By Tweed
Evaluating baseball players is something we all do, but also something that is probably a lot harder than most of us think. Comparing players to each other is even harder.
So, how do you do it?
We think the first step is to start with a question: What do you want to know?
Bill James used to rank players by their "peak" value and "career" value. That is, how good was a player while he was playing at the top of his game, and how good was he over the course of his career. But that's not the only way to skin this cat.
What if you had to choose a player for one game? How about a short series? Or a long series? How about the month of August? Or just the post-season? And what about a whole season? Or just a few years? Or who would populate your team if you had a 10-year window? How about 20 years?
Here's an example. Let's compare two pitchers.
Player A and Player B play in the same era, but different leagues. They have nearly identical lifetime ERAs, but player A has 3 extraordinary ERA seasons, 3 Cy Young Awards and an MVP. Player B has one Cy Young and a consistently good ERA. Player A was a strikeout leader, Player B struck out a lot, but never led the league. Player A led the league in ERA 5 times, player B, twice. Player A had three 20+ win seasons, Player B had two.
I randomly chose two years, four years and five years for each player (the only criteria being that the pitcher had to have pitched at least 100 innings). Player A has the edge:
Two-year period:
Player A: .708 winning percentage and a 2.67 ERA.
Player B: .817 winning percentage and a 2.98 ERA.
Four-year period:
Player A: .677 winning percentage and 2.44 ERA.
Player B: .721 winning percentage and 2.81 ERA.
Evaluating baseball players is something we all do, but also something that is probably a lot harder than most of us think. Comparing players to each other is even harder.
So, how do you do it?
We think the first step is to start with a question: What do you want to know?
Bill James used to rank players by their "peak" value and "career" value. That is, how good was a player while he was playing at the top of his game, and how good was he over the course of his career. But that's not the only way to skin this cat.
What if you had to choose a player for one game? How about a short series? Or a long series? How about the month of August? Or just the post-season? And what about a whole season? Or just a few years? Or who would populate your team if you had a 10-year window? How about 20 years?
Here's an example. Let's compare two pitchers.
Player A and Player B play in the same era, but different leagues. They have nearly identical lifetime ERAs, but player A has 3 extraordinary ERA seasons, 3 Cy Young Awards and an MVP. Player B has one Cy Young and a consistently good ERA. Player A was a strikeout leader, Player B struck out a lot, but never led the league. Player A led the league in ERA 5 times, player B, twice. Player A had three 20+ win seasons, Player B had two.
I randomly chose two years, four years and five years for each player (the only criteria being that the pitcher had to have pitched at least 100 innings). Player A has the edge:
Two-year period:
Player A: .708 winning percentage and a 2.67 ERA.
Player B: .817 winning percentage and a 2.98 ERA.
Four-year period:
Player A: .677 winning percentage and 2.44 ERA.
Player B: .721 winning percentage and 2.81 ERA.
Five-year period:
Player A: .727 winning percentage and 2.41 ERA.
Player B: .697 winning percentage and 2.62 ERA.
Although, overall, Player B had the higher winning percentage, for the most part, Player A had a lower ERA - and winning percentage is more effected by one's team than is ERA.
But here are some facts that may make Player B the preferred choice, depending on what question you're asking:
Player B had 14 100+ inning seasons, and each one of them was a winning season. His first seven years and second seven years split like this (the order of the seven-year periods may or may not be sequential):
One Seven Year Stretch:
2.6 ERA, 105 wins, 39 losses (.729 win. perc.) and 5 of the 7 years were 200+ inning years.
Other Seven Year Stretch:
2.9 ERA, 127 wins, 57 losses (.690 win. perc.) and 6 of the 7 years were 200+ inning years.
Player A had 10 100+ inning seasons, and one was a losing season and another went even. Here are his splits:
One Five Year Stretch:
1.95 ERA, 111 wins, 34 losses (.766 win. perc.) and 4 of the 5 years were 200+ inning years.
Other Five Year Stretch:
3.92 ERA, 50 wins, 47 losses (.515 win. perc.) and only 1 of the 5 years was a 200+ inning year.
Indeed, 14 years of consistency mark Player B's career, with an ERA fluctuating between 2.01 and 3.24 during those years; while Player A varied widely in ERA between 1.73 and 4.48.
If I needed a guy at his peak performance for a game, I'd pick Sandy Koufax, Player A. During his last four years of pitching, he was arguably the best to play the game. Certainly, he dominated the league, and both leagues in Cy Young awards (his were won when one Cy Young was awarded in major league baseball). But if I wanted a guy to build a franchise around for the next ten to twenty years, I would choose Whitey Ford, Player B. The Chairman of the Board was the defensive anchor of those great Yankees teams of the 50s and early 60s.
Player A: .727 winning percentage and 2.41 ERA.
Player B: .697 winning percentage and 2.62 ERA.
Although, overall, Player B had the higher winning percentage, for the most part, Player A had a lower ERA - and winning percentage is more effected by one's team than is ERA.
But here are some facts that may make Player B the preferred choice, depending on what question you're asking:
Player B had 14 100+ inning seasons, and each one of them was a winning season. His first seven years and second seven years split like this (the order of the seven-year periods may or may not be sequential):
One Seven Year Stretch:
2.6 ERA, 105 wins, 39 losses (.729 win. perc.) and 5 of the 7 years were 200+ inning years.
Other Seven Year Stretch:
2.9 ERA, 127 wins, 57 losses (.690 win. perc.) and 6 of the 7 years were 200+ inning years.
Player A had 10 100+ inning seasons, and one was a losing season and another went even. Here are his splits:
One Five Year Stretch:
1.95 ERA, 111 wins, 34 losses (.766 win. perc.) and 4 of the 5 years were 200+ inning years.
Other Five Year Stretch:
3.92 ERA, 50 wins, 47 losses (.515 win. perc.) and only 1 of the 5 years was a 200+ inning year.
Indeed, 14 years of consistency mark Player B's career, with an ERA fluctuating between 2.01 and 3.24 during those years; while Player A varied widely in ERA between 1.73 and 4.48.
If I needed a guy at his peak performance for a game, I'd pick Sandy Koufax, Player A. During his last four years of pitching, he was arguably the best to play the game. Certainly, he dominated the league, and both leagues in Cy Young awards (his were won when one Cy Young was awarded in major league baseball). But if I wanted a guy to build a franchise around for the next ten to twenty years, I would choose Whitey Ford, Player B. The Chairman of the Board was the defensive anchor of those great Yankees teams of the 50s and early 60s.
The Chairman of the Board
Was Helped by Tremendous
Run Support from the Likes of
Mantle, Berra and Maris
And even if I had only 5 years to choose from, I would rather take 5 randomly chosen Ford years then 5 randomly chosen Koufax years. On the other hand, if I could choose the 5 years, I might take Koufax.
Control Eluded Koufax
Early in his Career;
But Not From 1961 - 1966
Comparing Koufax and Ford, or any single or group of players, should be dependent on the question asked. Who do you want starting any particular game? Koufax at his peak. But what if you didn't know what you'd get? What if it were Koufax of 1958? Would you risk that for the possibility of getting the domination Koufax could provide? And how about a ten year period - who would you pick? Whitey Ford. But maybe Koufax's dominance for a few years is worth a few poor years?
It comes down to what you're looking for.
No comments:
Post a Comment