In 1968 for The Baseball Encyclopedia the publishers put together the Special Baseball Records Committee. The committee had to decide how to balance accuracy with history, and as a result ended up tinkering with some of the game's most historic numbers. As The Numbers Game says, in 1876 walks were considered outs (what that would have done to Kevin Youkilis!) and in 1887 they were scored as hits. The committee made the decision to treat walks in their modern fashion. There were other small changes, too, and as a result, Ty Cobb got one more hit, Honus Wagner lost 15, and Cap Anson, the first member of the 3000 hit club, ended up with just 2995.
But the changes to the rules overtime had more serious implications for baseball's records. Ground-rule doubles were considered homers once upon a time. And the committee's biggest controversy came when then tried to change Babe Ruth's home run total. In 1918, Babe Ruth hit a walk-off home run with a runner on base and was awarded a triple; games were ruled over when the runner scored, so Ruth ended the game at third base. This was the Babe's 715th homer—but everyone knew Babe Ruth had hit 714. Due to public outcry and the work of one member, the committee changed the decision to keep walk-off hits the same as they always were, and Ruth went back to 714.
A little over a decade later, Pete Rose started to work toward Ty Cobb's hit total. After one researcher stumbled on a game where Cobb's hits were recorded twice, The Sporting News examined all Cobb's hits and found he'd really hit 4,190 (as opposed to the 4,191 of history or the 4,192 of The Baseball Encyclopedia. Also, Cobb did not, in fact, deserve the batting title in 1910. But Bowie Kuhn, the Commissioner of Baseball, announced that, "The passage of years, in our judgment, constitutes a certain statute of limitations as to recognizing any changes….The only way to make changes with any confidence would be for a complete and thorough review of all team and individual statistics."
What's the answer? Where do you strike the balance between respecting the game's history and getting it right? Given all the difficulties we see throughout the book in knowing everything that happened during baseball's early years, is it possible to ever get it truly right? If you truly want to compare players, a ground-rule double should be the same throughout history, but if you start turning homers back into doubles, doesn't that even affect the outcome of games? Is there a statute of limitations on the records? Given the unreliability of old statistics and box scores, how many home runs should Babe Ruth have? How many hits should Ty Cobb have?
It seems to me that the Babe Ruth and Ty Cobb questions are inherently quite different. Miscounting the number of hits in a season is something that can, and should, be corrected. But changing the number of home runs because they were scored differently at another time doesn't seem right. The rules for scoring were different then, but that's when those homers were hit. You've got to go with the rules at the time, not the rules imposed later.
Posted by: PlayhouseGang at November 15, 2005 10:55 AMI'd have to agree with Playhouse. Especially with the question raised about how those Ground-Rule Doubles turned homers would affect the outcomes of games.
Posted by: Torhu at November 15, 2005 11:45 AMHowever, here's a good question. How about Ichiro's single season hit record? Considering he played however many more games than the previous record holder...
Posted by: Torhu at November 15, 2005 11:45 AMTo answer the question "How many hits should Ty Cobb have?" I must respond, "fewer than I do." Thus, Ty Cobb's hit total shall henceforth be set at -1.
Posted by: mbnovak at November 15, 2005 11:55 AMI don't think its right to go back and rescore a game. The offical score card or box score is just that, offical. Sometimes the break goes in your favour and sometimes it does not.
Comparing the eras is just tough to do when you consider how much the game has evolved, but I don't think it should held agaisnt the players of the game now that they play more games.
Ruth hit 60 hrs for a season, regardless of the amount of games, Maris hit 61 for the season. The people in charge of baseball changed the length of the season, and that should not be the players fault, but it's what you do during that season that matters not how many games in it you play.
If you start counting the number of games played, then you would also have to count at bats, and how many were in extra innings and so forth, and that's just to much to consider.
Let records be awarded by what takes place on the field.
Posted by: mike at November 15, 2005 12:20 PMI'm certainly all about the retroactive questioning/tweaking as an exercise in understanding the evolution of The Game.
The 1968 Committee's creation closely precedes Curt Flood. Coincidence? The importance of individual stats immediately preceding free agency? Not coincidence. Evolution.
In 1918, the rules that "penalized" Ruth showed not bias, but merely a time in the history of The Game where winning was the objective. The evolution of that specific rule leads us to present day and to the Mac/Sosa/Bonds long ball frenzy...and "Moneyball" collaterally.
In the future, a stat such as Walkoff Homeruns may be the next big thing. The Game continues to evolve into an entity less concerned, overall, with "winning."
No matter how in depth one tries to normalize the data across generations of The Game, the fact remains that the objectives of one were not necessarily the objectives of another.
Century 21 never sponsored a "Baltimore Chop" competition...nor will .OBP ever get anyone an A-Rod contract.
Posted by: BAT bandwagoner at November 15, 2005 12:21 PMI come down in agreement with most of the other opinions posted here. Statistics can't be retroactively changed simply because of rule changes or changes in the definition of what constitutes a "hit" or a "home run". At one time or another, walks were awarded after 3 balls.. or 5 balls. Do you go back and take away walks that didn't require at least 4? or award walks to someone who got to ball 4 (but not 5) and subsequently either got a hit or made an out? Of course not.
Correct obvious errors in record keeping but you don't revise history.
Naturally the 162 game schedule, the DH, domes, artificial turf and interleague play have had an effect on statistics. 50 years from now, when games are all played outdoors on grass and the DH is abolished, do you think they'll go back to the 1970-2010 era and adjust stats based on what might have happened? I would hope not.
Posted by: JimCrikket at November 15, 2005 12:41 PMAn observation, about baseball and about life, is that we often make important the things we can easily measure. Wins, RBI's, Batting Average, ERA (especially for relievers) are all problematic when evaluating the true perfomance of players. jiggling with Ruth's homerun total doesn't make any difference, he's still great almost beyond words. When new, more illustrative, stats are developed (OPS/WHIP and the like) the great players are still great. ARod was sixth in the majors in OBP so it certainly helps. The rational behind the whole Moneyball idea is not that there are great players we haven't notices but that if the other teams are signing mid level players based on BA and you're looking at OBP or OPS ahead of that then you're name's Billy Beane.
Posted by: dan in london at November 15, 2005 12:46 PMBy the way
Congratulations to Fat Albert and A-Rod for their MYPs. Both well deserved and I think htere stats won't be monkeyed with it.
Posted by: mike at November 15, 2005 01:15 PMThe compilation of historical statistics comes from a need to compare players from different eras. The impersonal nature of statistics cannot Which I argue is difficult at best, impossible at worst. Rule changes, dead/live balls, dilution of pitching from expansion... the number of factors one can throw in is innumerable. Who can say how many homers Babe Ruth would've hit today? He probably never saw a 100mph fastball or a splitter that fell off the table. The beauty of the Babe is that he was so head-and-shoulders above everyone else that he was a phenomenon. 714, 715, who cares? His most impressive stat was ticket sales, if you ask me.
Posted by: MikeinMA at November 15, 2005 01:36 PMI don't think there is any question that Babe Ruth is the greatest baseball player ever. As some Canuck used to say..... The best there is, the best there was, the best there will ever be.
Posted by: mike at November 15, 2005 01:53 PMI think mike said it very well:
"If you start counting the number of games played, then you would also have to count at bats, and how many were in extra innings and so forth, and that's just to much to consider."
In order to make it a "fair" comparison, you'd have to take into account *way* too many factors.
For instance: Maris played in 161 games and was 26 years old in 1961 when he hit 61 homers. Ruth played in only 151 games and he was 32 years old in 1927 when he hit 60. It would seem that Maris had it easier.
Then again, during the 1920s, 49.6% of all games started were completed by the starting pitcher, down to 25.2% complete games in the '60s*. Ruth had more at-bats against guys who'd pitch pretty much all game, every game, whereas Maris had more against guys who'd come in fresh from the bullpen. So it would seem that Ruth was not without advantages himself.
Parity in the record books would be a good thing, but it would be so impractical to achieve in a perfectly fair and agreeable way that it may as well be considered impossible. Especially when you consider all of the stink raised by the steroid issue- by the time anyone got all of the current records set straight, people would be clamoring for post-steroid records to be adjusted as well.
*I got that stat from a Baseball Prospectus article entitled "A Brief History of Pitcher Usage" (click my name) that also describes some of the crazy rules mentioned by JimCrikket:
"At various times in the first two decades of professional baseball, the distance from the pitcher to home plate was less than 50 feet; a walk required nine balls; bunts that landed in fair territory before skidding to the backstop were considered fair balls; hitters could call for a "high" or "low" pitch; pitchers could throw the ball from a running start; and curveballs and overhand pitches were illegal."
Posted by: let's go mets at November 15, 2005 05:20 PMTinkering with history is not a good idea. Once you do that, then you'll have to account for situations (batting averages were a whole lot higher in the early years when the pitchers had to pitch underhanded) and it's impossible to accurately compare situations.
It's fun to tinker with the statistics to try to get accurate comparisions, but it's impossible to do so. Even in the same year, it's impossible to compare players accurately, simple based on where they bat in a line-up (and how they're protected), injuries, etc. It's a good academic exercise to try to level the playing field to get accurate comparisons, but nothing but fighting can come from it--people will never agree with all the variables taken into account and how important they are. It's better to leave the statistics alone and note all the important records.
I like the idea of having both "modern-era" records (since about 1900) and historical records--just to give us some dividing line of how in baseball's infancy, they were still trying to work out the details and things were different. Changing the records simply denies that things have changed. There are enough books now that will detail the changes and inaccuracies of past records (such as Ruth's home run controversey--Hank Aaron broke it anyway, so there's no need to quibble when someone else gets close). As for the addition of the DH, the season being 162 games, inter-league play, etc.--all those things are just strategic moves, and don't effect the way the game is played enough to mention.
There will be no asterisks, because that was the dumbest thing, ever.
In my humble opinion, anyway.
Posted by: Just Beth at November 15, 2005 10:31 PMI didn't read the Numbers Game, so this might be totally irrelevant. Don't kill me. I did, however, recently read "Cy Young: A Baseball Life", and therein was where I read one of the craziest revisionist history stats problems... in confirming Cy Young's win count of 511.
Before 1950, there was no "official" rule for designating who the pitcher of record was, so even though Chadwick had been assigning wins and losses, there wasn't really an established way to handle the decision in that rare case where a starter didn't actually finish the game. So, the rules we know of today were established in 1950... and then in 1968 another committee met to address the issue of what to do with all of those games scored before 1950.
The decision? Pitching records established between 1901 and 1949 would stand as received, but games between 1876-1900 would be reevaluated under the new rules.
So, enter Cy Young -- a guy who played from 1890 to 1911, half in the old era, and half in the new era.
Researchers in the 1970's went back and examined all of the 906 games he played in, and here's the funny part -- after awarding three wins where they had not previously been correctly awarded, and taking away three wins that had been incorrectly awarded... Cy Young still had 511 wins. (Thank god. I mean, it's just one of those numbers, like Pi or E or 6*10^23. You just accept that Cy Young won 511 games because the universe works that way.)
Posted by: Deanna at November 16, 2005 01:22 AMJust as long as we don't start going back and reevaluating the way we measure Sass then I'm happy :-)
Posted by: dan in london at November 16, 2005 03:51 AMYou certainly can't fault the players involved because the rules were different at one time. The other problem is it is next to impossible to completly reconstruct a game from those eras simply because boxscores were not as complete as they are today.
As late as the 30's, fans were allowed to sit on the field and if a ball was hit there it would be a ground rule double.
Posted by: Annie at November 16, 2005 08:19 AM511 wins, that dude deserves an award for that or something.
Posted by: mike at November 16, 2005 10:10 AMThis is really interesting. Unfortunately, I haven't had a chance to read the book, but now I am regretting that.
I'm currently working on updating the statistics at MLB.com. We are re-coding things to fit with our current system.
I am finding LOTS of scoring errors as I go through the 1967 games (at the moment) and I am forced to wonder if the changes I am making are going to have long-reaching effects, as MLB.com is more and more becoming the official storehouse of statistical information for baseball.
Lots of RBI errors, in particular. But I don't know for sure if these are CODING errors, or actual SCOREKEEPING errors.
It will be interesting to see how it all pans out.
Posted by: NY-Brian at November 16, 2005 10:17 AM4.
Ty Cobb should have 4 hits.
Chuck Norris says so.
Posted by: el diablo at November 16, 2005 12:09 PMI look forward to the days when baseball games are played on the moon. Obviously, they will have to be under a dome, and "Astroturf" will take on a whole new meaning.
The new "Moon League" offshoot of major league baseball (MMLB) will soon have statistical ramifications that will affect the way sabermatricians look at the game, and in fact, the way the game will be played in general.
First, the lack of gravity in the Moondomes (1/6 of Earth's gravity) will definitely affect the length of balls hit. If we estimate that the average major-league hit travels 200 feet, even the smallest utility infielders will find that they are slugging 200-300 home runs per year. (Nicky Punto will be one of the first to sign up.) This will be impossible to be remedied by owners, since larger ballparks to accommodate the long balls could only be covered by center fielders with legs like Michael Johnson on some SERIOUS Jose Canseco moon-juice.
Second, pitching will take on a whole new realm. Due to the lessened gravity, pitchers expand their repertoires above the standard curves, sliders, and sinkers, into more contemporary "moon-style" pitches such as uppercutters, floaters, and the aptly-named "Armstrongs." Due to all the new ways to flummox the batter, batting averages fall by .043 points per player. Plus, Cuban pitchers defecting to the Moon League becomes near impossible due to finding a suitable lifeboat.
Future statisticians will soon figure new "Sabermoon" techniques to calculate Moon League statistics, using the 1/6 gravity as a base for some new and complicated math. Eventually, the Minnesota Twins are contracted to the Moon leagues, where their slugging confidence is bolstered and their 36-year veteran Johan Santana has found new gravity-free life in his amazing changeup.
So here's my prediction - Stats will always be changing and the Sea of Tranquility Twins will win the MMLB pennant in 2036!
Posted by: Neil at November 16, 2005 04:58 PMI think the most significant thing to keep in mind about 'The Numbers Game' is how it points out how 'the numbers' have influenced traditional baseball thinking in odd and scattershot ways rather than in a powerfully organizing sense. In that sense, it really is a prequel of sorts to 'Moneyball'; before we had Billy Beane, we had Branch Rickey, after all.
In a larger sense, though, there's an ironic realization that even the traditionalists who decry 'statheads' take much of their mythology about the game from the numbers. Without a statistician counting up the games and declaring that Lou Gehrig had set a new record for consecutive games played, traditionalist odes to the toughness of Cal Ripken aren't as meaningful. Without somebody showing how unlikely it is to get a bunch of hits in consecutive games, Joe DiMaggio's and Pete Rose's streaks mean less.
Ultimately, though, I think the question that leads into this discussion is misguided - the idea that baseball's statistical record provides an unbroken chain from the distant past to the present is an illusion, for exactly the reasons BG points out and more. For instance, it's all but impossible that someone will break Cy Young's 511 win record until pitchers find a way to win 40+ games per season again. The late Stephen Jay Gould postulated that the lack of .400 hitters in present-day baseball is evidence that baseball players have evolved into the equivalent of a more mature biological species. There are anywhere from a dozen to twenty or so players who are likely to pass the 500 home run 'plateau' in the next decade, and it's not impossible that after 2025 there may be a player passing Babe Ruth's career home run total every season.
The definitions of statistics change, the game itself changes and goes through cycles, and the ensuing statistical record only seems to be persistent, like a ship in which every physical part has been replaced but that we continue to think of as the same ship because it happens to have the same name.
Posted by: David Michael Wintheiser at November 16, 2005 06:07 PM