The Greatest of This, the Greatest of That

            The method that we introduced yesterday, evaluating players by their performance over a period of years, rather than in one year.  It is essentially an effort to stabilize data which otherwise jumps up and down. We concluded that Freddie Freeman is now the best player in baseball. 

            Of course, in that conclusion we are looking only backward, whereas in all the other years the system is able to look both backward and forward.   In the future we may assign the most recent years to someone else, not Freeman.  We’ll see.  But there are a lot of other things that we can do with this method or a similar/related method.   By the same method by which we identified the best player in baseball at each point in time, we can also choose the best player in the National League in each year, and the best player in the American League.   We could choose the greatest catcher, see how that evolves over time, or the greatest second baseman or whatever.  We could identify the greatest player under 25, or the greatest player over 30.   We’ll do some of that later.

            First, let’s deal with the "problem" of the system, in real time, being able to look only backward.   In retrospect, we can say that the moment in time at which Ty Cobb surpassed Honus Wagner to be the greatest player in baseball was 1909, but in 1909 it was impossible to know that.  It is a separate question, but I think still an interesting question, of who was the greatest player in baseball at each moment in the past, based only on what could have been definitely known at that time. 

            For that, we use essentially the same method, but just a little bit different.  I weight the last ten seasons for a player at 1-3-5-7-9-11-13-15-17-19.  That also totals up to 100 (100%), but the most recent season counts at 19% of the value, rather than 10% if we just chopped the 19-year window in half.   By that system, the greatest player in baseball (based on what could have been known at the time). .. the #1 guys are:

            As you can see that is pretty much the same list we had before, with some small differences and the dates are different.   We can also look at who the best players are looking FORWARD from a specific year:

Making the point that the same data looks very different if looked at from a different angle.   In the years 1911-1920 Walter Johnson is the greatest player in the game looking forward, but in the same years Babe Ruth is the greatest player in the game looking backward.  The method looks at the same ten-year span, but looking forward from 1911, 36% of the value is in the 1911-1912 seasons, only 3% in 1919-1920, whereas looking back from 1920, 36% of the value is in the 1919-1920 seasons, only 3% in 1911-1912.

            Yesterday I had a list of the best players in baseball over time.  We can also do this by the league, the best player in each league over time.  This is the chart for the American League:

            I dislike citing a player as the best player in the league for one year, as this is not a one-year look at the question, but sometimes there’s just no other way to go.  The toughest spot on the chart above is the 1964-1965 period. 

            The charts are based on the 19-year formula given yesterday, and I respect that formula and go along with whatever it says at least 95% of the time.  The formula says that Mantle is still the best player in the American League in 1964, which is a reasonable thing to say, but Mantle’s era of greatness ended in 1965, and he dropped to third on the 19-year list.   Occupying the 1-2 spots are Harmon Killebrew and Al Kaline. 

            But Kaline and Killebrew were both injured in 1965, and actually weren’t a lot more valuable in that season than Mantle was.  If I put one of them on the 1965 line, he would be a one-year holder of the best player in the league title, since Frank Robinson came over in 1966, winning the MVP and the Triple Crown, and scoring as easily the league’s greatest player based on that and his ten years of outstanding play before that in the National League.  I would be listing Killebrew or Kaline as the best player in the league for only one year, in which neither one of them was remotely close to being the best player in the league.  I can’t do that. 

            The 19-year system ranks the top four players in the league as 1. Killebrew, 2. Kaline, 3. Mantle, 4. Brooks Robinson.   I thought about just ignoring that, and listing Tony Oliva as the best player in the league in 1965, which I think he probably was; Oliva fits the profile of the dominant all-around star better than Brooksie does.  But I decided to list Brooks Robinson as the American League’s best player in 1964-65, since (1) he was the MVP in 1964, and (2) he had a very strong season again in 1965.  By Win Shares, Mantle was actually more valuable than Robinson in ’64, but by a kind of meaningless margin (34 to 33); it is reasonable to say that Brooks took over the number one slot in 1964, and it’s the best I can do. 

            I terminated these charts are 2018.   The 2018 ranking should be based on all the years from 2009 to 2027.  We’re missing 15% of the relevant data for 2018 (5% for 2023, 4% for 2024, 3% for 2025, 2% for 2026, 1% for 2027.)  Missing 15% is OK, but if we go to 2019, we’re missing 21%.   Let’s just mark the years 2019 to 2022 as TBD. 

            This is the National League chart:

            Andrew McCutchen, I think, has never quite gotten his due.  He did win an MVP Award, but then, so did Jeff Burroughs.  Andrew was, I think. . .well, he wasn’t Bonds; he wasn’t Pujols, he wasn’t Trout.  But he was best player in the National League, year-in and year-out, for five years. 

            Tomorrow, I’ll give you the lists for the greatest catchers, second basemen, third basemen and starting pitchers, over time.  Thanks for reading.