Were 2018 Stats Projections Correct? – Matt Harvey

In 2018, the Mets and Matt Harvey finally split. After another rocky start to the season, Matt Harvey was traded to the Reds for Mesoraco and now is heading out west to start for the Angels.

Harvey’s 2017 was terrible, 6.70 ERA, 1.694 WHIP over 92.7 innings. With that computers between Baseball Prospectus, ZiPS, Steamer, ESPN, MLB.com and Baseball Reference put out this combined average projection line. Let’s compare it to Harvey’s actual 2018 line:

2018 Average Projection: 115.8 IP, 4.66 ERA, 4.54 FIP, 1.402 WHIP, 7.47 K/9
2018 Mets: 27.0 IP, 7.00 ERA, 5.68 FIP, 1.556 WHIP, 6.7 K/9
2018 Reds: 128.0 IP, 4.50 ERA, 4.33 FIP, 1.250 WHIP, 7.8 K/9
2018 Total: 155.0 IP, 4.94 ERA, 4.57 FIP, 1.303 WHIP. 7.6 K/9

It’s impossible to look at Harvey’s numbers and not ruminate on how bad he was with the Mets, and how much he turned it around with the Reds. Granted, he wasn’t an ace with the Reds but his 4.50 ERA with them is his best season posting since 2015.

It is also hard to claim that any particular outlet was successful in predicting Harvey’s season. They all had him way under for season innings (Steamer was the closest at 135, Baseball reference the next at 115). The closest ERA was ESPN at 4.88, but that feels wrong because Harvey really had two different seasons based on what team he was on.

We knew last year it was going to be hard to predict Harvey’s year. He once an amazing ace. Had a horrific injury. That’s a recipe to be hard to predict.

Posted in Main Page | Leave a comment

Were 2018 Stat Projections Correct? – Jerry Blevins

Every spring I collect several projections sources and average them together to do a meta-projection. Last year I pulled projections from Baseball Prospectus, ZiPS, Steamer, ESPN, and Baseball Reference. Jerry Blevins had an off year last year, which is why he remained a Met throughout the season and why he was not part of the sell-off in July. He had a stellar 2017 where he posted a 2.94 ERA, 3.12 FIP and 1.367 WHIP. Projections were wise to see that he would regress off those numbers but they didn’t see how far he would regress.

2018 Average Projection: 50.86 IP, 3.64 ERA, 3.44 FIP, 1.33 WHIP, 10.4 K/9
2018 Actual: 42.2 IP, 4.85 ERA, 4.96 FIP, 1.359 WHIP, 8.6 K/9

Jerry just didn’t enjoy the same dominating performance he had in 2017 where he got hitters to strike out a whopping rate of 12.4 K/9. Computers saw him regressing down 2 points there and in reality he ended up doubling that regression.

Baseball Prospectus was the best at seeing he his regression where they saw him posting a 4.41 ERA, the only projection program giving him a 4+ ERA (the next closest was Steamer at 3.76). It’s worth noting that in their 2019 Projection, they having posting a 3.98 ERA, so better than his last projection despite his performance being worse than their last projection. This confirms what a lot of us believe to be true that this past season for Jerry was a blip not a trend.

Jerry right now is still a free agent which is a product of how slow this free agent market is and how his 2018 performance knocked him down to a third tier bullpen pitcher. He’ll still find a major league contract and with his personality I would be against him coming back to the Mets, even if it doesn’t make sense.

Posted in Main Page | Leave a comment

Were 2018 Stat Projections Correct? – Jeurys Familia

Every spring I collect several projections sources and average them together to do a meta-projection. Last year I pulled projections from Baseball Prospectus, ZiPS, Steamer, ESPN, and Baseball Reference. Jeurys Familia starter the season as the Mets closer, was traded at the deadline to the Athletics and last month found himself back as a Met for the upcoming season on a friendly 3 year, 30 million dollar deal.

Familia will always be unfairly criticized by small portion of the Mets fan base for his performance in the 2015 World Series, which was fine (outside of one game, most of the blown games were due to factors outside of his control), we aren’t going to get into that here.

You can find all of last year’s projections here:

2018 Average Projection: 59 IP, 3.31 ERA, 3.255 FIP, 1.29 WHIP, 8.9 K/9
2018 Mets: 40.2 IP, 2.88 ERA, 2.55 FIP, 1.230 WHIP, 9.5 K/9
2018 Total: 72.0 IP, 3.13 ERA, 2.65 FIP, 1.222 WHIP, 10.4 K/9

Familia performed better than his projections, and performed better as a Met than he did as an Athletic (despite a higher K/9 as an Athletic). This could be a combination of Familia playing with a new team for the first time ever, being traded to be “the man” and increase of usage in games that actually matter.

Shockingly, for a second time in this series, MLB.com is the most correct for predicting his season numbers. They saw him having a 3.18 ERA, 1.24 WHIP (the lowest predicted WHIP). The lowest predicted ERA and FIP was from ZiPS who saw a 3.00 ERA and a 2.86 FIP (no WHIP predicted), these numbers were close to Familia’s numbers with the Mets.

Familia joins Lugo, Matz and Gsellman who all performed their projections last year, lets see if this trend continues tomorrow!

Posted in Main Page | Leave a comment

Were 2018 Stat Projections Correct? – Robert Gsellman

Every spring I collect several projections sources and average them together to do a meta-projection. Last year I pulled projections from Baseball Prospectus, ZiPS, Steamer, ESPN, and Baseball Reference. At this point last year, were still trying to figure out who Robert Gsellman would be. In 2016 he was the other half of the pair of rookies (Seth Lugo) that led the Mets to the Wild Card. In 2017, the injuries in the rotation forced a lot of pressure on Gsellman and he didn’t deliver.

Similar to Lugo, there was debate as to what Gsellman’s role would be in 2018. Pen? Starter? Long Man? Set-up guy? (Which wasn’t even considered until we saw how effective Lugo was out of the pen). He ended 2017 logging 119.7 innings with a 5.19 ERA, 4.89 FIP, and a 1.504 WHIP.

Computers gave him the following average projection and he responded with the following stats:

2018 Average Projection: 104.4 IP, 4.37 ERA, 4.345 FIP, 1.383 WHIP, 7.0 K/9
2018 Actual: 80.0 IP, 4.28 ERA, 3.95 FIP, 1.300 WHIP, 7.9 K/9

Gsellman did well in the pen and significantly improved off of his 2017 numbers, although the computer projections also saw this improvement in Gsellman’s future, even if they weren’t sure if he was going to start or not. On the whole, Gsellman performed better than his average projected line as well.

No individual projection from last year was quite right on Gsellman though. They either had his ERA considerably lower (MLB.com at 4.07, Baseball Prospectus the closest at 4.2) or higher (ESPN 4.87, Baseball Reference 4.5). None of them had his WHIP so low (probably due to starter/relief question) with MLB.com going the lowest with 1.34, ESPN the highest at 1.48.

With the Mets new additions of Diaz, Familia next season, Gsellman will probably see a drop of time in high leverage outings. I’m curious to see how the 2019 projections handle the new Mets bullpen with playing time, and how they see Gsellman continues his journey.

Posted in Main Page | Leave a comment

Were 2018 Stat Projections Correct? – Steven Matz

Every spring I collect several projections sources and average them together to do a meta-projection. Last year I pulled projections from Baseball Prospectus, ZiPS, Steamer, ESPN, and Baseball Reference. After spending a few days looking at fringe relievers and players not on the team any more, today we finally look at a player who will have a significant roll on the 2019 squad, Steven Matz.

Knowing the season the Mets had last year, our article last year with projections seems incredibly dated. We wondered about Matz after his 2017 year that saw him struggle start to start, missing consistency completely when the Mets rotation was seemingly full. At the start of spring last year the rotation was a deGrom, Syndergaard with Harvey, Wheeler, Matz, Lugo, Gsellman and Vargas fighting for the last three spots. Matz seemed like one of the odd men out, even with Lugo and Gsellman going to the pen. (It was a long time ago, but remember Matz ended 2017 logging only 66.2 innings with a 6.08 ERA and a 5.05 FIP).

2018 Average Projection: 109.6 IP, 4.17 ERA, 3.95 FIP, 1.328 WHIP
2018 Actual: 154 IP, 3.97 ERA, 4.62 FIP, 1.247 WHIP

Holding FIP off to the side, Matz outperformed his projections majorly in the amount of work he saw and his ERA. His WHIP was considerably lower than projected. MLB.com was the closest in predicting this. They had Matz at 130 innings and a 3.95 ERA. This is also the first time I think MLB.com was the closest in any projection. Steamer had his WHIP the closest at 1.3 (all of the WHIP were between 1.3 and 1.36).

If we are looking at trends though, all of the projections saw that Matz would have a bounce back year, even if they didn’t get the innings correct. Matz ended up bouncing back better than projected. Let’s see what 2019 brings!

Posted in Main Page | Leave a comment

Were 2018 Stat Projections Correct? – Seth Lugo

Every spring I collect several projections sources and average them together to do a meta-projection. Last year I pulled projections from Baseball Prospectus, ZiPS, Steamer, ESPN, and Baseball Reference. In full transparency, I’m writing this article about Seth Lugo on the morning of 12/30 and I probably won’t have any time to edit it before it posts at 8:00 am during the New Year. As of yesterday, there were rumors flying that the Astros are looking at Seth Lugo in trade talks with the Mets. Right now when I hear that the Astros are looking at a pitcher, I want to hold onto to that pitcher because I trust their eyes more than most.

When I wrote the 2018 stats projection article for Lugo last year, the season just started and Lugo was already in the pen. At that point it was unclear if he was going to stay in the pen. Most of us at that point knew Lugo as the rookie arm that led the Mets to the 2016 Wild Card game and the ace starting pitcher for team Puerto Rico in the 2017 World Baseball Classic.

Lugo had a tremendously successful year last year. However, I believe most projection programs were still trying to crunch his numbers as a player coming off an injury trying to fight his way into a crowded rotation that had a lot of injuries. That led to 3 projections having him at 90-95 innings, 1 at 71, 1 at 109 and then ZiPS was all in on Lugo the starter having him log 199.7 innings. The computers spat out this average line:

2018 Average Projection: 110.1 IP, 4.52 ERA, 4.52 FIP, 1.34 WHIP, 7.35 K/9
2018 Actual: 101.1 IP, 2.66 ERA, 3.17 FIP, 1.076 WHIP, 9.1 K/9

Lugo did make 5 starts last year bumping his innings total to 101.1. He had a 3.91 ERA in 23.0 innings as a starter and 2.30 as a reliever. His WHIP as a starter was 1.304 (as a reliever it was 1.009)

Lugo outperformed his projections in every way. The only projection that was close to correct was Lugo’s WHIP as a starter, which I think is how most projections saw him anyway.

Additional detail, when he started and relieved, he was not typical. His starts were shorter than normal, probably leading to his a lower ERA. His relief appearances often went over an inning, leading to a higher inning total.

The man with a ridiculous RPM curveball did great last year and I’m curious to see how projection programs handle him for 2019.

Posted in Main Page | Leave a comment

Were 2018 Stat Projections Correct – Rafael Montero

Last year I debated writing the Stats projection article for Rafael Montero because by the time I eventually got around to it in April, Montero already was sidelined by Tommy John surgery for the entire 2018 season.

Ultimately, Hansel Robles and Rafael Montero last year in January were tied together. They had snapshots with the major league team (the latter having more than the former) and were both kept on the roster in while other players (Josh Smoker) were cut. For Montero, the Mets needed starting pitching depth which they just didn’t have at the time. Montero in 2017 wasn’t 100% terrible. His terribleness would rotate. He would start, be great, give us hope and then come back out next week and flop.

2018 Projections: 90.6 IP, 5.19 ERA, 4.71 FIP, 1.564 WHIP, 72.5 K, 35.7 BB

That looks hella generous for playing time, but with a year passing since we’ve seen Montero on the mound, it’s easy to forget that he logged 119 innings in 2017. The 2018 Mets were bad, but that’s how bad the 2017 Mets were, 119 Montero innings!

Montero signed a minor league contract with the Rangers this off-season thus officially ending his time with the Mets. Good luck in Arlington!

Posted in Main Page | Leave a comment

Were 2018 Stat Projections Correct – AJ Ramos

Every spring I collect several projections sources and average them together to do a meta-projection. Last year I pulled projections from Baseball Prospectus, ZiPS, Steamer, ESPN, and Baseball Reference. AJ Ramos projections will be hampered by injuries, which is a common theme for the Mets as some projections were able to correctly predict time loss due to injury, some did not, and some never attempted it (sometimes by design).

AJ Ramos struggled in every possible way for the Mets last year. We noted last year that his inclusion on the roster was a curious case. While the Mets started trading to rebuild in in 2017, they acquired Ramos, which went against the pattern of trading bullpen pieces to get prospects. Ramos inclusion gave Mets fans hope that the team would be trying to win immediately in 2018 otherwise, why get Ramos? But alas, between ineffectiveness and injuries 2018 was a year to forget for Ramos. Lets take a look at was projected and what actually transpired:

2018 Projected: 60 IP, 3.73 ERA, 3.86 FIP, 1.351 WHIP, 68.8 K, 32 BB
2018 Actual: 19.2 IP, 6.41 ERA, 5.19 ERA, 1.627 WHIP, 22 K, 15 BB

He would have had a solid year if he produced the projected line even though his projected line was worse than his career averages in FIP (3.41), WHIP (1.270) and ERA (3.07). What the projections were able to see was that he was going to regress, but they saw a normal regression, probably because they didn’t see the injury.

Baseball Prospectus was the only who saw a tremendous regression in ERA (4.02).

What’s surprising is his final 2017 line wasn’t particularly great with a 4.74 ERA, 4.47 FIP and a 1.632 WHIP which was eerily close to his 2018 WHIP. AJ’s 2017 actual stats was a better projection of his regression than any of the computer programs.

Posted in Main Page | Leave a comment

Were 2018 Stat Projections Correct? – Hansel Robles

Every spring I collect several projections sources and average them together to do a meta-projection. Last year I pulled projections from Baseball Prospectus, ZiPS, Steamer, ESPN, and Baseball Reference. I’m also a completionist, so if I wrote a stat projection article last year, I have to review it at the end of the year for accuracy. Accuracy and projections haven’t gone the Mets way the last two years due to the amount of trades and roster changes so some articles feel more awkward than others, like this one. So were the 2018 Stat Projections on Hansel Robles correct?

Hansel’s rookie season was great but his 2017 was not so the Mets were unsure what they would get from the relief pitcher who in his last season and a half in Queens spent more time pointing up at the sky for fly balls becoming homers than locking down the game. Below we compare the average projected line to his actually stats from 2018. You can find the line by line projections here.

2018 Projections: 56.26 IP, 4.51 ERA, 4.525 FIP, 1.39 WHIP, 59 K, 28.5 BB
2018 Mets: 19.2 IP, 5.03 ERA, 7.13 FIP, 1.576 WHIP, 23 K, 10 BB
2018 Angels: 36.1 IP, 2.97 ERA, 3.22 FIP, 1.294 WHIP, 36 K, 15 BB
2018 Total: 56.0 IP, 3.70 ERA, 4.59 FIP, 1.393 WHIP, 59 K, 25 BB

First thing that popped out to me – Robles had a tremendous season with the Angels. He returned to be the type of pitcher that he showed he could be in his rookie season.

Next, the projections were eerily close to his final, combined line. He pitched almost exactly the same amount of innings, his WHIP was insanely close to average projected line and he ended up with the exact amount of K’s.

Not a single individual projection program had Robles with an ERA lower than 4.28 (Baseball Reference) so everyone missed that. Steamer was the most accurate on WHIP at 1.39 (they also were close on FIP at 4.65).

Hansel clearly benefited from being traded mid-year. Can he keep that up into next year? Will his new, ex-Met teammate Matt Harvey rebound the same way? Will any of us see it since the games are on the West Coast and as Mets fans we all, including myself, suffer from a tremendous East Coast bias?

Posted in Main Page | Leave a comment

Were the 2018 Projections Correct? – Paul Sewald

Every spring I collect several projections sources and average them together to do a meta-projection. Last year I pulled projections from Baseball Prospectus, ZiPS, Steamer, ESPN, and Baseball Reference. Were the 2018 projections correct for Paul Sewald?

Before the 2017 season, Sewald was a non-roster invitee to Spring Training. It was wildly predicted by most Mets pundits that he would end up on the Mets roster at some point in the season. Before 2018, it was unclear how many games he would pitch in as his existence on the roster was dependent on the health of other pitchers. The 2018 Mets, similar to the 2017 Mets, struggled in that department, so Sewald saw some time.

You can see the original projections here.

2018 Average Projection: 56.9 IP, 4.33 ERA, 3.995 FIP, 1.309 WHIP, 59.4 K, 20.75 BB, 9.376 K/9

2018 Actual Stats: 56.3 IP, 6.07 ERA, 4.23 FIP, 1.509 WHIP, 58 K, 23 BB, 9.3 K/9

Outside of the ERA and WHIP, the average projectile line did remarkably well. Sewald got almost the exact amount innings that was predicted, his FIP almost matched his projected ERA, he got nearly the amount of K’s and BB’s as well.

Mainly, his 2018 numbers tell the story of someone who was slightly unlucky (much higher ERA than FIP) but found himself in situations to get hurt by luck due to his WHIP.

For a pitcher like Sewald though, I really look at WHIP to see how he’s doing and no one predicted his WHIP to be this high, the closest was ESPN and Baseball Prosectus who both had it at 1.34 (and Baseball Prospectus also predicted the highest ERA at 4.75 so they get the edge as the most accurate overall).

Tomorrow we’ll look at former Met, Hansel Robles.

Posted in Main Page | Leave a comment