The path to Cooperstown through the hot corner

For my fifth annual THT column on position players and the Hall of Fame, I’m taking an in-depth look at the third basemen. As with my prior looks at slugging outfielders/first basemen, middle infielders, leadoff men and my two-parter on catchers, I’ll be presenting offensive statistical profiles of these players with three adjustments:

1. I focus on the block of “prime” seasons of a player’s career, rather than career totals or a smaller number of “peak” seasons;

2. within those seasons, I adjust batting stats for offensive context, both era and park; and

3. I present those adjusted batting lines in per-162-team-games notation.

You can examine my offensive method in detail as discussed in the column on sluggers—I’ve used the same constant methodology and baseline across this series—but I will review it here briefly before we proceed.

Prime value

My view of the Hall of Fame is that it’s fundamentally about the stars of the game, and accordingly that the core of a player’s Hall of Fame candidacy should focus on the years he was a star. Thus, at least in evaluating non-pitchers, the key inquiry for the Hall of Fame should be neither “peak value” (how good the guy was at his very best) nor “career value” (the sum total of his career) but “prime value.” Prime value is, roughly, looking at the number of years a guy had when he was a legitimate star and how good he was in those years.

In other words, when I look at a potential Hall of Famer, the first question I ask is, “How many seasons did this guy have where he was a Hall of Fame-quality ballplayer?” And the second is, “How good was he in those years—just around or above the line, or way above it?”

Now, I wouldn’t argue that you should throw peak or career value entirely out the window, but both have their flaws. Peak value really doesn’t capture the way most of us think about the Hall: as a shrine to a player’s sustained accomplishments of a period of years, not his very best day. Career value, on the other hand, has two drawbacks.

One is that that looking only at career value ends up putting too much emphasis on which guy played passably well when he was 38 and playing out the string as a part-timer, rather than the years when he was doing the things we’ll remember him for. I’m a big believer that you don’t play your way out of the Hall in your old age, and neither should you get inordinate credit for padding the career totals with mediocre or part-time seasons. Some credit, if the seasons are productive, but only at the margins if it’s a close call.

Second, baseball is played in seasons. If you look just at career totals, you miss that—you miss the fact that, at least for a star player, two seasons of 600 plate appearances really are worth more than three seasons of 400 PA at the same level of production, because the 600-plate-appearance seasons move the team closer to winning championships, while that third 400-plate-appearance season blocks the team from fielding a full-time replacement (the value of that is debatable, of course, depending what the team’s options are, but if we’re talking major stars, you’d almost always prefer full seasons).

In-season durability is a very important measurement of value, as it minimizes the amount of playing time that needs to be given to weak second-stringers. Most efforts to evaluate players by running career totals miss this.

As a result, what I have tried to do here is excerpt out the consecutive series of seasons, ranging here from five to 16 years, when each guy was a star and weigh that chunk against other guys’ primes. As you will see, I do look in some cases at the years beyond the prime years, for purposes of distinguishing between the players who tacked on extra seasons as part-time contributors and those who just stopped hitting or stopped playing. But I zero in mainly on the prime.

The criteria

I varied slightly the criteria for picking players and seasons over the first four columns, so I’ll flag here what I did differently this time:

1. I define seasons listed under the “other” column as seasons at a certain level of OPS+ (on base plus slugging, relative to the league average, as calculated by over a certain number of plate appearances. I wasn’t completely rigid about defining “prime” in those terms—again I tried to isolate the part that let each guy put his best foot forward—but for the “other” column, I tallied up the non-“prime” seasons when the player had 500 or more plate appearances with an OPS+ of 100 or better, or 400 plate appearances with an OPS+ of 120 or better.

In addition to the number of seasons, the chart presents under the “Seas” column the number of prime years in seasonal notation, i.e., total team games divided by 162, which will be 1 for a 162-game season, 0.95 for a 154-game season, 0.67 for a 108-game season, etc. Players whose primes straddled the strikes, World War I or the dawn of standard schedules around 1904 will have a bigger gap between the seasonal notation and the number of seasons.

2. I divide the charts in two groups—the “long prime” players with a prime of eight or more years, and the “short primes” of five to seven years—to avoid mixing apples and oranges. In the usual case, you need an eight- or nine-year prime, at a minimum, to make a credible Hall of Fame case as a non-pitcher.

The Incompleat Starting Pitcher
The end of the nine-inning start and how we got here.

3. For the most part, there really weren’t many star third basemen who significantly straddled 1920, the breakpoint I’d used in prior columns on the theory that earlier players really aren’t a useful yardstick for evaluating contemporary Hall of Fame candidates who played in the lively-ball era. Regardless of the merits of comparing players from that era with sophisticated statistical analysis, the shape of their statistics is simply too different to make them meaningful yardsticks in real-world arguments about putting modern players in the Hall.

That said, since great third basemen were so rare before 1920—indeed, the great bulk of the list is from after World War II—I also include a separate list here of the pre-1920 third basemen going all the way back to 1871. I haven’t divided those by length of prime since the list is mostly short-prime guys.

On the whole, I looked at 61 third basemen, sweeping broadly to illustrate both the relative rarity of the players at the top of the pyramid and to give context to the relative ordinariness of the guys further down who are surrounded by people you don’t think of as immortals. The list consists of 13 Hall of Famers (two of whom, Cap Anson and Tony Perez, are in as first basemen, and a third—John McGraw—who is in as a manager), eight active third basemen (including Alex Rodriguez, whose years as a shortstop I examined in the column on middle infielders), and 40 players who are off the writers’ ballot; at this point I’ve about given up trying to figure out who will and won’t be on the next Veterans’ Committee ballot, although Ron Santo remains a plausible candidate.

It should be noted that third base is an odd position because a lot of Hall of Famers have put in time at third base without making a career of it. Some, like Carl Yastrzemski, were just brief experiments, but the following Hall of Famers spent one or more seasons in which they qualified for the batting title and spent at least half their team’s games at third base:

Five seasons: Harmon Killebrew, Paul Molitor, Perez, Joe Sewell, George Davis
Four seasons: Anson
Three seasons: Cal Ripken, Bobby Wallace
two seasons: Rogers Hornsby, Frankie Frisch, Travis Jackson, Buck Ewing
One season Honus Wagner, Jimmie Foxx, Mel Ott, Arky Vaughan, Ryne Sandberg (who won a Gold Glove at third; I believe he’s still the only player to win one at two positions), Luke Appling, Roger Connor, Jim O’Rourke

One can find similar examples among guys who are part of the Hall discussion or otherwise had long, successful careers: A-Rod (six, including two MVP Awards), Pete Rose (four), Dick Allen (four, including a Rookie of the Year award), Bobby Bonilla (four), Gary Sheffield (three), Edgar Martinez (three), Jim Thome (three), and Joe Torre (two, including an MVP Award).

The numbers

The first caution I would emphasize here is that this is just a study of batting stats. Having scrapped it in the catchers column, I’m not reviving here my rough, seat-of-the-pants defensive grading system from the prior columns, which didn’t add anything analytical but simply congealed conventional wisdom for convenience.

The batting percentages here (batting average, slugging average, on-base percentage—for the charts I stuck with listing slugging before OBP for consistency with prior columns, but used the Avg/OBP/Slg convention in text) are translated statistics. You can read a detailed explanation of the method in the slugging outfielders article; for consistency with the earlier pieces I used the 2005 National League as the average “season.”

All stats are from (I translate based on differentials in league Avg/Slg/OBP rather than runs per game, thus reducing the distorting influence of variation over time in the volume of unearned runs). The essential idea of translations is to show what a player’s performance in Year X and Park X was equivalent to in Year and Park Y, not to project what a guy would have hit under other circumstances, which is unknowable.

The other offensive numbers—plate appearances, steals and caught stealing, double plays—are actual, not translated (I included base stealing and GIDP figures because they’re the two main components of offense that aren’t captured by slugging and OBP), but are averaged per 162-game season, so as to put players whose careers were in the 154-game era or were interrupted by strikes on a common footing. (As discussed above with regard to seasonal notation, for example, 1981 is generally counted as two-thirds of a season in the averaging.)

The “Rate” column in the chart is simply (translated Slg)*( translated OBP)*( translated PA). It’s not any kind of scientific formula, just a handy metric to organize the data on the table by the three main variables. I prefer multiplying rather than adding slugging and OBP (as is done with OPS), since a single point of OBP is worth more than a single point of slugging. As you can see, this metric organizes the data very strongly in favor of guys who were very durable in-season and against guys with low OBPs.

Of course, besides not counting defense, the “Rate” metric doesn’t count the baserunning and double play numbers on the chart, so don’t treat the rankings as holy writ. But they’re a good rule of thumb.

With those lengthy preliminaries out of the way, let’s run the offensive numbers, starting with the longer-prime guys:

Prime  Oth Player               Age    Seas    Avg    Slg    Obp   Rate   G-3B     In?
    11  5  Eddie Mathews      21-31    10.6   .285   .557   .390  148.5    150    YES
     9  3  Wade Boggs         25-33     9.0   .338   .481   .425  144.1    151    YES
    14  0  Mike Schmidt       24-37    13.7   .267   .572   .375  140.4    141    YES
    13  2  Chipper Jones      24-36    13.0   .303   .529   .390  128.1    110 Active
    11  7  George Brett       22-32    10.6   .317   .547   .376  124.3    133    YES
    11  1  Ron Santo          23-33    11.0   .280   .507   .367  123.1    155     No
     8  3  Stan Hack          26-33     7.6   .293   .436   .374  116.5    153     No
    10  1  Sal Bando          25-34     9.9   .268   .474   .366  113.9    151     No
     9  1  Ken Boyer          25-33     8.7   .287   .484   .349  113.5    144     No
    10  2  Bob Elliott        25-34     9.6   .289   .487   .366  113.2    130     No
     8  2  Scott Rolen        22-29     8.0   .279   .502   .362  113.0    144 Active
     8  3  Toby Harrah        26-33     7.6   .272   .450   .376  110.6    108     No
    10  2  Ron Cey            26-35     9.7   .266   .487   .354  108.9    151     No
     8  0  Ken Caminiti       28-35     7.6   .284   .495   .355  106.3    142     No
    11  1  Eddie Yost         23-33    10.5   .257   .403   .385  104.4    145     No
     8  4  Graig Nettles      26-33     7.9   .260   .477   .332  104.1    157     No
     9  2  Robin Ventura      23-31     8.6   .277   .464   .358  103.7    142     No
    12  1  Brooks Robinson    23-34    12.0   .287   .463   .330  103.4    158    YES
     8  6  Buddy Bell         27-34     7.7   .287   .452   .349  100.7    145     No
    10  6  Darrell Evans      26-35     9.7   .248   .449   .355  100.0    117     No
    10  1  Bill Madlock       23-32     9.6   .310   .484   .365   99.2    111     No
     8  1  George Kell        23-30     7.7   .311   .458   .352   98.2    137    YES
    11  0  Pie Traynor        24-34    10.5   .290   .444   .331   97.0    151    YES
     9  0  Troy Glaus         23-31     9.0   .252   .492   .352   96.7    125 Active
    10  0  Matt Williams      24-33     9.6   .274   .514   .319   96.1    137     No
    13  1  Buddy Bell         23-35    12.6   .286   .441   .344   95.3    145     No
     9  1  Tim Wallach        24-32     9.0   .268   .466   .321   94.6    153     No
    10  0  Ken Keltner        21/31     9.6   .272   .474   .316   93.7    149     No
     8  0  Mike Lowell        26-33     8.0   .275   .454   .333   93.5    145 Active
    11  2  Carney Lansford    22-32    10.7   .297   .446   .346   90.7    127     No
    16  0  Graig Nettles      25-40    15.6   .254   .460   .332   89.9    142     No
    10  0  Doug DeCinces      26-35     9.6   .265   .483   .333   89.0    131     No

* – Data incomplete or unavailable.

I’d profiled Eddie Mathews in comparison to the other great third basemen when he died in 2001, so I wasn’t caught totally by surprise at how good he really was, but even I was surprised that he rates as decisively better in his prime than any other player at his position. Mathews beats Mike Schmidt offensively for two reasons: He hit for better averages, thus producing better OBPs, and he was amazingly durable, even by comparison to a guy like Schmidt who never had a major injury.

To cap it off, Mathews hit into the second-fewest double plays of anybody on the chart (unlike Schmidt, without the aid of a ton of strikeouts) and rarely made outs getting caught stealing. He was also an excellent defensive player, arguably as good as Schmidt. As Bill James has observed, Mathews doesn’t get the respect he deserves for a variety of reasons: because he peaked early and was overshadowed by the longer career of teammate Hank Aaron; because he played in Milwaukee, in a pitcher’s park in an age of bandboxes; because his best years didn’t quite match up with those of his teams, denying him any MVP recognition.

How underrated was Mathews after his retirement? His first year on the Hall of Fame ballot, in 1974 (the year Aaron broke Babe Ruth‘s record), Mathews got 32.3 percent of the vote. This for a man who, at the time, stood in 10th place on the all-time home run list with 512. It took Mathews until his fourth year on the ballot to crack 50 percent, and his fifth year to win induction. Yet the record is clear: Mathews was, at the time he retired, easily the best ever at his position, and more than 40 years later, it remains a fair debate if he still deserves that title.

Wade Boggs is also a little bit of a surprise, if only because we’re still unaccustomed to setting him side by side with the greatest ever at his position. Boggs was in some sense a creature of Fenway; for the years of his prime (1983-91), he batted .382/.478/.542 at home, compared with .308/.395/.406 on the road. Over those years, Boggs averaged 58 doubles per 600 at-bats at home, 29 on the road.

But then, the ability to uniquely exploit your home park is a useful skill; even if Boggs may not have been as valuable in another home park, he was a critical part of a team that won three division titles (and one pennant) in those years. The Red Sox from 1983-91 played at a 93-win pace (416-311, .572) at home, and a 77-win pace (346-384, .474) on the road; from 1986-90, a 101-win pace (251-153, .621) at Fenway, a 73-win pace (182-223, .449) on the road. Boggs was a major factor in their dominance at Fenway and must be valued as such.

Mike Schmidt is now the conventional choice as the greatest of all third basemen, and given that his prime years lasted three years longer than Mathews’ I might still take him as No. 1. (Then again, Mathews had six solid seasons outside his prime; Schmidt had none. It’s a close call.) He shows up here as averaging fewer games at third due to his one-season experiment in 1985 as a first baseman (in an abortive attempt to break in Rick Schu as a successor), but he remained a steady defensive player at third (and won the 1986 NL MVP) after returning.

Chipper Jones averages the fewest games at third per year for his prime on this list because his prime covers a two-year period in which he played every day in left field. Chipper wasn’t moved for being a butcher at third, although the move does reflect that he’s never been regarded as the kind of outstanding glove man several of the other elite third basemen were.

Adjusted for context, only Boggs tops him in getting on base, only Schmidt had a longer prime, and only Mathews, Schmidt and Brett were more powerful sluggers; Chipper lags behind the first tier only for being a bit less durable year-to-year than the top three. By now, it’s not news that Chipper should and most likely will skate into Cooperstown.

It’s hard for me to be dispassionate about George Brett, perhaps my favorite player ever and one of the very few guys whose resume of clutch hitting defies any effort to apply the usual statistical cautions about the elusive nature of “clutch” (although my own theory is that Brett’s career average of .337/.397/.627 over 43 postseason games is in part a reflection of him being healthier during postseason series than over the long seasons). It’s now 30 years since Brett’s magical 1980 season, when he batted .390/.454/.664, drove in more than a run per game, and had more homers (24) than strikeouts (22).

When placed in context, Brett nonetheless comes up short of his contemporaries Schmidt and Boggs on durability and walks, but as usual there are caveats. He had an impressive second career outside his prime years as a first baseman and then DH, including a .306/.389/.509 season in 1988 (103 RBI, 149 OPS+) and a .329/.387/.515 last hurrah in 1990 (AL batting title, 153 OPS+). The 1990 season was one for the books: The 37-year-old Brett was batting an anemic .256/.330/.324 on July 1, then suddenly flipped a switch and turned back into the Brett of 1980 for half a season, batting .386/.433/.663 with 36 doubles in 78 games the rest of the way.

Using the context adjustments above, Brett’s 1986-90 seasons (he moved to first in 1987, DH in 1991) score a rate of 111.0, with a batting line of .302/.382/.507. On the whole, I’d still rate Brett ahead of Chipper, but it’s close.

Ron Santo should be in the Hall of Fame. Period. When you break down the third basemen this way, you see immediately that there’s a distinct first tier: Mathews and Schmidt are the A+ students; Boggs with his slightly shorter prime is the A student; Chipper, Brett and Santo are the A- students. However you break that group out, it’s the immortals at the position, and Santo is one of them, and a fine glove man to boot (five Gold Gloves).

Santo was the total package—good batting averages, excellent power, patience, tremendous durability and consistency; the only chink in his armor is a slightly large number of double play balls. Like Boggs, Santo was especially well-suited to exploit his home park, batting .296/.383/.522 at home for his career compared with .257/.342/.406 on the road. In any event, the benefits Santo reaped from Wrigley Field are largely offset by the extreme pitcher-friendly conditions of his era.

That brings us to the guys who inhabit the borders of Hall-worthiness.

Stan Hack, the poor man’s Boggs, was also a fine third baseman who played in Wrigley, but there the similarities to Santo end; Hack’s prime years were shorter than Santo’s, ran through a high-scoring era (although the late 1930s were not nearly as hitter-happy in the NL as in the AL) and got a boost at the end from World War II (I count 1943 as his last prime season, but he excelled in 1945 as well); on the other hand, Hack wasn’t a power hitter of the type Wrigley generally favors. He certainly deserves to be remembered as a star of three pennant-winning Cubs teams, but Hack was not enough of an offensive force to bust down Cooperstown’s doors on the strength of an eight-year prime.

Sal Bando isn’t quite Hall of Fame material either, although the Hall would not be in any way embarrassed by his presence; it can and has done far worse. As you’ll see from the next chart, he was a better player at his peak than Tony Perez, and his prime as a whole holds up better than Perez’s when you consider their positions. He may well have been a better player than Jim Rice, too.

Bando is the classic underrated player: pitcher’s park, pitcher’s era, value concentrated in his power and walks rather than batting averages. But really, anybody who thinks Pie Traynor was a better baseball player than Bando … I want to be in a league with that guy. Despite his disadvantages, Bando’s value was recognized by the writers of his day—he was runner-up for the MVP to Vida Blue in 1971, fourth in 1973, third in 1974.

After being one of the anchors (along with Reggie, Catfish and Fingers) of the A’s dynasty that won five straight division titles and three consecutive World Championships (the only non-Yankees team to manage that feat), Bando arrived in Milwaukee in time to help spark the Brewers’ turnaround into a contender. Bando’s the only guy on this list with more than two World Championships to his name; Mathews, Boggs, Schmidt, Brett and Chipper each won one.

Ken Boyer was perhaps more respected in his day than Santo, winning an MVP Award and going on to manage, but offensively Boyer didn’t do anything quite as well as Santo besides hit for average, and his prime didn’t last as long.

Bob Elliott was a converted outfielder whose prime years include a couple of seasons of battering war-depleted pitching, but his MVP season for the Boston Braves in 1947, which he basically reprised for the 1948 pennant winners, shows that he was no fluke. Elliott was dumped when the Braves came up with Mathews, and he never had another good year.

Scott Rolen is the classic guy who could have been a Hall of Famer if he’d stayed healthy—he started early and well, but his prime years were dinged here and there by injuries that stole playing time and leached productivity, and then he ran off the bridge. He’s still a productive player now, but a shell of his old self.

Toby Harrah may look out of place in this company, but for eight years he was a fine offensive player and only missed time at third to play shortstop. Like a number of the guys on this list, he was a patient hitter in a lower-scoring era. Harrah’s also one of the few of the long-prime third basemen who was any sort of base thief.

Ron Cey was a slightly lesser Bando, a guy who had power and patience, was a key contributor to a lesser dynasty of sorts (four pennants, including one World Championship, with the Dodgers) while playing in a pitcher’s park in a low-scoring era, and helped his next team (the 1984 Cubs) over the hump. There are only two World Series-winning teams prior to the 1990s to not have a Hall of Famer (the 1981 Dodgers and 1984 Tigers), and the multiple just-short candidates from each team have a good deal in common. Cey, like Davey Lopes, suffers partly from having had a late start to his career.

Ken Caminiti’s story is a tragic one, whether or not you believe that his use of steroids contributed to the toll the cocaine and other drugs took on his heart, killing him at age 41. Certainly Caminiti was one of the poster boys of the steroid age, batting .260/.319/.378 his first five years as a regular, with a career high of 13 home runs, to .293/.382/.531 from age 31-37, including his 40-homer MVP season in 1996—a transformation that’s not explained simply by the 1994 offensive explosion and his departure from the Astrodome.

Caminiti just went bonkers down the stretch in 1996, batting .368/.457/.756 with 32 homers and 98 RBIs in 86 games from June 25 to the end of that season. But steroid asterisks or no, while Caminiti was very briefly a great player and for an eight-year stretch a very good one, he sustained greatness for far too little time to be a Hall of Famer.

Eddie Yost didn’t do much but walk, but for 11 seasons, mainly for the dead-end old Washington Senators of the 1950s, he walked enough to be a tougher out than anybody on this list besides Boggs, Mathews or Chipper. Not bad.

Graig Nettles is one of the guys I list twice here, because his best eight-year stretch rates him considerably higher, while if you stretch his prime over 16 seasons it’s of similar quality, albeit with diminished playing time. Nettles, of course, was also an astonishing defensive player as well as one of the wittiest, most quotable players in memory. My favorite Nettles quote, from 1986, when he was 41 and in his third-to-last season: “I play for two reasons. One, for the fun, the other for the money. When it stops being fun, I’ll play for the money.”

Robin Ventura, like Rolen, was tripped up by injuries in his prime; Ventura’s skill set as a relatively patient power hitter with a good glove gives him much in common with many others on this list, but over fewer years. Ventura will always be remembered by Mets fans for the “grand slam single” in the rain to win Game Five of the 1999 NLCS from behind in the bottom of the 15th, and in Texas as the guy who charged the mound on Nolan Ryan in a 1993 brawl and got trapped in a headlock and pummeled for his troubles.

Brooks Robinson was a better hitter, especially for average, than his raw numbers would suggest, but he simply doesn’t belong in a Hall of Fame discussion on the basis of his bat. That said, his durability (note the 158 games played at third base per 162 scheduled team games, over a 12-year stretch) and longevity, combined with solid offensive output, gives him a sound basis to be enshrined for his glove. Buddy Bell was a similar hitter to Robinson and also a good fielder over a long career, but not as good in any respect.

Darrell Evans was another third baseman in the Bando/Cey mold, and like Brett he effectively had a second career—in Evans’ case, arguably more successful than the first—at first base (.253/.362/.489, Rate of 102.8 from age 36-40). Evans had a few too many lost seasons in his prime, but at his best (two 40-homer seasons 12 years apart) he was a very dangerous slugger.

How much did Evans hit the ball in the air? In 1987, as a slow-footed 40-year-old batting in the middle of a lineup with enough baserunners to score 896 runs—Evans batted with a man on first 209 times—and striking out just 84 times, he hit into only two double plays all year.

Bill Madlock was a great hitter for average (four batting titles), and had a bit of power and—at times—speed (career high: 32 steals), but wasn’t the most patient hitter and was held back by missing time year in and year out to nagging injuries. Note that Madlock’s low number of games at third was due to a year-and-a-half experiment of the Giants playing him at second base.

Madlock was a notorious second-half hitter (career: .290/.350/.424 in the first half, including .256/.307/.360 in April; .321/.384/.462 in the second half, including .348/.412/.511 in August), and was acquired in-season in time to help the stretch runs of the 1979 Pirates, 1985 Dodgers and 1987 Tigers.

If you list the guys who are in the Hall of Fame as third basemen and whose prime years began during the years 1871-1950, you get a grand total of five players: Jimmy Collins, Frank “Home Run” Baker, Pie Traynor, Fred Lindstrom, and George Kell. The Hall opened its doors in 1936, but it didn’t have a third baseman until the Veterans Committee inducted Collins in 1945, followed by the writers inducting Traynor in 1948, the Veterans inducting Baker in 1955, the Veterans returning—after a two-decade hiatus—with Lindstrom in 1976, the writers finally relenting to let in Mathews in 1978, and then Brooks Robinson and Kell being inducted by (respectively) the writers and the Veterans in 1983.

I offer that list as a mild apologia for the poor quality of some of the early selections: These guys may not look like much when set in context next to their modern counterparts, but in a few cases they were the best of their eras. That said, Kell really isn’t defensible even giving a fair amount of credit for his glovework. As a hitter he was a lesser Madlock—he wasn’t a significant offensive force for most of his career (he drove in as many as 70 runs in a season just four times), and his prime years were not long.

The best AL third baseman between Baker and Brooks Robinson wasn’t Kell; it was Yost or Harlond Clift or Al Rosen, depending how much stress you place on the length of their primes. Kell’s prime runs only one season longer than Clift’s, and Kell was clearly the weaker hitter. I put little enough stock in the writers, but there’s a reason they never gave Kell more than 37 percent of the vote.

When I was a lad in the 1970s, conventional wisdom regarded Pie Traynor as the greatest third baseman of them all, on the basis of seven 100-RBI seasons and a .320 lifetime batting average. Conventional wisdom was insane, given the opportunity to choose Mathews instead.

Traynor from age 24-34 averaged 5 home runs and 39 walks per 162 games, plus his batting averages (unlike his OBPs) failed to reflect more than 20 sacrifice hits per year, including 77 in a two-year period (his batting average in 1927-28 drops from .340 to .318 if you include the sac hits as outs). When you run the whole thing through the sort of context adjustment I’m doing here, so much air lets out of the Traynor balloon that he ends up this far below the likes of Ron Cey, Toby Harrah and Darrell Evans as a hitter.

For all that, I’m not suggesting the voters were wrong to enshrine Traynor. Depending on how much credit you give for his defense—and the weak offensive numbers of pre-Mathews third basemen suggest how much of a defense-first position this was for so many years—he really does have a case to be the best third baseman of the 1900-1950 period, when you consider his 11-year prime, durability and consistency.

Collins and Elliott are the only other contenders whose primes are of comparable length and offensive quality, and Elliott isn’t regarded as a great glove man. But with the arrival of superior two-way third basemen after 1950, Traynor no longer has any business in any serious discussion of the greats at the position.

Troy Glaus had, in his his prime, genuinely elite-status raw power, but injuries and a low batting average put a damper on the quality of his career. Much the same could be said of Matt Williams, although in Williams’ case it was lack of walks, not lack of hits. Tim Wallach was another poor man’s Brooks Robinson, which of course is an excellent thing to have on your team, just not a Hall of Famer.

Ken Keltner is immortalized in baseball analyst circles for Bill James’ vivisection of the famously poor case made for his Hall candidacy by a committee of his friends. As I’ve done before, I just ignore the missing year that Keltner lost to war, but his low OBP in particular erases any doubt that James was right about his being too far back of the pack for it to matter.

The last three names—Mike Lowell, Carney Lansford and Doug DeCinces—are in the “just happy to be listed” category. Lansford comes out here as basically the same hitter as Traynor, but with more injuries.

Next, we move on to the shorter primes among the post-1920 third basemen:

Prime Oth Player              Age    Yrs   Avg   Slg   Obp   Rate  G-3B      In?
    5  1  Al Rosen          26-30    4.8  .295  .561  .374  142.6   146      No
    6  8  Alex Rodriguez    28-33    6.0  .292  .549  .389  139.6   145  Active
    7  5  Ron Santo         23-29    7.0  .289  .530  .378  137.6   160      No
    5  0  David Wright      22-26    5.0  .304  .502  .380  128.9   155  Active
    5  4  Sal Bando         25-29    5.0  .281  .500  .383  128.8   157      No
    5  9  Tony Perez        25-29    5.0  .288  .527  .345  122.7   152     YES
    5  0  Howard Johnson    26-30    5.0  .263  .527  .350  119.4   122      No
    7  1  Harlond Clift     23-29    6.7  .258  .467  .362  117.8   156      No
    5  0  Jim Ray Hart      22-26    5.0  .285  .525  .350  114.1   118      No
    5  0  Whitey Kurowski   25-29    4.8  .286  .519  .348  111.7   144      No
    5  1  Fred Lindstrom    20-24    4.7  .296  .474  .332    106   138     YES
    7  0  Eric Chavez       22-28    7.0  .267  .479  .342  101.4   144  Active
    7  1  Aramis Ramirez    25-31    7.0  .284  .503  .339   97.4   133  Active
    6  3  Richie Hebner     21-26    6.0  .290  .503  .354   93.8   126      No
    6  0  Bill Melton       23-28    6.0  .262  .478  .332   87.4   117      No
    6  2  Bob Horner        20-25    5.6  .272  .542  .329   85.8   108      No

* – Data incomplete or unavailable.

Al Rosen didn’t get a regular job until he was 26 (he batted .328 and slugged .563 in the minors while blocked behind Keltner), started breaking down with back trouble at 31, and was retired by 33—but in between, for five seasons, he was as good as the best ever at the position.

I’ve covered Alex Rodriguez already in the middle infielders column, but while he was already qualified for the Hall just on the basis of his time as a shortstop, he’s closing in on also compiling a full Hall of Fame career just as a third baseman, having won two MVP Awards at the position. In A-Rod’s case, of course, Cooperstown debates are all about the effects of steroids on sportswriters.

I’ve listed Ron Santo’s best seven-year prime here for context, as it really stands out in terms of quality. Ditto Sal Bando’s five best. Both are discussed above.

David Wright, like A-Rod, is listed here through 2009. Five seasons don’t a Hall of Famer make, and even a healthy Wright could fall off the pace of the real elite third basemen, especially if he’s unable to solve Citi Field as a home run park. But entering his age-27 season, Wright still stands an excellent chance to compile a no-questions-asked Hall of Fame resume as long as he can stay healthy.

Tony Perez came out rather poorly when I stacked up his prime against the first basemen and corner outfielders. If you sever out just the years of that prime when he played third, he looks much better, with outstanding power and a more acceptable OBP. Perez in 1970 was just the 10th third baseman to hit 40 homers (Mathews did it four times, Harmon Killebrew three, Rosen and Dick Allen once each).

But of course, even in that five-year period, Perez was at best erratic at the position, making 67 errors over one two-year period (actually, Perez was moved after the 1971 season, the first of his career as a third baseman when he’d topped the league average in both range factor and fielding percentage), and overall his years as a first baseman just don’t measure up to the Hall’s standards. Had he posted five more seasons of the same quality as the five at third, he’d be a much more legitimate Hall of Famer.

Howard Johnson earns, oddly enough if you think of him as a compact home run hitter with Popeye the Sailor Man forearms, the distinction of being by far the best base thief among the post-1920 third basemen considered here. HoJo, now the Mets batting coach, was famous for a couple of things.

One, he was a dead-fastball hitter who gained a reputation for big late-inning homers mainly because he owned closers—especially Todd Worrell—who relied heavily on their heat. Two, he had a highly pronounced pattern of playing better in odd-numbered years. Three, he was frequently deployed, sometimes with ghastly results, as a shortstop, mainly due to Davey Johnson‘s view that the team could afford a weak defensive shortstop on days when they expected fewer ground balls—especially when Sid Fernandez, an extreme power/flyball pitcher who generated few ground balls, was on the mound.

HoJo averaged 43 appearances at short per year from 1986-91 despite the fact that nobody argued that he was a good defensive shortstop, but he never complained about the duty (he was better defensively at third, but also very error-prone, as his lifetime .929 fielding percentage attests).

And four, his power/speed combination: There have only been 17 seasons of 30 homers/30 steals by an infielder in the game’s history, and 12 of those have come since 1996; HoJo had three of the first five (along with Tommy Harper and Joe Carter), and in fact averaged 31 homers and 32 steals over the five seasons of his prime. Anyway, for these purposes, Johnson was unquestionably a Hall of Fame-caliber player from 1987-91, but those five years were his only years as a productive regular.

Harlond Clift is sometimes touted as a Hall of Fame candidate, having been the original power-patience-glove prototype that became the template for later success stories at third. His defensive stats are outstanding, although influenced by context (playing for a team with a 6.24 ERA that strikes out 2.6 men per game, you have a lot of defensive opportunities), as were his batting stats in the wild offensive context of the late 1930s.

Certainly he was the best AL third baseman of the 1920s/30s/40s, but unrecognized by virtue of playing for horrible St. Louis Browns teams. Unfortunately, Clift hit the wall at age 30 in 1943, including suffering a debilitating case of mumps that spread to his testicles; he later suffered a horseback riding injury and was done as a major leaguer by age 32 (although his bat revived in the low minors after the war). As a result, there’s really only the seven seasons of Clift’s prime and not much else.

Jim Ray Hart‘s career with the Giants of the 1960s was finished off by injuries even more swiftly than Clift’s or HoJo’s (Hart never had 400 plate appearances in a season after his age-26 season), but Hart, too, was a tremendous hitter. Ditto for Whitey Kurowski, although like many of his 1940s Cardinals teammates, Kurowski’s prime numbers include three seasons facing weak wartime competition.

Context may explain the Hall’s selection of Pie Traynor and George Kell, but all the context in the world can’t avoid the fact that Fred Lindstrom is a ridiculous choice for the Hall, one of the three or four worst ever, along with his infield-mate George Kelly. Both remain notorious examples of Frankie Frisch’s influence in getting his old teammates elected by the Veterans Commitee.

Lindstrom got his career off to a solid start, and did finish second in the MVP balloting at age 22, but after his 25th birthday he batted .288/.326/.412, appeared in 100 games in a season just twice, drove in as many as 65 runs in a season once, and played his last game at age 30. And even in his early years, when Lindstrom rode the wave of an NL batting average that went as high as .303, his power wasn’t exceptional and he rarely walked. There are easily 20 better third basemen on the outside looking in than Lindstrom.

Besides being the man who proves that even Billy Beane can look dumb when he has money to spend, Eric Chavez is yet another high-quality, though not Hall-quality, third baseman cut down by injury after reaching his late 20s. Aramis Ramirez hopes to avoid the same fate. His Rate here would be 103.4 if you excluded his injury-plagued 2009, but if Ramirez is going to land a respectable place among the longer-prime third basemen, he’ll need to press on through the setbacks of 2009.

Richie Hebner ranks low on this list due to gaps in his playing time, and after age 26 he hit some off years and then moved to first base. But in his prime in the low-scoring early 1970s, Hebner was a fine hitter, with an OPS+ as high as 152 in 1972.

Similarly shortened primes, and unlike Hebner not followed by much of a second act, afflicted slugging third basemen Bill Melton and Bob Horner. Horner had Hall of Fame power and a short swing that made him hard to strike out, but he didn’t walk much, his batting averages were inflated by his home park, and his weight problems made him useless on the field and the bases and prone to injury.

In prior columns, I left off at this point, but for the third basemen I thought it would also be interesting to look at their predecessors over the game’s first five decades. Note that the seasonal notation reflects the fact that these guys played much shorter schedules—Cap Anson, for example, played just 2.36 seasons worth of 162-game schedules over a seven-year period dating back to the start of the National Association in 1871:

Prime Oth Player             Age   Yrs    Avg    Slg    Obp  Rate  G-3B    In?
  6    0  Levi Meyerle     21-26   2.0   .336   .579   .407 160.7    81    No
  7    18 Cap Anson        19-25   2.4   .326   .511   .409 158.7    88   YES
  6    4  Frank Baker      23-28   5.7   .322   .574   .362 138.7   156   YES
  7    1  Denny Lyons      21-27   6.0   .315   .532   .400 132.9   115    No
  6    3  Heinie Groh      25-30   5.5   .301   .481   .380 123.5   142    No
  14   1  Ezra Sutton      20-34   6.7   .294   .468   .357 113.7   122    No
  10   0  Ned Williamson   21-30   6.5   .256   .459   .361 110.4    93    No
  6    0  Bill Bradley     22-27   5.4   .296   .502   .337 109.7   151    No
  8    0  Bill Joyce       24-32   7.7   .269   .502   .396 106.9    95    No
  9    1  Jimmy Collins    27-35   8.1   .283   .479   .329 105.7   153   YES
  7    11 Deacon White     34-40   4.9   .301   .438   .362 104.3   130    No
  9    0  John McGraw      20-28   7.7   .306   .441   .428 104.0    99   YES
  7    0  Red Smith        22-28   5.6   .285   .463   .348 102.5   157    No
  8    0  Heinie Zimmerman 24-31   7.5   .300   .509   .327 101.5   106    No
  7    4  Tommy Leach      24-30   6.5   .276   .470   .324  96.2   109    No
  6    1  Art Devlin       24-29   5.8   .274   .421   .358  93.6   151    No
  10   2  Larry Gardner    26-35   9.3   .280   .445   .323  90.6   152    No

* – Stolen base data incomplete or unavailable. None of the pre-1920 third basemen had complete caught stealing data or any double play data.

Levi Meyerle is the kind of player who’s really impossible to remove from the context of 1870s baseball; he won the first batting and home run titles in organized baseball, hitting .492 and slugging .700 in 26 games in 1871 in the newly formed National Association (the .492 isn’t exactly a real record, but it’s the highest batting average ever recorded over a major league season), but after 1877 he moved on to other leagues that we no longer consider “major.” Like Anson, he illustrates the brevity of the schedules played at the time. That said, Meyerle was clearly one of a handful of the standout stars of the infant pro game.

Cap Anson was an offensive monster for two and a half decades, although the extent of his dominance is also exaggerated here by the fact that early baseball, being chock full of errors, saw a lot more plate appearances per game. Reflecting the convention at the time to treat players as generalists who moved all over the diamond, Anson went back and forth between third and first with occasional time as a catcher for his first seven seasons before spending a year as an outfielder, then settling in at first. He doesn’t really count as a third baseman in anyone’s evaluation, but since he spent just slightly more than half his games at the position for those years, I’m listing him here.

Frank Baker was really the first superstar third baseman the game saw – note that his context-adjusted slugging percentage beats even those of Schmidt, Rosen, Mathews, A-Rod and Brett—and as such I view him as a worthy Hall of Famer despite the relative brevity of his peak (he was still a good player but not at all the same after holding out for a year at age 29 and then again at 34, good enough to contribute to the Yankees’ first two pennants). He was famously deadly in October (.363/.392/.560 over 25 World Series games, including an OPS above 1.000 in each of his first three World Series, against some of the toughest pitching staffs of the dead-ball era).

It’s too late to bother fixing much in the way of the occasional injustices in the Hall’s treatment of 19th-century players, but as arguably the best-hitting third baseman of the 19th century, Denny Lyons could make a case. However, Lyons’ achievements need to be discounted just a bit, as he had his best seasons in the American Association in 1890-91, when the league was bleeding talent to the NL and to the Players League rebellion, his career wasn’t long and he missed a fair number of games even in his prime.

Heine Groh’s prime could be extended two more years if you included his time as a second baseman before moving to third. Groh, the third baseman for the 1919 Reds, hero of the World Series for the 1922 Giants, and a contributor to three other pennant winners famed for his “bottle” bat, was a genuinely big star at his peak: He led the NL in OBP twice, hits once, runs once, doubles twice, and OPS once in a three-year period from 1917-19.

Ezra Sutton, perhaps the most obscure third baseman of the lot, had some serious ups and downs (from age 28-32, he batted .255/.278/.313, an OPS+ of 91, yet somehow kept his job and had his bat come roaring back to be one of the NL’s best hitters the next three seasons). Sutton was the first of many great Braves third basemen (see also Mathews, Chipper Jones, Elliott, Horner and Red Smith), spending 12 seasons with the Braves. He moved frequently to other positions, mostly shortstop.

Ned Williamson, third baseman and later shortstop for the Cubs dynasty of the 1880s, is best known today for a fluke: After he hit 49 doubles and two home runs in 98 games in 1883, the ground rules were changed at tiny Lakefront Park, and the Cubs went homer-happy, hitting 142 home runs in 113 games, a staggering number for the 1880s; Williamson’s 27 stood as the single-season record until Babe Ruth broke it. Despite outhomering their opponents nearly 2-to-1, the Cubs had a bad year and they changed the rules again; Williamson led the league in walks anyway, but while he remained a solid hitter he never again cracked double figures in homers. In his day, though, Williamson was better known for his superior glove.

Bill Bradley, the third baseman for the Nap Lajoie-era Indians (then the Naps), is a prime example of how excerpting the prime seasons and adjusting for context can change the way you view a player—Bradley’s career slugging percentage was .371, but when you translate his prime into the modern context, you get .502.

I thought Bill Joyce would rate much higher here; his career OPS+ is the fourth-highest among third basemen with at least 2,000 career plate appearances. But Joyce missed far too much time, including the entire 1893 season at age 27, and his career lasted just eight seasons.

Jimmy Collins, third baseman for the first World Series winners (the 1903 Red Sox) was the first third baseman to have what we would modernly recognize as a long, stable career. Typical of his era, he was best known for his glove, but had some power.

Deacon White, another of the early game’s superstars, was on his second career at third base after finishing his catching years, but like Anson was good enough to still be a star at 40 even with the nearly non-existent training and medical practices of the 19th century.

John McGraw may have been the game’s greatest manager, and was certainly the best player among the game’s great managers (with the arguable exception of Frank Chance, whose best years as a manager coincided with his years as a star player). McGraw, of course, batted .334 for his career, ranks third all-time in career on-base percentage (.466; he led the league three times and posted a .500 OBP from age 24-28) and was a highly successful base thief. Unfortunately, the scrappy Irishman was frequently injured, missing most of the 1896 season at age 23, being relegated to half-time play by age 28 and essentially full-time status as a dugout manager by 30. The lost playing time, plus his lack of power, causes him to rate low here.

Red Smith’s main place in baseball history is the .314/.401/.449 line he put up after being acquired by the Miracle Braves on August 10, 1914, although he was unavailable for the World Series. Smith, too, ended his playing career at age 29 just before the lively ball arrived.

In 1912, Heinie Zimmerman led the NL in batting, home runs, hits, doubles, slugging and total bases, was third in RBIs and fifth in triples and OBP, and finished tied for sixth in the MVP balloting, behind teammate Joe Tinker, who batted .282/.331/.351 to Zimmerman’s .372/.418/.571. It would be that kind of career for the lead-gloved Zimmerman, who was eventually banned from baseball as part of the Black Sox-era corruption scandals (he was implicated in a plot with teammate Hal Chase to throw a game in 1919). Zimmerman also batted .120/.120/.200 and made some “boneheaded” defensive plays during the 1917 World Series, which the Giants lost to the White Sox 4 games to 2, and ended up being implicated in a racketeering and bootlegging operation a decade after he was banned from the game. Giants outfielder Benny Kauff, also later banned from the game when he was indicted for auto theft in 1921, batted .160/.160/.440 in that Series. Hal Chase’s high average and low walk totals reflected a talented guy who was believed to have co-conspirators around the league let him pad his stats with meaningless hits to avoid detection (a common practice in their circle—nobody paid attention to batter walk totals at the time). Was Zimmerman, who had a similar offensive profile, involved in the same kind of pervasive corruption as Chase? Was there something fishy about the 1917 World Series? We have rumors, but nothing more, so the cloud over Zimmerman’s career remains murky.

Tommy Leach had a very long career at multiple positions, but some of his best seasons came during his years as third baseman for the Honus WagnerFred Clarke Pirates. The third baseman for the rival Giants at the time was Art Devlin; both were interviewed in The Glory of Their Times. Larry Gardner, teammate of Tris Speaker on three pennant-winning teams in Boston and Cleveland, had a long and stable career at third, but wasn’t an especially productive hitter despite good batting averages.


By my reckoning, the Hall of Fame’s roster of third basemen, inducted as third basemen, should look like this:

1. Eddie Mathews
2. Mike Schmidt
3. Wade Boggs
4. George Brett
5. Chipper Jones
6. Ron Santo
7. Frank Baker
8. Brooks Robinson
9. Pie Traynor
10. Jimmy Collins

With the arguable, borderline cases being Lyons, Sutton and Bando.

Print This Post
Sort by:   newest | oldest | most voted
Northern Rebel
Northern Rebel
Thank you Mr McLaughlin, for an excellent article! I only have two points to make. First of all, the idea that George Kell is in the hall of fame, is bad enough, he was a good player, that’s all. The idea that he’s in, and Ron Santo isn’t, is borderline criminal! Also, thank you for mentioning Harlond Clift, a player sabermetrics fans overlook. He was the FIRST of the prototypical 3rd basemen, combining power, walks, and excellent defense. It’s a damn shame he didn’t play for a competitive team in his prime. If he played for the Yankees, he would… Read more »
andy rothstein
andy rothstein

I think a guy named ARod played some third also, if I remember…

Steve Treder
Steve Treder

Terrific stuff, Dan!  Takes me back to my old Crossroads series.

mark f
mark f

Darin Erstad won gold gloves at 1st and the outfield which might make him the only player to win a gold glove at more than one position as I can’t seem to find Sandberg list for a 3B gold glove at all.

thanks for the continuation of these articles

David in Toledo
David in Toledo

Well done!  Thanks for mentioning Rosen.  I hope this is taken as further evidence for Ron Santo.


Sandberg never won a gold glove at third – they were all after he moved to second.  He’s the first player to win a gold glove the first year after moving to a new position.

Northern Rebel
Northern Rebel

The gold glove is pure popularity. Gregg Nettles was a better third baseman, but Brooks Robinson was still living off the reputation of his younger years.

Gary Gaetti won 3 or 4 years too late, and Wade boggs was the better fielder when George Brett was winning them.

Northern Rebel
Northern Rebel


I believe third base has the least amount of hall of famers.


The interesting thing about Boggs is that he came up so late, after clearly demonstrating his hitting prowess for 3-4 years in the minors. If the Red Sox hadn’t been so stubborn because he wasn’t a power hitter, he might be on top of the list for value.

Dan, Great work. But I can’t agree with your assertion that Mathews was as good a defender as Schmidt. Mathews developed into a good defensive player after struggling in his first couple of years. He had a very strong arm and above average range. But he wasn’t nearly as valuable as Schmidt in the field. The historical fielding stats we have for both players have limited value. While total plays deserves some weight as a measurement of a fielder’s range, it is not dispositive. For example, Mathews had large disparities in Total Fielding Runs Above Average from year-to-year (and this… Read more »

Why’s Jimmy Collins in the HOF? 465 errors, wow!

Northern Rebel
Northern Rebel
jacob: Jimmy Collins played in an era of small gloves, and horrid field conditions. He was actually considered the greatest 3rd baseman of his time. He was the first 3rd baseman to charge a bunt, and field it with his bare hand, and throw guys out, and was an outstanding hitter, in a period of small ball offense. He was the unquestioned leader of a Red Sox team that won the first World Series, and would have won a second, if John McGraw wasn’t an ass and refused to play, because he hated the President of the American League at… Read more »