An approach to frame consideration of steroids era players for the Baseball Hall of Fame.
I believe that one key reason writers charged with selecting the new members of the Baseball Hall of Fame are having such a big struggle with determining how to deal with the steroids era is that in their vote, they are being forced to oversimplify a complex problem, reducing it to a single outcome – “yes” or “no”.
Given the current trajectory, it seems most unlikely that large numbers of voters will ever cast their ballots for players from the steroid era based solely on the numbers. If one assumes the steroid question will always remain, the issue moves to what to do about it.
While there is no way everyone’s value system will ever align completely, I think it would help matters greatly if a common measurement system was used as a base. Again, I am not so idealistic to expect all voters will see matters identically, but context is badly needed around a subject where there is considerable disagreement and confusion.
In approaching this particular problem, I see three major variables. They are level of suspicion as to steroids use, level of admission by the player and the extent that steroids use may have enhanced the player’s stats.
My proposal is to generate discussion about how much each of those variables should reduce a prospective Hall member’s appeal. Understand that even of you disagree as to whether any one of the variables should be considered at all, you could nullify it by assigning it a zero value.
Let’s start with a tabular representation of the concept.
| Suspicion level |
High | High | High | Medium | Low | None |
| Caught | Caught | Caught | Appearance | Appearance | ||
| No apology | Half apology | Accepted apology | ||||
| Career results | ||||||
| High affect | A | B | C | D | x | x |
| Moderate effect | E | F | G | H | x | x |
| Little to no effect | x | x | x | x | x | x |
The horizontal axis represents the level of suspicion about a player’s steroid use, starting with whether they were caught, suspected based on appearance or circumstantial data or not considered to have used. The next level for those caught is how they dealt with it. Did they out and out lie, did they immediately fess up or did they do something in between?
The vertical axis represents the perception of how much steroid use may have affected the results of the player on the field, high, medium or low.
While there are 18 possible cells in the table, only eight are really worth discussing. They are the ones in the gray area, both figuratively and literally, labeled “A” through “H” above.
My hope is that voters could agree on this basic framework. The next challenge would be to try to decide which players fit in each. They could then begin to separate players into those more likely and less likely to get into the Hall. Even better would be to gain agreement on the penalties for players that land in a particular cell.
I am going to stop here for a moment. My primary objective in introducing this topic is complete. I do not want what follows, which is obviously very subjective, to cause people to throw out the whole concept if they disagree with my personal assessment below. Yet I think it is necessary to take the next step to fully illustrate what I mean.
How the matrix might be applied
We will start by assuming all players begin on level ground, with nothing artificially holding them back from Hall consideration – at 100 percent. We then introduce a scale to lower their appeal based on the previously-stated variables.
Again, these percentages are a starting point. Some may think they should be higher; others might want them lower or to be zero. That is fine. The whole idea is to establish a discussion framework which will then generate considerable discussion and even negotiation, perhaps.
We’ll begin with suspicion level. If a player is suspected of having used, even if he has never been caught, he receives a 10 percent penalty in this example. Right off the bat, I know some will accuse me of being un-American as the accused are innocent until proven guilty. I understand that, but reality is something else.
The reality is that a player suspected of use is never going to get as many votes as one who was never suspected. The formula simply takes that into account. It does no good coming up with a model that some would discard out of hand, yet if some wanted to assign 0 percent to those suspected, they could. At least the discussion would be on the issue.
The players who were caught are assessed a greater penalty here than those who were only suspected or apparently did not use. Some may believe the penalty should be the same no matter what, with others likely in support of a zero value. OK, but in this example, I assessed a greater penalty for those who clearly evaded the truth and exacted the highest penalty from those who thumbed their noses at their accusers and the public.
| Suspicion level | Penalty |
| Low/no suspicion | 0 percent |
| Medium appearance | -10 percent |
| Caught-apology accepted | -20 percent |
| Caught-1/2 apologized | -30 percent |
| Caught-no apology | -40 percent |
Next is the impact on a player’s stats of their use or suspected use. Certain Hall candidates were more one-dimensional than others. Some were suspected of using for a long time, including during their peak period of productivity, while others may have used just for a short period or may have done so later in their career.
Like everything else here, this scale is subjective. In my example, a low penalty is exacted from those considered to have put together Hall-worthy stats even without steroids with a high penalty assigned when numbers are viewed to have been significantly enhanced by the player having been enhanced.
| Career results | Penalty |
| No effect | 0 percent |
| Moderate effect | -10 percent |
| High effect | -40 percent |
Put it together and subtract from 100 leaves you with the following view. The color coding of the cells are a grouping. Those in the red cells, from 0-30 percent, are considered to have almost no chance of becoming a Hall of Famer, the yellow in the 40-60 percent range may eventually have a chance and the green population, 70 percent or better, should have a good chance of induction.
| Suspicion level |
High | High | High | Medium |
| Caught | Caught | Caught | Appearance | |
| No apology | Half apology | Accepted apology | ||
| Career results | ||||
| High affect | A 20% | B 30% | C 40% | D 50% |
| Moderate effect | E 50% | F 60% | G 70% | H 80% |
The final step is to assign example names to the eight cells, an action that may generate as much or more debate than setting the percentages. Again, the names are simply examples to help illustrate my point.
| Suspicion level |
High | High | High | Medium |
| Caught | Caught | Caught | Appearance | |
| No apology | Half apology | Accepted apology | ||
| Career results | ||||
| High affect | Palmeiro, Sosa | McGwire | Giambi | Bagwell |
| Moderate effect | Clemens, Bonds | Ramirez | A-Rod | Pujols era |
A. 20%. These players were caught using, but despite that, insisted they were clean. Their use is viewed to have greatly affected their numbers. As such, their Hall appeal is the most limited. Rafael Palmeiro and Sammy Sosa are my examples.
B. 30%. These players were also caught, but may have lied until forced to come clean. In the case of my example, Mark McGwire, his eventual admission was further clouded by an unconvincing stance that his numbers were not improved by his steroid use.
C. 40%. These players fessed up and their apology was generally accepted without reservation, yet like in A and B, their numbers are still viewed to have been significantly enhanced. Jason Giambi is the best example I could come up with. I think McGwire also could have been here had he simply made the common-sense acknowledgement that steroid use helped his results.
D. 50%. These players were never caught using anything, but have been suspected of use. Again, I am not debating right or wrong, but reflecting reality. Like A-C, these players’ numbers are thought to perhaps not been Hall-worthy had they not been suspected of using. Jeff Bagwell seems a good example.
Bagwell and McGwire
Let me step aside again to make a point. Some writers have used the McGwire vote total to suggest that even if Bagwell used, he should never admit it. I disagree, not with the bottom line, but with the comparison.
McGwire and Bagwell started in different cells. Like it or not, Big Mac was already squarely in the red area. His 2010 admission was made to allow him to work in MLB as the Cardinals hitting coach, not to fix his Hall problem. If you follow this model, the best he could have done was move into cell “C”, likely not good enough, anyway.
On the other hand, even a believable admission from Bagwell would bump him down from “D” to “C” on this scale, a ground-losing proposition in voters’ eyes. Of course, best for him would be ironclad proof that he never used, but that is impossible to secure.
Along with the general passing of time, a release of more confidential lists and positive test results without his name on them would help Bagwell and others from his era potentially unfairly cast in cell “D”.
| Suspicion level |
High | High | High | Medium |
| Caught | Caught | Caught | Appearance | |
| No apology | Half apology | Accepted apology | ||
| Career results | ||||
| High affect | Palmeiro, Sosa | McGwire | Giambi | Bagwell |
| Moderate effect | Clemens, Bonds | Ramirez | A-Rod | Pujols era |
E. 50%. Like the players in A, these players have been caught using but would not admit it, despite considerable evidence to the contrary. What makes this group different is that their numbers may have been Hall-worthy even before their suspected steroid use began. Barry Bonds and Roger Clemens are among those who seem to fit here.
F. 60%. These players were caught and came clean, sort of. There is some concern their use impacted their career numbers, but perhaps not excessively. Suspended late in his career, Manny Ramirez may be a good example.
G. 70%. This is another group of players proven to have used steroids, but set apart in that their admission seems generally accepted and their career results may be good enough to stand on their own. Alex Rodriguez is my example.
H. 80%. The final group is pretty much everyone else in baseball that has excelled on the field since more rigorous testing began. I use Albert Pujols as a placeholder. The superstar never having failed a test doesn’t stop some from whispering and wondering. Is it fair? It doesn’t matter. It happens out there in the real world, so it needs to be accounted for.
In closing
I understand the limitations of this model. One of them is that the actual result of the Hall of Fame voting is the summation of several hundreds of individual “yes” or “no” votes. One person cannot vote 60 percent “yes”, for example.
Yet over time, I can see how some players in the 40-60 percent “yellow” area could receive enough support to reach the necessary 75 percent for induction. Others may have to wait for a future Veterans Committee to consider their case. Other players may move between cells as more information about the past comes to light, whether positive or negative.
Still, whatever values you believe in and wherever you might place these players in the grid, I believe that by using a framework now to assess players from the steroid era, everyone may better come to grips with the fundamental problem.
Follow me on Twitter.
Follow The Cardinal Nation Blog on Facebook.
Do you not think there could be a difference between those who admit it but weren’t caught (such as McGwire) and those who just plain got caught such as Manny Or Palmiero?
It gets down to a proven failed drug test vs. a heavy weight of evidence. Those are fair points to debate.
By the way, I am curious about your opinion. What in the above do you think might be sabermetrically-oriented? The use of percentages, tables and colors?
Nothing there looks sabermetric to me except the fact that you applied a little analysis — Something sportswriters might want to try to do more often
I’m sure the true saber guys would kick you out if you tried to reprsent that as sabermetric!
That’s quite an analysis Brian. How long did it take you to come up with that?
I do have a question. How did you decided which players performances were only moderately affected as opposed to highly affected? I for one would rather see Mark McGwire in the hall than Barry Bonds simply because I believe he is lying and that holds more weight with me than how it affected his playing. I don’t think character issues should have more weight than performance but they should be given significant consideration.
The idea had been around for awhile, but once I started typing on Saturday afternoon, it was basically done in a couple of hours.
It is kind of wonky as a CBS writer noted today on Twitter, but he also suggested that it is better than hysteria. He had commented in response to a Joe Strauss tweet that suggested this is “Sabes run amok”, a most odd statement.
Assigning players to moderately vs. highly affected was clearly a personal call. To use the two names you mentioned, my view that McGwire was a one-dimensional player with his primary Hall stat being his home runs, also one of the most likely stats to have been enhanced by steroid use. A number of observers believe that Bonds was on a Hall trajectory as a superb all-around player before he apparently began using. As such, on the results scale, McGwire was docked 40 percent and Bonds just 10.
On the suspicion scale, Bonds was docked 40 percent, McGwire 30. It would have been only 20 for Mac had his admission been more credible. When all is said and done, they both still used.
My initial approach was that results and suspicion were equally scaled 0-50, while I sense you may be in favor of increasing the suspicion scale and reducing the results scale. Others may feel the opposite, not caring who lied and caring more about the bottom line stats.
Of course, there is nothing magical or hard and fast about these percentages. They could and probably should be adjusted as more people get their heads around the concept.
Thank you for the explanation Brian. I am impressed with the idea though I would give more weight to the suspicion issue.
Oh, and Joe Strauss is a putz. I hope he reads this.
Good idea to put some thought into this discussion.
“On the suspicion scale, Bonds was docked 40 percent, McGwire 30. It would have been only 20 for Mac had his admission been more credible.”
I think McGwire’s admission was VERY credible and complete. Some people object to his statements about how much he believes they enhanced his performance. To me, that is independent from his admission of use. Why do I care how much he believes it helped his performance? It should be up to writers (in the case of the HOF) to weight the performance enhancing roll, totally independent of the player’s belief.
So in your version of the model, you would apparently eliminate cells B and F. Would you also remove C and G? In other words, do you not care if a player apologized or not?
One way to inject your beliefs without changing the base model would be for you to score those cells in question in your scoring the same as A and E, treating all known users the same. The downside is that some voters may not see it as such a black and white situation as you may. In fact, your approach may have an exception or two. On the main board, I think you mentioned that you would treat Bonds differently because you don’t like him. Not sure how that would be handled. Maybe the addition of a penalty box…
Again, no right or wrong answers.
I was actually joking about Bonds. Yes, I don’t like him, but I would not actually evaluate him differently because of that. I would just like to take 3 of MVPs and give them to AP.
As far as admissions, I would not give anyone much credit for an admission. I don’t think any admission is sincere if it is based on a positive test, or even a credible accusation. This is where I might give McGwire more credit than most, because there was no proof and little credible evidence against him. But to me a sincere admission would not be because he wanted to return to MLB and it would not have taken so long. IMO, only Caminiti deserves any credit for an admission.
“In approaching this particular problem, I see three major variables. They are level of suspicion as to steroids use, level of admission by the player and the extent that steroids use may have enhanced the player’s stats.”
Brian………………………………………. this is not your best work……….its seems ill advised to me. I know you have an affection for ” statistical analysis”, and have some very strong opinions/emotions surrounding this topic ……………….but your approach here seems to show some of the very weakest vulnerabilities of that approach………… I’d can it.
Between the time I started writing………….and hit the button……….this new post came up. To late? Or never to late?
Bottom line is would the Hall of Fame Directors in Cooperstown accept this and the answer is a resounding NO WAY .
Personally, I am way past being tired of all the steroid HOF columns being trotted out by many sportswriters.
As you all know, I disagree with penalizing baseball players for known or suspected steroid use by denying them entry to the HOF.
Brian, I can see where this article is coming from, but would you agree that we need a cocaine era matrix and an amphetamine era matrix (these would overlap!), a hot dog era matrix (Babe Ruth),
an alcohol era matrix (made the ball look bigger to some batters) ? etc., etc.
When exactly is the Pure as the Driven Snow (never looking for an advantage on the field or never cheating) matrix?
Looking forward to spring training!!!
chief, I could join the ranks of the complainers, which would be very easy, but what would that accomplish? More noise that leads to nothing but more noise.
I could join you and condemn the writers for not seeing the world your way, but what would that fix? Nothing.
Instead, I tried to take a realistic view of the situation. Not some ideal that is impractical and unattainable. It is pretty obvious that the writers are struggling about what to do with the steroid era. There seems no current hot dog problem.
To answer your last questsion, “Pure as the Driven Snow” is the cell at the far right bottom of the first graph, one of them with an “x” in it. I didn’t spend any time there because there is nothing to say…
Brian……………………..Your naivety concerning the Politics of this issue………. the direction and thrust of the investigation………….it resolution…………….yes…….in the most recent collusion agreement……
your methodology considers none of the relevant variables………… if it did…..you wouldn’t be publishing this piece……………….
Very likely………with a new commissioner……..at the very end of Bond’s eligibility, he will qualify, opening the door for others…………. the commissioners qualification at that point will effect the criteria………………………. this is………and always was about the MLB owners/ and the Union……… it isn’t about the sports writers………….. MM will never make it……… he crossed over. Wild guess that the grease throw at Jeff B. had more in it than cash…………….
WC, if what you are suggesting is that MLB, owners, union, whoever is dictating how hundreds of sportswriters cast their Hall of Fame votes, then I am not buying it. If it is something else, as far as I am concerned, we can just let it pass. I am ok with being naive and uninformed…………..
You think, after Bonds was kept from playing baseball……….is never prosecuted, or convicted, that he wouldn’t be in a position to file a civil action….considering what he knows about his case…… How the actions of MLB slandered and defamed him, preventing him from achieving his proper recognition????????????????????? these things have already been decided………. MLB is paying and they will work out the political campaign that will exonerate a few……………….as it will condemn a few…..MM………..Everyone has already been tried………its over………. negotiated…..and kept secret.
WC knows everything……….so don’t bother…….
So far, it does not appear the writers are struggling – they’ve spoken loud and clear. It will be more complicated as some of the media darlings become eligible.
Perhaps you’ve missed the scores of writers sermonizing to their peers (but all the rest of us have to read it, too) why they aren’t voting properly (which would be in agreement with them).
My point is that there has never been an era without players using illegal drugs or cheating to be a better ballplayer, and yes some of these users have been voted into the HOF .
If memory serves correctly there were a number of bicycle racers in the 1904 Olympics caught blood doping.
I just think it is hypocritical of todays writers, to be so high and mighty with their judging and speculation, when many of the same writers knew what was going on and wrote articles praising McGwire, Sosa etc. for “saving baseball” .
I pretty much agree with Joe Posnanski, Bill James and Rob Neyer on the steroid thing and think 30 years from now people are going to wonder what all the fuss was about.
Appreciate your effort Brian in this post and I’m not criticizing you specifically, I’m criticizing the whole hubbub over steroids etc.
No problem at all, chief. I just decided to try to look at it in a different way. Discussion is good.
Don’t forget the pine tar, resin, cork, and spit matrices!
Interesting. Maybe Sosa should get extra credit (or demerit) for being in multiple formulas.
What do you guys think of Ken Rosenthal’s answer to the steroid issue, saying he won’t vote for anybody their first year of eligibility.
Probably a fair position.
I haven’t read the article, but the concept doesn’t seem all that revolutionary. I can’t imagine any player who was caught using, apology or not, would receive enough votes to get in the first year.
On a tangentially-related topic, I find the whole first ballot tradition irritating. Just because the first voters 70 or 80 years ago did not unanimously select the first inductees, then no one forever more will be unanimous either, no matter how good of a player they were. (This has nothing to do with steroids, of course.)
It is silly that noone is voted in unaminously. Certainly there was a backlog of qualified candidates in 1935 , but it seems that is no longer a problem