Post by eric on Mar 7, 2017 13:49:05 GMT -6
This came up in shout yesterday: can we objectively define an NBA contender?
Yes.
First of all I restricted my analysis to the NBA from 1977-present. The league had at least 22 teams in this span, so Conference Finals has meant pretty much the same thing, at least compared to back in the 50s - being in the final four of an eight team league isn't exactly a mighty feat.
My initial idea was a team making two Conference Finals or better in a three year span. This turned out to be too granular and too harsh - teams moved into and out of contender status too much and there was one year where the method claimed there was only one contender. Neither of these fits what we mean by the word.
Then I tried making three CF or better in a five year span. Teams' contender statuses were more contiguous, but it turned out to be even harsher than the first method, as there were five years with only one contender.
Then I tried the happy medium of two in a five year span, and as is so often the case the happy medium turned out to be absolute garbage, just worthless. There were contenders everywhere, frequently seven in a given year.
Finally I tried two in a four year span, and this one I think really works. There was still one year with seven contenders but that was '94 and '94 was a weird time due to Jordan's retirement. On average there were 4 contenders a year, which seems right. As an added bonus 36 of the champions of the 37 spans were rated as a contender by this method, the highest percentage so far discussed.
.
Because we're using multi year spans, we can't define a contender until after the playoffs have happened. I think of this as a feature, not a bug. Prognosticators always claim they knew what was going to happen in retrospect anyway, completely ignoring the massively wrong predictions they just made. The goal of this is to measure the truth, not predict it. Here's the chart, and the way to read this is 2013 = "2013 through 2016", and it's not meant to state that the given team were contenders in that entire span. For instance look at Cleveland and Miami - Cleveland wasn't a contender in 2013-2014, Miami was not in 2015-2016. You get the idea.
I think it's a good sign that every span is immediately identifiable: Garnett's and Bird's Celtics, Kidd's Nets, Moncrief's Bucks, and so on down the line. Regardless of what happens this year three teams have already qualified as contenders in the next span: Cavaliers, Warriors, and Thunder(!). Other teams that would qualify with a Conference Finals appearance or better are the Hawks, Rockets, Pacers, Heat, Spurs, and Raptors, although obviously at most four of those teams can actually do so regardless of how the seeding plays out.
Yes.
First of all I restricted my analysis to the NBA from 1977-present. The league had at least 22 teams in this span, so Conference Finals has meant pretty much the same thing, at least compared to back in the 50s - being in the final four of an eight team league isn't exactly a mighty feat.
My initial idea was a team making two Conference Finals or better in a three year span. This turned out to be too granular and too harsh - teams moved into and out of contender status too much and there was one year where the method claimed there was only one contender. Neither of these fits what we mean by the word.
Then I tried making three CF or better in a five year span. Teams' contender statuses were more contiguous, but it turned out to be even harsher than the first method, as there were five years with only one contender.
Then I tried the happy medium of two in a five year span, and as is so often the case the happy medium turned out to be absolute garbage, just worthless. There were contenders everywhere, frequently seven in a given year.
Finally I tried two in a four year span, and this one I think really works. There was still one year with seven contenders but that was '94 and '94 was a weird time due to Jordan's retirement. On average there were 4 contenders a year, which seems right. As an added bonus 36 of the champions of the 37 spans were rated as a contender by this method, the highest percentage so far discussed.
.
Because we're using multi year spans, we can't define a contender until after the playoffs have happened. I think of this as a feature, not a bug. Prognosticators always claim they knew what was going to happen in retrospect anyway, completely ignoring the massively wrong predictions they just made. The goal of this is to measure the truth, not predict it. Here's the chart, and the way to read this is 2013 = "2013 through 2016", and it's not meant to state that the given team were contenders in that entire span. For instance look at Cleveland and Miami - Cleveland wasn't a contender in 2013-2014, Miami was not in 2015-2016. You get the idea.
year atl bos bro cha chi cle dal den det gs hou ind lac lal mem
2013 1 1 1
2012 1
2011 1
2010 1
2009 1 1
2008 1 1
2007 1 1 1 1
2006 1 1 1
2005 1
2004 1
2003 1 1
2002 1 1 1
2001 1 1 1
2000 1 1
1999 1 1
1998 1 1
1997 1 1 1
1996 1 1
1995 1 1 1
1994 1 1 1
1993 1 1 1
1992 1 1 1
1991 1
1990 1 1
1989 1 1 1
1988 1 1 1
1987 1 1 1 1
1986 1 1 1
1985 1 1 1
1984 1 1
1983 1 1
1982 1 1
1981 1 1
1980 1 1
1979 1 1
1978 1
1977 1
year mia mil min nop nyk okc orl phi pho por sac sas tor uta was
2013 1 1 1
2012 1 1 1
2011 1 1 1
2010 1 1 1
2009 1 1 1
2008 1
2007 1 1
2006 1
2005 1 1 1
2004 1 1 1
2003 1 1 1
2002 1
2001 1
2000 1
1999 1 1 1
1998 1 1 1
1997 1 1 1
1996 1
1995 1 1
1994 1 1
1993 1 1 1 1
1992 1 1
1991 1 1 1
1990 1 1
1989 1 1
1988 1 1
1987 1
1986
1985
1984 1
1983 1 1
1982 1 1 1
1981 1 1 1
1980 1 1
1979 1 1 1
1978 1 1 1
1977 1 1 1
I think it's a good sign that every span is immediately identifiable: Garnett's and Bird's Celtics, Kidd's Nets, Moncrief's Bucks, and so on down the line. Regardless of what happens this year three teams have already qualified as contenders in the next span: Cavaliers, Warriors, and Thunder(!). Other teams that would qualify with a Conference Finals appearance or better are the Hawks, Rockets, Pacers, Heat, Spurs, and Raptors, although obviously at most four of those teams can actually do so regardless of how the seeding plays out.
