Baseball for the Thinking Fan

Login | Register | Feedback

You are here > Home > Primate Studies > Discussion
Primate Studies
— Where BTF's Members Investigate the Grand Old Game

Wednesday, February 20, 2002

Baseball Survivor: Transcending Reality with Style

Voting off the top 100 players of all-time

As we get ready to stow away the harsh realities of the great
2001-02 “offseason of discontent,” we should take a moment to hand out
praise to those creative individuals who have helped us make it
through a cold, cold winter.

I’m speaking of Tom Hanrahan and Justin Kubatko,
whose twist on “reality TV” (an annoying boil on American
semi-consciousness that is, thankfully, contracting before our very
eyes) has been a joy to watch unfold.

Their website,


Baseball Survivor uses the “voting off the island” concept
to create a new take on that old chestnut of fan pastimes, the Top 100

Hanrahan’s concept-take a “top 100” list created by a consensus of
deep-dish baseball fans (in this case, nineteen members of the


Society for Baseball Research) and apply
the “voting off the island” concept incrementally until there is only
one player left-is something of a cross between the vote-off methods
familiar to us from Survivor and The Weakest Link. The
main difference-no ganging up, and no backroom deals a la figure
skating (for example).

Kubatko’s implementation of the concept on the Baseball Survivor
website is well-nigh perfect, with detailed voting results and a
master page which shows those who’ve been asked to leave the island
with lines drawn through their names.

The page also demonstrates how Hanrahan’s concept works in terms
of creating an ordered Top 100 list. As each player is voted off the
island (in weekly balloting that is now up to Week 26, and is
scheduled to end around the fourth of July), they get assigned to a
slot on the list., The first two players voted off the island?





There are currently twenty players still “on the island.” Take a
look at the list and see how well you think the voters are doing:

Hank Aaron

Grover Cleveland Alexander

Barry Bonds

Roger Clemens

Ty Cobb

Eddie Collins

Lou Gehrig

Lefty Grove

Rogers Hornsby

Walter Johnson

Greg Maddux

Mickey Mantle

Willie Mays

Stan Musial

Babe Ruth

Mike Schmidt

Tris Speaker

Honus Wagner

Ted Williams

Cy Young

The most recent players to be “voted off” (and thus assigned
rankings in the Top 100) are:

21. Nap Lajoie

22. Frank Robinson

23. Christy Mathewson

24. Jimmie Foxx

25. Joe DiMaggio

26. Joe Morgan

27. Tom Seaver

28. Eddie Mathews

29. Rickey Henderson

30. Mel Ott

Complete results can be found on the site, along with some
detailed statistical profiles for each player put together by Kubatko.

A similar project oriented around ranking current players might be
an interesting follow-up for either this group or another using its

A more detailed effort, working from a positional standpoint,
could also be attempted. This would make for an interesting corollary
to the rankings by position that are featured so prominently in Bill
latest edition of the Historical Baseball Abstract.

Aficionados of baseball history will enjoy this site-and they’ll
be especially interested in it now that the finish line is coming into
view. Kudos to Kubatko, Hanrahan, and their associates for creating a
marvelous diversion during a harsh winter of ill winds.

Don Malcolm Posted: February 20, 2002 at 05:00 AM | 51 comment(s) Login to Bookmark
  Related News:

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

   1. Robert Dudek Posted: February 20, 2002 at 12:23 AM (#604786)
Joe Morgan was voted off too early and there are too many outfielders (9 out of 14 position players, by my count).

Morgan should be in there at the expense of Tris Speaker. Mathews or Foxx instead of Cobb might be defensible and would even out the outfielders/infielders a bit.

   2. David Jones Posted: February 20, 2002 at 12:23 AM (#604787)
Did I miss something? Where is Eddie Murray?

   3. Charles Saeger Posted: February 20, 2002 at 12:23 AM (#604789)
Hornsby and Lajoie ahead of Morgan? Indefensible, pun intended.
   4. John (You Can Call Me Grandma) Murphy Posted: February 20, 2002 at 12:23 AM (#604790)
At this point in his career, Maddux should be behind Seaver (it's very close between Tom and Roger). Talk to me a couple years from now.
   5. Eric Posted: February 20, 2002 at 12:23 AM (#604791)
I love Greg Maddux and Roger Clemens, too, but voting off Christy Mathewson before either of them seems to be the result of some fairly modern-day biases. Frankly, because I think that the transition from pre-1901 to post 1901 baseball is a bit tricky, I think Mathewson ranks just behind Johnson as best all time, with Young, Alexander, Grove, Clemens, Maddux and Seaver fighting for third place.
   6. Mike Emeigh Posted: February 20, 2002 at 12:23 AM (#604792)

Look at the group of 103 players that these folks selected. Is there anyone in that group who "doesn't" deserve to be considered as one of the 100 best of all time? There are lots of other players (like Eddie Murray) who have an argument, but the list was limited to 100. If the list had included another 50, we'd still be having arguments, just at a lower level. I don't agree with every selection, but the list these 19 guys came up with is very reasonable.

Voting off Mathewson before Clemens and Maddux isn't all *that* unreasonable. Mathewson didn't really become Mathewson until the Giants got some talent to go with him, and he played in an era where the talent was much more concentrated among a couple of teams than it is now - the difference between the Giants/Cubs/Pirates and the rest of the league was quite large.

The one objection that I have with this group is that they chose not to send players from the Negro Leagues to the island. It'd be interesting to see how Josh Gibson and Oscar Charleston would stack up in this type of competition.

I probably shouldn't quibble, though - I was asked to participate, but declined.

-- MWE
   7. Alan Posted: February 20, 2002 at 12:23 AM (#604793)
How come so many people here are so low on Hornsby?
   8. Mike Emeigh Posted: February 20, 2002 at 12:23 AM (#604796)

How come so many people here are so low on Hornsby?

It's not that people are low on Hornsby, exactly; it's more a matter of being much higher on Joe Morgan than most non-statheads are.

Hornsby was a heckuva hitter, although he played in a heckuva hitter's era. He was not very good defensively, and he had by all accounts an extremely grounchy and prickly personality, which led to at least one trade (and probably contributed to a couple more) during his playing days, and also led to his termination from a number of managing jobs after his career was over. Morgan was, within the context of his era, close to as valuable as Hornsby with the bat, and far more valuable with the glove.

-- MWE
   9. Rob Wood Posted: February 20, 2002 at 12:23 AM (#604797)
In the early going, five players were voted off each week. When we got down to 50 players, we decided to be more "cautious" (and allow more discussion) and only vote two players off each week. I believe, now that we have reached the top 20, we will be voting only one player off per week going forward.

Each voter submits an ordered ballot each week with five names on it, with their least favorite player (the one they want to evict the most) getting 5 points, the next 4 points, ..., down to 1 point. The number of voters submitting ballots has varied from week to week, but the top N vote-point getters that week are evicted no matter how many ballots are received (where N was described above).
   10. Rob Wood Posted: February 20, 2002 at 12:23 AM (#604800)
I am not really an apologist for Rogers Hornsby, but Hornsby's offense is head and shoulders above the other all-time great second basemen. According to OWP, Morgan is 740, Collins is 751, and Lajoie is 755. These are fabulous OWP's. Morgan's 740 translates into a 120-42 record.

Hornsby blows that away. Hornsby's OWP is 822, good for a 133-29 record, or 13 additional wins every year over Morgan. The OWP company that Hornsby keeps is pretty lofty. Following Ruth (865) and Williams (849) we have Hornsby (822), Mantle (822), Cobb (818), Gehrig (816), Bonds (809). So in Hornsby you have a 2B who hits like Mantle, Cobb, Gehrig, and Bonds. Not too shabby.

Thus you would have to make the argument that Hornsby's defensive and personality deficiencies were tremendously harmful for him to be considered in the same group with (or even inferior to) Lajoie, Collins, or Morgan. I just don't buy that argument.
   11. John (You Can Call Me Grandma) Murphy Posted: February 21, 2002 at 12:23 AM (#604803)
Dan said:
Hornsby was also praised for his glove back in his day, which is why I totally disregard the opinion of contemporaries.
I wouldn't go that far. You have to take any comment with a grain of salt, but sometimes there is something worthwhile within the opinion that statistics do not convey. We haven't come to the time yet that we can rely solely on the statistics themself (though we are getting closer).

17 years old? When I was 17, I just picked up my 1982 Bill James Abstract at B Dalton! :-(
   12. John (You Can Call Me Grandma) Murphy Posted: February 21, 2002 at 12:23 AM (#604804)
Paul B. said:
I'm generally suspicious of the sort of argument I'm about to make, since it calls to mind the dread 'intangibles', but with Hornsby I buy into it: Hornsby was a great hitter, but a number of teams decided they'd be better off without having him on the team. There's little evidence that this was a bad decision, either: neither the Cardinals, Giants nor Cubs fell out of contention after ridding themselves of Rogers, with each team winning a pennant within 2-3 years of the transaction.
I'm all for a discussion of intangibles, as long as there is some way that we can estimate what the percentage is for a particular player. Hornsby was a jackass. Can we have a show of hands to see if anyone disagrees with that assessment? I read the great Charles Alexander biography of the Rajah a couple years ago. Every time I was about to start to like him, I'd read something new to dissuade me.
With that said, how does that affect his team's winning percentage? If he had an icy disposition one day, does that lower everybody's OPS twenty points? Thirty points? I'm not trying to say there is no affect on a team, but I would have to think it's minimal. His behavior is not going to change how I hit or field.
That doesn't mean, though, I necessarily believe he's the greatest. I don't think the competition was nearly as strong as today. There IS a question mark after his name.
   13. John (You Can Call Me Grandma) Murphy Posted: February 21, 2002 at 12:23 AM (#604805)
Following my train of thought from before: Why does Hornsby hurt his team so much but nobody ever mentions anything about Cobb. Oh, we always read what a "lovely" fellow he was, but does anyone deduct behavior points from his ranking? Sam Crawford loved Ty's sparkling personality so much, he stopped speaking to him.
Maybe the difference between Cobb and Hornsby is that the Tiger management knew that Cobb was worth a headache for the wins he put up on the scoreboard, while the Cardinals, Giants, Braves, etc. just gave up on Hornsby (and gave up a truckload of runs)!
   14. Dash Carlyle Posted: February 21, 2002 at 12:23 AM (#604806)
"(The)numbers show that Lajoie was a marvelous fielder, with faboulous assist and putout totals."

Lajoie's assist totals were nothing special, but his PO totals were off the charts, even well into his 40s. Bill James makes a convincing case in the new Abstract that the reason for this was that Lajoie was making basically all of the *discretionary* plays at 2B, the plays where either the 2B or SS can cover the bag (force plays, throws on steals, etc.). Second basemen in Lajoie's era were moving off the bag and letting shortstops handle more of the plays at 2B. Lajoie didn't adapt, though. His shortstops were often last in the league in POs, which would bear this out.

Lajoie's unusual number of POs has a lot to do with his overrating by the Linear Weights system (Total Baseball), which pegs him as the 2nd or 3rd greatest player of all time. This ranking, more than anything else, is responsible for the recently adopted notion that Lajoie was a historically great fielder. Lajoie was a great hitter and a good fielder, but he wasn't even the best player of his own era, Honus Wagner was.
   15. tangotiger Posted: February 21, 2002 at 12:23 AM (#604811)
What a great concept!

The cross-era comparisons I don't think are fair at all. People really have a tough time doing comparisons for SAME era (even same YEAR... see Ichiro v ARod v Giambi), much less cross-era. For the NHL's 75th anniversary, they did all-star teams by chunks of 25 years. So, 1917-1941, 1942-1966, 1967-1991, or some such.

For Baseball Survivor part 2, I would suggest the same kind of thing, so that you have say 5 chunks of 25 years, and you can add 1 chunk for the Negro Leagues. In those 6 chunks, you can select your top 20 players, and then begin the voting off 1 player per era (so, say 6 votes off). Then you are down to 10 players per era (at the same time), and keep proceeding. You can then "reward" the 6 best players at the same time.

Maybe at that point you can debate the merits of cross-era, as you have left Matthewson, Ruth, Williams,Schmidt,Bonds,Gibson, or whatever.

Anyway, great job!
   16. John (You Can Call Me Grandma) Murphy Posted: February 21, 2002 at 12:23 AM (#604814)
I was the author of the post before Greg's.

Greg said:
(What I'm saying is that you have three levels of analysis:

1) Teams' records 2) Observations of contemporaries 3) Teams' behavior in trading

One by itself may not prove anything (Banks/Williams), even two in tandem may not mean much, but when a player hits the jackpot (and in Hornsby's case adds on the factor that the teams that traded him almost always improved and the teams that got him got worse), it's something to consider. And considering this combination doesn't mean you have to devalue players who meet one or two of the standards.)

Hornsby wasn't traded because he was hurting his team. He was traded because he pissed off the owner of the team at the time.

As for his teams getting worse when he joined them, are you suggesting that he has negative worth? I think one has to (and I haven't done this myself) analyze the team for injuries, players hitting their peak, etc., before we can determine that Hornsby should be saddled with the decline in winning percentage.

As for Bonds vs. Musial, I agree that Bonds would go now. A couple of years from now, Musial should be going.
   17. John (You Can Call Me Grandma) Murphy Posted: February 21, 2002 at 12:23 AM (#604816)
Now that I look at Hornsby's stats compared to Collins, I would say it's a dead heat. Here we can through in the intangibles. :-)

Matty is not so "clearly overrated;" he just happens to be a victim of adjustments. Maddux looks better than Mathewson because a 2.20 ERA in 1998 is so much better in comparison to the average ERA of that year. There's just so low an ERA can go, folks. And if the batter can't hit the ball, there ain't no amount of juice or weight-training or bringing in the fences that will drive up your ERA.

I'm not exactly sure if you are making this claim, but are you saying that Christy is as good as his statistics say without allowing for context? Then I can say that the hitters of that time are horrible compared to today. You can't go by absolute statistics because you will fall into that trap.

Bill James once had a great analogy: Milk cost much more than it does now. Does that mean it's that much better? There is no way that Mathewson would have his statistics in today's game (and the same goes for today's hitters during the Deadball Era). One other thing: How many black ballplayers did he pitch to?

If I misunderstood your argument, please let me know.
   18. David Jones Posted: February 21, 2002 at 12:23 AM (#604819)

The more I look at it, the more it seems that James' argument about Lajoie's defense is fraudulent.

He says that Lajoie's defense wasn't rated that highly by his contemporaries. That assertion is blatantly false, as anyone who has drudged through reels of deadball-era microfilm will tell you.

Further, to what extent is it true that Lajoie's putout totals are out of whack with his assist totals? Lajoie led the league in assists four times. He led the league in putouts four times.
I just don't see much evidence to support his claims.
   19. John (You Can Call Me Grandma) Murphy Posted: February 21, 2002 at 12:23 AM (#604822)
Chris R:
The way I would try to evaluate a pitcher would be to find out the percentage that pitcher was compared to the league. Then, I would take the standard deviation for that pitcher and then apply it to the league you are comparing. You're right - the greater the offense is in a league, the greater the spread between the player compared to the league is. Standard deviation corrects that (it doesn't help to find out if the average player of 1905 is the same as the average player of 2001 though).

Paul B.
I've never been what you would term "pro-Hornsby", but I just want to be fair to him anyway. I still think Lajoie is below him though (I'm staying out the defense question for Nap until the Win Shares book comes out). Thanks for allowing me to come on aboard!
   20. Dash Carlyle Posted: February 21, 2002 at 12:23 AM (#604823)
"James's argument against LaJoie's defense is certinly not convincing, at leat to me. Interesting yes, but convincing, well let's wait to see the defensive win shares system before we make a ruling on that. "

You don't need to know the finer points of the Win Shares system to see that Lajoie was handling loads more POs at 2B than his shortstops were. What's more likely, (1)that Lajoie, a superstar and his team's manager for a good while, was taking nearly all the plays at 2B, or (2) that Lajoie was the greatest defender who ever lived, and all his shortstops were lousy defensively? Linear Weights is responsible for the latter notion, and it's completely unsatisfying.
   21. Rob Wood Posted: February 21, 2002 at 12:23 AM (#604824)
Regarding Nap Lajoie's defense, the win share numbers that are reported in Bill James' NHBA indicate that Lajoie was a good, not great, defensive fielder. On a per 162 game basis, among the second sackers for whom the data is reported, we have Fox (7.7), Frisch (7.4), Grich (7.4), Sandberg (7.3), Randolph (7.1), Alomar (6.9), Biggio (6.3), Collins (6.3), Lajoie (6.2), Gehringer (6.1), Whitaker (5.9), Morgan (5.6), and Hornsby (4.3).
   22. John (You Can Call Me Grandma) Murphy Posted: February 22, 2002 at 12:23 AM (#604831)
Bill Maugans said: If he had the benefits of modern nutrition and workout programs (and the disposition to use them) Bonds would have to play 10 more years to get close to him.

I'll buy the argument that he would be in better shape in our era. I don't necessarily buy the argument that he would be any where near as dominating in our era than he was in his own. The one aspect of his game that he destroyed his competition was the home run. How many people were practicing the uppercut in the early twenties? The early thirties? He was able to stand head and shoulders above the pack because the vast majority of the hitters were of the "hit them where they ain't" school, while the Bambino's style of game was much greater in that offensive environment.
That's not to say he wasn't great. He would be my pick for last man off the island. He did help create wins more than any other player in history. I just don't feel he would be the baseball God now as he was then.

   23. David Jones Posted: February 22, 2002 at 12:23 AM (#604836)
Could somebody please post some thorough data concerning Lajoie's putout per game rates for his whole career?

As I said before, I believe Lajoie led the league in putouts four times, the same number of times he led the league in assists. I also believe (I don't have the stats in front of me) that he first led the league in putouts when he was with Philadelphia, not Cleveland.

Citing here and there examples is one thing. Does anyone have a comprehensive list that I could look at?


Second, if it is true that Lajoie's fielding stats are inflated, how much is that true, and how much should that impact his value? James used it to put Biggio ahead of Lajoie...I just don't see much of a reasonable justification for that. I think there IS a decent argument that Lajoie was as good as Collins, though getting him past Hornsby is difficult. I'll take all three over Morgan, however.
   24. John (You Can Call Me Grandma) Murphy Posted: February 22, 2002 at 12:23 AM (#604837)
Does anyone know how many defensive innings at second Nap Lajoie had for his career? I estimate he had about 94 Win Share for defense, so I need his estimated defensive innings to figure out his Win Shares per 1000 innings. James has a list of the top ten second baseman ranked according Win Shares per 1000 innings, so we can get a gauge to see where he feels Nap should be.

James does have Nap as winning one Gold Glove in 1914 (he doesn't say if Lajoie would or wouldn't have won any before 1910).

   25. Mike Emeigh Posted: February 22, 2002 at 12:23 AM (#604839)
Team totals for 2B and SS, 1908 AL (G/2B PO/2B A/SS PO/SS A; primary 2B in parens after stats)

Bos: 155/293/463/375/577 (McConnell)
Chi: 156/313/509/285/570 (G. Davis/Atz)
Cle: 157/453/544/256/535 (Lajoie)
Det: 154/333/460/351/509 (Downs/Schaefer)
NYH: 155/316/422/320/517 (J. Delahanty)
Phi: 157/351/424/301/459 (D. Murphy/Collins)
StL: 155/361/466/321/567 (J. Williams)
Was: 155/334/452/372/568 (Niles)

avg w.out Cle: 155/329/457/332/538

Even if we accept for the moment that Lajoie was taking away putouts from the SS (which I'm not entirely convinced that he was really doing), looking at the numbers above it looks like he was taking away AT MOST about 80 PO - and he STILL would have had more PO than every other 2B in the league, when you subtract 80 from his total. The 2B+SS PO total for Cleveland was 709, the best in the league, and only Washington (with George McBride, the Rey Ordonez of his era, at SS) was particularly close to that.

Realize, too, that double plays in this era were still a relatively rare occurrence. Realize that Cleveland pitchers allowed the fewest runs per game in the league in 1908, and allowed fewer H+W than any other team in the league except for the White Sox, so they didn't have a lot of runners on. Then realize that Lajoie led the league in DPs for 2B with 78 - 25 more than any other team's 2Bs turned. That's a *huge* advantage, especially when there aren't that many runners on base to start. Cleveland had 95 DPs overall, tied with Detroit for second behind St. Louis's 97. The Browns had 38 more runners on in 27 fewer innings than did the Indians, and the Tigers had 231 more runners on in 50 fewer innings.

Lajoie was, indeed, making a lot of plays. Some of those plays *might* have been discretionary plays that the other fielders on the team *might* have made (I should note that Cleveland SS did turn 50 DPS, an average number). But I think that, even if we penalize him as much as we can for the number of discretionary plays he might have taken from others, Lajoie was still quite a bit better than his contemporaries.

-- MWE
   26. John (You Can Call Me Grandma) Murphy Posted: February 23, 2002 at 12:24 AM (#604842)
Paul B. said: ( In the 1999 MLB 20th Century team I picked Morgan and Robinson because... well, Jackie was a terrific player, obviously, and I do believe on things like a 20th Century team there are reasons to give him a huge subjective boost. I'm sure others on this board might disagree. )

Do you know if anyone has compiled a list of the career stats of second basemen after age 28 throughout history (such as Bill James did with Minnie Minoso). I don't think it's unreasonable at all to say he might be in that select 1 through 3 greatest second basemen of the 20th century.
   27. John (You Can Call Me Grandma) Murphy Posted: February 23, 2002 at 12:24 AM (#604843)
Nappy said: (To me its like discounting Ruth's numbers because he swung for the fences at a time when the conventional wisdom preferred small ball, like saying Ruths HR (Naps POs) are an anomaly because they were outliers and thus must be discounted.)

I don't know if you were referring to something I had said earlier about Ruth in the thread, but I'll jump in anyway. I would never discount what Ruth did -- he was as great as you're going to get.

With that said, Ruth is not going to hit 150-200 home runs with today's ball (as Keith Olberman idiotically points out all the time). He wasn't born on Krypton.

If we compare Tommy Leach's 6 home runs to the average home run hitter of 1902, this would be equal to a great Mike Schmidt season. Since Tommy weighed only 150 pounds and his homers were of the inside-the-park variety, this is (to say the least) preposterous. In fact, almost all of the home run leaders before the lively ball would project to be greater than any Babe Ruth season.

The Lajoie PO problem is different. Babe Ruth's homers unquestionably helped win games. Lajoie's POs might have relative worth or they might not. I'm still undecided on the issue.

   28. David Jones Posted: February 25, 2002 at 12:24 AM (#604845)
I don't think Jackie Robinson should get much of a boost for starting older than other players. Unlike other black players who jumped to the majors, Robinson had only been playing professional baseball since 1945, so it's not like the color line likely caused him to miss much of his major league career. Of course, there are stories about Robinson trying out for teams, but the fact remains that he also served in the war and was a star in several other sports, so I'm not convinced that Robinson should be given a subjective boost for time missed.
   29. John (You Can Call Me Grandma) Murphy Posted: February 25, 2002 at 12:24 AM (#604846)
(Of course, there are stories about Robinson trying out for teams, but the fact remains that he also served in the war...)

Serving in the military during WWII should also be a factor. He shouldn't be penalized for fighting for his country (which I don't think you would disagree with). I do think you have a point in terms of how much of a boost he should get. I think, at the very least, that he should be rewarded by getting credit for the years 1945-1947 while he was with the Monarchs and the Royals.
   30. John (You Can Call Me Grandma) Murphy Posted: February 25, 2002 at 12:24 AM (#604848)
(On that subjective factor for Jackie: I didn't think I needed to spell it out explicitly, but I was mainly referring to his rather, um, significant role in baseball history. And also strictly in terms of voting for the All Century team. I remember having a long thread haranguing the voters for choosing Hornsby and Jackie and sort of rolling my eyes; you don't have an Official MLB All Century Team without Jackie Robinson.)

I think he belongs with or without his contributions to integration. He was just plain good!

(One reason to not penalize Ruth for the advantage his uppercut swing had is that it had a noticeable effect on other players--both anecdotally and rather obviously upon the statistics of the game.)

I don't want to penalize him either. He was the Pied Piper of home run hitting. I just don't think if he were playing today that he would be miles ahead of Bonds, Sosa, etc.

   31. Mike Emeigh Posted: February 25, 2002 at 12:24 AM (#604849)
Clay starting adjusting middle infield POs for plays made by the other MIF last year, FWIW, before James suggested doing the same thing in the NHA.

I'm not sure I buy the argument that some MIF putouts are discretionary and that the chances aren't evenly divided for reasons not related to fielding skill. Without knowing the exact distribution of balls in play and the situations in which SS and 2Bs are getting putouts, one can't really make that judgment. The Orioles might have had some 4-6 forceouts where the 2B, because of Cal's arm strength, opted to go for a marginal DP rather than taking a sure out at 1B. Baltimore 2Bs might have deferred to Cal on pops to the OF with runners on base because they knew the runners were less likely to take chances w/ Ripken handling the ball. If you adjust POs together, you penalize a good fielder who covers a lot of ground for playing next to a statue. I'd rather overrate a good fielder than underrate one.

-- MWE
   32. Fadeaway: The Baseball History Podcast Posted: February 26, 2002 at 12:24 AM (#604852)
David, the fact that Robinson had only been playing professionally since 1945 doesn't mean he didn't lose more years than that to segregation. One of the results of segregation was that black kids did not have the same incentives and encouragement to perfect their baseball skills that white kids did. Joe Black tells a story about his childhood, the day he found out that blacks weren't allowed in the major leagues, he ran home in tears and tore up all his baseball cards of Hank Greenberg and Joe DiMaggio. Black, of course, ended up going into baseball anyway, but for every Joe Black there were probably many kids who just gave up on baseball.

If being a major league baseball player had been a viable career goal when Jackie Robinson was growing up, I have no doubt that he would have been in the major leagues by age 21 or 22 at the latest.
   33. David Jones Posted: February 26, 2002 at 12:24 AM (#604855)

When Jackie Robinson was 22 years old, he was playing baseball, for the UCLA Bruins. According to Arnold Rampersad, that year he batted .097, and led his team in errors. This is a major league player?

Second, your argument about how he could have grown up differently had he been white seems to call into play similar arguments for many, many other players. Why not the say the same thing for every good black athlete of the day? All of them "could have been" major leaguers. Damn, if I had grown up African-American and living in the inner city, my 6'7" height could have made me an NBA player by now, right?
   34. Mike Emeigh Posted: February 26, 2002 at 12:24 AM (#604864)
My personal take is that I'd rather promote a more nuanced view of fielding--the general public thinks Roberto Alomar is the finest glove at 2B today, when the fact is Boone or Durham or any number of guys have been just as good (or better)--so I'd put more trust in a system that doesn't show one guy towering above everybody else. But I can see the problems with that position, given that just as every so often a hitter or pitcher comes along who dominates the game, so too have there been truly dominant defensive players.

My take is that, when one guy towers over the field after we've taken into account all of the factors for which we can logically account, and there is no clear, indisputable evidence that he's there because of factors unrelated to his ability, we should assume that he's there because of his ability. In the case of Lajoie, all we have is James's speculation - with no clear evidence to support his speculation. Lajoie may have taken more balls than his SS because he was more likely to convert them into outs; I see that stance as having just as much support in the historical data as James's stance that he was taking discretionary plays away from the SS who could just as well have handled them.

-- MWE
   35. John (You Can Call Me Grandma) Murphy Posted: February 26, 2002 at 12:24 AM (#604867)
I agree with most of your analysis concerning Hornsby. What I think you're wrong about Hornsby is concerning his deadball stats. He was the best hitter in the National League at 21 and the second best at 23. He was a great hitter before and after 1920 (age 24).
I may be wrong, but I think there is a misconception about Bill James's evaluation of Lajoie as a defensive player. He feels he's not a towering defensive player such as Maz, Ozzie or Brooks, not that he wasn't a fine second baseman. He compared him to Gehringer, Herman and Sandberg as a defender. Not too shabby!
   36. Rob Wood Posted: February 26, 2002 at 12:24 AM (#604868)
The Lajoie debate will continue to thrive for years to come, I reckon. Could someone post an explanation of the "false normalization of fielding statistics" phenomenon? Thanks much.
   37. Mike Emeigh Posted: February 26, 2002 at 12:24 AM (#604869)
The thing about Lajoie is that you can't say he's better than the other great second baseman because of his freakish putout totals. That's saying he's greater than Collins, Morgan, Hornsby etc. because he took all the throws to second and liked to call the SS and 1b off of popups.

You don't know that's what he did. You only have James's speculation on that. You don't know that he wasn't chasing down popups hit down the foul lines and out into short center, or diving to his left to snare line drives headed for right. You weren't there.

Rob Wood:

"False normalization" of fielding stats occurs because the number of outs varies little from team to team, and players on bad teams generally have more opportunities to get putouts and assists because of the extra baserunners their teams allow. Example:

Team A: ground ball hit to 2B, runner out 4-3
ground ball hit to SS, runner out 6-3
ground ball hit to 2B, runner out 4-3.

Team B: ground ball scoots by 2B into RF for a hit
ground ball hit to SS, fielders' choice 6-4-3
ground balls scoots by 2B into RF for a hit, runner advances to 3B
ground ball hit to SS, DP 6-4-3

Same number of outs and runs, but the 2B gets 2 putouts and an assist and the SS gets two assists, whereas in the first inning the 2B gets 2 assists and the SS one. The plays "not" made by the 2B allow both the 2B and SS to pick up an extra chance each. When you normalize chances to league average, as in Fielding Runs, the 2B and SS will "look" better relative to league average not because the 2B was a good fielder, but because he wasn't (in this context). Fielding Runs adjusts for this to some extent by double-crediting assists, but then gives it back with the way that double plays are handled.

What you need to do to address this is (a) to consider hits as failed fielding plays, and (b) to make an adjustment for double plays based on the number of runners on base (and outfield assists as well). James addresses this indirectly (from what I can tell) by his team adjustments, but the direct approach of treating hits as fielding failures is more intuitive to me.

-- MWE
   38. John (You Can Call Me Grandma) Murphy Posted: February 27, 2002 at 12:24 AM (#604874)
I'm suprised nobody has mentioned the George Sisler ranking concerning his fielding? James has him down as a POOR defensive first baseman. Lajoie, at least, is considered a great defender (if not a historic one).

Just thought I would ignite another 100 page thread. :-)
   39. J. Lowenstein Apathy Club Posted: March 03, 2002 at 12:24 AM (#604877)
Is it just me, or does anyone else think this the _worst_ possible way to come up with a top 100? (Other, that is, than picking guys at random).

   40. Rob Wood Posted: March 03, 2002 at 12:24 AM (#604880)
Is it just me, or does anyone else think this the _worst_ possible way to come up with a top 100? (Other, that is, than picking guys at random).

It is hard to know how seriously to take this comment. In case it is meant to be a substantive comment as opposed to a mindless flame, I'd like to reply. First, the Baseball Survivor group exercise has been a lot of fun for the members involved, and has been extremely informative. While the rankings are the ultimate result of the exercise, there is a lot more to it than that. The group engages in intelligent discussions about a myriad of issues, both analytical as well as qualitative, about players, about teams, about history, etc. Many interesting and new analyses have been conducted and posted to the group to "inform" the discussions.

Second, every list that I have ever seen of the all-time greats has been flawed to some degree. This is true of all the lists, be they directly a result of statistical methods or not. Every individual observer and every statistical method is subject to idiosyncrasies and biases. In such cases, group methods often lead to better and more robust results. We have found that to be the case in the Baseball Survivor group. While no individual voter agrees with where every player is ranked by the group consensus, I think we all agree that the group's ranking is an excellent reflection of our views.

Third, baseball analysis, opinions and expertise evolves over time. People can consider the Baseball Survivor's ultimate ranking of the all-time greats to be a snapshot of the views of a group of experts convened at this time. Our rankings can then be compared to other rankings made by others in the past, or possibly in the future.
   41. scruff Posted: March 05, 2002 at 12:24 AM (#604884)

I hope this doesn't get lost in here now that this thread has gone on so long.

Someone this missed my radar the last two weeks. I looked at the site and just missed the comments, forgetting to drop back.

I've read a lot of what was posted, but not every single word.

"I am not really an apologist for Rogers Hornsby, but Hornsby's offense is head and shoulders above the other all-time great second basemen. According to OWP, Morgan is 740, Collins is 751, and Lajoie is 755. These are fabulous OWP's. Morgan's 740 translates into a 120-42 record.

Hornsby blows that away. Hornsby's OWP is 822, good for a 133-29 record, or 13 additional wins every year over Morgan. The OWP company that Hornsby keeps is pretty lofty. Following Ruth (865) and Williams (849) we have Hornsby (822), Mantle (822), Cobb (818), Gehrig (816), Bonds (809). So in Hornsby you have a 2B who hits like Mantle, Cobb, Gehrig, and Bonds. Not too shabby.

Thus you would have to make the argument that Hornsby's defensive and personality deficiencies were tremendously harmful for him to be considered in the same group with (or even inferior to) Lajoie, Collins, or Morgan. I just don't buy that argument."

Rob, I think you are overstating Hornsby, for a couple of reasons. First, his career is much shorter than most of the people you compare him to. He through by the time he was 35.

Second, I get Hornsby at .789 for his career, not .822. I have him at 177-49. I use newRC, and I adjust for league and park as well. I came out with .836 for Williams and .849 for Ruth, so my numbers tend to be a little lower (are you using old RC?), but for Hornsby the difference is much larger.

I get Eddie Collins at 219-80, .734. Hornsby at his peak was .863 (3 year) or .855 (5 year). Collins was .834 and .824, with significantly more playing time. Collins' D was light years ahead of Rogers as well.

Hornsby's edge widens on the career OW% only because he was done by the time he was 35. Collins was the more valuable (and in my opinion better at their peaks even) ballplayer. It's not an especially difficult choice.

Even if the difference was as large as you would say, it wouldn't be 13 wins per year, it would 13 wins over the course of their careers, I'm not sure if that was just a case of what you said being different from what you meant. Barry Bonds was about 13-14 wins above a replacement level player last year, there's no way Rogers was that much better than these guys every year.

Paul B --

I agree w/what you wrote on 2B pre-1920. Because the DP skill was not as valuable (hit and run always on, etc.) 2B was a lot like 3B is today, and 3B was a lot like 2B was. If you look at the top 3B prior to 1920, their hitting stats are not impressive in terms of all-time greats. Yet there were all of these 2B that could hit.

Fielding the bunt at 3B, as well as having a good arm (more speed in the game) were very important. 2B was not a terribly difficult position in the lively ball era.

Robert Dudek has done a lot of work on replacement level by position pre-1920 and the numbers strongly confirm this (there will be an article about this on the Hall of Merit in the near future) suspicion.

For historical purposes, I really think you need revise the comparisons of 2B and 3B as positions. Instead of 2B and 3B, you should have 3B pre-1920 + 2B post-1920 as one position and 3B post-1920 + 2B pre-1920. Lajoie and Collins should be compared to guys like Schmidt and Santo and Mathews; Home Run Baker should be compared to guys like Morgan, Grich, Gehringer and Sandberg.

If they were grouped like this, it wouldn't look like 3B (as we know it today) was so under-represented in the Hall.
   42. scruff Posted: March 05, 2002 at 12:24 AM (#604885)
One other thing, I think this is an outstanding project, and I love player data pages, especially the most similar at peak years tables.
   43. Rob Wood Posted: March 05, 2002 at 12:24 AM (#604890)
A couple of replies. Steve Rohde asked about whether anyone is thinking of publishing the discussions since they could make interesting reading. I do not know if this has been discussed. Tom Hanrahan or Justin Kubatko, the group leaders, may know.

Scuff asked about the OWP numbers I posted for Hornsby and others. These originally appeared in another member's post to the Survivor group, so I do not know how they were calculated.

Scruff also caught me in an error. When I converted OWP's to "wins above average", I neglected to divide by nine (duh). A player's OWP estimates the win pct of a team made up of nine hitters identical to the player, with average pitching and defense.

Let me do the calculation again. Using Scruff's numbers, Hornsby's .789 OWP means that he contributed (.789-.500)*162 = 5.2 offensive wins above average per 162 games played. Collins' .734 OWP means that he contributed 4.2 offensive wins per 162 games played.

But Collins played many more seasons than did Hornsby. Let's do the career calculation, and see if I do it right. Again using Scuff's numbers, Rogers Hornsby had a .789 career OWP in 2259 games and Eddie Collins has a .734 career OWP in 2826 games. To find the number of wins each is better than average over their careers, multiply the OWP by the games, subtract half the games, and divide by 9. For Hornsby, this is (.789-.500)*2259/9 = 72.5; and for Collins this is (.734-.500)*2826/9 = 73.5. So Collins is slightly ahead. Note that if you use the original OWP numbers I posted, Hornsby is slightly ahead.

Scruff's point is valid. Due to their longer careers, Collins, Lajoie, and Morgan likely contributed as much or more to their team's offense than Hornsby did. Considering defense and other factors would seem to push these guys ahead of Hornsby from a career value perspective.

The remaining vestige of a case for Hornsby is that his peak value may have been higher than that of the other contenders for all-time best second baseman. But Lajoie, Collins, and Morgan each have strong peak-value cases as well.

In light of the above discussions, it may be of interest to note that Joe Morgan (26) and Nap Lajoie (21) have already been voted off the Baseball Survivor island. And Eddie Collins is on the bubble to be voted off this week; if so he would be the 18th greatest player ever. Rogers Hornsby has received a few votes to date, but seems destined to come in around the 10th spot.
   44. scruff Posted: March 06, 2002 at 12:24 AM (#604892)
Ken -- a few notes on how I get my numbers (if you use them I'm sure people will ask).

I take each season, and adjust for park and league (and DH when relevant, using a .95 factor for the league).

But then I add an adjustment for team games to make all seasons equivalent of 162 games. The reason I choose this is because every season (except 1994 :-)) has a pennant, so it seems like the logical cutoff. If only 125 games were played in a season, those 125 have more impact on the winning of a pennant than 162 games in a typical season.

Because of this, the numbers probably won't matchup with others. I'll gladly share my spreadsheet with those who ask, but it's about 7 MB, so you'll need to send me a CD (I have done this in the past). The method has some flaws (I use a 2.00 exponent, and Stats Park Factors, which are far from perfect) and I'm working on fixing them before the Hall of Merit kicks in.

However I don't think the flaw would really bias the numbers one way towards Collins or Hornsby.

The other thing Ken, was that I'm not sure using games played is a good context because of partial games. Hornsby hung around for a lot of partial games late in his career, you might be overstating his value a little. Plate appearances / a weighted (on season PA) avg of (league PA / total team G / 9) would probably be a better estimate for career length, although it is a little time consuming.

Maybe you can take some of this back to the group and help keep Eddie on the Island a little longer and get Rogers the boot . . .
   45. scruff Posted: March 06, 2002 at 12:24 AM (#604894)
Rob Wood -- I apologize for calling you Ken. I don't know how that happened. Maybe I was thinking of this guy or something. Sorry . . .
   46. John Posted: March 09, 2002 at 12:25 AM (#604932)
I have seen the list and refuse to comment on anything that is not taken seriously? Griffey over Rose? Where is Sandburg? etc, etc, etc.
   47. John Posted: March 09, 2002 at 12:25 AM (#604933)
I have seen the list and refuse to comment on anything that is not taken seriously? Griffey over Rose? Where is Sandburg? Mike Piazza stands before Bench? etc, etc, etc.
   48. ERROR---Jolly Old St. Nick Posted: March 12, 2002 at 12:25 AM (#604956)
A good case, Andrew. But what would stop Ruth from pitching every 4th or 5th day and playing 1B the rest of the time? There might be a defensive liability on the first day after he pitched, but I doubt that would come close to negating the huge added value he would give his team as a starting pitcher.

Another way of looking at it is this: Who would win in a game between nine Barrys and nine Babes? No contest.

The only real case for Bonds would be to emphasize the huge difference in the quality of the competition: the talent pool of a multicolored world (pop. 1,000,000,000+ adult men) vs that of the white U.S. of Ruth's era (pop. less than 50,000,000 adult men). This is a factor which oldtimers tend to discount out of misplaced nostalgia, and SABR types often dismiss because of the unquantifiable aspects of the comparisons, but there is no question that the game of today is light years above the ML of 1914-1935.

But Ruth dominated in a way Bonds never has (for obvious reasons; still...), and Ruth was one of the better pitchers of his era. With Ruth's combined talents, I just can't see how Bonds, or any other position player, can rank over him.

I realize, of course, that this is an argument with no "correct" answer.

Now if you had brought up Martin Dihigo....
   49. ERROR---Jolly Old St. Nick Posted: March 16, 2002 at 12:26 AM (#604980)
Since Ruth was able to beat Walter Johnson about two-thirds of the times they faced one another, I somehow suspect that most of the home runs hit in a game between nine Babes and nine Barrys would be hit by the Babes. The final score would probably be somewhere along the lines of 45 to 3, maybe 45 to 4.

Of course, with Barry's contract, he could afford to hire a few designated pitchers. He could start by signing up Hub Pruett.
   50. John (You Can Call Me Grandma) Murphy Posted: March 17, 2002 at 12:26 AM (#604981)
The problem with the Babe Ruth as superhuman power hitter is that he was doing this when he was basically the ONLY true power hitter in the AL. Except for Gehrig and Foxx at the end of the twenties, he was the only real slugger. It took a while before the uppercut was mastered, so the Babe looks like Gulliver against the Lilliputians. Most of the hitters that were able to have good home run totals from that time were greatly helped by their parks.

Bonds, on the other hand, has a plethora of sluggers to be compared to. That DOESN'T mean that Bonds would out-slug the Bambino. I do think it would be a good show!
   51. Rob Wood Posted: March 21, 2002 at 12:26 AM (#604999)
I am not a big fan of Greg Maddux, but it is perfectly reasonable to consider him among the best pitchers ever. Maddux's 1992-2001 period of sustained excellence is arguably the best consecutive 10 years ever. Kevin B. first says that where Maddux ranks makes the entire endeavor "totally invalid", and then he closes by saying that it is "just my opinion". Maybe Kevin should be more open-minded and consider the views of experts in the field.

You must be Registered and Logged In to post comments.



<< Back to main

BBTF Partner

Dynasty League Baseball

Support BBTF


Thanks to
for his generous support.


You must be logged in to view your Bookmarks.


Page rendered in 0.5512 seconds
60 querie(s) executed