Will Clark for Cooperstown

[Brad Harris]

Take a peek in the Stats/Sabermetrics forum at the thread entitled "Bill James and the NHBA Ratings" for more information.

After running some preliminary numbers, if you remove James' opinions as expressed in the subject element (and, often, in the player comments) from the equation, then his ratings system as expressed in the Abstract shows Will Clark to be the 5th-greatest first baseman in history (just behind Gehrig, Dick Allen, Foxx and Bagwell.) I'll say that again.

According to Bill James' formula, Will Clark is the 5th greatest first baseman in history.

That shouts "Hall of Fame" to me, especially considering he's already a solid borderline candidate. If you put any faith in win shares, you've got to agree that Clark is a Hall of Famer.

I was an unsteady supporter of Clark's already, but now that my faith has been reassured, my doubts are gone.

Hop on board the bandwagon now. Destination Cooperstown, starting in 2006!

[Ram Man]

My opinion? Not a chance.

[rockin500]

no way. he doesnt deserve to be in. i dont care how you crunch bill james' numbers, will clark does not deserve to be in. He was a good player for a number of years, but definitely not hallworthy.

[Biggtone23]

He will be the 3rd best firstbaseman on the ballot that year. . Mattingly and Hernandez are waiting until they make it on to veterans committie ballot and Clark will have to the same and may have to wait until the middle or end of this century when some special committie on late 20th century players takes a look at him

[Brad Harris]

Originally posted by Biggtone23
He will be the 3rd best firstbaseman on the ballot that year. . Mattingly and Hernandez are waiting until they make it on to veterans committie ballot and Clark will have to the same and may have to wait until the middle or end of this century when some special committie on late 20th century players takes a look at him

Whoa...wherever you want to draw the line for entry to Cooperstown, Will Clark is definately ahead of Don Mattingly in the line. Mattingly has a weaker case.

[The Commissioner]

Chancellor, while I wouldn't necessarily argue with the idea of Clark being a Hall of Famer, the fact that James ranked him as "the 5th greatest first baseman" in history shouldn't be used as a factor. That speaks more to what kind of an "expert" James is than to how great of ballplayer Clark was.

[BoSox Rule]

Originally posted by Chancellor
According to Bill James' formula, Will Clark is the 5th greatest first baseman in history.

Damn, I wan't James as far away possible from the Red Sox after reading that.

[cubbieinexile]

Originally posted by Chancellor
Take a peek in the Stats/Sabermetrics forum at the thread entitled "Bill James and the NHBA Ratings" for more information.

After running some preliminary numbers, if you remove James' opinions as expressed in the subject element (and, often, in the player comments) from the equation, then his ratings system as expressed in the Abstract shows Will Clark to be the 5th-greatest first baseman in history (just behind Gehrig, Dick Allen, Foxx and Bagwell.) I'll say that again.

According to Bill James' formula, Will Clark is the 5th greatest first baseman in history.

That shouts "Hall of Fame" to me, especially considering he's already a solid borderline candidate. If you put any faith in win shares, you've got to agree that Clark is a Hall of Famer.

I was an unsteady supporter of Clark's already, but now that my faith has been reassured, my doubts are gone.

Hop on board the bandwagon now. Destination Cooperstown, starting in 2006!

How did Bill James ranks Will Clark as the 5th greatest first basemen of all time? If you look at just Win Shares then Will ranks 11th or 12th. I forget which. While his personal opinion has him ranked 14th.
Plus I think you are doing the Win Shares wrong. If the subjective element is a 50 point system than all of your win shares totals should be within 50 points of the totals Bill James gives. Yours do not. For instance you have Lou Gehrig listed at 160.92 win shares. While James gives him 489 Win Shares. A difference 328 Win Shares. So like I said I think you are calculating wrong.

[cubbieinexile]

One more thing,
The subjective element is part of Win Shares it had to do with the rankings. Meaning your win shares should match up to the ones Bill lists in the book. So for Will Clark you should have gotten 330 not 140. Because the subjective element was something that Bill added after the Win Shares compilation to flesh out the players. For instance a team leader who motivates his team would be more valuable than a sullen player who stays in the back even if they have the same stats. Or if a player lost time to the war this is his way of keeping them from being unfairly penalized.

[Brad Harris]

Originally posted by cubbieinexile
How did Bill James ranks Will Clark as the 5th greatest first basemen of all time? If you look at just Win Shares then Will ranks 11th or 12th. I forget which. While his personal opinion has him ranked 14th. Plus I think you are doing the Win Shares wrong. If the subjective element is a 50 point system than all of your win shares totals should be within 50 points of the totals Bill James gives. Yours do not. For instance you have Lou Gehrig listed at 160.92 win shares. While James gives him 489 Win Shares. A difference 328 Win Shares. So like I said I think you are calculating wrong.

Either I you haven't read how his rating system (in the Almanac) worked or I have failed to communicate as much in the other thread.

I went through the first base position - the top 100 players - as rated by James in his Almanac.

Though his ratings are based on the win shares method, they are not based on career win shares.

In fact, James' system heavily favors (to the tune of around 75%) great performers with shorter careers over merely good performers with longer careers (which isn't necessarily wrong.) This is before he makes subjective adjustments for time and place and for the "intangibles."

James uses four criteria, pretty much equally weighted, for the objective side of his ratings. They are:

1. Career value.
2. Average value of three best seasons.
3. Average value of best 5-year run.
4. Career average.

The numbers you are thowing at me are the career totals, not the number James uses to establish his rankings, but only a small part in that equation.

The final number James arrives at doesn't really have a name. He doesn't publish it anywhere. I had to do all the calculations myself based on the component data James did provide in the Almanac.

I'll do Lou Gehrig here so you can understand better.

James ranks Gehrig as the greatest first baseman in history in his Almanac. Let's look at how he arrived at that conclusion.

James has six factors which went into these ratings.

1. Career Value
Lou Gehrig does, indeed, have 489 win shares. As described in the Almanac, James divides that number by ten, with a result of 48.9 and finds the harmonic mean between 25 and 48.9; this gives Gehrig's career value a weight of 33.09.

This doesn't mean Gehrig's career value is 33.09 win shares. It's simply 33.09 in terms of this formula. That number (33.09) represents his career win shares, but it isn't win shares as a unit of measurement.

2. Best 3 Seasons
Found on page 476 of the Almanac, if you don't feel like taking the time to pour through Gehrig's season-by-season totals in Win Shares, Gehrig's 3 best seasons netted totals of 44, 42 and 41 win shares.

That's a total of 127. Divide by 3 and our value for the second component is 42.33.

3. Best 5-year Run
Same page, total of 193 in shares in five consecutive years. This translates into an average of 38.60.

4. Career average
Again, this value is listed on the same page in the Almanac and is given as 36.61.


Add all these components together and you arrive at 150.63; there are two components remaining.

5. Timeline adjustement
James defines this as year of birth minus 1800, divided by 10.

Gehrig was born in 1903. That would give him a timeline adjustment of 10.3.

Add that to our sub-total and we see that Gehrig (minus the sixth and final, subjective element) has a score of 160.93 (I mis-typed it on the other post, where I put 160.92).

This does not say that 160.93 represents 160.93 win shares, but just a rating number of 160.93 which is based on win shares (plus the timeline adjustment.)

Now...what I'm able to do, thanks to what information James has provided is to calculate that 160.93 represents all the data James put into his rankings except for the subjective element (which we know is between 1 and 50 points.)

All I can do is calculate the individual elements and the sub-total without the subjective element.

Personally, I think adding the subjective element to this list was something of a mistake, at least without attempting to defend (in the player comments) what those scores were and why he selected them.

Now...empirically (as this particular win shares-based system suggests), Will Clark rates as the 5th greatest first baseman in history.

What infuriates me is how Will Clark rates significantly higher than Don Mattingly, yet Mattingly is given a higher subjective element to the point that James can say, in the Almanac that Mattingly was better than Clark, but not bother explaining why.

The only conclusion one can draw, from James' own numbers, is that James just thinks/feels he is. And that's too subjective to be included in a stats-based system.

As someone else pointed out (and this is the whole point of my doing this exercise), this should demonstrate, not that win shares is a crummy system per se, but that James' rankings are as biased as yours or mine would be.

However, these are James' numbers.

[yellowdog]

My opinion on Will Clark is that he had a very nice career. He was also a fierce competitor who gave everything he had. A clutch hitter who had one of the sweetest swings for a power hitter I`ve ever seen.

But as far as the HOF, my gut tells me...close, but no cigar.

[cubbieinexile]

Now I see where you got your ratings. But I don't think you can stop before the subjective part to make your case for Will. I understand that Bill leaves out that critical info but to leave that part out in your arguement for Will is wrong. The elements that are subjective in nature are in fact very important to the Hall of Fame voters. For instance World Series performance and Leadership two key qualities for voters. Evidently if your ratings are correct then Bill feels that 6 or 7 players below Will were better at the 5 or 6 elements he was looking at.

Basically what I am saying is that the whole formula should be looked at and examine not part of it. That would be like stopping half way in a algebraic formula. It just doesn't work. For Bill James Rating system it A+B+C+D+E=F, and your stopping at D and calling it a day. I understand why you are doing it, but we also just as easily take A away or B if we wanted too and get different results and use that to make a case for some other player.

I understand that Bill makes it difficult to fully understand his ratings, but I don't think that means his partial answer should be used as evidence for Will's induction.

[Brad Harris]

Originally posted by cubbieinexile
Now I see where you got your ratings. But I don't think you can stop before the subjective part to make your case for Will. I understand that Bill leaves out that critical info but to leave that part out in your arguement for Will is wrong. The elements that are subjective in nature are in fact very important to the Hall of Fame voters. For instance World Series performance and Leadership two key qualities for voters. Evidently if your ratings are correct then Bill feels that 6 or 7 players below Will were better at the 5 or 6 elements he was looking at.

Basically what I am saying is that the whole formula should be looked at and examine not part of it. That would be like stopping half way in a algebraic formula. It just doesn't work. For Bill James Rating system it A+B+C+D+E=F, and your stopping at D and calling it a day. I understand why you are doing it, but we also just as easily take A away or B if we wanted too and get different results and use that to make a case for some other player.

I understand that Bill makes it difficult to fully understand his ratings, but I don't think that means his partial answer should be used as evidence for Will's induction.

The only part I'm leaving out is that which cannot be calculated without James providing more information than what he has.

Personally, I believe the criteria of "greatest" should be determined by stats, primarily. Win shares does a great job of removing the relativity of time and place from those numbers and putting everyone on a level playing field. If you've got two players who are 5% apart, say, in the final ranking...then you might apply the subjective content.

James, however, wraps all "intangibles" into one - everything from clutch hitting to post-season performance to time lost to WWII - and gives it a number of his choosing (which, by the way, has the potential of being a larger component than 99.99% of all other individual components for all players so rated).

The point is that Bill James isn't any more qualified than you or I to assign a 1-50 point value to the "subjective" criteria he uses for each of those players. Perhaps James placed a little more emphasis on the value of World Series play while you or I might place a little more emphasis on the importance of being perceived in your own era as being the best in the game? Perhaps James felt that winning a gold glove was more impressive in this regard than leading the league in defensive win shares? Who knows? And that's the point.

Going solely by the evidence that is provided is all I can do and nothing more.

And that makes a whole lot more sense to me than just accepting that "Bill James ranks Will Clark 14th all-time."

[cubbieinexile]

I understand what you are saying, and I agree that the rankings are Bill Jame's opinion. But at the beginning of this thread you used Bill Jame's formula to state that Bill Jame's formula has Will Clark as the 5th greatest 1B. You then proceeded to use the partial information to support Will's case for the HOF and blur the lines between (which is where you got me) win shares and ratings. The subjective element is part of the rating system, it is that what cannot be statistically measured by Win Shares. And those things are important. If anything you should fill in blanks with your own opinion on those things that Win Shares do not measure. Follow Bill's criteria for the subjective element and then see where these players land.
Bill never intended his rankings to be based on only the win share element. If he did he would have published that part or put it in the book somewhere. He did not because Bill understands that Win Shares does not cover all the ground that is required to fully rank players.

Finally to look back at your original statement you said that if you have any faith in Win Shares then you got to agree that Will Clark is a HOF'er. I disagree because the ratings even without the subjective element is not Win Shares. It is a formula based on Win Shares. Which is a big difference. The formula itself is a matter of opinion. Why is 5 consecutive seasons used? Why is it three best seasons overall? Why use the Harmonic Mean? Plus like you said Bill Jame's favors peak over career and a lot of times his step for best 3 seasons will be higher than his number for career total (since he uses Harmonic mean and the baseline of 25 it lowers the importance of a long productive career). Which as you probably have noticed tends to hurt players like Eddie Murray. Consistently good players who don't have a lot of peaks and valleys are hurt the most in his formula. Bill's formula rewards greatness but limits the penalty on mediocrity by minimizing offseasons and injuries. Which is why the subjective element is so important.

If we look only at win shares and in no-formula ways. Such as total, yearly total, and seasonal average it shows Will Clark to be a very good player though I don't think one that should go into the hall. Especially since he played first base.

In a nutshell what I am saying is for you to finish the equation and then make the case.

[cubbieinexile]

The study of Bill Jame's rankings have intrigued me.

Bill's formula really likes peak over career and I tend to disagree with that to a point. For instance if we break down Bill's formula into part likes you did I see some things I disagree with.

Take Ty Cobb for instance.
Ty Cobbs Career total numbers comes out to 37.19
His three best season average comes out to 47.67
Right there I see something I disagree with. Bill's formula weights three seasons of a players career more heavily than the players entire career.
The next step is 5 consecutive and Cobbs comes out to 43.00
Now I don't totally disagree with this one in that he is measuring sustained greatness or lack there of. So in some ways this step helps reward players that are continually good instead of the peaks and valley type player. But why stop at 5? If we are measuring the greatest of all time shouldn't we hold them to a higher standard? Also in a lot of cases this step doubly awards players from there 3 best season step.
4th step is Career Wins per season and Cobb comes to 38.76
I have some problems with this step but overall I like it. The only problem with this step to me is that it minimizes the penalty a player has by missing time. The first step penalizes players with short careers or players that missed chunks of time, but this step can give some of those points back by factoring lost time out. Since this step only measures total Win Shares/ total games. Perhaps a better method would have factored in total seasons into the equation. Perhaps an average of the two.
The Final Non-subjective step is Time-Line adjustment is 8.6
I understand this part, but this part is really subjective. He just happens to have come up with a formula for it. So that we can all see it instead of having him lump it into the misc pile. The only problem I have with this is we really don't know if it is all that accurate. For instance becuase of this Griffey gets twice the bonus that Cobb did on this. Is the competition and difficulty level in Grffey's day twice as competitive and twice as difficult then Cobb's day? I don't know but I think I would like to see some statistical work to go along with part of the formula. Perhaps a plateau at some point. Players who played after 1970 are treated like they were born in 1940 or something like that. I can't really believe that the game gets more and more difficult as years go on. Nor can I believe that there were not periods in time when baseball was less competitive then in years gone by. For instance what about the expansion era 60's? Was that decade really more difficult and more competitive than say the 20's? Should players from 60's get a 4 point advantage over players from the 20's? Also as years go by the youth bonus is going to get so large as to be higher than the other steps! For instance a player born in 1980 is getting a 18 point bonus in his total.
The last part is the unknowable on our part. And I would agree with Chancellor and say that not knowing the total for this part makes it hard to really get behind these rankings. For instance is each element weighted evenly? Does he have full possession of the facts to be able to properly identify true leaders and disrupters? For instance he thinks Dick Allen was a clubhouse cancer. Some of Dick former teammates and coaches disagree with that view. Plus who really knows how in detail he got on all these players for this part. Does he really know if a player from 1892 was a clubhouse leader or not? Then of course there is the fielding element he puts in there. Bill James has been wrong before about fielding and there is no reason to think he can't be wrong now.

That is about all, If I were to create a system of ranking based on Win Shares I think I would increase the consecutive seasons part from 5 to 8 and change the three best seasons to 5. Also I think I would get rid of the first part (Like Bill himself said, to get that part he had jump through a lot of hoops. Which to me means it shouldn't be done) and try to find some other way. The more I think about it I don't know why this is necessary anyway. While I haven't crunch the numebrs for all the players looking at the numbers Bill provides the harmonic mean and the WS/162 games closely resemble each other. Perhaps just using WS/162 and then in the subjective element removing points for injuries would be best.
By doing it this way I think you would better balance the two type of players measured. The ones with great peaks and the ones who are consistently good.

[Biggtone23]

Anyone else hate Bill James for trying to turn baseball into Math class?

[The Commissioner]

I do, but not so much for trying to turn it into a math class as much as for trying to turn it into a math class using flawed, skewed, and subjective numbers. Professor James seems to believe that his is the only correct answer key.

[J W]

...except I keep reading disclaimers quoted BY HIM that his word is not the gospel.

Could be that what you don't like is the use of his formulas as the #1 basis for everything in the sport.

I think cubbie's points ring a lot of truth. the numbers... 3 and 5, and the harmonic mean as opposed to straight averages--they can be tinkered with to produce different results.

But I like Bill James' work... I think the Win Shares system is better than nothing (though some other systems have advantages elsewhere--I haven't thoroughly studied the others).

...and I LIKE MATH. It's the universal language. So there.

[Cougar]

I like math too.

James doesn't get everything right. No one does. What James does best is ask good questions. He then tries to answer them, and usually comes up with a pretty good response. Sometimes he doesn't, but no one who takes intellectual risks bats 1.000, and that's OK.

I agree that his subjective rankings should be a little more transparent. That is a fair criticism. I also think Chancellor did a terrific job deconstructing James's method, adjusting it to trim it of most of the obvious subjectivity, and then scrutinizing the results.

And yeah, it's a heck of a statement when you're fifth all time on such a scale. This takes into account things that get overlooked easily. Clark put up his best seasons in Candlestick -- a truly lousy hitters park. And, he was an incredibly solid player. First basemen who hit for average and power, plus field their position well, and do it over a fairly long career, just don't come along that often. When you look at the record, there's very few guys like that.

I thought Clark belonged before, but I thought he was borderline. This tells me that he's a solid pick for the Hall.

That said, a system that ranks Dick Allen ahead of Jimmie Foxx sort of lacks some face validity. With all due respect to Richie, there's just no way you can defend that to anyone who knows the history of the game. I don't blame James for adding an extra wrinkle to fix results like that -- the system needed it. But you get unintended consequences with a fix like that, like nudging Mattingly ahead of Clark. (Of course, Mattingly & Clark is a heck of a lot closer call than Allen & Foxx; I'd rather err on the former than the latter.)

[rockin500]

i have little use for bill james.

[BigRedMachine]

I don't think Will 'The Thrill' Clark has a prayer of a chance to get in the HOF.

What impresses me the most about Clark is his lifetime .303 batting average. He just did not have enough of an impact on the game to be a serious HOF candidate.

The following players are not in the HOF as well:

Dick Allen
Don Mattingly
Keith Hernandez
Al Oliver
Bill Buckner
Steve Garvey

Clark can't be in the HOF without these guys going in too.

[Fuzzy Bear]

Since this thread, Will Clark has been relegated to "one and done" status by the BBWAA. He did not get enough support to stay on the ballot.

There were several reasons for this, IMO:

(A) Clark did not play in 140 games after age 29, except for 1998.
(B) The reason for this was that Clark became frequently affected with injuries.
(C) Clark was not a home run slugger; he was Mark Grace with more power.
(D) Clark retired a bit early; he had a good 2000 season, but packed it in after that. Had he been able to stay at his career norms a bit longer, and keep his BA over .300, he might have been able to stay on the ballot.

I have come to regard Clark as superior to Mattingly, all things considered, although my perception may stem from Mattingly finishing poorly.

[538280]

Will Clark should be in the HOF. He was a truly great hitter at his peak, and a very, very good one through his career. Bill James addresses in his book Win Shares why exactly Clark's 1989 season (44 Win Shares) rates so well. Because looking at it in the Encyclopedia it looks like a very, very good but not truly great season. I think his explanation makes sense. Basically Clark starts out with MVP type numbers (.333 average, 23 HRs, 111 RBI, .407 OBP), and every extra step you take to evaluate him more closely he comes out just a bit better. Let's say Clark starts out at what James says is the borderline MVP candidate level (30 WS), and with each extra step he jumps up about 2 more WS.

-He only grounded into 6 double plays, a great total for a guy who played 159 games

-He was 8 for 11 as a basestealer

-The 1989 NL, especially in Candlestick Park, was an extreme pitchers' environment. The league averages were .250/.317/.373, reminescent of the "2nd deadball era" in the 1960s. We don't tend to think of that when looking at the season, because the 1980s in general were not really a tough decade for hitters. Nor were the seasons around 1989 particularly extreme (offensive levels jumped up quite a bit the next year in 1990). Candlestick in 1989 was also particlarly tough on hitters, James says the park factor was 91 (for some reason BBRef lists 97, but James definitely says in his book it was 91).

-Clark did particularly well in clutch situations (James doesn't think that clutch hitting is a real skill, but he does think that players who do better in the clutch in a particlar season are more valuable in that season). Clark batted .389 with runners in scoring position and hit 13 of his 23 homers with men on base.

-Clark was an outstanding defensive 1Bman.

I think when you factor all of those things in, you find Clark raises significantly above the marginal MVP level, and becomes a truly great player for that season. 44 WS may be slightly overstating his greatness, but I think Clark was a legitimately tremendous player in 1989.

-

[thecoach]

Will (The Thrill) Clark would get my vote. He was a team player and gave it his all. And he had the sweetest swing in memory!

[Fuzzy Bear]

I would vote for Clark ahead of Mattingly, but behind Keith Hernandez and Dick Allen.

Keith Hernandez has a surprisingly strong case, IMO, and I am surprised that his candidacy generates so little interest.