Projecting Corey Seager
To say the Dodgers have a surplus of infielders on their roster would be an understatement. Justin Turner, Howie Kendrick, Jimmy Rollins, Chase Utley, Enrique Hernandez, Alex Guerrero and Jose Peraza all have their merits as major leaguers, and all happen to play either second base, third base or shortstop. Not all of them have had the best of seasons this year, but still: that’s seven players capable of playing the infield for just three spots.
Unfortunately for the Dodgers, the injury bug has taken a huge bite out of that infield depth. Kendrick, perhaps the team’s best infielder, has been out for nearly a month with a hamstring strain. Peraza and Hernandez, who’s been on a tear of late, also went down with hamstring strains last week. Suddenly, the Los Angeles didn’t look all that deep in the infield, especially considering how poorly Utley and Rollins have hit this year.
Enter Corey Seager. In an effort to shore up their ailing infield, the Dodgers summoned their top prospect to the big leagues last week, providing them with another option at shortstop and third base. Seager was the consensus top prospect left in the minor leagues at the time of his call-up. He topped just about every outlet’s mid-season prospect list this summer, and his .300/.346/.464 performance since August 1st certainly hasn’t diminished his case. Overall, Seager hit .293/.344/.487 in 125 minor-league games this year, with all but 20 of those games coming at the Triple-A level.
Seager’s a very well-rounded hitter. Both his strikeout rate and isolated power were markedly better than the Pacific Coast League average in Triple-A this year, which played no small role in his posting an above-average wRC+. But while his Triple-A numbers were good across the board, they were a notch or two less than great.
Being less than great isn’t really the norm for Seager. In fact, his four-month stint in Triple-A marked the first time in his pro career that his hitting had come particularly close to league average. At every other stop, he hit much better than that.
Due to this drop-off in production, KATOH isn’t enamored of Seager. Well, it kind of is, but it doesn’t consider him to be one of the very best prospects in baseball, like most human analysts do. KATOH pegs Seager for 8.9 WAR through age 28, which would have landed him 18th overall on KATOH’s most recent list. He graded out slightly less favorably last season with a forecasted WAR of 7.9.
In sum, Seager was an excellent hitter between High-A and Double-A last year, but was a bit far away from the big leagues. This year, he was close to the big leagues, but his offensive performance was merely good. In both cases, the result was a very good, but less than elite, KATOH forecast.
Let’s break out the Mahalanobis machine and generate some comps.
| Rank | Mah | Name | PA thru 28 | WAR thru 28 |
| 1 | 0.68 | Arquimedez Pozo | 80 | 0.0 |
| 2 | 0.82 | Maikel Franco* | 384 | 1.0 |
| 3 | 0.86 | Cole Liniak | 33 | 0.0 |
| 4 | 0.89 | Scott Cooper | 1,850 | 6.9 |
| 5 | 0.92 | Damon Hollins | 15 | 0.0 |
| 6 | 1.11 | Ryan Klesko | 2,779 | 12.8 |
| 7 | 1.19 | Olmedo Saenz | 311 | 1.2 |
| 8 | 1.24 | Darrin Fletcher | 1,569 | 3.5 |
| 9 | 1.31 | Pat Cline | 0 | 0.0 |
| 10 | 1.31 | Richard Hidalgo | 3,011 | 20.0 |
| 11 | 1.34 | Trevor Plouffe | 1,933 | 3.2 |
| 12 | 1.38 | Josh Reddick* | 2,394 | 14.4 |
| 13 | 1.38 | Giovanny Urshela* | 250 | 0.2 |
| 14 | 1.45 | Gary Scott | 198 | 0.0 |
| 15 | 1.46 | Jeff Mathis | 1,360 | 0.0 |
| 16 | 1.58 | Daniel Descalso* | 1,552 | 0.0 |
| 17 | 1.60 | Jose Ortiz | 498 | 0.0 |
| 18 | 1.61 | Conor Gillaspie* | 1,259 | 1.0 |
| 19 | 1.82 | John Mabry | 2,166 | 1.5 |
| 20 | 1.90 | Brendan Harris | 1,639 | 2.4 |
2B,3B and SS marked in yellow
There are a few great hitters in that bunch, but overall, this group is pretty underwhelming. Pozo, Linak and Saenz all fell off the map in their mid-20s despite having very similar age-21 seasons at the Triple-A level. In fairness, though, most of these guys probably lacked the physical tools that Seager brings to the table.
It’s not immediately clear how much, or where, Seager will play for the Dodgers down the stretch and in the postseason. In the five games since he was called up, he’s played two at short, two at third and came in as a pinch-hitter once. But it remains to be seen how Don Mattingly will deploy him once Kendrick, Hernandez and Peraza are back in the fold. However, considering Utley is the only other left-handed hitting infielder on the roster, I’d imagine Seager will play pretty regularly against right-handed pitching these next few weeks.
In any event, there’s no denying that the Dodgers are a better team with Seager on the roster. Even if he’s used purely in a bench role, he’s immediately one of their better pinch-hitters, and is likely better than many of the team’s other infield options. For 2015, 21-year-old Corey Seager will serve as a useful part-time player on a playoff-bound team. But going forward, and possibly as soon as next year, he figures to be a productive regular on the left side of the Dodgers infield.
Chris works in economic development by day, but spends most of his nights thinking about baseball. He writes for Pinstripe Pundits, FanGraphs and The Hardball Times. He's also on the twitter machine: @_chris_mitchell None of the views expressed in his articles reflect those of his daytime employer.
Woah, that list of comps is ugly – Daniel Descalso? Jeff Mathis? Where’s the Tulo comp I was expecting?
It looks like a bunch of guys who made it to AAA at age 21 but only hit around league average. I’d love to see the Mahalonbis comparison to Seager’s 2014 stats – bet a few more promising names show up.
To be fair, he was merely average in AAA. He didn’t exactly do anything to garner good comps from a CPU. lol
“…especially considering how poorly Utley and Rollins have hit this year.”
Since joining the Dodgers, Utley has a 132 wRC+. April 6-May 8, Utley had a wRC+ of -11. May 9 on, he has a 113 wRC+. Likely all driven by terrible BABIP luck for that first month (.079 BABIP in that bad stretch vs .304 BABIP outside of it, career .301 BABIP).
Phillies fan defending Utley here, that’s all.
“.079 BABIP” Ouch. It makes me wonder what a guy off the street would have for a BABIP against major league pitching.
Mostly said guy would have a ton of strikeouts and rarely make contact (or fair contact, at least). But he would certainly manage SOME infield dribblers at least, given enough plate appearances.
Ben Sheets was the worst hitting pitcher I could think of with enough PA to be useful (I ignored Jon Lester). Sheets had 212 Ks to 19 BB in 500+ PA, and still managed a .143 BABIP. He managed 31 singles and 3 doubles, so I’m guessing many/most were infield hits.
If we assume Sheets was a decent hitter somewhere along the way (college? high school? Little League?), I think random guy off the street might have even more Ks, but I wonder if the BABIP would be similar.
So Utley was basically running a BABIP (in an extremely small sample size, obviously) probably worse than anyone commenting here would manage. Now, Utley had 7 walks to just 12 Ks in April, along with 2 home runs and 2 stolen bases, so he was waaaaay more productive than someone random would be. But I think I could run a .080 BABIP, given enough chances.
Utley’s batted-ball quality was almost exactly as good in April as it was in May, just with extremely poor luck. For example, on flies and line drives:
– Utley’s average batted-ball velocity was 86.9 MPH in April, 86.8 MPH in May.
– However, 35% of his FB/LD were in the 80-89 MPH “donut hole” in April, vs. just 18% in May.
– And even on those hit over 90 MPH, Utley only had a .429 SLG in April vs. 1.500 in May (through August the MLB average SLG on FB/LD at 90+ MPH is 1.115)
Pretty silly to only generate results based on one season as opposed to the entire field. Not sure why it would generate fair comps that way. Especially considering even if you average them all out or take them all into consideration you get a much better picture than just his AAA stats
I’m pretty sure the results weren’t based on one season.
I was assuming that both comp lists were based on his comprehensive lines at the time the were generated, i.e. after 2014 and to date in 2015––just that the new projection is revising for his good-but-less-than-great line in AAA this year.
Is that not the case?
Yes, same question here. What is the weight of Seager’s AAA numbers from this season in the KATOH calculation? Most recent seasons seem to be weighted too heavily. Comparing Seager’s KATOH with Lindor, for example, I’d expect to see stronger comps for Seager (at least as strong as Lindor, which included McCutchen) since he’s younger than Lindor and destroyed AA. Lindor had a stronger AAA (118 wRC+), but the overall offensive body of MiLB work by Seager seems to be missing here.
You might be right that most recent seasons are weighted too much, but it does make sense to pay more attention to the 100 games he played at AAA than the 60 games he played at AA.
That and not doing any weighting based on number of PAs. KATOH is pretty silly. It’s a nice amateur project, but I’m not entirely sure why fangraphs is running with it like this.
It’s just another alternative for how to judge projections. Honestly though I’m pretty unimpressed with it’s projections so far in terms of rating players, maybe I’m biased but I trust Kiley’s opinion more than this rating system.
Ask yourself why you trust Kiley’s opinion. If he has a track record of accuracy when it comes to prospects, then that trust makes sense. If it’s because he has worked for front offices and talks to scouts, that means something, yet Top 100 prospect lists are littered with players that had similar backing and yet failed. Most of those 100, in fact. I imagine your trust comes from being more familiar and therefore comfortable with that type of evaluation, plus the general preference for human subjective judgment over statistical analysis (why we ended up with a playoff committee rather than computers even though the computers do a better job than the humans).
There’s no question that Mr. Mitchell’s model is flawed, but that has more to do with the limitations on information available to him than his methodology. When we have PITCHf/x data for every minor league stadium, analysis like this will take a major leap forward in predictive accuracy. But even now it has utility, and Mr. Mitchell has been repeatedly clear that this is intended as a supplement to scouting rather than a replacement for it. You’re supposed to look at the relatively pessimistic results for someone like Corey Seager and take a closer look to see if the scouts might be overrating him. Maybe they are and maybe they aren’t (I obviously believe they are), the point is to utilize this information for an alternate perspective rather than a defining one.
I mean, we all know that the usefulness of a projection system for minor leaguers based ONLY on minor league stats is going to be limited right? This isn’t news.
And, if you think this is in any way an alternative to Kiley’s ranking and not a complement to them, then you are interpreting this super wrongly.
You can’t necessarily say that Seager will do well as a pinch-hitter since coming off the bench to hit is markedly different than being in the lineup. Bench players tend to be more experienced.
Rollins is a switch hitter. Turner splits significantly better against righties than he does lefties. So not necessarily anywhere for Seager to automatically play against a right handed hitter, the question really becomes is he better than Rollins? Donnie loves his vets, so Rollins likely gets the playoff starts this year, but I’m not convinced that is the correct call.
I’m going to be the contrarian yet again and say that I’m not as fond of Corey Seager as the better known prospect analysts, primarily because I don’t like his swing. Watch the lower half of his body in particular: https://youtu.be/Wfh4hq7QLDc?t=60 The way the back leg collapses and the lead foot rolls over means that all his power has to be generated from the upper body. His hips also have a tendency to fly open which is why he’s so pull-happy. Frankly I’m impressed that he makes contact as often as he does considering that his torso and eye level are moving downward rather than being stable leading up to impact. Without serious mechanical adjustments, I don’t see him having the power you would typically associate with a third baseman and I expect him to have difficulty with the outside part of the plate.
Pull-happy?
http://i.imgur.com/5TYvx6i.png
Yes, pull-happy. All but four of Seager’s home runs in 2015 have been to right field, and his ground-balls are overwhelmingly to the right side. He does hit a lot of infield pop-ups to the left side and similar looping fly-outs to left field, which is characteristic of a left-handed hitter whose bat head dips below his hands. You can see what I’m talking about on this more detailed spray chart: http://mlbfarm.com/player.php?player_id=608369
I have to say, I don’t find that spray chart that convincing. Maybe I haven’t looked at enough of them, but it doesn’t look like someone I would describe as “pull happy.” Every hitter tends to get more hits to his pull side, right? So if Seager is totally average in this regard, we would still see a pull tendency.
>Every hitter tends to get more hits to his pull side, right?
Doesn’t seem like it.
http://www.fangraphs.com/spraycharts.aspx?playerid=1101&position=OF&type=battedball
Sorry, don’t know why I said that, meant to say a typical hitter, not every hitter
No?
It basically looks like he hits the ball to all fields. A 20-year-old without plus-plus raw power homers more to pull than oppo? That’s not rare.
Chad, almost all of his home runs have been to right field with none to center and three to left. As noted, the vast majority of his hard-hit balls in the minor leagues were also to the pull side. He pops the ball up to the left side of the field due to the nature of his swing. So yes, he is pull-happy.
If what you say is true, the fact that’s he’s done as well as he has done with all of his mechanical flaws is mighty impressive – it would seem to indicate he has a great deal of raw talent and power. If he makes these adjustments, and I’ve heard nothing of him that indicates he’s a Javier Baez tough-to-work with type, you’d imagine his ceiling would be quite high. Let’s not forget he’s 21. He has plenty of time for adjustments.
Doesn’t Paul goldschmit do that too?
His front foot does roll over, but Goldschmidt has a fairly stiff front leg whereas Seager’s front knee is bent and flying open. Goldschmidt’s torso and eye level remain fairly constant throughout the swing. His head begins and ends in roughly the same place after moving backward a bit during his load. Seager, on the other hand, ends up significantly farther forward at the point of contact than when his swing began.
Jd. Okay thanks for that information. I’m a Dodgers fan. So do you think Seager can fix that easily?
Furthermore, Vin Scully, perhaps one of the all-time baseball Junkies praised Seager. He rarely hypes prospects (save the Clayton Kershaw Sandy analogy). He said that the game slows down for great players and Seager plays with ease.
So here’s to hoping Scully is right
How common is it for a 21-year-old prospect to have a flaw in his swing, which he eventually corrects? Is it the kind of thing where it actually adds somewhat to his projectability, that if he can perform despite the flaw, he could fix the flaw and be even better? Like a tall-skinny pitcher that is projected to fill out, or tighten his mechanics?
Arquimedes Pozo! While these seem like terrible comps, at least I can thank you for reminding me of one of my all-time favorite baseball names.
Chris,
What code in R do you use to calculate the Mahalanobis distance to generate your player comps? I love the additional highlights in the table by the way.
This a good example why scouts or the eye ball test is better for projecting prospects than computer generated algorithms are.
Let’s save the conclusion until we see the results.
a lot of confirmation bias in your comments for this article. it looks like you’ve already come to a conclusion
Why does everything have to be either or? Why are people, still fighting a stats vs. scouting war?
Computer algorithms + scouting > scouting
Computer algorithms + scouting > computer algorithms
luckily we don’t have to choose between them, we can use both.
So, I would guess that Seager is a relatively low #1 overall guy, meaning in a stronger prospect class he might be closer to #5-#10, which he was pre-season. He’s in a class that has had the most MLB promotions, like, ever, so I get that calling him the #1 overall prospect inflates him a bit. Still though, I would think being even an average hitter at 21 in AAA would get more appreciation from KATOH, but I guess not.
His performance at the big league level, while SSS, has been promising already.
Today, I learned what the Mahalanobis distance is. I love fangraphs.
Lets not forget, he was sick for at least week and lost a bunch of weight/muscle while in AAA. His numbers before and after (starting two weeks after) were much stronger. He likely had mono or something.