The playoffs are a tiny sample size roller coaster, perhaps better enjoyed in one’s viscera than in the analytic mind. And I hope to get around to writing a more personal post about what the 2014 World Series was like for me, but I can’t entirely turn off the rational part of my Royals fandom that often informs my writing and understanding of team history. As we finally added some new post-season games to team lore this past season, I’ve thought about how to best objectively quantify playoff performance, particularly how it should factor into my top 100 Royals rankings. In the first version of my rankings, I looked at each player’s playoff performance and gave it a sort of gut reaction rating between one and 10. (Then I added that number to half their wins above replacement, plus wins above average. After sticking with that for a bit, I decided the playoff factor was making an over-sized impact on the ratings and multiplied the 1-10 ratings by 70%, thus giving playoff participants somewhere between a .7—7 point boost.) But the inconsistencies of my gut ratings started to bug me, and I’ve come up with convoluted playoff rating formulas so that all players are being judged roughly the same way.

The things I wanted to measure were playing time, win probability added (WPA), runs created (RC) for position players, and ERA for pitchers. I also decided to include a small subjective adjustment (ADJ) for positional players that tries to take into account other important factors including positional difficulty, defense, and baserunning. The raw adjustment number going into the formula can be -1, -.5, 0, +.5, or +1. And for pitchers, the ERA rating is 100-((ERA/3.83)*100). (The average ERA has been close enough to 3.83 every year the Royals have been in the postseason that I’m not adjusting for era.)

After fooling around with formulas to get ratings that were generally on a 1-10 scale, I’ve come up with these disgusting beauties:

For position players: (plate appearances-60)/50)+(WPA*1.5)+ADJ+(RC/10)
For pitchers: ((IP-11/11)+WPA+(ERA rating/10))*.6

(Some players end up with a negative number, but those are all converted to a +1 in my player rankings. And George Brett’s number is over 10, but I cap the rating at 10, so that’s what he gets for my rankings. Pitchers must pitch at least 10 innings to earn a rating over 1. Blergh.)

Here are the top 10 playoff ratings:

1. George Brett 13.2
2. Willie Wilson 6.0
3. Danny Jackson 5.8
3. Wade Davis 5.8
5. Hal McRae 5.7
6. Eric Hosmer 5.6
7. Amos Otis 5.4
8. Greg Holland 5.3
9. Frank White 4.5
10. Alex Gordon 4.0

Top 10 win probability added:

1. Willie Aikens 1.5
2. Eric Hosmer 1.4
3. George Brett 1.4
4. Wade Davis 1.1
5. Danny Jackson .9
6. Greg Holland .9
7. Alex Gordon .9
8. Kelvin Herrera .9
9. Paul Splittorff .7
10. Amos Otis .7

Top 10 runs created*:

1. George Brett 81
2. Hal McRae 46
3. Willie Wilson 41
4. Frank White 38
5. Amos Otis 34
6. Eric Hosmer 29
7. Willie Aikens 26
8. Lorenzo Cain 25
9. Alcides Escobar 21
10. Alex Gordon 21

And here is a link to a spreadsheet with all ratings and the numbers that went into them. (You should be able to sort any column you like.)

*runs created formula I used: (BB*.707)+(HBP*.74)+(1B*.909)+(2B*1.301)+(3B*1.653)+(HR*2.145)+(SB*.2)+(CS*-.391)