Compiling the Greatest MLB Team of All Time
As baseball fanatics, one of the questions that we love to debate is, “What was the greatest team ever?” Was it Murderer’s Row from the 1927 Yankees? How about the Big Red Machine in the mid-70s? Maybe the 2001 Mariners, despite not winning the World Series, were the best?
We could debate this all day long, and while it's certainly interesting, I aim to tackle a slightly different question with this article. What I'm interested in is not the greatest actual team to ever exist, but the greatest hypothetical team to ever exist.
What would happen if all of the greatest individual seasons by players at each position came in the same season by members of the same team? Who would be on this team and how many games would they be expected to win?
This means that we can put Babe Ruth and Barry Bonds in the same outfield, and Pedro Martinez and Cy Young in the same rotation. We have statistics that place players from different eras on an equal playing field, which makes such comparisons reasonable and practical.
How did I choose the team? Well, for the hitters, I used the statistic Adjusted Batting Runs, which, using linear weights, takes the contributions of each offensive player, adjusts for things like era, park effects, and so forth, and arrives at a final number that represents the number of batting runs a certain player provided during a certain time frame in a context-neutral environment. In layman’s terms, this metric gives us the total offensive contribution of a player in runs added that can be used to compare hitters from different eras.
For each starting position player, the only requirement was that they spent at least 50% of their games at their respective positions. The DH was the next best available hitter.
For the reserves, I shaped the roster like a typical 25-man roster in the modern era. I was sure to include a backup catcher, a utility infielder, and a fourth outfielder. I was more lenient on playing time limits with these players, requiring that they only spent 20% of their games at each respective position. As with any current MLB roster, all positions are covered with a capable reserve.
Choosing the pitchers was similar, as I used the statistic Adjusted Pitching Runs to determine the staff. For the five starters, I simply took the five highest totals among pitchers with at least 60% games started, while the membership in the pen required that at least 80% of appearances came in relief. I also formed the pen based on handedness, requiring at least two lefties to be on the squad, and capped the innings for short relievers (everyone but the long man) at 100. This prevents a pen full of guys from the pre-WWII era whose quantity, not quality of innings escalated their totals.
Without further ado, here is the 25-man roster, beginning with the position players.
|Position||Name||Year||Age||Adj Batting Runs|
|Reserves||Name||Year||Age||Adj Battting Runs|
Upon first glance at the position players, Arky Vaughn’s name immediately jumps out. Did he really have the highest single season Adj. Batting Runs by a shortstop in the history of the game? And if so, why is he not more of a household name?
While he was more than just a one year wonder, Vaughn did have a season for the ages in 1935. The Pittsburgh shortstop hit a robust .385/.491/.607 to go along with 97 walks and only 18 strikeouts. This high slash line and absurd K/BB ratio pushes Vaughn above other candidates such as Honus Wagner and Alex Rodriguez in his shortstop days. He was legit, and it's mere oversight that this Hall of Famer is not more of a household name.
Mike Piazza may be another surprise, but a .362/.431/.638 slash line with 40 home runs and 124 runs batted in was far above guys like Joe Mauer in 2009, Buster Posey in 2012, and Johnny Bench in 1972. Piazza’s 40 home runs, .362 AVG and .638 SLG are all within in the top five totals for catchers in a single season, making him more than deserving of his starting spot.
The corner guys, first baseman Lou Gehrig, third baseman Miguel Cabrera, left fielder Barry Bonds, and right fielder Babe Ruth, may come from different eras, but have one uniting principle: the ability to mash. The home run totals from these four sluggers are as follows: 47 for Gehrig, 44 for Cabrera, 73 from Bonds, and 54 from Ruth. The lowest RBI total from any member of this quartet is 137 and Gehrig had the most at 175.
Can these guys do more than just mash? Yeah, they hit for a decent average, too, as everyone finished with a mark between .328 and .373. Bonds had the lowest average of the quartet, but coupled it with a .515 OBP, which was higher than any hitter outside of Ruth.
While this is obviously impressive, I haven't even mentioned my two favorite statistics from this group.
First, the slugging percentages of the corner outfielders are simply inhumane. Ruth posted a .847 mark, but Bonds takes the crown with a .863 total in 2001. To put that number in perspective, a .863 figure would typically be close to the league leader in OPS, yet Bonds’ .863 mark does not even include OBP!
Second, the strikeout-to-walk ratios of this group are extremely impressive. Power hitters typically have a reputation for striking out frequently, but this quartet firmly contradicts against this label. Below are the strikeout to walk ratios of these four sluggers.
Arky Vaughn is the poster boy to elite K/BB ratios, but these sluggers are quite good in their own respect. Cabrera is the only slugger with more Ks than BBs, and it was very close. Bonds and Ruth again lead the charge with almost twice as many strikeouts as walks, which shows that these hitters were masters at every aspect of their craft.
The remaining two hitters, Rogers Hornsby and Mickey Mantle, are certainly no slouches in their own right. Both players had OBPs over .500, as Mantle walked 146 times and Hornsby hit a healthy .424. These guys also offer a good amount of home run power. Mantle hit 34 long balls and Hornsby hit 25, which are great totals for up-the-middle players.
One thing these players lack is tremendous base-stealing ability, but when the center fielder and second baseman are getting on base over half of the time and hitting a combined 59 home runs, speed can be somewhat overlooked. Mantle did swipe 16 of 19 bags, but if I was coaching this team, I would not be letting Hornsby run as he posted a mere 5-17 mark on the basepaths.
The bench is a diverse group, containing high average hitters like Joe Mauer, Nap Lajoie, and Stan Musial, and sluggers like Alex Rodriguez and Mark McGwire. The high average guys would be primarily used for spot starts, but Rodriguez and McGwire would be used often as pinch hitters. Since much of the thump in the lineup comes from the left side in Bonds, Ruth, and Gehrig, the two right-handed power bats from the bench will be useful in neutralizing the effect of lefty specialists.
The next question to answer is the batting order. Though I’m sure this will be controversial, here is the lineup that I would use.
The (S) means switch-hitter while the (L) signifies a lefty. All other hitters are righties.
|1||Mickey Mantle (S)||0.365||0.512||0.665||28||6||34||146||75|
|2||Barry Bonds (L)||0.328||0.515||0.863||32||2||73||177||93|
|3||Babe Ruth (L)||0.376||0.532||0.847||36||9||54||150||80|
|5||Lou Gehrig (L)||0.373||0.474||0.765||52||18||47||109||84|
|7||Arky Vaughan (L)||0.385||0.491||0.607||34||10||19||97||18|
There are four hitters with an OBP of at least .500 in the lineup, and it's not a coincidence that they are the first four hitters in the lineup. As the best hitter in the lineup, Bonds would hit second while Ruth, the second-best hitter in the lineup, would bat third. For those unfamiliar with the concept of hitting one’s best hitter second, the increase in plate appearances for the best hitter outweighs the potential for more runners on base during their plate appearances.
Mantle goes before Hornsby mainly because his high OBP is largely walk-driven, while Hornsby’s is driven by his .424 AVG. Additionally, Hornsby has a higher SLG, so it makes sense to put Mantle in the leadoff spot and the more powerful Hornsby fourth. Mantle did hit nine more home runs, but Hornsby had eight more triples and fifteen more doubles, which more than make up for the deficit in home runs.
I obviously don't believe in the notion that the guy with the most home runs ought to necessarily bat fourth, as Hornsby's 25 long balls pale in comparison to the respective 47 and 58 dingers by Gehrig and Foxx. Instead, I think the best hitters, especially those with high-OBPs, should bat near the top of the lineup, even if that means batting a less powerful hitter fourth. He may lack the power profile of Gehrig and Foxx, but Hornsby's high average will allow him to drive in plenty of runs from the cleanup spot.
Power threats Gehrig and Foxx follow Hornsby, with Gehrig hitting in front of Foxx because he is a slightly better hitter. Foxx’s 58 home runs are great, but as with Hornsby and Mantle, Gehrig gains the edge by hitting significantly more doubles and triples than Foxx.
Arky Vaughn in the seven hole may surprise people, but I could not justify putting his .491 OBP any lower in the order. Vaughn is not all-OBP either, as his .607 SLG was only about 30 points lower than Cabrera and Piazza’s marks.
Cabrera and Piazza fill out the lineup as power threats with relatively lower OBPs. Tony La Russa would be disappointed with the lack of a second leadoff hitter, but when your eight and nine hitters combine for 84 home runs, I would like to think that he could put up with this lineup.
|Role||Name||Year||Age||Adj Pitching Runs||IP||ERA||K|
The rotation features two guys from the steroid era and three guys from the pre-WWII era, with no one in between.
Martinez’s 2000 season is my favorite pitching season of all time. In the height of the steroid era, Martinez posted a 1.74 ERA, which was the best in Major League Baseball that year. The second place finisher, Roger Clemens, posted a 3.70 mark, which was a full 1.96 runs worse than Martinez!
Pedro also led the league with 284 Ks, a 0.737 WHIP, 5.3 H/9 and 0.7 HR/9 on the way to his third Cy Young Award in four years. While these numbers are impressive in any season, those statistics are even more impressive considering that run totals were extremely high during this time. It's fun to dream about what type of numbers Martinez would have posted in 1968, where offense was low.
The number two starter is the all time leader in Wins with 511. Cy Young threw an astounding 371.1 innings in 1901 with a 1.62 ERA on his way to a 33-10 record. Young also led the league in strikeouts with 158, fewest BB/9 with 0.9, highest K/BB ratio at 4.27, and fewest H/9 at 7.9.
The stat lines of Young compared to Martinez demonstrate the differences among eras, as Young did not record many strikeouts while Martinez did not log as many innings. Nevertheless, both pitchers were the best of the best compared to their peers during the time in which they played.
Walter “Big Train” Johnson from 1912 is our number three guy. A member of the inaugural class of inductees to the Hall of Fame in Cooperstown, Johnson’s best season is not far behind Young’s 1901 campaign. In 1912, Johnson threw 369 innings with a 1.39 ERA and 303 strikeouts, the latter of which was the best mark in the league by a long shot. The Big Train also led the league in WHIP, fewest H/9, K/9, and K/BB ratio.
Our fourth starter is one of the faces of the steroid era, but amidst all of the controversy surrounding his name, let’s not forget that Roger Clemens was really good at baseball. The Rocket’s best season, according to our metrics, was 1997, where he led the league with both 264 innings pitched and a 2.05 ERA. Clemens also led the league in strikeouts with 292, WHIP at 1.03, wins with 21, complete games with nine, and shutouts with three.
Again, these numbers are even more impressive considering Clemens pitched in the height of the steroid era where offense was at a high. Finally, this seems like a good time to say that no, I do not care whether members of this team used steroids. Bonds and Clemens put up numbers deserving of membership on this team so they will be on this team.
Rounding out the rotation is Lefty Grove, who is in my opinion the most underrated starting pitcher from the pre-WWII era. Grove’s 1931 season was one for the ages, where the lefthander compiled a 2.06 ERA in 288.2 innings with 175 strikeouts while taking home an MVP Award. In addition to ERA and Strikeouts, Grove also led the league in complete games, shutouts, wins, winning percentage, WHIP, and K/BB ratio. He was no slouch and is absolutely deserving of his role.
|Role||Name||Year||Age||Adj Pitching Runs||IP||ERA||K|
Selecting the bullpen was necessarily arbitrary in order to prevent a pen full of old guys who threw significantly more innings than a typical reliever today. Since the roster is constructed to replicate the current norm, with a five-man rotation, at least a six man bullpen, and so forth, it seemed appropriate to choose short relievers who excelled in roughly 60 to 80 innings over those who were pretty good in a different era over substantially more innings.
The only exception to that principle is long reliever Mark Eichhorn, who is also the player that you most likely have not heard of. For the long reliever, I allowed any relief pitcher to fill the role, ideally one with a substantial amount of innings. Eichhorn fits this description perfectly, as he threw 157 innings with an ERA of only 1.72 on his way to a sixth place finish in Cy Young Award voting and a third place finish in Rookie of the Year voting in 1986.
Eichhorn neither led the league in any statistical category nor started a game, but posted solid numbers across the board. For a guy who averaged 2.28 innings per appearance, this is an elite long reliever who will solidify a spot typically held by mop up guys who are neither good enough to crack the rotation nor good enough to pitch high-leverage innings in relief.
For the short relievers, I didn't factor saves into roster decisions in any way, but it's not a coincidence that all of these relievers earned saves. I find the save statistic to have little relevance so I did not appoint a closer, although the relievers are ranked in order. This does not mean that Jonathan Papelbon will always throw the ninth for my team, but rather that I would use Papelbon in the highest leverage situations.
The three right-handers are very close, as Papelbon, Eric Gagne and Fernando Rodney posted respective ERAs of 0.92, 1.20, and 0.60. In addition, all three were All Stars while Papelbon finished second in AL Rookie of the Year voting, Gagne won the Cy Young Award, and Rodney finished fifth in AL Cy Young Award voting during their respective seasons. Furthermore, all three had BB/9 at or under 2.2 with a K/9 of at least nine and as high as 15.0.
The lefties in the pen, B.J. Ryan and Damaso Marte may be somewhat of a surprise, but they each edge out Billy Wagner for a roster spot. Both lefties had a good season, though neither was as good as any of the three righties listed above. Nevertheless, almost all teams carry at least two left handed relievers, so the roster spots of Marte and Ryan, he two best left handed relievers according to our metric, are justified.
How Many Games Would They Win?
To answer this question, I began with an average MLB team in terms of run differential, adjusted the totals for multiple factors, and then calculated the winning percentage using the Pythagorean Wins Formula.
For the initial totals, I took the median number of runs scored and runs allowed from MLB teams during the 2013 season, which turned out to be 653 runs scored and 677 runs allowed. This makes the initial run differential slightly negative, but this will not affect the data very much.
From this, I split the pitchers and the hitters and calculated the runs scored and runs allowed of the team separately.
For the hitters, I summed the Adj. Batting Runs totals from the members of the starting lineup and one-fifth of the sum of the total Adj. Batting Runs from members of the bench. One-fifth is admittedly arbitrary, but this allows bench guys to contribute without being unrealistically influential.
Since this led to an excessive number of plate appearances, I took the total number of Adj. Batting Runs and averaged them over the median number of plate appearances for a 2013 MLB team. Adj. Batting Runs is a statistic that measures a contribution above or below average (zero Adj. Batting Runs is average), so we can add this total to the initial runs scored total from an average MLB team and arrive at the total number of expected runs scored by the greatest team. This total is 1,518.14, but we will cut off the decimal and call it 1,514.
For the pitchers, a similar formula was used to find the runs allowed total for the team. To begin, I summed the Adj. Pitching Runs of the entire staff and then adjusted the total to spread out over an appropriate amount of innings, which again was the MLB median from 2013. There was no need to divide the innings from the bullpen since they already throw substantially fewer innings than the starters.
Upon having the total number of Adj. Pitching Runs spread out over an appropriate number of innings, I subtracted this total from the median number of runs surrendered by an MLB team in 2013. This gives us the total number of expected runs allowed by the pitching staff, which is 280.1, but we can call it 280.
This brings the total expected run differential to 1,518 runs scored and 280 runs allowed. Using the Pythagorean Win Formula, we can expect a team with this run differential to win 95.519% of their games, which when projected over a 162-game season, results in a record of 154.74 and 7.26.
The 2001 Mariners won 116 games, as did the 1906 Cubs, who hold the highest winning percentage of any team since 1900 at .763. The projected winning percentage of the greatest team, .955, blows that figure away. The 1906 Cubs lost 36 games, but the greatest team would only be projected to lose between seven and eight.
Another fun stat is the difference in wins between the 2001 Mariners, who currently hold the record for most wins in a season and the greatest team, which is 38. Those 38 wins, if added to the 2013 Marlins, would take them from the second-worst team in the league to the best team in the league by three games. In 2012, 38 additional wins would have turned the Houston Astros into a playoff team. Finally, 38 wins is the difference between the 2012 Pittsburgh Pirates (who were below .500) and sole possession of the title of most wins in a single season.
The fact of the matter is that this compiled team would be the best team in MLB history, and it wouldn't be remotely close. A 154-8 record is almost impossible to fathom, especially in a game such as baseball where weird things tend to happen and the saying “anyone can beat anyone on any given day” holds relevance. If you have never heard this before, there's a high chance that the Astros' broadcasters will repeat this phrase many times whenever the Tigers are in town this season.
Unless new players come along and perform better than the current members of this roster, no team better than this could ever exist. This is the upper limit of team success in Major League baseball using real, historical players. We can still dream about a team that could go 162–0, but it will require future players that are better than the best players to have ever existed thus far and extreme luck in outperforming one’s Pythagorean record.
So yeah, maybe there's a chance. Maybe a team gets lucky and happens to have 25 players on their active roster that collectively have better seasons than the 25 players on the roster of the greatest team. It is logically possible, right?
Yes, it's logically possible, but the chance of that happening is so small that it is almost inconceivable. A roster full of 13 Mike Trouts and 12 Clayton Kershaws would not get the job done, and they are widely considered to be the best pitcher and position player in the game today. This is fun to dream about, but realistically, getting to 117 wins is difficult enough, and that is very far away from the 154 that this team would be expected to win or the 162 that would comprise a perfect season. Dream on, but let’s be realistic too.