Latrinsorm
11-24-2014, 05:16 PM
http://www.basketball-reference.com recently rolled out an updated model for their Hall of Fame probability metric, which is best described as the similarity any given player has to the current membership of the Hall. It takes into account the following criteria...
height
championships
traditional leaderboard
peak Win Shares
All Star selections
...and returns a value between 0 and 1, where 0 is no similarity (and thus presumably no chance) and 1 is perfect similarity (and thus presumably a lock). All values are NBA-only, so no other leagues or college or international or je ne sais quoi is relevant to the method. Also note that it is impossible to take away from one's Hall probability, no matter how putridly one plays in the twilight of one's career.
Also recently I read several articles about the top draft classes in NBA history, and I thought it would be interesting to sum the Hall of Fame probabilities of all their members and see what happened. We can compare a lot more years this way than simply counting actual Hall of Famers: consider that Shaquille O'Neal is not yet in the Hall of Fame but was drafted in 1992, whereas we can be relatively confident that the 2003 draft class for example is very near if not at its eventual value. We also don't have to worry about openly ridiculous Mitch Richmond situations.
Now, the NBA draft has been by no means static in its history, so right off the bat I decided to start from the merger (1977 draft) and also include only those taken in the first two rounds. It turned out that there were only a handful of players who were even All-Stars from later rounds, one of whom (Mark Eaton) was selected in multiple years anyway, so this selection works well. For convenience, I also only counted up players with at least .007 Hall of Fame probability because that's how far the b-r.com leaderboard goes (to David Lee, as it happens).
The graph:
http://img.photobucket.com/albums/v456/johnnyoldschool/NBAHallbyDraft_zps883e0cdd.png
Pretty good!
It is commonly accepted that 03, 96, and 84 are the best classes in some order and that's what we've got here. 96 is the best combination of depth and breadth: nine players is the most of any year, and four of them (AI, Ray Allen, Kobe, Nash) are at or near 1. Compare to 03, where three players are at or near 1 (LeBron, Wade, Bosh) and Carmelo rounds out the group with .7144, but then there's nobody and nobody likely to join them. 85 benefits from breadth with 7 but doesn't have quite the depth, although it is with 84 the only year with four inducted Hall of Famers so far.
It is also commonly accepted that 00 was garbage, and in fact nobody made the leaderboard for HoF probability. Zero players. The closest was Michael Redd's .004. Also interesting is that while there is better cumulative talent in the surrounding drafts, for whatever reason there are no single players who seem destined for the Hall. There is at least one actual inductee every year from 77 to 88, then we have...
89 Tim Hardaway .792
90 Gary Payton inducted
91 Dikembe Mutombo .969
92 Alonzo Mourning inducted, Shaq soon to join
93 Anfernee Hardaway .287
94 Jason Kidd 1.000
95 Kevin Garnett 1.000
96 several 1.000
97 Tim Duncan 1.000
98 Dirk Nowitzki 1.000
99 Shawn Marion .756
00 Michael Redd .004
01 Tony Parker .939
02 Amar'e Stoudemire .729
03 several 1.000
(It's also worth noting that Ginobili was drafted in 99 and Yao was drafted in 02, both of whom are absolute locks to make the Hall on their international merits.)
If we establish a Mitch Richmond line at .700, which happens to be his actual HoF probability, then we can sneak in someone from every year except 93 and 00, and nobody there is even CLOSE. You generally get odd wrinkles when you introduce a ceiling (real or effective) on a function, which is to say Jason Kidd isn't really on the level of Tim Duncan, but even still. The drop-offs in those years are astonishing and inexplicable. A drop-off in 06 makes sense because of the age limit introduction, but nothing like that happened in 00 or 93. You never know!
height
championships
traditional leaderboard
peak Win Shares
All Star selections
...and returns a value between 0 and 1, where 0 is no similarity (and thus presumably no chance) and 1 is perfect similarity (and thus presumably a lock). All values are NBA-only, so no other leagues or college or international or je ne sais quoi is relevant to the method. Also note that it is impossible to take away from one's Hall probability, no matter how putridly one plays in the twilight of one's career.
Also recently I read several articles about the top draft classes in NBA history, and I thought it would be interesting to sum the Hall of Fame probabilities of all their members and see what happened. We can compare a lot more years this way than simply counting actual Hall of Famers: consider that Shaquille O'Neal is not yet in the Hall of Fame but was drafted in 1992, whereas we can be relatively confident that the 2003 draft class for example is very near if not at its eventual value. We also don't have to worry about openly ridiculous Mitch Richmond situations.
Now, the NBA draft has been by no means static in its history, so right off the bat I decided to start from the merger (1977 draft) and also include only those taken in the first two rounds. It turned out that there were only a handful of players who were even All-Stars from later rounds, one of whom (Mark Eaton) was selected in multiple years anyway, so this selection works well. For convenience, I also only counted up players with at least .007 Hall of Fame probability because that's how far the b-r.com leaderboard goes (to David Lee, as it happens).
The graph:
http://img.photobucket.com/albums/v456/johnnyoldschool/NBAHallbyDraft_zps883e0cdd.png
Pretty good!
It is commonly accepted that 03, 96, and 84 are the best classes in some order and that's what we've got here. 96 is the best combination of depth and breadth: nine players is the most of any year, and four of them (AI, Ray Allen, Kobe, Nash) are at or near 1. Compare to 03, where three players are at or near 1 (LeBron, Wade, Bosh) and Carmelo rounds out the group with .7144, but then there's nobody and nobody likely to join them. 85 benefits from breadth with 7 but doesn't have quite the depth, although it is with 84 the only year with four inducted Hall of Famers so far.
It is also commonly accepted that 00 was garbage, and in fact nobody made the leaderboard for HoF probability. Zero players. The closest was Michael Redd's .004. Also interesting is that while there is better cumulative talent in the surrounding drafts, for whatever reason there are no single players who seem destined for the Hall. There is at least one actual inductee every year from 77 to 88, then we have...
89 Tim Hardaway .792
90 Gary Payton inducted
91 Dikembe Mutombo .969
92 Alonzo Mourning inducted, Shaq soon to join
93 Anfernee Hardaway .287
94 Jason Kidd 1.000
95 Kevin Garnett 1.000
96 several 1.000
97 Tim Duncan 1.000
98 Dirk Nowitzki 1.000
99 Shawn Marion .756
00 Michael Redd .004
01 Tony Parker .939
02 Amar'e Stoudemire .729
03 several 1.000
(It's also worth noting that Ginobili was drafted in 99 and Yao was drafted in 02, both of whom are absolute locks to make the Hall on their international merits.)
If we establish a Mitch Richmond line at .700, which happens to be his actual HoF probability, then we can sneak in someone from every year except 93 and 00, and nobody there is even CLOSE. You generally get odd wrinkles when you introduce a ceiling (real or effective) on a function, which is to say Jason Kidd isn't really on the level of Tim Duncan, but even still. The drop-offs in those years are astonishing and inexplicable. A drop-off in 06 makes sense because of the age limit introduction, but nothing like that happened in 00 or 93. You never know!