By Roger Weber

Click here to download a clean Word version of this study

A version of this study completed in December 2005 is available in the online Baseball Almanac. It is generally regarded as one of the most conclusive studies of its kind and is the most visited study on great baseball teams on the internet.

Throughout the years, there have been many great baseball teams, and it seems that each year there is another. Based on just memories, it is difficult to compare teams between years, and without any on field competitions possible between these teams, there is really no way of determining an all time champion. As a result, there have been conflicting comparisons of teams through different statistical means.

For all the attention some of the more famous comparisons get, I feel there are a few areas where they are inaccurate. So in this study, I looked to explain both sound reasoning and thorough math to determine the best team ever.

It is important for a baseball team to perform far better than its competition to be considered great. But giving so much emphasis to the number of games ahead of second place seems to not only be grading the team, but the competition they played as well. Perhaps a team who finishes 30 games ahead of second place in a division is playing weak competition. It doesn't seem right to reward them for that.

In my comparison, I use number of games ahead of second place, but don't give much emphasis to it. Some may argue that it isn't fair to older teams that I count a margin of games ahead of second place within a division equal to games ahead of second place in a league. The divisions were created for a reason, though. As the league expanded, each division became in itself a miniature league, comparable in size to the entire league from the early 1900s. It would be even more unfair to count a team's games over second place in a league of six teams equal to a team's games over second place in a league of 16 teams.

Some comparisons count out teams from before 1920. These teams tend to have a higher final rank, for many reasons. To me, though, it doesn't seem fair to discount a team due to their existence prior to an arbitrary date, but it also seems unfair to include teams that may skew the rankings. In my analysis, I include all teams from 1902-2005, but I will also include in this article a ranking of teams not including 1902-1919.

Selecting teams for the comparison

I used fairly basic criteria for selecting teams for the comparison. All World Series winning teams were included, except for a few whose circumstances I will explain. Also, teams like the 1954 and 1995 Cleveland Indians were undoubtedly great teams, yet neither won the World Series. For that reason, all teams with regular season winning percentages of at least .670 were included. While this number is quite arbitrary, it worked as a good cutoff point between including some great teams who didn't win the World Series, and allowing too many teams to be included in the scoring. There are a few exceptions to this rule, though.

The 1998 Atlanta Braves are included in the rankings although their winning percentage was just .654. They played dominantly, and if not for a very "lucky" San Diego Padres team, these Braves may have won the World Series.

The 1994 Montreal Expos are also included. Since there was no World Series in 1994, it seemed logical to include the team with the best record from the regular season. Since part of the scoring is based on postseason play, the 1994 Expos, 1904 Giants, and 1902 Pirates, all of whom never played in playoffs or a World Series, are included and are given arbitrary but fair 50% postseason winning percentages.

There are a few teams not included in the final rankings. Mainly, the period from 1942-1945 was not included because it was the time of World War II. Most teams lost many good players to the war effort, and young, inexperienced, and probably inferior quality players were called into major league action. The talent level of major league baseball went down, as did that of the minor leagues as a result, creating a few years of mostly inexperienced players. It isn't fair to compare Babe Ruth's teams to teams made up of mostly minor or lower league players.

As stated before, teams from before 1920 are sometimes excluded from all time rankings. While it isn't fair to exclude them completely, I include them along with a ranking for teams post-1920 excluding teams from pre-1920.

The formula

Most importantly, a team must win the majority of its games. Winning is what determines the champion, so winning should be the number one basis for how to decide a great team. Winning, though, is determined by a team's ability to produce and defend against runs. For that reason, I view a team's scoring dominance over the opponents as a similarly important aspect of greatness.

While it is given great importance to some historians, I don't view games ahead of second place as a key determining factor, but it does deserve some merit, since a great team should be able to separate itself from the competition easily. A team's dominance over the rest of the league or division defines its unique ability to win in the year it played. Great teams don't finish second. The Florida Marlins have twice won a World Series despite never winning their division. Had they finished with the record they had prior to 1994, the year in which the Wild Card was created, the Marlins would have missed the postseason.

A great team should finish atop both leagues, with the trophy in the display case. While not winning a championship doesn't necessarily take away from a team's winning ability, a team like the 1954 Indians, who finished the regular season 111-43, but lost in the World Series 4 games to 0 should not be placed on the same level as a team who finished with a 107-47 regular season record but won the World Series 4 games to 0, even though both finished with the same final record of 111-47.

A large determinant of a team's greatness is how it performs in the postseason. Great teams don't choke. They continue to play as they had all season long, or better in the postseason, dominating their opponents in the World Series, and any Playoffs.

The regular season versus the postseason

It is greatly debated as to how much the regular season should count compared to the postseason in a comparison. Here is how I determined this aspect:

I will give 85% of the overall score to a combination of regular season winning percentage, run scoring dominance over opponents during the regular season, and postseason winning percentage.

Since the length of the regular season and the playoffs has changed frequently throughout history, there is no uniformly fair way to determine how much weight to give the postseason. I will try to determine the relative importance of the postseason, in essence, how many regular season games the postseason should count for using the current setup of the season.

First of all, the regular season is 162 games. The playoffs are usually an average of about 15 games for any given team that plays in the World Series. But, for the first 162 games, all 30 MLB teams are involved. In the first round of the playoffs, just eight are involved. Therefore, I think games in the first round of the playoffs should be considered 3.75 (30/8) times as important as regular season games, since about one fourth the number of teams plays in them. Games in the second round of the playoffs see only 4 teams playing, so it seems those games should be about 8 (30/4) times as important as regular season games. Games in the World Series match just two teams, so they should be about 15 (30/2) times as important as a regular season game.

Inserting the numbers

It would seem to make sense just to multiply a team's totals by the desired percentages, but that would leave far too much emphasis on games above second place and postseason winning percentage. For that reason, I tried dividing the desired percentage by the average count for a certain category. This seemed to work, but there was a problem. With the categories like winning percentage, there is a very small span that these teams cover, yet with games above second place, there is a large span between best and worst, which left games over second place counting far more than regular season winning percentage, even with the applied desired percentages, and this was not my intention.

This led me to consider standard deviations, since they are an accurate span of difference from the average. What I finally found to be most accurate was to divide the desired percentage by the standard deviation for that category. By doing this, I made it so the possible difference between the best and the worst teams in that certain category is equal to the desired percentages of overall grade. This is somewhat confusing, but here is a chart to help define how the formula was ultimately created. (All winning percentages and % of runs scored are divided by 100. They are a fraction of 1.)

Category, Standard Dev., Multiplied by, to yield desired percentage

Winning Percentage, 0.04695, x, 605.414, =, 28.427

Avg. Runs - ERA, 0.5909, x, 25.0547, =, 14.214

% of runs scored, 0.03869, x, 367.402, =, 14.214

Games above second place, 6.94399, x, 1.44009, =, 10

Postseason Win %, 0.17662, x, 159.36, =, 28.1457

Won W.S.?, 0.47848, x, 10.4498, =, 5

To get the resulting score for a team, I multiplied its season totals by the numbers in the column labeled "Multiplied by". If this seems crazy to multiply winning percentage by 605, and games above second by just 1.4, remember that the goal was to make the difference created between the best and worst teams be equal to the desired percentage. Most of the 605 is guaranteed for every team. Only a small portion of that is different between teams. Every team on the list has a winning percentage of at least .530, but none has a winning percentage above .741. Yet, the span between games above second goes from -10 to 30. If you don't understand it, just trust that I spent a great deal of time figuring this out mathematically.

Scoring the teams

If you have figured out how the scoring works, you can figure out the team rankings.

Before I expose the team by team rankings, I should give some basic data about the overall findings. (All winning percentages and % of runs scored are divided by 100. They are a fraction of 1.)

For all teams:

Category: average, standard deviation

Winning Percentage: 0.63285, 0.04679 

Avg. Runs – ERA: 1.68763, 0.59228    

% of runs scored: 0.60386, 0.03875      

Games above second place: 8.23874, 6.91568    

Postseason Win %: 0.6891, 0.17716      

Average Scores:

Teams: average, standard deviation

Pre-1920: 804.8443, 53.89875

Post-1920: 744.5767, 60.97084

All: 755.979, 64.017

Because the average score of teams from before 1920 is so much higher than the average score of teams from after 1920, it is probably a good idea to eliminate all teams from before 1920 to insure a more accurate rating system. Certain lurking variables skew the results form before 1920. Mainly, these are that the competition was poorer, and since the league was smaller, one team could attain more great players. Also, the postseason was shorter, so it was easier for a team to win all their postseason games. Also, with a smaller league, a team had a better chance to finish many games ahead of second place.

In my rankings, I will rank all teams in order of score, but I will also give post-1920 teams a ranking among all teams from that era.

Here is another interesting feature comparing the decades.

Years: average score

1902-1909: 832.4726

1910-1919: 784.1231

1920-1929: 781.5494

1930-1939: 804.41

1940-41, 1946-49: 755.2973

1950-59: 740.8345

1960-69: 735.5162

1970-79: 740.9028

1980-89: 703.8307

1990-99: 726.9237

2000-05: 701.387

Obviously, this is not an accurate comparison of all baseball between the decades. What I believe accounts for the recent decline is the addition of playoffs and the Wild Card, which allows for weaker teams to win the World Series, and since these rankings include all World Series winners, this may have an impact.

From the 1930s to the 1960s, there is a steady decline. This may be because the leagues were getting larger, yet there was still just one division in each league, which left for smaller annual "games above second place" counts. This idea is supported by the average increase in the 1970s, as divisions became a factor.

It is interesting to note that the 2000s have been the weakest years. This can be explained by the fact that three Wild Card teams have won the World Series. These were weaker teams that would not have been in the World Series in prior years.

The teams

110 teams are included in the Rankings.

Note that disparities in the numbers of games played were taken into account in the formula. Also note that teams from seasons in which there were no playoffs are given an arbitrary, but fair 50% winning percentage in the playoffs, since it can be safely assumed all teams on the list would have made the playoffs.

Click here to download report with full team listing