@Philabaseball
A few days back I released some information regarding a project I had been working on that ranks every baseball club in Japanese history dating back to 1950. The project actually was a follow-up to a similar research study I completed a few months back regarding a ranking system of the best clubs in Major League history dating to 1900.
Since I've completed both projects and have had some time to comprehend the data, I wanted to note a few observations.
Before going any further, here's a quick recap in case you missed some of my recent work related to these projects.
My goal was to find an objective statistic (the Greatness Number) that would allow for fair comparison of baseball teams across eras, thus providing an objective ranking of the best teams in baseball history. Winning percentage is based on dominance (and some luck) in a single season, so it can't carry too much significance when comparing teams from various years and eras. Wins is another tricky statistic because the number of games played has changed several times over the past century.
We already know the best team does not always win the World Series, so looking at those clubs alone is a meaningless exercise.
To figure out what contributes the most to a winning club I ran a regression analysis using MLB data from 1900 to the present. I used winning percentage as my y-variable and of approximately 15 various team related statistics found run differential, WHIP, and OPS to be the three most statistically important items tied to winning.
The result (just based on MLB data, not Japan) is a formula that offers a predictive winning percentage based on the three most meaningful statistics: (.503)+(RDiff*.000)+(WHIP*-.053)+(OPS*.095).
To level the field so run differential was fair (remember, everyone did not play 162 games) I had to alter the formula for MLB clubs to include run differential per game: (.503)+(RDiffPG*.096)+(WHIP*-.051)+(OPS*.091).
I went and ran similar models for every Japanese baseball club dating back to 1950. Being that the game has some minor alterations not found in America, I knew the coefficients would have to differ a bit.
In full disclosure, data does not exist (in English, at least) to allow me to run a regression of various team-based statistics versus team winning percentage, so I am making an educated assumption that run differential, WHIP, and OPS are also the three most statistically important metrics in relation to winning in Japan.
The Japanese formula: (.492)+(RDiffPG*.088)+(WHIP*-.077)+(OPS*.151).
So, what does the data show?
Well, if we look at the MLB and Japan formulas as being equal (which we should), than the top two greatest baseball seasons belong to Japanese clubs: the 1951 Yomiuri Giants and the 1950 Shochiku Robins.
Here is how the top 10 clubs rank in MLB and Japan (remember, each is using their own formula as noted above):
Overall, here is a combined top 10 list:
As for the worst clubs in history, Philadelphia has several contenders - seven to be exact.
However, the 1950 Hiroshima Carp rank as the worst baseball team in history:
Combined, here is a look at the worst teams in baseball history:
In my opinion, the results of the models are accurate because the variables are all encompassing. It isn't as if a club could luck out for a season and receive a high Greatness Number because the data set is massive - for MLB it included 2,340 teams and for Japan it included 771.
Rob Neyer and Eddie Epstein put together a valiant effort in their ranking system in 2000 using standard deviations, but they consider the 1970 Orioles to be on par with the 1927 Yankees and I simply can't wrap my head around that logic.
For those wondering, the '70 Orioles are 50th in Greatness Number ranking, ahead of the 1905 Cubs and behind the 1941 Dodgers.
I'll have more on my findings over the next few weeks, including a detailed look at the top teams in the histories of MLB and Japan. I'll also provide insight into how Philadelphia ranks in MLB history.
Also, later this week, I'll explore the '50 Yomiuri Giants and explore why the Greatness Number deems the club to be the best in baseball history.
I'll have more on my findings over the next few weeks, including a detailed look at the top teams in the histories of MLB and Japan. I'll also provide insight into how Philadelphia ranks in MLB history.
Also, later this week, I'll explore the '50 Yomiuri Giants and explore why the Greatness Number deems the club to be the best in baseball history.
- Patrick Gordon is Managing Editor of the Philadelphia Baseball Review. Follow him on Twitter @Philabaseball or contact him at pgordon@philadelphiabaseballreview.com. Comments are welcome, so feel free to leave your thoughts below.