suedehead6
Registered User
- Jan 7, 2019
- 5
- 5
I've manually done some calculations that I like for the Rangers 1926-27 team that I'll explain and show below. It would be way too time-consuming for me to actually figure all this for many teams/years, but thought I'd throw it out there just in case somebody out there has the interest and know-how to automate most or all of these things.
I started with game-by-game results for all games not involving the Rangers that season and found each of the remaining teams' goals for/game and goals allowed/game in those non-Rangers games. Which I then applied to the Rangers game-by-game as "expected goals for" and "expected goals allowed" specific to each opponent. Then I took the Rangers actual goals for and allowed versus those expectations for a rating of "goals for above expectation" (GF Ax) and "goals allowed below expectation" (GA Bx). (Positive numbers being good in both cases.)
Those numbers were then divided up into credits to the individual goalies, defensemen, and forwards that actually played the specific games in this way:
forwards:
(GF Ax*.76)/3
(GA Bx*.25)/3
defensemen:
(GF Ax*.24)/2
(GA Bx*.35)/2
goalies:
GA Bx*.4
(Those proportions are part-arbitrary, part-based on Tom Awad's calculations behind Goals Versus Threshold.)
Getting the game-by-game player lineups would be, for me at least, a huge undertaking. Would be great if someone has that in a database somewhere, but I sort of pieced together this one Rangers season using the NHL.com player game logs. Because the starters played almost the entire games at that time, I didn't worry about parsing credit between the full roster and am giving all of the forward credits to Bun Cook, Frank Boucher, and Bill Cook because they all played the vast majority of the minutes in all 44 games. Likewise for defenseman Clarence Abel. The other top-pair d-man, Ching Johnson, missed a good chunk of games due to injury, so I manually matched up which games he appeared in or not. (So far I haven't tried to fill in the blanks of who was playing in Johnson's spot when he was out). Anyway, here are the results:
[TBODY]
[/TBODY]
I started with game-by-game results for all games not involving the Rangers that season and found each of the remaining teams' goals for/game and goals allowed/game in those non-Rangers games. Which I then applied to the Rangers game-by-game as "expected goals for" and "expected goals allowed" specific to each opponent. Then I took the Rangers actual goals for and allowed versus those expectations for a rating of "goals for above expectation" (GF Ax) and "goals allowed below expectation" (GA Bx). (Positive numbers being good in both cases.)
Those numbers were then divided up into credits to the individual goalies, defensemen, and forwards that actually played the specific games in this way:
forwards:
(GF Ax*.76)/3
(GA Bx*.25)/3
defensemen:
(GF Ax*.24)/2
(GA Bx*.35)/2
goalies:
GA Bx*.4
(Those proportions are part-arbitrary, part-based on Tom Awad's calculations behind Goals Versus Threshold.)
Getting the game-by-game player lineups would be, for me at least, a huge undertaking. Would be great if someone has that in a database somewhere, but I sort of pieced together this one Rangers season using the NHL.com player game logs. Because the starters played almost the entire games at that time, I didn't worry about parsing credit between the full roster and am giving all of the forward credits to Bun Cook, Frank Boucher, and Bill Cook because they all played the vast majority of the minutes in all 44 games. Likewise for defenseman Clarence Abel. The other top-pair d-man, Ching Johnson, missed a good chunk of games due to injury, so I manually matched up which games he appeared in or not. (So far I haven't tried to fill in the blanks of who was playing in Johnson's spot when he was out). Anyway, here are the results:
| name | p | tot | GA Bx | GF Ax |
| Chabot | G | 6.2 | 6.2 | |
| Abel | D | 3.5 | 3.2 | .4 |
| Bu. Cook | F | 2.3 | 1.5 | .8 |
| Boucher | F | 2.3 | 1.5 | .8 |
| Bi. Cook | F | 2.3 | 1.5 | .8 |
| Johnson | D | 1.2 | 1.5 | -.4 |
| other | D | 2.3 | 1.6 | .7 |
| Winkler | G | 1.0 | 1.0 | |
| total | 21.0 | 18.0 | 3.0 |
