Your Ballot for the 2014 Player of the Year
A couple of weeks ago, we announced the creation of the FanGraphs Player of the Year Award, which we created to recognize the single best performance in a given season, regardless of the position the player plays or the league that he plays in. Just one award for one player, with everyone eligible, and all players measured on the strength of their own performance.
In that post, we announced that the voting panel would include 11 members of the FanGraphs staff, myself included. Today, we’re amending that slightly, as I am recusing myself from the voting panel this year, as our plan to reveal the winner — and all of our balllots — on October 20th means that my participation would reveal some of the order of the NL MVP ballot which I cast over the weekend. In order to avoid that issue, I’m surrendering my vote for the Player of the Year Award this season. To replace me on the panel, we have chosen to give you guys a ballot instead.
Before I link to the ballot, though, I need to explain the grading process that we have attached to the system. While the ballot does contain a normal 1-10 ranking system, we have added an additional evaluation layer, with a 0-100 grade also being recorded for every player you list on your ballot. The grade is actually the more important number than the ranking on the ballot, though both will be tabulated, and the player’s average ranking will be used as a tie-breaker, so position on the ballot matters as well. But the grade is the more important evaluation, so let me walk through how the grading scale works.
The grade represents your estimate of the percentage of years that the player’s performance in that season would result in his placement at the top of your ballot. The baseline, or a grade of 100, is the kind of season that would result in a Player of the Year Award in almost any year throughout baseball history. These kinds of grades should be reserved for types of performances that go down in baseball history as the best seasons of all-time: Babe Ruth in his prime, Ted Williams hitting .400, or turn-of-the-century Pedro Martinez, for example. These performances would result in a Player of the Year award in nearly 100% of the seasons in baseball history.
In more normal seasons, the best performance might belong to a player whose performance would only win the Player of the Year Award 60 to 80 percent of the time; it’s still a great performance, but would get beaten out by a player having a legendary year. It is perfectly acceptable for the top player to get a grade that is not particularly close to 100, as this scale is not designed to follow the traditional grading scale used in education. It is an estimate of the number of of times, out of 100, you’d expect the player to be the best player in the league.
By the time you get to the bottom of your ballot, you may placing very low grades on players, perhaps even as low as zero. If a player is legitimately the 10th best player in a season, it is perfectly rational to believe that it would be extremely rare for his performance to result in a Player of the Year award, and a grade of 0, 5, or 10 does not reflect that he had a poor season; you are recognizing his strong performance by simply including him on the list in the first place. Do not be afraid to use the entire scale if it reflects your belief about the strength of the player’s performance.
The grade will allow us to not only see the differences in your opinion beyond what a numerical order will show, as well as allowing for comparisons across years. In some years, the player who finished second or third in the voting may very well have a higher aggregate grade than a player who wins in a different year. Consider your grades carefully for all 10 positions on the ballot, as the system is designed to recognize even those who happened to have fantastic seasons in a case where someone just happened to be even better in that year.
So, in effect, we are asking you to not only measure a player against his peers this season, but against players throughout baseball history. In some years, we might even have a winner with a grade of 30 or 40, if it’s a particularly weak crop of elite players and no one has a truly outstanding year. In other years, we might have multiple strong candidates, with several players receiving very high grades, and we can acknowledge the greatness of even the runner-up by utilizing this system.
We also understand that this is an estimate, and it’s an unfamiliar scale, so perhaps we all won’t be very good at estimating probabilities to begin with. It’s possible that our scales could change over the years, and maybe we’ll become more lenient or more harsh with our grades as time goes on. We don’t expect this system to be perfect, especially in year one, and we’re not claiming that this is going to be the definitive way to compare players across different seasons. We just think it’s worth measuring the extra data point of how strongly you feel about a player’s performance in that season, and we think this is an interesting way to gather that information.
The winner will be the player with the highest aggregate grade across all of the ballots, with your ballots counting as one collective vote. In the case of a tie, we will use the highest aggregate ranking, so if you’re putting the same grade on several players — a completely valid thing to do — use the order of the ballot to determine your preference among players with the same grade. And remember that you are free to use the entire scale; a grade of 0 is not a knock against a player’s performance, as you are rewarding that player simply by placing him on the ballot, putting him ahead of the hundreds of players who didn’t make your top 10.
We will tabulate all the ballots and announce the full results of the FanGraphs Player of the Year Award on Monday, October 20th, the day between the League Championship Series and the World Series. We hope you enjoy participating, and we hope you put some real thought into your ballot. Feedback is always appreciated, especially as we learn about how this process will work together, so feel free and provide comments below, and we will take those into account.
And now, for the ballot. Most of the text above is included on the page, as well as a chart which shows the frequency of seasons per year in which a certain WAR threshold is achieved. Of course, this is not the WAR of the Year award, and you do not have to consider WAR in any way in your voting if you do not want to. My theoretical ballot had many disagreements with our published WAR, for instance, and we strongly encourage you to base your decision on more than just decimal differences in WAR. The chart is there as a guide, however, to give you an idea of how often a certain level of accomplishment is reached, so you can have context for how rare your evaluation of a player’s performance is in the historical context.
Alright, have at it. The FanGraphs Player of the Year Award ballot can be found here. Make your vote count.
Dave is the Managing Editor of FanGraphs.


Congratulations Eric Sogard
Glad I made the cut. At least I have a fighting chance now.
I’m looking forward to Munenori Kawasaki’s acceptance speech and his post-ceremony interview with Eno.
This is the first, and likely only, time that I’ve mentally compared Dave Cameron to Dan Lebatard.
Will we be able to see our personal past votes?
That Royals-A’s game deserves like 18 well written posts. So far, like 2 crappy ones.
Boy you showed em, champ. Feel free to write one up in the community section.
Or, barring that, leaving some constructive criticism.
No? Just wanted to bitch? Ok.
He’s actually right though. There hasn’t been nearly enough about that game.
And “do it yourself or shut up” is just a lazy cop out. Last I checked I’m not a professional baseball writer and he probably isn’t either.
Trout of the year
Campaign Time!
Let’s talk Raul Ibanez. He obviously was not the best player this year. If he was, he wouldn’t have been glued to the bench last night. But dude deserves some recognition.
I’m not asking you to vote him first. This is not a campaign of sabotage. I wouldn’t want to sabotage a really good idea like this Fangraphs Award. This is a campaign of love for a guy who shouldn’t be playing at age 42.
I’m asking you to vote him in tenth, with a rating of zero. That last vote for Adrian Beltre or David Price will likely not be all that relevant.
Raul Ibanez. 2014.
I’m going to do that right now.
Support Jaack in his goal. But if you want to vote Beltre for 9th place, that’s ok too.
I still can’t believe Wendy Thurm was included on the panel to vote for this award. Rants against pink gear, insight into TV ratings, and multiple posts on fan engagement….is that really why we come to FanGraphs?? NOOOOO.
Wendy does awesome work.
Totally agreed. And I think I know what you’re really getting at. Only men should be allowed to vote. No matter how analytical or insightful Wendy’s posts are, she’s still a woman and just shouldn’t be given a ballot.
Seriously, you’re complaining about Wendy Thurm (who writes extremely intelligent, in-depth, and well-written articles) getting a vote, but you have no problem with Carson Cistulli (fuck that guy) wasting his votes on Charlie Blackmon or Sam Fuld or Bruce Chen or some 28-year-old rookie no one’s ever heard of?
I’m sure Wendy’s gender has nothing to do with your complaint.
Hey, now. Cistulli called Corey Kluber before anybody else. And last night Brandon Finnegan.
Lets not go nuts praising Wendy, I find her posts often lacking in critical thought. She generally presents things (like finances) at face value without challenging the legitimacy of unofficial information.
Definitely. Don’t want her getting too uppity.
I personally have no problem with Thurm getting a vote. I do have a problem with a reasonable critique (see Hurdles, Pirates) of her being straw manned as sexist.
Take note, cass
Do you think we could get along without the f-bombs? Probably most people don’t care, but for some reason I do. Just a request.
Wendy’s work is easily some of the best on the site. Easily.
yeah hearing women talk is not why we come to mangraphs amirite bros??
I love Wendy’s articles about tv conctracts/territories, and legal issues. I do wonder how much I would listen to her speak though…not that I know what her voice sounds like
I want to upvote this because it is what I’d say sarcastically, but I’m hesitant to do do so because people might think I wasn’t being sarcastic.
I was definitely sarcastic, so go ahead
Watch out, she might just write a research-less article about how your favorite franchise is racist.
Everyone knows the braves franchise is racist. That chop thing is just freaking unbearable as well as being racist as hell. It takes no research to figure that out. Just a bit of common sense.
Also, I’ve met Wendy. Her voice is pretty normal.
Dude, YOU get a vote, and you’re complaining about an actually FG contributor getting a vote?
Accidentally left Kluber off my ballot. Any way to edit it?
Clear your cookies and vote again? If the website is smart enough not to fall for that, just vote from a different computer/device.
Democracy.
I think it would be nice to have 99 and 1 (or similar) available as option for grades. As it stands, I only gave 5 guys non-zeros.
ya i kind of felt like my grades were high…quick someone do a statistical analysis on the likelihood that the average 5th best player in any year would be the 1st in another year.
For this I’m using rWAR as a way to rank players. I’m not saying that’s the best or even OK way to rank players but it’s a simple shortcut that will serve to get some data. I used rWAR rather than fWAR because that’s what I had quick access to.
In the last 100 years, the median top rWAR for the season has been 10.3. Also, in the last 100 years, there has never been a fifth best rWAR that high.
If you took a random year’s 5th best rWAR and compared it to a random (other) year’s top rWAR there is a 2% chance that the 5th best would be greater. (205 chances out of 9900)
This year was a very low year for the top rWAR players. In fact, at 8, that beats only one other year’s top rWAR (7.7 in 1981).
If you take the scale literally and go only by rWAR to decide the rank (which, again, I do not endorse), then Trout/Kershaw would each would get a 2 out of 100 and everyone else would get a 0.
Long story short, in the average year, point distribution would follow this general distribution for players 1-10 respectively:
50,25,10,5,2,1,0,0,0,0
Can NotGraphs post the silliest joke ballots? This is an earnest request.
A termite walks into a pub and asks, “Is the bar tender here?”
An alley cat crawls into a hole in the ground and asks “Any molasses left?”
Seconded
Whilst moving Clayton Kershaw, I noticed that it is possible to list the same player multiple times. I expect at least one person to submit a ballot of ten Jeff Francoeurs.
Additionally, now I want to see Patrick Dubuque try his experiment again, but with teams of 25 Jeff Francoeurs, 25 Domonic Browns, 25 Aaron Crows, and 25 Jason Marquises.
Where’s the love??!??
Breaking RBI Baseball, brought to you by Jon Bois
Hard to vote for Frenchy when Kendrys Morales put up a .271 wOBA in 400 PA’s primarily as a DH. Plus, there are multiple Kendrys to vote for.
i only game three guys non-zero grades. if i look at bbref’s WAR, which is higher than fangraphs’ this season, then kershaw’s 8.0 WAR is the lowest WAR total for a player to lead hte majors in WAR since mike schmidt (7.7) in 1981, and the second lowest going back all the way to cy young’s disappointing 7.3 WAR to lead the majors in 1900.
WAR isn’t everything, sure, but he performancs this year were… muted to say the least. 7 more starts for kershaw, of course, and itd be a different story.
Or if Trout’s defense rated higher of if Victor Martinez was Andrelton Simmons at shortstop or if Andrelton Simmons was Victor Martinez at the plate. Or if early 90s Bonds’ defense and base running combined with early 00s Bonds’ offense and with Nolan Ryan’s badassery.
i just have no idea what you mean to say here…
it simply seems grossly unrelated to my comment/observation that the performances this season were historically underwhelming.
i guess maybe you’re latching onto my final line about kershaw being perhaps historically relevant if only he’d had 7 more starts. and while, sure, that presupposes that he would be able to maintain the same level of performance as he had in his actual 27 game starts as he would in his imagined 7 additional starts, that’s hardly less of a stretch than to create centaurs and griffens.
a full season of tulo might also have been historically relevant and relevant to the conversation. a minotaur of juan lagares and josh donaldson, with jacob degrom’s hair, far less so.
This seasons also exists… had your top three not played, than your 4th would get the award, leading me to believe that there should be at least a couple more non-zero values on your ballot.
It says I have to log in to vote. I didn’t know Fangraphs even allowed people to register their username. Is this new?
I’ve had mine for a year. Does not affect the way your comments appear, however; only your ability to submit things to Community Research or probably other features I am too drunk to understand.
The only annoying thing about this is when you want to post a comment under a joke pseudonym (Ruben Amaron Jr.) you have to log out. I, too, have this name registered, but I normally comment without logging in.
It’s kinda funny when I post under a joke pseudonym and then forget to switch back to my regular username.
For instance.
don’t forget about fangraphs: the game!
everyone forgets about fangraphs: the game… even fangraphs forgot about their short football experiment
The feature I like best you get for logging is having a custom default leaderboard on player pages. Now, when I go to a pitcher’s page, instead of seeing W, L, K/9, and BB/9, I see K%, BB%, FIP-, ERA-, stuff I am more interested in.
Love the idea. However, much more helpful than the “number of players per season with WAR >=” table would be a “number of seasons with at least one player with WAR >=” table.
I feel like this 0-100 grades are asking for trouble. I know Dave explained that they generally correspond to the percentage of years in which that player’s performance would translate to a player of the year, but that’s a little hard to conceive of. I know I don’t have a very good sense of what a typical winner would be like. I honestly have no idea if Trout this year would be POY in 80% of years or 20%.
And, as many have said, it’s probably the right thing to do to give players 5-10 on a ballot a 0. It’s pretty rare, I imagine, that even the 5th best player one year is ever the best player of another year. I can’t imagine the intention of the scoring system is to only give score to the top three or four candidates.
If I hadn’t really read and thought about the rating, I’d probably have just gone by feel and given Trout a 90 and Kershaw an 80, probably based on what percentage of the best possible performance they gave, though by the discussed methodology that would be quite wrong. I get that fangraphs audience isn’t precisely the general public but it’s not not the general public either. I’m just imagining how thoroughly the audience of a site like ESPN would fail to use this rating correctly.
If only there was some tool that could help you compare seasons over time…
Snark aside, I get what you’re saying and I really hope that most of the voters take the time to read what Dave means by 0-100.
This might be more helpful. I looked at WAR for this year’s leaders: Trout 7.8, Kershaw 7.3, Kluber 7.2 (essentially the same number and same initials for those two). I then looked at historical leaders (for batters and pitchers separately) in a couple arbitrary year ranges to see where these numbers generally fell. Knowing WAR isn’t perfect, I gave each player a error band of roughly 0.2 WAR. Keep in mind this was just to give me a very rough reference historically.
From 1935-2014: MT ranked from 175-215, CK ranked 100-150
From 2000-2014: MT ranked 30-50, CK ranked 15-25
From 2006-2014: MT ranked 15-25, CK ranked 4-12
I then considered what I know about their seasons, which stats are more important to me than others, and how I felt these players compared to other players I voted for. In the end, I felt that there was a low probability of the seasons that these players had in 2014 winning a player of the year title in other seasons, around 10% chance. Finally, I weighted this season and the more recent seasons more heavily than past seasons, as that would be the best indicator of how these 2014 performances would hold up moving forward, so I bumped my #1 player (Kershaw) up to 25%.
It is not perfect, but I gave it some thought.
That sounds like the right way to do it. I basically used the top-30 WAR leaderboard as my nominees, went down the list and ordered it by personal bias, such as weighting higher peak performance over accumulated value, then somewhat haphazardly gave Trout a 30, Kershaw a 10 and Kluber a 5. Done.
Well as the barely-cogent masses we’re definitely going to have a skewed, less-educated view of relative historical value, but since our collective vote only counts as one ballot overall I don’t think we’ll screw everything up too badly. I imagine collectively we’ll be at least as objective as Communist Cistulli after his third glass of French wine.
I agree. It’s not that it is a terrible concept, but there is no way it will actually be followed. In half the years, the top guy should get a 50 or worse. I gave Kershaw a 50 as my top guy, but now I’m thinking that was too high.
Yeah, part of my objection is that, particularly this year, there are certainly fewer than 10 players who have any argument for any year’s player of the year award. There were 3 players this year with more than 7 WAR, and I don’t know if there is any year when the best player in the league had less than 7 WAR. So, basically that means that the 4-10 guys of the final ballot should be tied for 0 points, or people voted wrong.
I saw 8 players that could’ve been hypothetical Players of the Year, but gave no player more than a 35 grade. I couldn’t give Trout a high grade knowing that he wouldn’t have come close to beating his own last two seasons.
I had Betances and Wade Davis in there because, for relievers, they had amazing years, even if not as valuable as a good starter or position player. And for that reason I ranked them higher than, say, Stanton, even though he is obviously more valuable, and even though I think he’d be the best player in more years (which I reflected in the point grade). Went Kershaw, Trout, Kluber top 3.
Here is a bit of problem that I have. Are we suppose to take into consideration just overall performance or peformance in historical context?
One of the problems I’m running into is I look at Jonathan Lucroy, who posted a 6.3 fWAR, good for the 30th best catching season on all time. But we almost know he was worth 23.7 runs because of framing. If you give him full credit for that, he comes out at 8.7, which is a top 5 season for a catcher. But don’t have framing data for Gary Carter or Johnny Bench, so I am unsure how it fits in historical context.
Clayton Kershaw had one of the best seasons in recent memory on a per inning level, but he didn’t crack over 200IP, only good for 168th on RA9-WAR since 1945. Also it was an impressive season but this is no turn of the century Pedro because the offense is simply so depressed.
I feel like poor Lucroy gets lost in the mix. If you split the difference on framing, he is in the Trout/Kershaw tier. I wouldn’t mind giving him credit for the fact that it is impossible for a catcher to play 162 games and “intangible” such as game calling that could make him Fangraphs Player of the Year.
A top 30 all time catching season is still crazy historical. Only other players in the past 20 years are Pudge,Mauer,Posey, and Piazza(struggle to call him a catcher)
I clearly didn’t read the instructions, but sort of like my grading system better anyway. I graded it as “in an average year, what percentage of players would this person be better than”. Thus, my lowest grade was a 70. This is completely different than the instructions, but makes a lot more sense to me.
For the purposes of this exercise though you can probably argue that no one should have a higher grade than a 70, and even that is optimistic. (35-40? Even lower? Feels like most years there’s someone that puts up much better numbers than this one)
I agree, that is why I like my grading scale better. If I would have done it correctly, I probably would have given a 40 as the top score and multiple zeros.
I think your objection to the scoring system is reasonable.
Your system makes sense. Another alternative is to imagine the best possible season (the kind that Dave describes as being worthy of a 100) and give the player a score that represents what percentage of the level a player achieved. So Trout this year might be about 75% of the best possible season, and maybe 85-90% in 2012 and 2013. It’s a little more difficult to quantify than your method.
On this day, we are all Dave Cameron.
As an optimist with very little net perspective (read, I’m young) on baseball history, I probably scored everyone too highly. So I adjusted everyone down a bunch, and they’re probably all scored a bit too highly.
With enough voters, though, I do think those issues should sort of fall out, and the ballot should still end up being reasonable.
I think the scale ends up with a bit of a problem as well. My top 3(Lucroy,Trout, and Kershaw) were only at 35% or so.
According to the bottom of the directions on the actual ballot page, there are 2 or 3 players EACH YEAR (on average) better than Trout (highest WAR this year).
So are there any math people here who can tell me, if you average three 8 WAR players per year, how likely it is that a single year has no 8 WAR players, given a sample of, say, 100 years?
It’s probably less than 10% right? I gave Trout, my #1 pick, a 15% (which I think is generous), Kershaw 10%, and Kluber 5%. Everyone else got 0. I feel that, if this were done right according to the directions, we should have anyone with under 7 WAR (although you can go either way with catchers and framing) at 0. So any spot below 3rd on the ballot should be a 0 point tie, right?
TL;DR: We need a different ranking scale than the historical winner because how often is the fourth best player in a given year going to be the best in another year? Not often.
P.S. I would love to see an article that showed me the numbers on actually how often the third, fourth, fifth, etc player in a given year would have been the best in another year. Query that database!
You would also need the standard deviation of 8 WAR players per year, though I think it could be estimated using the number of players in the league.
Some numbers:
The last time Trout’s 7.8 WAR was the best position player WAR was 1999. Of course Pedro Martinez blew past that on the pitching side.
In 1994 Trout’s 2014 would be tied for the best WAR… in the strike shortened season.
He would have been the best position player in 1986… but two pitchers were worth than him that same year.
Now for the moment you have all been waiting for: The last time that 2014 Mike Trout would have been the best player in MLB was 1981. Even then, he would have tied Mike Schmidt also with 7.8 WAR.
I’m not going back any farther. The point is that the best player this year would be the best player in any other year closer to 0% of the time than the next lowest delineation of 5%. We should have a ballot of all 0’s out there.
I lied. I went back farther.
In 1952 he would have tied (again) at 7.8 with Jackie Robinson.
And, finally, in 1951, He would have, for the first time (counting backwards), been the best player in that year, edging out Eddie Stanky by just .1 WAR. And Mike Trout played in 5 more games this year than existed in the season at that time. Would Stanky have accrued another .1 WAR in five more games? The world will never know.
New question: Someone who has better database skills than I (I just went back one by one looking at the leaderboard), can we see what the lowest league-leading single season WAR ever was?
I gave one player a 5, and everyone else 0.
I gave Trout 30% and Kershaw 10%, the remaining 8 all received 0.
Why do we have a historical context scale in a ballot based on a single season award?
Will be fun to figure out and vote, but it seems like a bit of a curious choice.
Whew, I almost voted for the wrong McCutchen. I always get them mixed up because they look so much alike.
And, uh, yeah. The 0-100 grade scales will lead to inconsistency if people are near as generous as I was with it. I had Trout (he’s Mike Trout) and Kershaw (narrative/great) at 80, as well as 2 others at 50. High.
When I try to view Fangraphs’ 2014 overall pitching leaders, I get only starters. The buttons say All, Starters, and Relievers, but what is shown are Starters, Starters, and Relievers. I tried with WPA and ERA, both of which should have lots of relievers among the leaders.
Set the IP limit to 0. Qualified is only going to show players with at least 162 IP pitched.
When do we need to submit our ballot by? I’d actually like to spend some time thinking about and looking into it, but I’m not going to have time to do that right away.
Gotta say, I think just going by the WAR leaderboard is against the spirit of both WAR and the award.
Sure Trout and Kluber had slightly more fWAR than Kershaw (Kershaw beats them both in rWAR, and RA9-WAR), but can you really say that they were more outstanding than a guy who was within the margin of error, despite missing six starts at the beginning of the season?
Also, if you aren’t giving Kershaw credit for outperforming his FIP by now, after 6 years of doing so, you’re thinking wrong.
I’d argue for Lucroy because of his absolutely ELITE pitching framing. Borderline 9 WAR player including framing. Even giving him half the credit, he still stands with Trout/Kershaw.
As Matt said above, Lucroy has a top 30 catching season of all time. Since only 1990, Kershaw is ranked only 45th. Esteban Loaiza has had a better season by RA9-WAR.
I love Lucroy, but I do share Dave’s trepidation at giving him full credit for those numbers.
There’s an argument to be made that, rather than “stealing strikes,” Lucroy creates an expanded strike zone for his pitchers, who still have to hit their spots within that zone.
Also, how does StatCorner define their zone? Is it personalized by umpire, or is it an aggregate?
Obviously, different umps have different strike zones, and that should be accounted for.
That’s just as unreliable as UZR in player value. If he’s such a great pitch framer, would it be ok to say you can add the difference between Aj Ellis’ pitch framing and Lucroy’s to Kershaw if he had Lucroy as his catcher? Obviously not, but that’s kinda the point. How do you determine what value that actually brings to the player or the team overall. There is absolutely no way Lucroy was the best player in baseball this season. It was either Trout or Kershaw, and my admittedly biased vote goes to Kershaw, with Trout just behind him.
NL Russell Martin
AL Victor Martinez
There are no wrong opinions.
However,your opinion is wrong.