Tracking the Hall of Fame Tracker – 2019 Version
If you’re a Baseball Hall of Fame aficionado, one of the great joys of this season is diving into "The Baseball Hall of Fame Tracker", which is diligently maintained by Ryan Thibodaux and his merry band of ballot trackers. If you’ve never seen it before, it’s essentially an online Excel file that keeps track of writers’ Hall of Fame ballot selections as they become known. Sometimes a writer posts a column announcing his or her selections, sometimes the writer may verbally communicate the selections through some other medium (like Twitter), etc. In any case, Thibodaux and his team collect the results and track them here.
I always find it interesting to see how the different players are tracking and to see if there are any trends or observations that can be made. Now, having looked at this tracker for a few years, there are some caveats about looking at the results, especially early on in the process. One general observation is that, typically, most (but not all) players’ voting %’s tend to ride kind of high relative to their final totals, in part because the writers who make their choices known ahead of the deadline tend to vote for more players per ballot than those who don’t. (Note – for future reference, I’ll refer to those who announce their choices as "public" voters, and those who don’t as "private" voters. When I refer to "public" votes, I am also including any anonymous voters who have made their selections known to Thibodaux)
A Little Bit of History
Before diving into the 2019 results to date, here’s a little bit of background/history.
The "tracker" shows detail going back to 2009, although Thibodaux has made the tracker a lot more robust over the past few years. Here’s what I was able to calculate based on the information contained in the tracker. (I should say that these might not be exact, as not every candidate is listed in the detail sections, and in the earlier years the tracker doesn’t show the same level of summary information as it does now. However….I feel that these are pretty close.)
Year
|
Total # of Ballots
|
Public & Anonymous Ballots
|
Public % of Total
|
Private Ballots
|
Private % of Total
|
Average Selections per Ballot -Overall
|
Average Selections – Public & Anonymous
|
Average Selections - Private
|
Avg. # of Selections - Public vs. Private
|
2018
|
422
|
317
|
75.1%
|
105
|
24.9%
|
8.5
|
8.7
|
7.6
|
14.3%
|
2017
|
442
|
314
|
71.0%
|
128
|
29.0%
|
8.1
|
8.4
|
7.4
|
13.5%
|
2016
|
440
|
311
|
70.7%
|
129
|
29.3%
|
7.9
|
8.2
|
7.3
|
12.3%
|
2015
|
549
|
331
|
60.3%
|
218
|
39.7%
|
8.4
|
8.7
|
8.0
|
8.7%
|
2014
|
571
|
302
|
52.9%
|
269
|
47.1%
|
8.4
|
8.6
|
8.2
|
4.9%
|
2013
|
569
|
168
|
29.5%
|
401
|
70.5%
|
6.6
|
6.5
|
6.7
|
-3.0%
|
2012
|
573
|
114
|
19.9%
|
459
|
80.1%
|
5.1
|
4.8
|
5.2
|
-7.7%
|
2011
|
581
|
122
|
21.0%
|
459
|
79.0%
|
6.0
|
5.8
|
6.0
|
-3.3%
|
2010
|
539
|
91
|
16.9%
|
448
|
83.1%
|
5.7
|
6.0
|
5.6
|
7.1%
|
2009
|
539
|
60
|
11.1%
|
479
|
88.9%
|
5.4
|
5.4
|
5.4
|
0.0%
|
A few observations:
- In 2015, several writers had their voting privileges removed as the BBWAA trimmed the ranks, requiring voters to meet requirements as "active" members covering the game, so the base of writers participating in the Hall of Fame vote over the past few years (beginning with the 2016 election) is about 100 voters lower than it had been.
- The percentage of "public" voters has soared over the past several years, and has increased every year (except for one) since 2009. If you go back to 2009, only 11% voters made their ballots public. Last year, it was up to 75%, the highest mark ever. It has definitely become a much more common practice for a Hall of Fame voter to announce his selections (and to trigger the inevitable "feedback" that is sure to come his or her way via social media).
- The other upward trend has been in the number of selections made per ballot. If you go back about 10 years, the average per ballot was generally between 5 and 6 names per ballot. We have now essentially had 5 straight years where that figure has been 8 or higher (2016 was just under that figure, at 7.94), and last year reached 8.5, the highest figure to date. On average, voters are including about 60% more names on their ballots, than they did just a decade ago.
The interesting thing about 2013 is that, even though the average # of selections per ballot increased sharply from 5.1 to 6.6, that was also the infamous year that no candidate achieved 75% or higher. If you recall, that was the year that Barry Bonds, Roger Clemens, Mike Piazza, Craig Biggio, Curt Schilling, and Sammy Sosa (not to mention Kenny Lofton) debuted on the ballot. The next year, we saw the quantum leap up to 8.4, and it’s been hovering around that figure ever since, as the ballot continues to be well stocked with solid candidates.
- Coinciding with the recent trend of more selections per ballot has been the widening gap between "public" and "private" voters. The last column in the grid posted above compares public vs. private average names per ballot. Last year, public voters included about 14% more players per ballot than private voters.
This is one of the reasons why, as you watch the public votes tallied each year in the tracker, you have to take them with a grain of salt, as public voters have tended to be more "generous" than private voters, and the final percentages typically end up lower than what’s in the tracker, although that doesn’t necessarily hold true for every candidate. Some candidates do poll better among private voters. For example, in 2017, Fred McGriff received about 23% of the overall vote, but he was named on about 29% of the private votes, and only about 21% of the public votes. Lee Smith, who was just elected to the Hall of Fame by the Today’s Game Era Committee, was another candidate who consistently polled better among private voters for several years. The same was generally true for Jack Morris when he was on the ballot.
Going the other direction…..in 2018, the twin towers of Barry Bonds and Roger Clemens each received 56-57% of the overall vote. However, they had a significant split - they received over 60% of the public vote, but only 42% and 46% (respectively) of the private vote.
Mike Mussina was similar…..he received 69% of the public vote last year, but only 47% of the private vote.
Edgar Martinez, received enough support from public voters (76%) last year to get in, but got only 52% of the private vote, a 24-point gap.
Curt Schilling was the player with the largest gap last year. He got 57% of the public vote last year, but only 32% from the private voters, a 25-point gap.
In recent years, it’s been generally true that more players poll better among public voters vs. private voters, which is why it’s important to look at the results of the tracker with that in mind.
Another dynamic in the voting is that more and more voters are using the full 10 slots. In 2018, almost 55% of the public voters used the full 10 slots on the ballot, up from 52% the year before. Here’s the 2018 distribution by the number of candidates selected.
|
# of Votes
|
% of Total Public Votes
|
10-Player Ballots:
|
173
|
54.6%
|
9-Player Ballots:
|
47
|
14.8%
|
8-Player Ballots:
|
28
|
8.8%
|
7-Player Ballots:
|
25
|
7.9%
|
6-Player Ballots:
|
20
|
6.3%
|
5-Player Ballots:
|
10
|
3.2%
|
4-Player Ballots:
|
8
|
2.5%
|
3-Player Ballots:
|
4
|
1.3%
|
2-Player Ballots:
|
1
|
0.3%
|
1-Player Ballots:
|
0
|
0.0%
|
Blank Ballots:
|
1
|
0.3%
|
Total Public Votes
|
317
|
100.0%
|
In addition, the "tracker" also indicates that there were at least 63 public voters (or about 1 out of every 6 or 7 of total voters) who expressed the notion that they would have voted for more than 10 had they been allowed to, and that figure undoubtedly underestimates the number who would have done so, as not every voter who voted for 10 players comments on whether or not they would have voted for more. I believe that, as we go along, there will be more and more pressure applied to do away with the 10-player limit.
2019 Results So Far
OK. Enough history. How are the early 2019 results looking?
As I write this (on December 15), 45 public and anonymous ballots have been tracked, which (per the tracker) is about 11% of the total 412 votes expected to be submitted (by the way, don’t be surprised if that total number of votes changes the next time you check the tracker, as it changes frequently).
Also at this early date, the ballots have been coming in with an average of 8.6 names selected. That’s actually down a bit from this time last year. When I did my first check-in last year, it was tracking at about 9.1 before eventually settling at 8.7 (for the public votes). That implies that the average # of names per ballot that has been trending up for several years could come back down a bit this year.
Here are the results as of this moment among players who have received at least one vote:
Player
|
% of Ballots
|
Mariano Rivera
|
100.0%
|
Edgar Martinez
|
91.1%
|
Roy Halladay
|
88.9%
|
Mike Mussina
|
82.2%
|
Curt Schilling
|
73.3%
|
Roger Clemens
|
68.9%
|
Barry Bonds
|
66.7%
|
Larry Walker
|
66.7%
|
Omar Vizquel
|
44.4%
|
Todd Helton
|
31.1%
|
Fred McGriff
|
31.1%
|
Manny Ramirez
|
22.2%
|
Scott Rolen
|
22.2%
|
Andy Pettitte
|
15.6%
|
Sammy Sosa
|
13.3%
|
Jeff Kent
|
11.1%
|
Gary Sheffield
|
11.1%
|
Andruw Jones
|
6.7%
|
Billy Wagner
|
6.7%
|
Lance Berkman
|
4.4%
|
Roy Oswalt
|
2.2%
|
Michael Young
|
2.2%
|
Miguel Tejada
|
2.2%
|
So, at this point, if these percentages held up, the top 4 players (Rivera, Martinez, Halladay, and Mussina) would be elected. In our Bill James Online prediction contest, I predicted those 4 to get elected.
Mussina will probably be very close one way or the other. Even though he’s at 82% support so far, keep in mind that he’s one of the players that has a track record of polling much stronger among public voters vs. private voters, so one would expect his overall figure to come in lower than what it is tracking at so far.
Here’s Mussina’s public vs. private gap over the last few years:
Year
|
Public
|
Private
|
Gap
|
2018
|
69.1%
|
46.7%
|
22.4
|
2017
|
57.6%
|
37.5%
|
20.1
|
2016
|
48.2%
|
30.2%
|
18.0
|
2015
|
30.2%
|
16.1%
|
14.2
|
So, if Mussina maintains 82% support among the public voters (and if public voters once again comprise about 75% of the total votes cast), I estimate he’ll need to receive around 54% of the private votes in order to get 75% of the overall vote. I think that’s do-able, but it’s not a lock. According the Tracker, Mussina was about 49 votes short last year, and he’s had a net gain of 6 votes so far from returning voters (7 voters who didn’t vote for him last year have flipped to a "yes" this year, and then he also lost support from one voter who voted for him last year). So, with about 11% of the vote in, he’s gained 12% of the votes he would need. I think it’s going to come right down to the wire for Mussina.
Let’s dig a little deeper into some of the early results, and I’ll also refer back to our Bill James Online "prediction contest", where we invited our members to submit their predictions. I’ll provide our "consensus" prediction (which is the average of all of the submitted ballots), and I’ll also include the prediction that I submitted, just for additional reference.
Mariano Rivera
Actual % to Date
|
Consensus Member Prediction
|
My Submitted Prediction
|
100.0%
|
95.8%
|
98.0%
|
Not much drama here, other than whether or not Rivera would become the first player to ever be elected unanimously. I would not expect that to happen, as there are normally a handful of voters who won’t vote for a first-year candidate, and those are usually (but not always) found among the private voters. A high 90’s figure seems inevitable.
Welcome to Cooperstown, Mariano.
Edgar Martinez
Actual % to Date
|
Consensus Member Prediction
|
My Submitted Prediction
|
91.1%
|
78.8%
|
82.0%
|
Edgar’s looking like a lock in his 10th and final time on the ballot. Even if he keeps the same gap as last year (where he polled 24 percentage points lower among private voters), he would still end up with around 85% of the vote. In fact, even if he only draws 52% of the private vote (like he did last year), I estimate he would end up with 81%.
He’s as good as in.
Mike Mussina
Actual % to Date
|
Consensus Member Prediction
|
My Submitted Prediction
|
82.2%
|
70.7%
|
76.5%
|
I covered Mussina a little earlier in the article. Basically, I think he’s got a really good shot, but it will probably be close since he tends to poll much lower among the private voters.
I think he’ll squeeze in.
Roy Halladay
Actual % to Date
|
Consensus Member Prediction
|
My Submitted Prediction
|
88.9%
|
60.7%
|
79.0%
|
A really strong showing so far for Halladay in his first year on the ballot, gaining support from 40 of the 45 votes so far. Here’s what that means…..if he maintains that level of support among the public voters, he would only need about 33% of the private vote to get 75% overall.
He’s looking awfully good unless we see a real dropoff in the public support. If I were able to revise my prediction, I’d say he’s good for at least low 80’s in overall support.
Barry Bonds and Roger Clemens
Always seems natural to look at them together…..
Player
|
Actual % to Date
|
Consensus Member Prediction
|
My Submitted Prediction
|
Bonds
|
66.7%
|
60.9%
|
58.8%
|
Clemens
|
68.9%
|
61.9%
|
58.5%
|
This is actually slightly down to where Bonds and Clemens were polling in the early results a year ago. I think they’re kind of stuck in neutral at this point, and I think there is a connection, at least in part, to Joe Morgan’s letter to the writers last year where he encouraged them to not vote for alleged steroid users.
According to the tracker, Bonds and Clemens have each gained one vote from a writer that did not vote for him last year (as it turns out, it’s the same writer in both cases).
They have virtually no chance to be elected this year, especially since the private vote for these 2 tend to run so much lower than the public.
Curt Schilling
Actual % to Date
|
Consensus Member Prediction
|
My Submitted Prediction
|
73.3%
|
55.5%
|
57.0%
|
Schilling is up about 10 points vs. where he was in the early polling a year ago. He has a net gain of 2 votes (gained 3, lost 1) among returning voters.
It’s possible that the strong feelings about Schilling’s social media statements and his viewpoints may be growing less intense as time goes by, and voters may be more inclined to vote for him now. I think a final tally of high-50’s or low-60’s this year seems about right.
Omar Vizquel
Actual % to Date
|
Consensus Member Prediction
|
My Submitted Prediction
|
44.4%
|
39.6%
|
40.0%
|
Vizquel was one of the surprises and biggest stories of last year’s early balloting, being named on the first 7 ballots that were tracked, and he was at about 53% at this point last year. He’s down about 9 points vs. this time last year, although he has picked up 4 votes among returning voters. He finished at 37% overall last year, so an uptick seems likely.
Vizquel has kind of taken the place of players like Lee Smith and Jack Morris as the type of player who tends to end up more significantly more popular among private voters. Last year, he polled almost 12 points higher among private voters vs. public voters. I think he’s likely to stay in the 40-50% range for a while, as he’s a very polarizing candidate, and I’m not sure either camp is likely to yield much. We should have the pleasure of several more years ahead of us to debate the pros and cons of his candidacy.
Larry Walker
Actual % to Date
|
Consensus Member Prediction
|
My Submitted Prediction
|
66.7%
|
40.5%
|
45.0%
|
Walker was one of last year’s big movers, improving from 21.9% in 2017 to 34.1% in 2018. Our group predicted another step up this year, but he’s looking like he’ll surpass our expectations. He’s already gained a whopping 9 votes from voters who didn’t vote for him last year. He only received 24% of the private vote last year, but even if he holds at that level with that group, he’s got a good shot at finishing well over 50% this year. That’s still a long way from 75%, and next year will be his final year on the ballot, so he’s running out of time, but it’s been an impressive run for him the last couple of years. Next year should be really interesting, and, at the very least, I think he’ll set himself up well for strong consideration on a veteran’s committee ballot.
Lance Berkman
Actual % to Date
|
Consensus Member Prediction
|
My Submitted Prediction
|
4.4%
|
11.2%
|
3.5%
|
Berkman’s not a strong candidate, but I’m including him here because I think he’s an interesting one. I (along with a few others) slated Berkman in the 5% and under category.
Why did I think he would get such little support? Let’s see….how to put this? He didn’t "seem" like a Hall of Famer. Or, at least that’s the sense I got. He was a good, valuable player. But he wasn’t a star. But there was another reason.
Do you remember an article I did last year on "Hall of Fame Metric Similarity Scores"? It was a method I developed to compare similarity of players, not across offensive categories, but across the "Hall of Fame" metrics found on baseball-reference.com (Hall of Fame Standards, Hall of Fame Monitor, Black Ink, Gray Ink). I was kind of proud of it.
To review, here’s Berkman’s list of top 10 comps by that method:
Score
|
Name
|
HOF Std
|
HOF Monitor
|
Black Ink
|
Gray Ink
|
Position
|
1,000
|
Lance Berkman
|
44
|
98
|
8
|
107
|
OF/1B
|
942
|
Carlos Delgado
|
44
|
110
|
8
|
124
|
1B
|
942
|
Fred McGriff
|
48
|
100
|
9
|
105
|
1B
|
926
|
Pee Wee Reese*
|
39
|
100
|
7
|
102
|
SS
|
914
|
Jason Giambi
|
44
|
108
|
13
|
115
|
1B
|
912
|
Jackie Robinson*
|
38
|
98
|
8
|
121
|
2B
|
896
|
Matt Holliday
|
41
|
110
|
13
|
107
|
LF
|
894
|
Bobby Doerr*
|
41
|
96
|
5
|
128
|
2B
|
894
|
Johnny Damon
|
45
|
90
|
6
|
77
|
CF
|
890
|
Bill Dahlen
|
48
|
94
|
4
|
96
|
SS
|
886
|
Magglio Ordonez
|
42
|
114
|
6
|
86
|
RF
|
Here’s what I wrote about Berkman last year in that article:
"I anticipate that Berkman will have a tough time getting support, although he was a tremendous player. He did have some good years, and had six top-10 MVP finishes. His career OPS+ of 144 is outstanding…it’s roughly in the same neighborhood as Edgar Martinez (147)….and his career OBP is over .400 (.406). However, he didn’t even reach 2,000 hits, and his career total of 369 HR’s is fairly low for someone primarily thought of as a power hitter, and won’t inspire a lot of excitement among the voters.
Jackie Robinson is not a good comp for a variety of reasons, so subjectively I would remove him. The #11 comp, to replace Robinson, would be George Van Haltren. If Van Haltren doesn’t do it for you, the next one down would be Will Clark (similarity 876), who I think is a decent comp.
So, excluding Robinson, the only 2 players on Berkman’s comp list that are in the Hall are Doerr and Reese, strong defensive middle infielders who were elected by Veterans Committees many years after their playing days were done. They really aren’t great comps for Berkman either. Delgado and McGriff are Berkman’s top 2 comps, and neither one made much of an impact on the writers’ ballot (McGriff, of course, is still on the ballot, but treading water).
I was a big fan of Berkman’s, but I think he’ll draw very little support."
I think that’s still a pretty accurate assessment. Berkman was a good player, but I just don’t think he struck many voters as a "Hall of Fame type". He’ll be fortunate if he draws enough support to stay on the ballot.
Todd Helton
Actual % to Date
|
Consensus Member Prediction
|
My Submitted Prediction
|
31.1%
|
33.3%
|
35.0%
|
We had a pretty wide range of predictions for Helton in our contest, from a low of 15% to a high of 62%. Overall, though, the average was 33.3%, and so far he’s tracking close to that. It’s a nice debut for Helton, and it’s shaping up as a better debut that his Coors Field buddy Larry Walker, who received about 20% in his first year on the ballot. I can see Helton as the type who could gain momentum over time….his rWAR is above 60, he has good "traditional" stats, and I think Walker’s progress in recent years is eroding some of the potential anti-Coors Field sentiment that’s out there. He should have an interesting ride on the ballot in the coming years, and it wouldn’t surprise me at all to see him get in on a future ballot. I thin there could be some future ballots where he will be considered one of the stronger candidates, and that always helps.
Scott Rolen
Actual % to Date
|
Consensus Member Prediction
|
My Submitted Prediction
|
22.2%
|
21.5%
|
19.0%
|
For many, Rolen had a bit of a disappointing debut on the ballot last year. I think a lot of people saw the 70 rWAR, the 8 Gold Gloves, the top-10 JAWS ranking, and assumed he would make a strong dent on the ballot.
It didn’t turn out that way. He ended up with only 10% of the vote (12% public, only 5% private). Using the Hall of Fame Metric Similarity Score technique I used above on Berkman, Rolen is part of a group that, for the most part, has come up short at the ballot box:
Score
|
Name
|
HOF Std
|
HOF Monitor
|
Black Ink
|
Gray Ink
|
Position
|
1,000
|
Scott Rolen
|
40
|
99
|
0
|
27
|
3B
|
950
|
Lou Whitaker
|
43
|
93
|
0
|
31
|
2B
|
950
|
Edgar Renteria
|
38
|
109
|
0
|
22
|
SS
|
934
|
Omar Vizquel
|
42
|
120
|
0
|
25
|
SS
|
918
|
Alan Trammell*
|
40
|
119
|
0
|
48
|
SS
|
904
|
Jim Edmonds
|
39
|
89
|
0
|
60
|
CF
|
890
|
Aramis Ramirez
|
39
|
85
|
2
|
53
|
3B
|
882
|
Willie Randolph
|
34
|
92
|
2
|
39
|
2B
|
880
|
Chase Utley
|
35
|
94
|
3
|
42
|
2B
|
870
|
Dave Concepcion
|
29
|
107
|
0
|
25
|
SS
|
868
|
Buddy Myer
|
41
|
86
|
6
|
45
|
2B
|
Trammell crashed the party last year with the Veterans Committee vote, but a lot of the others - Whitaker, Edmonds, Randolph - are very similar types, accumulating a lot of rWAR, but often lacking in other areas that tend to get voters’ attention. All three of those players were one-and-done on the writers’ ballot. Utley, who will be up for election in a few years, is of that same general type, and it wouldn’t surprise me if he struggles for support as well. Of course, the landscape could be different by then.
In any case, Rolen has swung a few votes his way this year among returning voters (he’s plus 4 so far), but I still think he’ll struggle to surpass 20% this year. He’s still only on his second ballot, but I think he’ll need some type of advocacy campaign (like Tim Raines and Bert Blyleven received) if he’s going to get enough support.
Andy Pettitte
Actual % to Date
|
Consensus Member Prediction
|
My Submitted Prediction
|
15.6%
|
20.5%
|
25.0%
|
Another first-year candidate, Pettitte looks safe to stay on the ballot, but other than that he’s receiving moderate support. I may have been too optimistic on Pettitte with my 25% prediction. I’m not sure yet whether Pettitte will be the type to do better on public ballots. The steroid part of his story makes me think he could do worse among private voters (who tend to be tougher on the steroid candidates), but, on the other hand, Pettitte has a fairly similar case to Jack Morris – had about 250 wins, was on a lot of champions, had a lot of personal postseason success, named to a few All-Star games, had a few top-5 Cy Young finishes (though never winning one), all of which helps offset a relatively high ERA (although Pettitte’s 117 ERA+ is significantly better that the 105 posted by Morris). That type of résumé might resonate with private voters. In any case, I think my prediction is ultimately going to end up too high.
Fred McGriff
Actual % to Date
|
Consensus Member Prediction
|
My Submitted Prediction
|
31.1%
|
26.4%
|
24.5%
|
McGriff is in his 10th and final year on the BBWAA ballot, and he’s doing well compared to prior years. He already has a net gain of 5 votes (6 gains, 1 loss) from returning voters, who are clearly giving him a bit of a "final year" bump. He only got 21.5% support from the public voters last year, and 23.2% overall, so he has a decent chance to stay above 30%. Obviously, he’ll finish well short, but I think he will make for a very strong candidate on a veteran’s committee ballot in the upcoming years. I think he will carry a great deal of appeal for that forum, and I fully expect him to eventually be elected.
I’ll post another update on the tracker down the road.
Thanks for reading.
Dan