Going Forward: Where Fans and Numbers Disagree

The other day, I got the idea to compare UZR data against data taken from the results of the Fan Scouting Report. Though there are certain things I’d change about the methodology were I to repeat the study, I’m still content with what I found, and I think it’s interesting to look at where the fans and where the numbers don’t see eye-to-eye. For example: Juan Uribe, when he’s playing in the field. The numbers have loved him. The fans have tolerated him. That’s interesting, even if I don’t know exactly why — yet. It might just have to do with the way Uribe looks, but there could be more to it than that.

Anyhow, once I compared and contrasted fans and numbers in the past, I felt the urge to do a similar sort of thing looking forward. FanGraphs hosts a few different projection systems; among them are the Fan projections and the Steamer projections. Soon, we’ll also have full ZiPS data, but we don’t have that uploaded yet. But we can make do with those two. Many fan opinions are in, and all the Steamer evaluations are in. Which players and pitchers do the fans like more? Which players and pitchers do the fans like less? Is there anything we can learn from what we find?

A few things, first. Instead of WAR, I’m going to compare only wOBA for position players and FIP for pitchers. For these purposes I’m not real interested in things like base-running or defense. I included position players the fans project for at least 300 plate appearances, and pitchers the fans project for at least 60 innings. This gives me 286 total position players and 190 total pitchers. Unsurprisingly, so far the fans are slightly optimistic: For the same players, the average wOBA is four points higher, and the average FIP is 17 points lower. Maybe that’s just Steamer being pessimistic. It’s probably the fans being optimistic.

I know we’re dealing with some small sample sizes with fan projections. I know posting this could very well change those very fan projections, because the projections are still live and you’re still free to enter your own. And I encourage you to enter your own! That’s how we get bigger sample sizes! But let’s check where things stand as of this this writing. You’re going to see four tables, three of them include 10 players and one of them includes a dozen because of a three-way tie. On to the data!

Position players fans like more than Steamer

Name Fans Steamer Difference
Avisail Garcia 0.359 0.320 0.039
Darin Ruf 0.364 0.328 0.036
Xander Bogaerts 0.359 0.325 0.034
Grady Sizemore 0.333 0.303 0.030
Nick Franklin 0.335 0.307 0.028
Carlos Quentin 0.371 0.344 0.027
Anthony Rendon 0.349 0.322 0.027
Bryce Harper 0.389 0.363 0.026
Chase Headley 0.354 0.328 0.026
Jordy Mercer 0.334 0.309 0.025

For the most part, it’s a pretty young table, with Quentin and Sizemore being the old guys and with Quentin and Sizemore being a surprisingly identical 31. It’s no surprise fans can be a little more bullish with youth, and from the looks of things, White Sox fans are pretty excited about Avisail Garcia, who had an encouraging debut in Chicago after arriving in a trade. Steamer sees him repeating last year’s 97 wRC+, but the fans are more fond of his power ability and figure he can sustain an elevated BABIP. I’ll side more with Steamer here but this post isn’t supposed to contain much in the way of my own editorial content.

Ruf is coming off an interesting 2013 in which he hit better in the majors than he did in Triple-A. Fans believe more in what he did in the bigs. Bogaerts has almost everyone flipping their lids. Sizemore is interesting here, and someone you don’t see is Franklin Gutierrez, who missed because the fans (rightly) didn’t project enough playing time. But the fans have been a lot higher on both Sizemore and Guti, suggesting that people are optimistic about the frequently injured. It’s a warm and easy feeling.

You can wade through the rest. I’ll note the fans like Bryce Harper more than Steamer, but Steamer likes Mike Trout a little more than the fans. It seems the fans insist on continuing the debate we’ve all been having for two years. I’ll still pick Trout. That’s me.

Position players fans like less than Steamer

Name Fans Steamer Difference
Henry Urrutia 0.305 0.329 -0.024
Daniel Descalso 0.272 0.296 -0.024
Brian Roberts 0.283 0.306 -0.023
Lonnie Chisenhall 0.299 0.322 -0.023
J.P. Arencibia 0.278 0.300 -0.022
Joaquin Arias 0.271 0.291 -0.020
Jonny Gomes 0.321 0.338 -0.017
Mitch Moreland 0.315 0.332 -0.017
Carlos Ruiz 0.315 0.331 -0.016
Omar Infante 0.311 0.326 -0.015
Gerardo Parra 0.311 0.326 -0.015
Wilin Rosario 0.342 0.357 -0.015

This is a table of less-interesting players, and it’s of interest to see Henry Urrutia tied for first — or last — since you probably know next to nothing about him. He’s currently in line to bat a lot for the Baltimore Orioles! The fans would prefer that not happen. He’s tied with Daniel Descalso, and what you don’t see is the fans are also more down on Pete Kozma than Steamer. The impression I get is Cardinals fans really, really don’t care for those guys and must be more than ecstatic to have Jhonny Peralta on the payroll, despite the issues.

Chisenhall comes as a bit of a surprise, since it would be easy to figure the fans would like him more, as a talented 25-year-old. But maybe Cleveland Indians fans welcome the Carlos Santana/third-base experiment. I couldn’t be less interested in talking about Joaquin Arias, although just Thursday I was reminded he was part of the Alex Rodriguez trade. The Rangers selected him instead of Robinson Cano. In 2004, Arias was part of a trade for Alex Rodriguez. In 2010, Arias was traded for Jeff Francoeur. He hasn’t had the career he could’ve.

With Arencibia, people can’t un-see the 2013 disaster. Fans think Moreland will get worse. Infante’s projected to go back to his 2012 self. Just missing the bottom of the table: Justin Morneau.

Pitchers the fans like more than Steamer

Name Fans Steamer Difference
Sean Doolittle 2.42 3.54 -1.12
Greg Holland 1.73 2.64 -0.91
Darren O’Day 3.11 3.99 -0.88
Sean Marshall 2.35 3.16 -0.81
Archie Bradley 3.76 4.53 -0.77
Sergio Romo 2.49 3.26 -0.77
Joe Nathan 2.72 3.48 -0.76
Danny Farquhar 2.66 3.35 -0.69
Carlos Martinez 2.99 3.66 -0.67
Jered Weaver 3.64 4.30 -0.66

I apologize for the blend of starters and relievers, but it is what it is. Doolittle blows away the competition. And the fans think he’ll blow away the competition… again. Fans have him projected for basically his career FIP. Steamer thinks he’ll walk more guys and cough up more dingers. That’ll send the FIP skyrocketing in a hurry, all the way up to the mid-3.00s, which still is not bad.

After Doolittle you find a few more relievers. I’ve just been reminded that Marshall has a 2.11 FIP since 2010. Then you get to you first starter in Archie Bradley, who the fans think is just about big-league ready. Steamer’s more cautious, and therefore more supportive of Arizona signing Bronson Arroyo.

The fans see Romo getting some strikeouts back. They’re not as concerned as Steamer with Nathan’s age. Farquhar makes sense here because Steamer doesn’t know as well as the fans do that Farquhar somewhat recently picked up a new pitch that changed him as a reliever. The difference with Martinez basically comes down to projected walk rate. With Weaver, it’s all about home-run rate. Fans and Steamer see the same strikeouts and walks. Steamer sees the worst home-run rate of Weaver’s career. Seems unlikely and overaggressive.

Pitchers whom fans like less than Steamer

Name Fans Steamer Difference
Ryan Dempster 4.46 3.78 0.68
Josh Beckett 4.23 3.72 0.51
Bronson Arroyo 4.62 4.18 0.44
Brett Anderson 3.97 3.56 0.41
Jake Arrieta 4.34 4.02 0.32
Ervin Santana 4.02 3.71 0.31
Jeff Locke 4.23 3.93 0.30
Mike Leake 4.19 3.89 0.30
John Lackey 3.98 3.69 0.29
Jason Hammel 4.21 3.94 0.27

And now we close with a handful of guys who are Boston Red Sox players or who have been Red Sox players. It’s a table including both the younger and older, with some fairly interesting names in it. Sox fans haven’t been able to forget Dempster’s dinger problems, so they see those continuing. The same goes for Beckett, who also has some health questions. The fans think Arroyo will repeat last season, whereas Steamer figures he’ll experience some positive dinger regression. Yet, oddly, Steamer projects Arroyo for a higher ERA than the fans, so go ahead and make sense of that one.

Anderson’s shifting to what might still be an easier league, but a far more difficult ballpark environment. The fans like Arrieta’s chances of improving, but not enough to actually be good. It’s interesting to see the fans basically project Santana to be a league-average pitcher. If front offices agree with that, it’s no wonder he’s still available and difficult to justify signing what with the draft-pick sacrifice.

Locke and Leake make for a fun pair to say out loud. Throw in Lackey and it’s an enjoyable trio, and Steamer likes Lackey’s chances of dinger suppression more than the fans. In the end, we get Hammel. So far the fans aren’t loving his odds of getting back to what he was in his breakthrough 2012 season with the Orioles. That’s understandable, since 2013 happened, and since front offices around the league are uncertain about Hammel’s elbow health.

Overall, I wouldn’t say there’s a ton we can learn from this exercise. At the least, we could use some bigger samples of fan projections for each player so we can talk about the fan projections with a greater degree of confidence. As the sample sizes grow, the differences between the fans and Steamer take on greater significance. But it’s still interesting to see where people disagree the most, and it will be interesting to revisit in a month or so to see how things might have changed over time. I encourage you to submit some fan projections for players on your favorite teams. If you agree with what Steamer or ZiPS, that’s fine. That matters. If you disagree, that’s interesting. That can be something to investigate.



Print This Post



Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.


Sort by:   newest | oldest | most voted
Franco
Guest
Franco
2 years 4 months ago

If nothing else, we now know Sox fans hate their pitchers.

jb
Guest
jb
2 years 4 months ago

we didn’t know that already?

EthanB
Member
EthanB
2 years 4 months ago

Dempster is interesting. If he’s a reliever, which he projects to be to start the year, the fans look pretty pessimistic there. Otherwise, the fans look about right, and ZiPS agrees – it has him at a 4.44 FIP next year.

Billy
Guest
Billy
2 years 4 months ago

Also (and I know I’m stating the obvious here), I’m sure that a factor is what type of fans contribute to these projections. Since this is fangraphs, it’s reasonable to imagine they may be a bit better than the average fan when it comes to this sort of thing, and stick to principles popular among the SABR-y crowd. How would other types of fans affect these projections? How would reasonably savvy, more-than-casual-but-not-expert fans move the needle? How about casual fans? How about intelligent, passionate, old-school fans who are not very SABR inclined? How about stupid, passionate, old-school fans who are not very SABR inclined? If fans tend to do entries for players they feel strongly about, how would these change if they were told to just do whichever players are in front of them?

I’d be interested in all of these questions.

RC
Guest
RC
2 years 4 months ago

“it’s reasonable to imagine they may be a bit better than the average fan when it comes to this sort of thing, and stick to principles popular among the SABR-y crowd.”

Meh, I’m not so sure. SABR people are equally as likely to get their hate on as anyone else. There are still a lot of Red Sox fans out there who think Lackey is some sort of bridge-troll who has never been able to pitch, despite the fact that his career has been a very consistent 3.5 or so ERA aside from the year he blew his arm out.

tz
Guest
tz
2 years 4 months ago

Maybe if we compare the rankings here to preliminary fantasy baseball draft rankings, we can see part of how the highly interested but non SABR-y crowd may view things differently.

MDL
Member
MDL
2 years 4 months ago

The Rangers selected [Joaquin Arias] instead of Robinson Cano.

Ouch.

d_i
Member
Member
d_i
2 years 4 months ago

Funny, I had just exported the fan projections to start my valuations for my draft and Avisail Garcia stuck out like sore thumb so I tweaked his. I see he just has 6 submissions so I’m guessing one or two of them are from his dad or something.

Schide
Member
Schide
2 years 4 months ago

I guess I’m one of those 6 but my projection has him with a .329 wOBA, so I have no idea why (most of) the other 5 would be so much higher.

GiantPain
Member
GiantPain
2 years 4 months ago

Heh. 1700 words, culminating in “Overall, I wouldn’t say there’s a ton we can learn from this exercise. ”

I love fangraphs

Anon
Guest
Anon
2 years 4 months ago

The overabundance of useless words is a Jeff Sullivan effect. If he writes an article, I only look at the charts, graphs, gifs, etc. to avoid wasting my time. Some people like his style, but I’m not one of them.

tz
Guest
tz
2 years 4 months ago

cool story bro

tz
Guest
tz
2 years 4 months ago

Suggest that “chance of dinger suppression” become its own stat (CODS?)

fly eli and tony plush
Guest
fly eli and tony plush
2 years 4 months ago

Or a great fantasy team name

Sir William Fauntleroy
Guest
Sir William Fauntleroy
2 years 4 months ago

Egad, sire! Mayhaps I shall compose thee a piece on said topic! Indeed, ‘twil be a codspiece, like none other!

ben
Guest
ben
2 years 4 months ago

Interesting. I agree with the numbers for the hitters but with the fans for the pitchers.

Sports Enthusiast
Guest
2 years 4 months ago

I think one thing that pops out is that fans expect underperforming players to play at their underperforming averages and for injured players and younger players to play at their career averages. That seems pretty rational, I wonder if its true?

Eric Feczko
Guest
Eric Feczko
2 years 4 months ago

These are interesting comparisons, however, what’s missing here is the variance.

An implication here that the fans express the same level of confidence in their projections for every player. The problem is that they are not, and that the variance in the fans’ reporting may be higher for some players than others. For example, the 0.359 for Xander Bogaerts is not necessarily the same as the 0.359 for Chase Utley. Though the central tendency of the fans is the same for both players, fans may be more certain that Chase will produce a 0.359 wOBA than they are that Xander will produce a 0.359 wOBA.

This uncertainty can be interpreted from the variance measures. Similarly, projection systems also have a degree of uncertainty that varies from player to player. It would be interesting to see whether these player-by-player differences in projections are really that significantly different between Steamer and the fans. It is possible that both Steamer and the fans are expressing really similar numbers, but have slightly different distributions of confidence for players (resulting in different central tendencies).

As a side note, it would also be interesting to look at Oliver’s numbers. By my eye Oliver’s projections look far off (especially for playing time and defense…less so for rate statistics) than either steamer or the fans.

M W
Guest
M W
2 years 4 months ago

I would have set the standards at 250 PA’s and 80 IP. That should get you about 300 offensive players and approximately 150 starting pitchers. Relief pitchers bounce around too much year to year to really be that concerned with in such an analysis IMO.

I used 250 and 80 because they are approximately half of the required totals to qualify for the batting and ERA titles.

M W
Guest
M W
2 years 4 months ago

I’d also wait and re-run this prior to opening day, in the hopes that more fans make projections.

wpDiscuz