I.   The Issue

The Over-Achievers

            There is an obvious correlation between the strength of a team’s roster and the number of games that the teams win.   We have seen this, for example, when we looked at the best teams of all time, the worst teams of all time, and the 1961 American League.    The ten best teams had an average record of 96-62, the ten worst teams had an average record of 59-96, and the American League teams generally finished in approximately the order predicted by the strength of their rosters.

            At the same time, obviously, some teams do over-achieve, and some do under-achieve.    Over-achievement can be broken down into two types:

            Natural over-achievement, which is a result of playing against weak competition, and

            Dynamic over-achievement, which means that the team, due to its internal dynamics, simply plays better than they ought to have played.

            The 1945 Dodgers were natural over-achievers.  They had the third-weakest roster that I have found, but finished 87-67 because the competition was even worse.   They may also have been dynamic over-achievers, I don’t know; I haven’t studied enough of the war-time teams to have a good grasp on the relative quality of the rosters.

            The 1961 Yankees were both natural and dynamic over-achievers.   They played in a weak league, but even knowing that, they still played much better than you would expect, looking at the roster of the team.  

            We can generalize about over-achieving and under-achieving teams just by eye-balling the records, but to do more than that we need to be able to measure over-achievement and under-achievement.   In order to do that, we need to pin down the expected wins for each team, given the strength of their roster.  And to do that, we need a formula.

            Let’s take the 250 teams that I have figured so far, and sort them into five groups, sorted by the season score:

            The top 50 teams we have studied had an average won-lost record of 95-63, and an average score of 271.   The bottom 50 had an average won-lost record of 61-93, and an average score of 118.

            We have surveyed more good teams than bad, although not by as wide a margin as this chart might suggest.  We’ve actually studied 159 winning teams, 71 losing teams, and 20 teams that finished at .500.   These 250 teams averaged a .536 winning percentage and a score of 201.179, so we can start with the assumption that .536 = 201.2.    .536 divided by 201.2 is .002 664, so if there was a straight-line relationship between the score and the winning percentage, we could predict winning percentage by simply multiplying the score by .002664.   Let’s try that:

            Score               271      Expected Winning Percentage   .721

            Score               118      Expected Winning Percentage   .316

            The actual winning percentages associated with those scores are .603 and .397, so that obviously doesn’t work; it’s not a straight-line relationship.   We need to flatten out the line.

            We can flatten out the line by adding “ballast” or “dead weight” to a team’s score, thus moving every team closer to .500.   Let’s add 10 points of dead weight to each team, and repeat the process.    That means that .536=211.2, which means that expected winning percentage equals (Score + Ballast) * .002538. 

            If you’re not following the math don’t worry about it; I’m just making this shit up as I go along.   If you are following the math, then by this formula the expected winning percentage of the top group would go down to .713, while the bottom group would go up to .326.   These numbers are still far too dispersed, but we’re moving in the right direction.  We need to add more ballast, and keep checking our progress.

            It turns out that the “ballast” approach gets its best results (meaning it reaches the minimum error) at 118.   We thus settle on this formula to approximate the team’s expected winning percentage:

            (Score + 118) * .001 679 = Expected Team Winning Percentage

            That formula doesn’t work great, in that it produces expected winning percentages which are significantly too high at the top end of the scale.   I controlled the effects of that by adding a rule that if your expected winning percentage comes out greater than .640, we’ll treat it as .640.    That’s still not great, but this is just a jerry-built heuristic that won’t stand the test of future research anyway, so let’s move on.  

            With that in place we can now say what each team’s expected winning percentage is (Score + 118, times .001679, except never greater than .640.)   And, having reached that marker, we can now draw up a list of the greatest under-achieving and over-achieving teams of all time.

So Give Me the List, Already

            Now we hit another one of those annoying glitches that make real-time research so frustrating.  Almost all of the greatest over-achieving teams we have identified so far either

            a)  Over-achieved for some obvious reason, like World War II, or

            b)  Are cases where maybe the process doesn’t really work.

            World War II reduced the quality of performance not only from 1942-1945, but also in the years just after the war.   In the years just after the war the scores are very low, as teams kind of went back to zero and had to start over.   Let’s begin, then, by presenting (and then ignoring) the over-achieving teams from 1942-1947:

            With those out of the way, that makes the #1 over-achieving team identified so far the 2001 Seattle Mariners.   That’s not too helpful, of course, because we have acknowledged already that our method is not reliable with teams from recent seasons.   I think that team did over-achieve by some good margin.   I don’t think a team with Mark McLemore in the lineup and Paul Abbott in the starting rotation is normally going to win 116 games.   But on the other hand, several players on that team are still active, and there may be a couple of Hall of Famers there who haven’t been honored yet, so we don’t really have a useful measurement.

            So with that nonsense out of the way, these are the top over-achieving teams that I have identified so far:

            Let’s look at those one at a time:

            1)  Pittsburgh Pirates, 1909.   This was a fascinating team, and in my view there is no doubt that they did over-achieve by a wide margin.  Maybe half of that is natural over-achievement.   The National League in that era—1904 to 1911—had three strong teams, and five really awful teams.    This allowed the good teams to just run away from the league, posting ridiculous won-lost records which systematically overstate the greatness of those teams to some extent.

            Beyond that, the Pirates just had a year.   Their lineup included Bill Abstein, whose expectation score is 2, and Jap Barbeau, who was another “2”.   You don’t normally find players like that on 110-win teams, no matter how weak the league might be.

            2)  Cleveland Indians, 1954.    A famous team, of course.   With a historic pitching staff and being chased by the Yankees, they went 22-7 in May and raced to 111 wins with a lineup including George Strickland, Bill Glynn and Dave Philley.   Our analysis suggests that this lineup would more normally win about 90 games—about the number the Indians usually did win in this era.

            3)  Chicago Cubs, 1906.   Another very famous team; they finished 116-36, and featured the double-play combination of Tinker to Evers to Chance.   Like the 1909 Pirates, they benefited from playing in a league which had been rent apart by the founding of the American League in 1901, and which would not fully recover until the 1920s.

            4)  Cincinnati Reds, 1939-1940.   One of the questions I posed earlier was, what causes teams to over-achieve?   Are there characteristics of teams that win more (or less) games than you would think they would, based on the strength of the roster.

            I think there is one characteristic of these teams that we can pin down.   Infield defense. 

            Remember, in our analysis we paid no attention whatsoever to batting statistics, so we can’t be discriminating in the common way against strong defensive players who are weak defensively.   The discrimination must be more circuitous:  that defensive specialists have shorter careers than more offensive players of the same value.   Otherwise, why would our system—which is based on age and career length—discriminate against defensive specialists?   Our system would discriminate against glove men only if baseball itself discriminates against glove men.

            Look at this list.  The 1939-1940 Cincinnati Reds had a brilliant defensive infield.  I’ve written about it in other places and won’t repeat that, but all four regular infielders were outstanding defensive players.   Their manager, Bill McKechnie, was a fanatic about defense, to the point of scorning hitters.

            Look at the team above them, the 1906 Cubs.   The most famous defensive infield of all time.  The 1954 Indians had two regular infielders, Bill Glynn and George Strickland, who were very weak with the bat, but who may have been outstanding defensively. 

            Bill Abstein, Jap Barbeau and Bill Glynn all disappeared from the major leagues quickly after playing their role on these remarkable teams.   Their teams didn’t believe enough in the value of their defensive contributions to keep them around—but on the other hand, those teams never performed at the same level after letting go of those players, did they?

            Am I saying that infield defense is under-rated?   Well. ..yes, if not exactly.   It may be that the “side benefit” we see to infield defense is unpredictable, that it manifests itself in certain geometric combinations, but has less impact in isolation.   It may be that this benefit is unreliable from year to year.  It may be that I’m imagining this benefit, which doesn’t really exist at all.

            But it may also be this:  that things which can be measured are often over-valued in comparison to things that are difficult to measure.   The 1954 Cleveland Indians replaced Bill Glynn with Vic Wertz, because Wertz offered very tangible benefits.   He could hit.  

            What Glynn did well was very difficult to measure (and still is, although we have gained a lot of ground.)  It is difficult for people to pass up benefits that they KNOW to be real to gain benefits that they cannot measure and cannot tally.   It is difficult emotionally to do that, and it is also often unwise to do that.   It is one thing to say that “I think infield defense is really important.”  It is a very different thing to make actual playing-time decisions based on that faith.   If you’re wrong, you lose, and if you lose you get fired—and you really don’t have any way of knowing how many runs Bill Glynn or George Strickland may have been saving with his glove work.

            And. . .I promise I’ll stop bloviating in just a moment.   But this is, in a sense, the central theme of my career:  the urgency of documenting unmeasured elements of the game.   What baseball statisticians normally do—what they normally did 30 years ago, and what they normally do now—is to try to pull conclusions out of the statistical record.   This is what people assume that statisticians do, and this is what people who don’t understand assume that I do.

            What I actually do is essentially the opposite:  I try to figure out ways to put things in to the record that are actual parts of the game, but which are missing from the statistics.  That’s what I’m doing 90% of the time; that’s what I’m doing right now.   I’m trying to find a way to get something that is a real and meaningful part of the game into the record books, so that other people can take it out of the record books.   I’m trying to add this concept, Team Talent, into the game—and yes, it’s speculative, and yes, it involves a great deal of guesswork and supposition, and yes, my math skills are slightly above “laughable”, and yes, people will be irritated with me because I have insulted the magnificent 1961 Yankees and insulted the heroic Jackie Robinson Dodgers, based on this Mulligan stew of invented approximations.   I understand that.   They’re just like the people who are mad at me because my Young Talent Inventory doesn’t list Joe Mauer or Ryan Zimmerman or whoever among the 25 top young talents in the game.  People have been mad at me about crap like that for 30 years, and I’m pretty well used to it.

            But it’s important to do it.   It’s important to do it for this reason:  that what you cannot measure, it is difficult to understand.   People would not argue about the existence of God if we could measure the length of his fingers.   Her fingers; whatever.   If we could measure God, we’d know.   People cannot argue about the power of hope if you can measure it, if you can document it, if you can predict it.   It is this conviction that gets me out of bed in the morning—or sometimes in the afternoon.  Let’s move on.

            5.  Cincinnati Reds, 1961.

            This was one of the most interesting things I found while doing this research.   The 1961 Cincinnati Reds were a Miracle Team—perhaps the least famous of baseball half-dozen or so Miracle Champions (the 1914 Braves, 1967 Red Sox, 1969 Mets, 2008 Rays.)   The Reds finished 67-87 in 1960, and went into the 1961 season as a consensus sixth/seventh place team.   They won 93 games, and won the National League pennant.

            The Reds substantially restructured their team in the winter of 1960/61 and in the first month of the 1961 season, trading their catcher for a second baseman, sending their first baseman to right field, moving their third baseman to shortstop, and inserting into the lineup a new catcher, new first baseman, new third baseman, and two new starting pitchers.   Some of you may remember that I discussed these adjustments in the last Historical Abstract.

            But here’s what I understand now that I didn’t understand two weeks ago.   The 1961 Cincinnati Reds were actually a weaker team, on paper, than the 1960 team had been—significantly weaker.   The 1960 Reds should have been, on paper, a contending team, finishing about 81-73.   They under-achieved by about 14 games.   The Reds shook up the team, making a much weaker team on paper—and then over-achieved in 1961 by about 19 games.  

            Why is that so fascinating?   Because it suggests that the Reds understood something about their players that is not apparent in the record books.   They understood, on the scene, that they had a reasonably talented team that was just going through the motions, and needed to be shaken as well as stirred. 

            The over-achievement of the 1961 Reds is essentially explained by an unusual number of players having career years.    First baseman Gordy Coleman, third baseman Gene Freese, center fielder Vada Pinson, and starting pitchers Jim O’Toole, Joey Jay and Ken Hunt all had the best years they had ever had—or were ever to have.   Right fielder Frank Robinson won the MVP Award, although he was later to have two even greater seasons (1962 and 1966).  

            And this also introduces the role of the manager in over-achieving teams.   It is common sense that over-achieving teams must over-achieve, to some extent, because they are well managed, and under-achieving teams must under-achieve in part because of a lack of direction.   The manager of the 1961 Reds, Fred Hutchinson, became a media darling after the 1961 season, and remained one of the iconic figures of that era.   When he died a few years later an award was established in his honor, the Hutch Award; it is still given annually, the 2008 winner being Jon Lester.    

            Fred Hutchinson had been a Navy officer in World War II, discharged at the age of 25 as a Lieutenant Commander (equivalent to a Major in the Army.   The ’61 World Series was a matchup between Hutchinson and Ralph Houk, who had been a Major in the Army in the war.)   There used to be a managerial style, the gruff but lovable boss.   Lou Grant.   Bear Bryant.  Dick Williams.  I’m sure there are still some of those guys around, but in the 1950s, early 1960s, they would actually slap you silly if you didn’t behave.  I suppose this is hard to believe now, but this was 50 years ago, that was the way things were at that time.   Hutchinson was one of those guys who would get right in your face and explain to you what you weren’t doing that you were supposed to be doing, and you were supposed to be intimidated by him and a little bit afraid of him, and you were supposed to have good reason to be a little bit afraid of him, because he was entirely willing to punch your lights out if the need arose.   It sounds harsh now, but you have to remember World War II.   That generation of men was forged in the fires of World War II, where other young men were dying violent deaths all around them.   A fist fight. ..what’s a fist fight compared to an early grave?   Stop whining.

            Most of those guys were more gruff than they were lovable, but Hutchinson was one manager who really made the gruff-but-lovable thing work.   The press always liked him, most of his players liked him, and they did play hard for him.  I think that Hutchinson does deserve some credit for the over-achievement of the 1961 Reds—as Fred Clarke, Al Lopez, Frank Chance and Bill McKechnie probably deserve some credit for the over-achievement of the four teams listed before. 

            But also, sometimes it just happens that several players have big years together.   The “infield defense” theory doesn’t really work here, as the ’61 Reds infield defense was probably below average. 

            For the 1998 Yankees, it is too soon to say for sure whether they over-achieved or whether they were one of the most talented teams of all time.  Well, they certainly were one of the most talented teams of all time; it’s too soon to know how high they were on the list.   We have Derek Jeter, for now, entered as a “19” in 1998.   Once he finishes his career and is voted into the Hall of Fame he will score at 30-something.   

            The 1961 Yankees we have already discussed, but they fit the profile of over-achieving teams with outstanding infield defense.   The Yankees’ infield defense was the strongest part of the team, other than the power. 

            6.  The Chicago White Sox, 1977.

            It was actually the ’77 White Sox who triggered this project, two months ago.   At that time it was already getting late in the free agent signing season, and it was apparent that the big hitters were going begging for jobs.    Pat Burrell was available then, Adam Dunn, Bobby Abreu, Manny Ramirez, Barry Bonds if you’re into that kind of thing and several other lower-level sluggers.   It was apparent that you could, if you wanted to, sign five of these players pretty cheap, and put out a lineup that would score a lot of runs but might have a fairly horrific defense. 

            And then I thought of the 1977 White Sox, who did essentially that; they picked up a lot of cheap hitters, sacrificed defense, and just tried to beat people up.   It worked spectacularly well.    They took a team that had no stars and several obvious weak spots, won 90 games, and also they played very entertaining baseball.

            Well, I’ve written about that team before, several times, and about the magnificent managerial performance of Bob Lemon.   But this time I got to wondering:   can you document that?   Can we measure the extent to which this team over-achieved?  How could we measure the extent to which the team over-achieved?   I really started this project by wondering if we could measure to what extent the 1977 White Sox over-achieved. 

            The ’77 White Sox certainly did not have a great infield defense, or a good infield defense, or an average infield defense, or a good outfield defense, or a good anything defense.   They just didn’t care.   They’d give you six runs and score seven.   They had about four DHs in the regular lineup—Oscar Gamble, Richie Zisk, Jorge Orta and Ralph Garr—and they had an outfielder at shortstop.   And they had no really good pitchers. 

            By our system we have them over-achieving by 18 games, and that may be conservative, because it was a pretty good league with a lot of strong teams.   In other words, the “dynamic over-achievement” of the team may have been off-setting a certain amount of natural under-achievement.   

            7.  Detroit Tigers, 1968.

            The Tigers for much of the 1960s under-achieved—or at least I believe they did; I haven’t studied the team for all of the relevant years.    There is an era there between the collapse of the Yankee dynasty (1965) and the rise of the Orioles in 1969 when the American League had no outstanding team.   It is my view that the Tigers had the most talented roster in the league in those years, and should probably have come out of it with more than one pennant.      

            The ’68 Tigers, on paper, were probably the weakest Detroit team of the 1960s.   They had Ray Oyler in the lineup, who couldn’t hit, and their relief ace was somebody named Daryl Patterson, who I don’t think is in the Hall of Fame.    The Tigers’ Hall of Famer, Al Kaline, was injured in ’68 and is not listed as a regular, which reduces our expectations for the team.    But Denny McLain won 31 games—obviously an over-achievement—Willie Horton hit a career-high 36 homers, Bill Freehan had career highs in homers and RBI, Jim Northrup drove in a career-high 90 runs, and the Tigers had a good year.