Post by eric on Jun 13, 2018 19:56:49 GMT
Let's look at every NBA MVP based on how many NBA playoff rounds they won vs. how many playoff rounds they could have won.
We'll get into more nitty gritty in a bit, but this is clearly a pretty solid methodology from the get go.
-The five greatest players of all time are Jordan, Russell, LeBron, Magic, and Kareem, and they're the top five here.
-The bottom five are all infamous losers.
-By far the best is by far the most winningest player of all time.
-By far the worst is by far the most infamous MVP of all time.
-The distribution is extremely well defined: 16 above average, 16 below, 4 above one SD above average, 5 below (10 total expected).
Can't really ask for more.
.
Let's get into some details. First a quick history lesson.
In 1951 (Cousy's rookie year) the NBA had four teams from each conference make the playoffs. (They called them divisions then, but they functioned as conferences do now.) With eight playoff teams it was simple, you had four first round matchups, two second round, and one Finals.
The league brutally contracted in the years following the ABL-BAA merger, leading up to 1954 when each conference was reduced to three playoff teams, but how can you set up a bracket with six participants? Obviously you set up a round robin for each conference where each team played the others twice each except when they didn't sometimes for some reason, the top two advanced, and then you had a standard four bracket.
This was immediately done away with and the NBA instead gave the top team in each conference a first round bye. The NBA's last contraction occurred in 1955 and this model was kept until...
1967. Having very gradually expanded back to ten teams, the everyone gets a trophy generation demanded that the overwhelming majority of teams still make the playoffs, so each conference again sent four teams each, and the straightforward eight team bracket returned.
But expansion escalated in response to the ABA, and by 1974 byes returned when five teams from each conference made the playoffs and the top three received a first round bye.
But expansion escalated even further in absorption of the ABA, and in 1977 the field was again expanded to six teams each with the top two getting a bye.
Finally in 1984 the current playoff structure was established: eight teams from each conference, no byes, four rounds. Hilariously there were only 23 teams in the NBA at this time so you could stumble around to a 30-52 win record and still make the playoffs. The more competitive ensuing generations have ensured that even as the NBA expanded all the way to 30 teams the playoff structure has remained intact.
.
So byes were around for almost half of the MVPs in this analysis. How do we count them? Let's look at a hypothetical standings:
58-24 Celtics (1 seed)
42-40 Royals (2 seed)
38-44 Knicks (3 seed)
52-30 Warriors (1 seed)
52-30 Lakers (2 seed on tiebreaker)
35-47 Hawks (3 seed)
The Royals featuring Oscar Robertson beat the Knicks and lose to the Celtics. They therefore won 1 out of 3 possible rounds.
The Warriors featuring Wilt Chamberlain get a bye in the first round and lose to the Lakers. They therefore won 0 out of 2 possible rounds.
1/3 > 0/2
But this feels wrong. If the Warriors had lost one more game, they would have gotten the cream puff Hawks and almost surely have won - that's the whole point of a bye, after all. And we should avoid models that reward teams for losing more games in the regular season.
What if we a) count a team receiving a bye as having gone 1 of 1 and b) count a team that missed the playoffs entirely during bye years as having gone 0 of 1, in each case on top of whatever else they had? When we do that our list instead looks like:
The big winners are Bob Cousy and Bob Pettit, who played on great teams for long periods when byes were a thing. (Bill Russell did too but when you're already at ninety friggin' percent it's hard to get much higher.) The only people who drop at all are Willis Reed, Bill Walton, and Oscar Robertson.
Overall I gotta say I like the first list better straight up, and once you throw in all the work you have to do to explain how the bye reward/penalty system works it doesn't seem worth it to me.
series won pct name
30 27 .900 Bill Russell
49 32 .653 Magic Johnson
60 35 .583 LeBron James
68 36 .529 Kareem Abdul-Jabbar
60 30 .500 Michael Jordan
49 24 .490 Larry Bird
39 18 .462 Wilt Chamberlain
76 35 .461 Tim Duncan
35 16 .457 Bob Cousy
44 20 .455 Kevin Durant
36 16 .444 Stephen Curry
41 18 .439 Julius Erving
76 33 .434 Shaquille O'Neal
80 33 .413 Kobe Bryant
28 11 .393 Willis Reed
34 12 .353 Dave Cowens
.343 ***average***
47 15 .319 Bob McAdoo
35 11 .314 Bill Walton
26 8 .308 Bob Pettit
56 17 .304 David Robinson
43 11 .256 Wes Unseld
76 19 .250 Karl Malone
40 10 .250 Russell Westbrook
72 16 .222 Hakeem Olajuwon
41 8 .195 Oscar Robertson
64 12 .188 Charles Barkley
73 13 .178 Moses Malone
84 14 .167 Kevin Garnett
80 13 .163 Dirk Nowitzki
72 11 .153 Steve Nash
56 8 .143 Allen Iverson
36 3 .083 Derrick RoseWe'll get into more nitty gritty in a bit, but this is clearly a pretty solid methodology from the get go.
-The five greatest players of all time are Jordan, Russell, LeBron, Magic, and Kareem, and they're the top five here.
-The bottom five are all infamous losers.
-By far the best is by far the most winningest player of all time.
-By far the worst is by far the most infamous MVP of all time.
-The distribution is extremely well defined: 16 above average, 16 below, 4 above one SD above average, 5 below (10 total expected).
Can't really ask for more.
.
Let's get into some details. First a quick history lesson.
In 1951 (Cousy's rookie year) the NBA had four teams from each conference make the playoffs. (They called them divisions then, but they functioned as conferences do now.) With eight playoff teams it was simple, you had four first round matchups, two second round, and one Finals.
The league brutally contracted in the years following the ABL-BAA merger, leading up to 1954 when each conference was reduced to three playoff teams, but how can you set up a bracket with six participants? Obviously you set up a round robin for each conference where each team played the others twice each except when they didn't sometimes for some reason, the top two advanced, and then you had a standard four bracket.
This was immediately done away with and the NBA instead gave the top team in each conference a first round bye. The NBA's last contraction occurred in 1955 and this model was kept until...
1967. Having very gradually expanded back to ten teams, the everyone gets a trophy generation demanded that the overwhelming majority of teams still make the playoffs, so each conference again sent four teams each, and the straightforward eight team bracket returned.
But expansion escalated in response to the ABA, and by 1974 byes returned when five teams from each conference made the playoffs and the top three received a first round bye.
But expansion escalated even further in absorption of the ABA, and in 1977 the field was again expanded to six teams each with the top two getting a bye.
Finally in 1984 the current playoff structure was established: eight teams from each conference, no byes, four rounds. Hilariously there were only 23 teams in the NBA at this time so you could stumble around to a 30-52 win record and still make the playoffs. The more competitive ensuing generations have ensured that even as the NBA expanded all the way to 30 teams the playoff structure has remained intact.
.
So byes were around for almost half of the MVPs in this analysis. How do we count them? Let's look at a hypothetical standings:
58-24 Celtics (1 seed)
42-40 Royals (2 seed)
38-44 Knicks (3 seed)
52-30 Warriors (1 seed)
52-30 Lakers (2 seed on tiebreaker)
35-47 Hawks (3 seed)
The Royals featuring Oscar Robertson beat the Knicks and lose to the Celtics. They therefore won 1 out of 3 possible rounds.
The Warriors featuring Wilt Chamberlain get a bye in the first round and lose to the Lakers. They therefore won 0 out of 2 possible rounds.
1/3 > 0/2
But this feels wrong. If the Warriors had lost one more game, they would have gotten the cream puff Hawks and almost surely have won - that's the whole point of a bye, after all. And we should avoid models that reward teams for losing more games in the regular season.
What if we a) count a team receiving a bye as having gone 1 of 1 and b) count a team that missed the playoffs entirely during bye years as having gone 0 of 1, in each case on top of whatever else they had? When we do that our list instead looks like:
series won pct name
39 36 .923 Bill Russell
52 35 .673 Magic Johnson
60 35 .583 LeBron James
42 23 .548 Bob Cousy
75 41 .547 Kareem Abdul-Jabbar
52 27 .519 Larry Bird
60 30 .500 Michael Jordan
42 20 .476 Wilt Chamberlain
43 20 .465 Julius Erving
76 35 .461 Tim Duncan
44 20 .455 Kevin Durant
36 16 .444 Stephen Curry
76 33 .434 Shaquille O'Neal
80 33 .413 Kobe Bryant
40 16 .400 Dave Cowens
33 13 .394 Bob Pettit
30 11 .367 Willis Reed
.357 ***average***
54 18 .333 Bob McAdoo
46 14 .304 Wes Unseld
56 17 .304 David Robinson
40 11 .275 Bill Walton
76 19 .250 Karl Malone
40 10 .250 Russell Westbrook
72 16 .222 Hakeem Olajuwon
75 15 .200 Moses Malone
42 8 .190 Oscar Robertson
64 12 .188 Charles Barkley
84 14 .167 Kevin Garnett
80 13 .163 Dirk Nowitzki
72 11 .153 Steve Nash
56 8 .143 Allen Iverson
36 3 .083 Derrick RoseThe big winners are Bob Cousy and Bob Pettit, who played on great teams for long periods when byes were a thing. (Bill Russell did too but when you're already at ninety friggin' percent it's hard to get much higher.) The only people who drop at all are Willis Reed, Bill Walton, and Oscar Robertson.
Overall I gotta say I like the first list better straight up, and once you throw in all the work you have to do to explain how the bye reward/penalty system works it doesn't seem worth it to me.
