Hall of Merit at Baseball Think Factory, there seems to be something of a consensus that Rolen is a better candidate than Jim Thome. I don't agree with that, but I do agree that he is well qualified for the Hall of Fame.
That said, Rolen's case is fairly subtle. He reached no major career milestones - he retired with a .281 career batting average, 2,077 hits, and 316 home runs. Rolen made seven All-Star teams, received MVP votes four times (finishing top 10 once), and won eight Gold Gloves. The first and last of these are good, although hardly strong enough to guarantee Hall-of-Fame induction.
I had a lot of trouble trying to come up with close comps for Scott Rolen. So, I decided to see what Player won-lost records say. I have constructed a methodology to determine the players most similar to a particular player as measured by Player won-lost records. My concern in using this to pick comps for Rolen (or anybody else here) is that the issue here is to identify players who Hall-of-Fame voters will see as similar in value and if the premise is that Hall-of-Fame voters will not appreciate Scott Rolen's statistical case, my concern was that my system would identify players who are perceived as much better than Rolen and, hence, did better in Hall-of-Fame voting than I expect Rolen to do.
Here, then, are the five players most similar to Scott Rolen in career value as measured by Player won-lost record, who debuted in 1947 or later and whose last game played was in 2011 or earlier (meaning they will have already appeared on at least one Hall-of-Fame ballot).
Much to my surprise, it turns out that the players most similar to Scott Rolen fall into the very group that I anticipate Rolen falling into: players whose statistical cases for greatness are likely to be too subtle to be fully appreciated by Hall-of-Fame voters.
The first four players on this list have all been elected to the Hall of Merit - as I expect Rolen to be elected very soon, probably this year. While none of the five players has been elected to the Hall of Fame.
Graig Nettles is, I think (and not surprisingly), a very good comp for Scott Rolen. An excellent defensive third baseman, he won only two Gold Gloves, because he shared a league with Brooks Robinson. Nettles was also a very good offensive player, hitting 390 career home runs. His offense was perhaps underappreciated because he hit for a relatively low batting average (.248 for his career) although he also drew over 1,000 walks in his career giving him a .329 career on-base percentage, which was slightly above league-average during his career, which took place largely in relatively low-scoring environments. Despite his low batting average, Nettles was named to six All-Star teams in his career (one fewer than Rolen) and received MVP votes four times, the same number of times as Rolen. Nettles received 8.3% of the vote in his first appearance on the Hall-of-Fame ballot (in 1994), lasting four seasons before falling below 5% and dropping off the ballot.
The second-most similar player to Scott Rolen was Rolen's teammate for 5-1/2 seasons with the St. Louis Cardinals, during which time the Cardinals won two National League pennants and one World Championship. Jim Edmonds bested Scott Rolen somewhat in triple crown stats: .284/393/1199 to .281/316/1287. Edmonds received MVP votes slightly more often than Rolen (6 to 4), and matched Rolen's 8 Gold Gloves but played in only 4 All-Star games (vs. 7 for Rolen). Jim Edmonds debuted on the Hall-of-Fame ballot two years ago. At that time, I anticipated that he would be under-appreciated, but I badly mis-estimated precisely how unappreciated. I predicted that Jim Edmonds would receive 13.6% of the vote that year, comparing him favorably to Larry Walker. Edmonds turned out to not even receive 13 votes; he received 11 votes, good for 2.5% and a summary dismissal from the Hall-of-Fame ballot. The next year, the Hall of Merit elected Edmonds in his second year of eligibility.
The third-most similar player to Scott Rolen is something of the poster boy for this type of candidate: the well-rounded, good-hitting, good-fielding infielder who ends up summarily dismissed by Hall-of-Fame voters. Bobby Grich made six All-Star teams, won four Gold Gloves, and received MVP votes five times with two top-10 finishes. He tied for the league lead in home runs in 1981, but with only 22 in that strike-shortened season. Grich received the same number of Hall-of-Fame votes in his first year (in 1992) as Jim Edmonds would 24 years later. There were slightly fewer voters in 1992, so Grich's debut percentage is a tick better than Edmonds, at 2.6%.
Like Scott Rolen, Ken Boyer had his best seasons playing third base for the St. Louis Cardinals, for whom he won a World Series. Boyer won only five Gold Gloves, but bested Rolen in All-Star appearances (11) and years receiving MVP votes (8). Boyer, in fact, won the National League MVP award in 1964. Ken Boyer's Hall-of-Fame voting history is somewhat odd. In his first five years on the ballot, 1975 - 1979, Boyer's support never rose above 4.7%. After being dropped from the ballot, he was reinstated six years later, in 1985, where he re-debuted (along with several other players) at 17.2%. He then remained on the ballot for ten more seasons, with vote percentages that ranged between 11.8% and 25.5%. For the purposes of this exercise, I am going to use Boyer's second "first" ballot, 1985, when he received 17.2% of the vote.
The fifth player in the above table is the only one who has not been elected to the Hall of Merit. And, frankly, I'm not entirely sure how he ends up as a comp to Scott Rolen. In many ways, George Foster has the best traditional Hall-of-Fame credentials of any of these five players, having led his league in RBI three times, home runs twice, runs scored once, with four top-10 MVP finishes, including three top-3 finishes, winning the award in 1977. Foster debuted with 5.6% of the vote, surviving for four ballots before falling below 5% and being removed from the ballot.
The simple average of the first-year vote totals for these five players - 8.3%, 2.5%, 2.6%, 17.2%, and 5.6% - is 7.2%.
At one time, Andruw Jones seemed to be on pace for election to the Hall of Fame. After the 2006 season, which was Andruw Jones's age-29 season, Jones won his tenth Gold Glove. That was something of a down season offensively for Jones, but he nevertheless hit 26 home runs with 94 RBI, giving him 342 home runs and 1,023 RBI. At that time, Jones's 342 home runs through age 29 tied him with Hall-of-Famers Hank Aaron and Mel Ott for the fifth-most home runs through a player's age-29 season.
Unfortunately, Jones's "down season" in 2006 was not an anomaly. His age-30 season was even worse, and his defense collapsed shortly thereafter. Jones's tenth Gold Glove was his last and he ended up with 434 career home runs.
The question, then, is how many voters might think that Jones was not merely on a Hall-of-Fame trajectory through age 29, but that he had already qualified for the Hall of Fame.
On the one hand, Andruw Jones's career is, I think, largely viewed as something of a disappointment. On the other hand, ten Gold Gloves is something of a magic number. There have been 16 players, including Jones, who won at least ten Gold Gloves. One is still active (Ichiro Suzuki), and one debuts on this ballot with Jones (Omar Vizquel). Of the remaining 13 players with at least 10 Gold Gloves, 11 are in the Hall of Fame. The exceptions are Jim Kaat and Keith Hernandez, who play arguably the two least important fielding positions (pitcher and first base).
What might that mean for Andruw Jones?
Looking at centerfielders with multiple Gold Gloves, none of them seem like particularly close comps to Andruw Jones. Stretching somewhat - both in terms of similarities as well as in terms of how far back in history one is willing to go - Paul Blair was perhaps something of a "poor man's" Andruw Jones. Blair won eight Gold Gloves and was, I believe, considered the finest defensive centerfielder of his generation. Like Jones, Blair played for very good teams and had considerable postseason experience. For his career, Blair appeared in eight postseasons with 52 postseason plate appearances compared to Jones's 11 postseasons and 76 plate appearances. Blair beats Jones in World Series winners 4 to 1. Blair did not have anywhere near Jones's power, retiring with only 134 career home runs, exactly 300 fewer than Jones. Blair had somewhat more speed, beating Jones in stolen bases, 171 to 152. The two players had similar batting averages (.254 for Jones, .250 for Blair) with Blair having played in much lower-scoring contexts throughout his career.
Jones has the stronger Hall-of-Fame case, so I am somewhat skeptical about how insightful Blair's Hall-of-Fame performance might be. Nevertheless, Blair appeared on one Hall-of-Fame ballot, receiving 1.9% of the vote.
Moving away from the Gold Gloves and back to the home runs, there is one player in the top 10 in home runs through age 29 who has appeared on a Hall-of-Fame ballot and has not been elected to the Hall of Fame: Juan Gonzalez, who through the age of 29 had hit 340 home runs and won two MVP awards. From age 30 on, Gonzalez hit 94 home runs, receiving MVP votes once (in 2001, when his .325/35/140 was good for fifth place in AL MVP voting). For comparison, Jones hit 342 home runs through age 29, 92 thereafter.
On the one hand, Gonzalez has a clear edge on Jones with his two MVP awards versus none for Jones. On the other hand, Gonzalez never won a Gold Glove and he may have lost some Hall-of-Fame votes due to PED rumors. Perhaps these balance out; perhaps not. Either way, Juan Gonzalez received 5.2% of the vote in his first year on the Hall of Fame ballot.
The simple average of Blair and Gonzalez's first-year vote totals is 3.6%, which is my prediction for Andruw Jones's first-year vote total.
Jamie Moyer had a very interesting career. He pitched until he was 49 years old (my current age, which just makes this aspect of his career even more astounding to me) and amassed 269 traditional pitcher wins. That said, Moyer's career is almost entirely that of a "compiler". He was named to only one All-Star team and received scattered Cy Young votes three times (he had one first-place vote in his career, in 2001, when he went 20-6 with a 3.43 ERA for the 116-win Seattle Mariners).
Moyer's sheer number of pitcher wins could be enough to at least get him into some Hall-of-Fame conversations. But they are at a level at which you probably need some kind of hook to actually get more than a handful of Hall-of-Fame votes. Moyer does potentially have the hook of pitching until he was 49. On the other hand, his career ERA was 4.25 and he holds the major-league record for career home runs allowed (522).
Per Baseball-Reference, the pitcher most similar to Jamie Moyer in career value was Dennis Martinez, who pitched until he was 44 years old and finished his career with 245 pitcher wins. Martinez did slightly better than Moyer in All-Star appearances (four) but did similarly poorly in Cy Young voting (two fifth-place finishes). Martinez received 1.6% of the vote in his only year on the Hall-of-Fame ballot.
Using Player won-lost records, the pitcher most similar to Jamie Moyer was Frank Tanana. Tanana only pitched until he was 39, and only won 240 games in his career. On the other hand, his career ERA more more than a half-run lower than Moyer's, 3.66 vs. 4.25. Frank Tanana appeared on the 1999 Hall-of-Fame ballot, but failed to receive a single vote.
The simple average between Martinez and Tanana would be 0.8%, which is my prediction for Jamie Moyer's vote percentage.
Johan Santana won two Cy Young awards in three years and probably should have won a third between 2004 and 2006. He preceded that with one very good - albeit relatively low-inning - season (12-3, 3.07 in 158.1 IP, 18 games started, 27 relief appearances) and two solid seasons. In the latter of these, 2008, he led the National League in games started (34), innings pitched (234.1), and ERA (2.53), while amassing a traditional won-lost record of 16-7 in his first season as a New York Met. Unfortunately, that would be the last season in which Santana started 30 games or threw 200 innings. He was very good his next two seasons - a combined 24-18, 3.05 ERA, but in only 365.2 combined innings. And then he was done.
The Hall of Fame has historically been willing to induct pitchers with relatively short careers who were great for a few years. The epitome of this type, of course, is Sandy Koufax, but other examplars include Dizzy Dean and Catfish Hunter. Going back in time to before I have calculated Player won-lost records, Addie Joss is in the Hall of Fame despite having pitched only nine seasons. His "career-ending" injury was a bit more serious than those of Dean or Koufax or Santana. Joss died after his ninth major-league season in April, 1911.
So, there is some precedent for a pitcher with a resume similar to that of Johan Santana to get elected into the Hall of Fame. But how likely is it?
Given Santana's specific resume, I would say not very likely.
Johan Santana ended his career with 139 pitcher wins. To the best of my knowledge, no starting pitcher in the Hall of Fame has that few. The aforementioned Addie Joss had 160 wins, on the strength of four consecutive 20-win seasons. Sandy Koufax had 165 wins.
Johan Santana is the sixth pitcher to appear on a Hall-of-Fame ballot who won exactly two Cy Young awards. Three of the previous five are in the Hall of Fame: Bob Gibson, who won 251 games, Tom Glavine, who won 305 games, and Gaylord Perry, who won 314 games. None of these three pitchers seem like a particularly good comp for Johan Santana.
The other two longer-retired two-time Cy Young winners do, however, seem like good comparables. Bret Saberhagen won Cy Young awards in 1985 and 1989. Saberhagen had a bizarre habit of alternating good and bad seasons through the early part of his career and battled injuries through the latter part. Because of this, Saberhagen ended up with only 167 pitcher wins in his career. In his first, and only, year on the Hall-of-Fame ballot, Saberhagen received 1.3% of the vote.
One has to go back another two decades for the second such pitcher. Denny McLain won 31 games, earning himself both a Cy Young and MVP award the year I was born, 1968. He followed that up the next season by going 24-9 with a 2.80 ERA, leading the American League in shutouts (9), games started (41), and innings pitched (325) to win a second consecutive Cy Young award. Injuries limited McLain to only 91 innings the next season and injury-induced ineffectiveness produced a 10-22 record in 1971, the last season of McLain's career in which he pitched more than 77 innings. McLain threw his last major-league pitch at the age of 28, finishing his career with 131 traditional pitcher wins. McLain received one vote for the Hall of Fame in his first season on the ballot, good for 0.3% of the vote.
McLain had some fairly serious problems outside of baseball, which ultimately landed him in prison for a time. I don't know the extent to which these may have affected his Hall-of-Fame vote total. And, of course, his time on the Hall-of-Fame ballot was nearly 40 years ago now.
Which is a shame, because McLain is, I think, perhaps the best comp to Santana in terms of going from the consensus best pitcher in his league to out of baseball in a few short years. While Saberhagen is more recent, he also had a longer career, although his greatest seasons were more spread out, so that, perhaps, Santana was more clearly the best pitcher in all of baseball at his more concentrated peak.
Given Saberhagen's greater recency and his somewhat better performance in Hall-of-Fame voting, I decided to weight Saberhagen's total at twice that of McLain. The resulted weighted average is 1.0%, which is my prediction of Johan Santana's vote percentage this year, which feels low to me, but Santana's career really was quite short for a Hall-of-Fame candidate.
If I lived in a bubble and read no baseball writing, I am not entirely sure if I would have even bothered to include Omar Vizquel in my list of players "whose candidacy warrants at least some comment."
Omar Vizquel was named to only three All-Star games in his career. He was named on an MVP ballot exactly once in his career, receiving a single eighth-place vote in 1999. The only offensive category in which Vizquel ever led his league was sacrifice hits, which he led in four times. Those are not the traditional statistics of a Hall-of-Famer.
Vizquel does, however, have one item on his resume that is of Hall-of-Fame quality. He won 11 Gold Gloves, more than any other shortstop except for Ozzie Smith. As I noted above in discussing Andruw Jones, 10 seems to be something of a magic number in terms of Gold Gloves and the Hall of Fame.
And, having been reading a lot of baseball writing over the years, it is clear that there are a number of Hall-of-Fame voters who intend to vote for Omar Vizquel for the Hall of Fame. It is less clear, however, exactly how many such voters there are.
Omar Vizquel's Hall-of-Fame case basically boils down to this: he played shortstop very well (11 Gold Gloves) for a very long time (2,968 career games; 2,709 of them at shortstop). There are, I think, two obvious comps to Omar Vizquel, both of whom are in the Hall of Fame.
Ozzie Smith played 2,573 games in his career, 2,511 at shortstop (and none at any other fielding position), and won 13 Gold Gloves. Smith was elected to the Hall of Fame in his first year of eligibility with 91.7% of the vote.
Luis Aparicio played 2,599 games in his career, 2,581 of them at shortstop (and, like Smith, none at any other fielding position). Aparicio won 9 Gold Gloves, received 27.8% of the vote in his first year on the Hall-of-Fame ballot, and was elected in his sixth year of eligibility with 84.6% of the vote.
It seems fairly obvious to me that Smith and Aparicio have additional markers that set them above Vizquel. Smith made 15 All-Star teams and received MVP votes six times. Aparicio made the All-Star team in 10 seasons (he made 13 All-Star teams - there were two All-Star games for several seasons in the 1950s and 1960s), received MVP votes ten times, and led the American League in stolen bases nine (consecutive) times. As mentioned above, Vizquel made three All-Star teams, received one (8th-place) MVP vote in his career, and never led his league in any offensive category other than sacrifice bunts.
The next table compares Smith, Aparicio, and Vizquel as measured by Player won-lost records.
When I look at a table like that, I am drawn to the measures of pWins over positional average (pWOPA) and replacement level (pWORL) - as, I think, you generally should be, since these put the numbers in context - compare wins to losses, compare against averages, etc. But what if we just look at the first number: raw pWins? In terms of raw pWins, it's pretty easy to see the similarity across these three players.
In fact, let's go one step farther. The next table shows the top 10 players who have made their major-league debut since 1947, earned at least half of their player wins as a shortstop, and have appeared on at least one Hall-of-Fame ballot. The players in this table are ranked by career pWins.
The next table repeats the same players in the same order, but instead of showing pLosses, pWOPA, and pWORL, the table shows MVP awards and Gold Gloves won as well as the player's Hall-of-Fame vote percentage on his first ballot.
So, where does all of that leave me? I don't know. So I'm going to just go with Luis Aparicio's debut total, 27.8%, and leave it at that. This may well come in low, but I just have trouble believing that a third or more of Hall-of-Fame voters are going to look at a guy who got a single 8th-place MVP vote over the entirety of his 24-year career and say, "Yup, that one's a Hall-of-Famer". Remember, MVP voters are the same folks who vote for the Hall of Fame. We shall see.
There are several other players debuting on the ballot who had very good careers, but none of whom seem likely to get more than perhaps a handful of courtesy votes.
There are 14 players who are appearing on the 2018 Hall-of-Fame ballot who appeared on the 2017 ballot. Their vote totals in 2017 and 2016 are shown in the next table.
The total number of ballots was virtually identical between these two seasons: 440 in 2016, 442 in 2017. Assuming a similar growth in voters this year, let's assume 444 votes in 2018.
In 2017, 442 voters cast a total of 3,595 votes which works out to 8.13 votes per voter. If the average ballot size stays constant between 2017 and 2018, that would work out to 3,611 expected total votes in 2018.
The estimated vote percentages for new candidates in the last section add up to a total of 956 votes. Returning candidates earned a total of 2,322. Adding those up, that totals 3,278 votes.
If the average ballot size remains the same, then, in 2018 as in 2017, this leaves 333 additional votes to be spread among the returning candidates.
Last year, only two players lost votes: Billy Wagner, who lost a single vote, and Curt Schilling, who lost 31 votes. Wagner's -1 is functionally equivalent to Jeff Kent, who gained a single vote, both of which are functionally equivalent to a constant vote total from 2016 to 2017. Schilling's decline, on the other hand, was definitely a real decline. The reason for his decline was likely due to controversial remarks he made on Twitter. To be perfectly honest, I do not follow Curt Schilling on Twitter, so I do not know if he has made additional controversial remarks this year. But even if he has, it seems likely to me that voters who are likely to hold such views against Schilling would have done so in 2017. It's hard to say how many of his lost votes he might regain with the passage of a year from his original remarks, but I would be at least a little bit surprised to see him lose additional votes barring anything new.
So, for a starting point, let's zero out the numbers in the above table for Wagner and Schilling.
The next step, then, is to figure out values to plug in for the two players who debuted on the 2017 ballot: Vladimir Guerrero and Manny Ramirez. It seems to me that the most logical starting point for estimating such things would be to focus on players who have been in similar position on recent ballots.
In the case of Vlad Guerrero, this is fairly straightforward. Guerrero debuted at a level very close to, but just short of induction. In the past five seasons, two players have debuted with similar percentages as Guerrero (71.7%) who failed to be elected.
In 2013, Craig Biggio debuted at 68.2%. He increased that in 2014 to 74.8% (two votes short of election) before being elected in his third year of eligibility. That works out to an increase in Biggio's second-year vote total of 6.6%.
Well, that would seem to make things simple. An increase of 6.65% times 444 expected voters would translate into an expected 30 (29.53) additional votes for Guerrero. Guerrero debuted at a higher percentage than either Biggio or Hoffman. Hence, a similar gain to them in the second year should be enough to elect him this year.
In 2016, Trevor Hoffman debuted at 67.3%. In 2017, Hoffman received 74.0% of the vote (four votes short of election). That works out to an increase in Hoffman's second-year vote total of 6.7%.
Manny Ramirez is harder to figure. Ramirez debuted at 23.8%.
Some candidates who debut in the low 20's see their votes steadily increase. On the current ballot, Mike Mussina is an example of this. Mussina debuted in 2014 at 20.3%. His vote total more than doubled by 2017 when he received 51.8% of the vote.
On the other hand, some candidates who debut in the low 20's remain there or even may see their support gradually decline. On the current ballot, Larry Walker is an example of this. He debuted in 2011 at the same level as Mike Mussina's debut (three years later), 20.3%. In the six years since then, Walker's peak vote total was 22.9% in his second year on the ballot. He declined as low as 10.2% in 2014 (Mussina's first year) before recovering those losses in 2016 and 2017. Last year, Walker was named on 21.9% of all Hall-of-Fame ballots.
Neither Mussina nor Walker are a particularly good comp to Ramirez, though. The problem with Mussina is that his Hall-of-Fame case is subtle and took some time to register with voters. The problem with Walker is that voters know that his stats have to be adjusted for his having played in Colorado but are unsure exactly how large of an adjustment to make.
The best comp to Manny Ramirez on the current ballot may be Gary Sheffield. Both Sheffield and Ramirez hit over 500 career home runs with career slash lines close to .300/.400/.500 (Sheffield hit .292/.393/.514; Ramirez hit .312/.411/.585). Both players were also associated with PEDs during their career and had negative reputations with at least some members of the press. Ramirez was the better hitter and probably had a better relationship with the press, but his connections to PEDs are also much stronger. The 2017 ballot was Sheffield's third ballot and the first ballot in which his vote percentage increased - from 11.6% to 13.3%, an increase of 1.7%. Sammy Sosa, another outfielder with over 500 home runs and PED associations, saw a very similar increase in 2017, from 7.0% to 8.6%.
An increase of 1.6% - 1.7% for Ramirez would translate into an additional 7 votes for Ramirez.
Plugging in these numbers for Guerrero and Ramirez and zeroing out the numbers for Schilling and Wagner, then, produces the following table.
The total increase in votes for returning players predicted here, then, is 308. As noted above, if voters match the average number of names per ballot as last year, the combined vote totals of returning candidates could increase by 333 votes. That leaves 25 additional potential votes to be distributed among returning candidates, which would work out, for example, to an additional three votes for each of the top eight players in the above table.
Putting all of that together, then, the next table shows my first guess at how BBWAA voters may vote for the players on the 2018 Hall-of-Fame ballot. Predicted percentages for first-year players have been converted to whole numbers based on 444 votes this year with the resulting percentages then re-calculated.
Players in bold are predicted by me to receive the 75% necessary for election. Players in italics are predicted to receive fewer than 5% of the vote, rendering them ineligible for future ballots.
Wow! I worked through all of the above and left the math in the final table to fall where it would. And how about that. Four Hall-of-Famers this year with Edgar Martinez falling three votes short of producing a five-player class. And then three more players over 60%.
That's what the math tells me, so I'm going to go with it. We'll know in a few months how close I am.
Performance-Enhancing Drugs and the Hall of Fame
Unfortunately, lately, it has become impossible to talk about the Hall of Fame without bumping up against the subject of performance-enhancing drugs (PEDs). This is not a topic that I'm particularly keen on discussing here. My interest is in the data and, as far as the data are concerned, a home run is a home run and a player win is a player win, regardless of what a player did to hit that home run or earn that player win. I gave my opinion on PEDs in baseball in my first Hall-of-Fame ballot article about the 2013 Hall of Fame ballot if anybody cares about my opinion.
The Individual Players on the 2018 Hall of Fame Ballot
Over the next several weeks, I will write up an article about each of the 33 players on the 2018 Hall of Fame ballot. For the most part, these will not be advocacy articles: plenty of other people will post plenty of those. But hopefully, they will be interesting articles that may reveal something new and/or interesting, or at least a little fun, about these players, using Player won-lost records. I hope you enjoy them.
Links to these articles will be added to this article as these articles are posted.
2018 Modern Era Hall of Fame Ballot
2018 BBWAA Hall-of-Fame Ballot
2018 Hall of Fame Results
All articles are written so that they pull data directly from the most recent version of the Player won-lost database. Hence, any numbers cited within these articles should automatically incorporate the most recent update to Player won-lost records. In some cases, however, the accompanying text may have been written based on previous versions of Player won-lost records. I apologize if this results in non-sensical text in any cases.
List of Articles