2018 Positional Power Rankings: Catcher

Earlier today, Jeff Sullivan re-introduced this year’s positional power rankings. Hopefully, you are familiar with our methods for producing these rankings, but if not, a brief reminder: all 30 teams are ranked based on projected WAR from our depth chart projections. With those in place, our writers endeavor to provide additional commentary, some of it useful and some less so. We begin this year’s rankings with catchers.

The catcher rankings start, as they tend to do, with Buster Posey and the Giants. After a gap, the rankings cover the next 28 teams on a slow steady decline, and then there’s the Phillies. While Gary Sanchez and Willson Contreras represent a youthful contingent, the top of the list is still dominated by veterans, with Posey, Jonathan Lucroy, Brian McCann, Salvador Perez, Yadier Molina, and Russell Martin all placing their teams in the top 10. The last-place Phillies could make a big move up this list if former prospect Jorge Alfaro can take a leap forward.

One aspect of these rankings somewhat unique to the catching position is the gap between the quantified and the less quantifiable. The power rankings seen here don’t take into account framing, which can make a big difference in a player’s value. Calling a game and interacting with pitchers is nearly impossible to put into a run value of any sort. That can make ranking catchers more subjective than other positions where value is more easily captured. Where possible, these aspects are mentioned in the rankings. Let’s begin with a face that should be very familiar.

We hoped you liked reading 2018 Positional Power Rankings: Catcher by Craig Edwards!

Please support FanGraphs by becoming a member. We publish thousands of articles a year, host multiple podcasts, and have an ever growing database of baseball stats.

FanGraphs does not have a paywall. With your membership, we can continue to offer the content you've come to rely on and add to our unique baseball coverage.

Support FanGraphs




Craig Edwards can be found on twitter @craigjedwards.

newest oldest most voted
mikejunt
Member
Member
mikejunt

I love this series but thvis particular edition deeply suffers from fg not using a defensive stat that includes framing. With framing data, the Dodgers duo is the first or second most valuable position in baseball, competing with Angels CF.

It’s hard to take any analysis that there’s a team in baseball that would rather have the As catchers than the Dodgers catchers for 2018 seriously.

Rex Manning Day
Member
Rex Manning Day

I admit that I don’t know much about framing metrics, but if there are framing metrics that say that Barnes + Grandal are even approximately as valuable as Mike Trout, then I am confident in saying that those metrics are not trustworthy.

Rex Manning Day
Member
Rex Manning Day

I will add, though, that if there are metrics that give Barnes + Grandal almost 4.5 WAR from their framing alone, and those metrics are accurate, then that is perhaps the single best argument for robo-umps that I can imagine. If it’s possible to add over 4 wins for your team just by manipulating ball and strike calls, then the mechanism for calling balls and strikes is broken.

mikejunt
Member
Member
mikejunt

Trout only projects at about 6 or 7 wins

Grandal and Barnes were worth 8 by WARP last year, but that was with Barnes being Buster Posey for 260 PA.

However top end framing is generally 15 to 25 runs above average and last year Barnes was 2nd on rate basis and Grandal 5th in MLB. In 2016 Gransal was first (and as you know Travis gave him an MVP vote downballot)

Projection wise with mean regression there’s probably 1.5 to 2.5 wins missing for that duo due to composite framing value. They are really good at it.

Those numbers arent out of line with what with previous framing highs and lows.

I guess this is where the discussion devolves into the 40% of FG commenters who find framings existence offensive and the other 40% who just want the most important aspect of catcher defense reflected in our value stat.

Rex Manning Day
Member
Rex Manning Day

There are four projection systems available on FG, and they project Mike Trout for 8.2, 8.4, 8.8, and 7.8 wins, or an average of 8.3 WAR.

Giving Barnes and Grandal 2.5 WAR from framing bumps their combined projection up to 5.7 WAR. Which is very good indeed, but it’s not within spitting distance of Trout. And I suspect that it doesn’t put them in 2nd place behind him, either, though they’d be up there.

As I said, I don’t know a whole heck of a lot about framing metrics, and it seems like this is more an issue of underestimating Mike Trout than overestimating the Dodgers’ framing. I have no doubt that the Dodgers catchers are excellent framers. But framing alone can’t make them as valuable as Mike Trout, and if it can, then something is broken.

mikejunt
Member
Member
mikejunt

Last year it was 34.3 combined framing runs vs mid-6s war from trout missing 6 weeks.

Rex Manning Day
Member
Rex Manning Day

Ok, you got me. Framing might be enough to make two otherwise-merely-good catchers equivalent to an injured Mike Trout.

mikejunt
Member
Member
mikejunt

Well really its evidence of why comparing win values across different WARs is far from effective. WARP evaluates Trout differently than Fwar; perhaps a fwar that includes framing would conclusively say, yes, it is less, perhaps it would do the opposite. A UZR formulation for framing might estimate its value higher or lower than BP does.

Just jamming BP’s FRAA stats onto catchers is going to be pretty inexact; WARP projects Trout only about 6.3 wins, making the comparison more accurate; perhaps UZR would project both groups to 8+, perhaps it would not.

That is my original point; fWAR needs to incorporate something so potentially valuable so we can compare apples to apples, because comparing apples to oranges in another valuation system is going to be extraordinarily inaccurate and unfair to one party or another.

WARP had C-LAD as the most valuable team position in baseball last year (ATL-C, COL-CF and CIN-1B the closest competitors). It projects it to be the most valuable position in 2018.

Their formulation is going to be different than what fWAR provides, because they’re not the same metric; you can see this by comparing even very good players across metrics (Giancarlo Stanton, for instance, was +1 win higher in WARP than fWAR due to different positional standards).

I would rather do an apples to apples comparison than only evaluate catchers with a metric I spend less time with and have to try to ‘convert’ it back to fWAR.

johansantana17
Member
johansantana17

If that were the case, Trout would still be more valuable, especially on a team (like the Dodgers) with a 26th man who is better than that of other teams’, because Barnes and Grandal take up two roster spots and Trout only takes up one.

scooter262
Member
Member
scooter262

Add me to the robot-umps camp. I hate the artificial manipulation of the strike zone by good framers. And similarly, I hate the fact that good pitches can be called balls just because the catcher didn’t frame it well.

IMO, a pitch thrown over the plate at the correct height is a strike, and one that isn’t is a ball. The sooner framing becomes moot the better, as far as I’m concerned.

jcutiger
Member
jcutiger

Maybe Roboumps will help get better hitters behind the plate.

rp_savage
Member
rp_savage

I think the A’s over the Dodgers is misleading here. Looking at Lucroy’s page, it seems to me that his ZIPS projection is still assuming he’d be a Rockie, but then in the depth charts, the two are averaged and adjusted for half his games at the Coliseum, which makes him look like a better hitter than he actually is and explains the higher projection

London Yank
Member
London Yank

If framing really is that important then you also shouldn’t be able to take the forthcoming pitching edition seriously since it will be based on FIP.

mikejunt
Member
Member
mikejunt

FIP is still more predictive than ra/9. Whether it’s 5% less accurate or not isn’t that significant.

FIP isn’t a perfect predictor, it’s just the most consistent one we have. It’s still got fairly wide error bars, just less wide than ra/9 or era.

And the variance isn’t 15 to 20 runs

London Yank
Member
London Yank

My point is that the wins that are missing from catcher framing are currently distributed to the pitchers because pitchers are given all the credit and all the blame for their walks and strikeouts with FIP. If you want to add WAR to the catchers you have to take it away from the pitchers.

mikejunt
Member
Member
mikejunt

This isn’t actually true because we use FIP and not RA/9, and not a hybrid stat that also considers contact quality like DRA.

Lets take a 1-1 pitch. There’s significantly different outcomes league-wide for 2-1 counts and 1-2 counts. Those differences aren’t driven exclusively by the changes in walk and K rates; though they are a significant factor.

There’s also the factor of players swinging at different pitches and making different quality of contact with worse outcomes (2 strike approach) to avoid the strikeout at 1-2 that doesn’t come into play at 2-1. Players have less 2 strike approach now than they did 20 or 30 years ago, but they still do.

Since FIP is based purely on TTO statistics, any change to batted-ball outcomes by count (which are considerable) will not be reflected in FIP-based WAR or projections.

The overall value of strikes that BP calculates to determine the value of framing accounts for overall PA results, not just the changes in BB and K rates.

While a stolen strike framed for strike 3 is maybe the most obvious benefit the framing, it’s more frequently pitching from ahead in the count that provides the cumulative benefit of superior framing over a very long season.

To point at the best catchers by CSAA, Austin Hedges, Tyler Flowers and Barnes, those catchers got about 3% more strikes than average. They each saw thousands of pitches.

It may only be 1-2 strikes per game, but over the 25,000 pitches thrown in a season, it puts their pitching staff in more advantageous positions fairly often, and only the K% increase and BB% reduction will be reflected in a FIP based WAR.

We use FIP because its more predictive, but it doesn’t encompass all of value, which we see in pitchers who consistently beat their FIP (Kyle Hendricks types) who do have a skill for consistently generating suboptimal contact.

It should be viewed as no coincidence that the Dodgers lead MLB in contact quality metrics for pitchers – this is the effect of having had good pitchers and great framing at the same time (contrast the other two leaders in framing value, who did so for sub-par pitching staffs in San Diego and Atlanta).

Its not just the Dodgers who benefit: Gary Sanchez is a well above-average framer and should probably project as a 5 win player at catcher, not 3.6.

Jonathan Lucroy, on the other hand, has fallen off a cliff from good to average to awful over the last few seasons (well documented by several writers on this website, including Sawchick and Sullivan) and was a sub-replacement level catcher by WARP last year due to being 20 runs below average as a framer last year,

There’s a reason no one wanted to sign Jonathan Lucroy, and it’s that teams think he’s a replacement level player. If he was as good as this projection, the Nats would have signed him months ago. They didn’t because he’s not clearly more valuable than Wieters given his substantial decline as a receiver.

And Wieters is uncontroversially awful.

Jetsy Extrano
Member
Jetsy Extrano

Does Fangraphs have a principled belief that “all framing is zero” is the most useful approximation currently available? Or is it more a matter of “haven’t built one we like enough”?

If you think a given model needs regression, you can regress it. If you don’t trust a single one, you can blend two. You can take the closer-to-zero estimate from each model if you like. Just seems like something’s got to be better than zero.

mikejunt
Member
Member
mikejunt

It’s just that UZR doesn’t currently include framing value and UZR is fangraphs chosen defensive metric.

It unfortunately makes it hard to use fwar to assess catchers without consulting other measurements to augment understanding.

mikejunt
Member
Member
mikejunt

I should have been more clear here, but Fangraphs doesn’t create its own metrics. Fangraphs hosts metrics and compiles a WAR from them. They aren’t the developers of UZR, DRS, FRAA or whatever that other framing metric is called. They host the ones people agree to let them host (eg, BP keeps its proprietary metrics in-house). If Fangraphs is going to have a framing metric, one of their hosted metrics is going to have to develop that data. FG and its writers don’t do that.

The alternative is hosting that data from someone who already produces it, but those companies keep their work proprietary, because it’s a lot of work.

Its unfortunate, though, to have multiple evaluation systems because it makes discussing players infinitely more complex, and particularly troublesome when there’s just one position where you have that issue.

Some people prefer to use Baseball reference for pitchers because they prefer RA/9 to FIP, though, so there are other situations. FG has the RA/9 WAR equivalent to help those folks out, but no access to any framing runs data to incorporate it.

It isn’t as simple as FG doesn’t care about framing; they have to pay for it or get someone who makes free metrics to develop it, and its a ton of data processing.

Jetsy Extrano
Member
Jetsy Extrano

Fangraphs has made a point of adding particular metrics to their stable, I think mostly by licensing them for money or viability (have they commissioned any new development?)

Would somebody be willing to put together a decent metric in exchange for being linked as “the Fangraphs framing metric”?

Scott
Member
Scott

I think it’s a little simplistic to just say we can add framing data and the Dodgers’ catchers are the 1st or 2nd most valuable position in baseball.

First, they probably wouldn’t even be the best catching unit, because Buster Posey exists and he would get credit for framing as well. While Posey took a statistical dip in his framing numbers last year, I’d bet the projections would put him at a similar level to Grandal/Barnes, or a little below. Posey has the advantage in offense and the other aspects of catcher defense (blocking, fielding, controlling the running game, etc.).

Second, if you’re going to include framing, then you also need to take into account the other important aspect of catching that isn’t included in these rankings: game-calling. Harry Pavlidis did research a few years ago on game-calling: http://www.espn.com/mlb/story/_/id/12880518/mlb-best-game-caller-dodgers-catcher-aj-ellis and the fact that Jeff Mathis is able to keep getting jobs despite his offense is another data point that shows the importance of game-calling. I remember reading an article from Baseball Prospectus about AJ Ellis’s game-calling skills, and also about how Grandal rated as one of the worst game-callers, but I can’t find it now. Anyone know what article I’m referencing?

Anyway, I think it’s disingenuous to say you can slap on some framing stats and say the Dodgers have maybe the most valuable position group in baseball. That isn’t taking the other catchers into account, and it’s not taking the whole picture of catcher defense into account.

mikejunt
Member
Member
mikejunt

Posey has been a league average framer for 2 seasons running and I don’t think he projects at the top of the league anymore. Posey’s also giving away tons of value by playing 1st so frequently.

Their competition for most valuable C in baseball are Sanchez (above average, great offense) and Flowers (great framing, good offense).

The core reason that the Barnes/Grandal combination has been so ridiculously valuable is that they manage to give all their catcher PAs to good catchers. There’s a reason that catchers are typically projected to 450 PAs instead of 600; catchers get a ton of days off, and the other guys usually suck (and that is indeed the case for both NYY and SFG: the guys playing when Sanchez and Posey don’t play are muuuch worse by both offensive and defensive metrics).

I’d love to see someone put a run value on game calling, though its even more difficult than framing given the differences in pitcher personalities (Grandal’s game calling may effect a rookie pitcher a lot, Clayton Kershaw not much (since he’s making most of the decisions about pitch sequencing) and Kenley Jansen literally not at all (no game calling when you throw the same pitch).

In any case, you might think that is a simplistic analysis, but it’s also exactly what WARP both says for 2017 and projects for 2018.

The big advantage of Dodgers C over the other catcher positions is giving all the PAs to guys who are good; Sanchez and Posey have to be so good they make up for 150 PAs being wasted on Romine and Hundley, who aren’t very good on either side of the game. And in fact, their biggest competition in 2017 by WARP was ATL, which had Flowers/Suzuki to the same effect (Suzuki had a bounceback year and was an above-average framer, and Flowers was the best framer in MLB during an offensive breakout season).

Sanchez is a legit 5 win player though, it’s not only the Dodgers duo not getting their due here.

Scott
Member
Scott

Where are you getting your framing numbers? I’m looking here: https://legacy.baseballprospectus.com/card/58548/buster-posey and Posey’s framing numbers look like this:

2012: 19.1
2013: 18
2014: 23.1
2015: 14.3
2016: 29.3
2017: 5.5

I’m not sure how that looks like a league average framer for 2 seasons running.

mikejunt
Member
Member
mikejunt

No, you’re correct, in my brain 2016 is already 2 years ago even though we haven’t had a 2018 season yet.

Given the path of guys who steeply decline like Lucroy I would not expect a bounceback to same levels, but nothing between 0-15 would surprise me from Posey in 2018.

Scott
Member
Scott

Okay, that makes sense. I’m not sure if there’s enough of a data set to show that catcher framing falls off a cliff so quickly, like with Lucroy. With Posey, my guess would be that he generates a lot of his framing runs by working with pitchers with plus command like Bumgarner and Cueto, and with both of them injured/not as effective last season, it hurt Posey’s numbers. I’m not sure if they have framing numbers by pitcher, but that might be slicing the data too thinly.

mikejunt
Member
Member
mikejunt

BP does determine CSAA for both pitchers and catchers in order to isolate the catchers contribution and only credit them for FRAA.

The pitchers’ contribution is a factor in DRA as it’s indicative of command.

Its true that staff+catchers are a closed ecosystem but teams have a lot of pitchers these days and its pretty easy to identify low-command types on the staff and see the differences between Clayton Kershaw and Pedro Baez, for example.

If teams only used 8-9 pitchers this would be more of a problem than it is now when we have a loot of pitchers on each team and more opportunity to see what happens overall.

frazier15n
Member
frazier15n

Yeah the A’s bit essentially broke this piece for me. Did not read past that. Because life has enough lies.