really tell the story of a guy who is learning to pitch more effectively to his quirky repertoire and take advantage of a league whose hitters have evolved to hit low pitches. his career innings totals are not crazy despite being 32, and he’s been quite healthy in his career too.
One thing I noticed about Estrada this year is his increased cutter usage. Brooks has him at 8.5% cutter usage (+8.4% over last season), and hasn’t been used over 1% of his pitches since 2011 (before he was a consistent starter). At first blush, his cutter doesn’t have much value by itself (-3 runs above average) but all of a sudden the change goes from +6 to +13.2 and the four seam from -11.9 to +9(!). Wondering if using the cutter has an effect on the four seamer, especially with Estrada’s documented release point consistency.
Maybe this is a dumb question and I have just missed the explanation but what makes “inducing soft contact” an actual skill instead of just kind of random luck?
I know guys seem to be able to do it consistently which indicates it isnt luck but I just dont get how a pitcher can influence what type of contact is made in a meaningful enough way that we can project them to keep doing it.
i think guys on either extreme end of the GB/FB spectrum generally have this skill
For Young, it seems to be something about his release plane (he’s 6’10”) and pitching up in the zone that induces a lot of popups, which account for most of his FIP-beating. Weaver is only slightly less tall (6’7″), but I imagine it works the same way, plus he has a wicked changeup like Estrada (who’s only 6′). My hypothesis is that having a solid changeup and and a high-rise fastball not only keeps hitters from generating solid contact off the changeup but also often causes them to swing often under the high fastball, leading to a lot of popups.
Young’s a great example. His four-seamer has a ton of rise to it, and would probably sail over bats if the threw it in the low 90’s. But because he throws it in the 80’s, guys are able to get a piece of it, popping it up.
As Kevin points out, extreme GB or FB pitchers have a skill in almost missing bats, generating weak contact.
Hitting is timing – pitching is disrupting timing. Guys with an extreme difference in velocity or break from similiar looking pitches out of the hand don’t miss many bats , but they do miss barrels – trust me ; i lived by it
Mostly it seems to be about different speeds and different heights on the pitches, but without overpowering stuff. A guy with those “inducing soft contact” skills who can throw 95 with movement would be a big-time strikeout pitcher instead. In fact he would be Clayton Kershaw. Guys with less “stuff” who can do this instead induce weak contact.
This is probably the best explanation I have heard. That actually makes alot of sense.
I’m still not totally sold on the concept of soft contact being a skill but that at least does have some logic to it.
I agree, best explanation of why some guys get weak contact. That’s what I was trying to get at in my comment about Chris Young, but Momus nailed it.
Explaining Estrada is pretty simple. He throws high fastballs that stay up. Batters aren’t used to such a pitch and they hit lazy fly balls. If Estrada maintains this skill, he will continue to beat his FIP. If he starts missing with those fastballs, loses velocity or spin rate, he will get destroyed.
This reminds me a bit of the early days of Hellickson when people looked for a reason to believe he would not regress so heavily. Slightly different as he was younger and it was only 2 years of data, but there was a lot of handwaving to hope that the latest Rays pitcher could sustain beating peripherals.
A lot of “induces soft contact”, “changeup might be a key” and “maybe his BABIP won’t regress as much as expectd” explanations.
A factor from the Jays side has to be throwing a bone to the fan base in the PR mess they are in with AA gone – Memories of his last start against a pretty good contact team in a hitters park are still pretty strong . Deal should sell a few season tix no?
I think there is a middle ground here.
Does Marco Estrada possess the fly ball, Z-Contact% skill-set that allows him to consistently beat his peripherals? Yes.
Does that skill-set mean we should expect a .216 BABIP and 3.13 ERA in the future? Absolutely not.
My ERA Estimator, which takes into account pop-ups (but also normalizes HR rate), pegs his 2015 efforts at 4.54. Better than his 4.93 xFIP? Certainly. I’d also imagine that his K/BB numbers will improve slightly coming off of career worst numbers in 2015.
Does that mean his 3.13 ERA is repeatable? Absolutely not.
Fun article, August. Toronto has a solid analytical staff. They found support to resign Estrada at a moderate (1.5 RA9-WAR/yr) price. Your quick overview touches on several reasons why they might have signed him. Plus, you tweaked a few who believe the FIP model is more real than the game. FIP is, on average, an effective predictive model. It is not the whole story. Otherwise, MLB front offices wouldn’t need to hire PhD algorithmic traders.
Consistently low BABIP? Declining K-rate producing increased pitch efficiency?
Reminds me of Roy Halladay.