I noticed that LeBron James was the top rated player last on Win Shares per 48 and PER last season, but that his ratings on these metrics were actually the lowest for the league leader in 25 years (since Bird in 1985-6). His marks in 2008-9 were the highest since Jordan's in 1990-1 and you'd have to go back to early 70s Jabbar to find the previous guy. Jabbar is the only guy to top James more than once on WS/48, he did it 3 times. Wilt also did it once. Jordan also topped .300 3 times (to tie Jabbar in that regard) but topped James' best only once.
http://www.basketball-reference.com/lea ... early.html
Will James top .300 on WS/48 again or was he one and done? He just missed a second one in 2009-10. Obviously this will affect his championship prospects / total some but it is so rare it is not a championship requirement. Might even be a negative to concentrate so much of a team's strength in one player as opposed to have 2-3 strong stars.
Kobe's top performance on WS/48 was the 195th best in NBA history at .224. http://www.basketball-reference.com/lea ... eason.html
It came in 2005-6, when Kobe and the Lakers got bounced in the first round. In the title seasons he was near .200 and only .160 in the last one (was it the last one ever?) In all the title seasons Shaq and Gasol were better on WS/48, Shaq by a larger margin.
Wade's best was .239 in his title season with Shaq, but he was the clear leader on WS/48. Last season he was at .218. Bosh was at .177 (with a career best of .200 4 seasons ago).
In his 4 peak seasons with the Cavs, James never had a teammate above .180 (Varejao was the 2nd best on WS/48 several times).
Jordan averaged a bit less than 1 other player above .200 on WS/48 in their title seasons and once had 2 with him.
It takes a combination of enough top player excellence and key support. The mix varies.
Dirk had a .200+ sidekick in 4 seasons (and in 2002-3 he had two- Nash and Bradley) but this season Chandler was his highest rated sidekick ever on WS/48 at .218.
In 2000-1 there were 8 guys over .200 on WS/48 in the league. Utah had 2 in Stockton and Malone but Shaq was league's best in the regular season and maintained it in the playoffs while Malone had a really terrible playoffs.
In 2001-2 Shaq again was best in the regular season and only slipped a little in the playoffs. The Spurs had 2 over .200 but Robinson slipped pretty badly.
In 2002-3 Kobe joined Shaq over .200 just in time to face Dallas with 3. Unfortunately for Dallas all 3 slipped, Dirk by a modest amount, Nash badly and Bradley completely.
In 2003-4 the Wolves were the only team with 2 over .200 but both Garnett and Cassell underperformed and Detroit took the prize with 4 guys over .170 in the playoffs.
In 2004-5 the Spurs were the only team with 2 over .200 and won it.
In 2005-6 the Suns were the only team with 2, Dirk was highest, but Wade carried the day. Shaq faded quite a bit but Mourning went over .200 in the playoffs to provide enough support.
In 2006-7 the Spurs smashed to the title with 3 over .200 (Ginobli, Duncan and Barry)
In 2007-8 both Boston and San Antonio had 2 over .200 but the Spurs top guys slipped further than Boston's.
In 2008-9 both the Lakers and Magic had 2 but both Lakers maintained that in the playoffs while the Magic only had 1 do so.
In 2009-10 the Spurs were the only team with 2 regular season but both faded badly in the playoffs. Meanwhile Kobe picked it up in the playoffs to .190 and he and Gasol were enough to win it again.
In 2010-11 there were actually 4 teams with 2 in the regular season. In Chicago both faded, in Orlando 1. Miami and Dallas both had 2 over .200 for the full playoffs but Dallas played better in the series.
2 over .200 on WS/48 doesn't a guarantee of a title. You have to maintain it and often have to beat it at least once. But it appears to be a pretty good shorthand regarding what you usually need to win one.
Exploring with WinShares per 48 minutes
Re: Exploring with WinShares per 48 minutes
There were only 39 players with over 1000 minutes played last season and .150+ on WS/48 in the regular season. http://bkref.com/tiny/P5zOd
All but 4 played for playoff teams, for an average of 2.2 per playoff team. The 8 teams who got out of the first round had 21 for an average of 2.6. The 4 conference finalists had 11 for an average of 2.75. The Lakers and Thunder had 4, the Heat 3.
If you don't have a really super superstar who gives large edge over other #1 players then the strength of your 2nd, 3rd and 4th players is the next main way to distinguish yourself.
The Bulls had 2 over .200 and 5 just under .150 (but over .140) and 2 more over .100. Influenced by the great shot defense and the shared credit for it.
Non-playoff teams averaged just 2.1 guys over .100 with 1000+ minutes played. Influenced by only 1 non-playoff team with an above average defense. On non-playoffs almost everyone should be available in trade discussions if any upgrade option can be found.
The playoff teams averaged 5.8. (The Bulls had 9.)
The view with WS/48 is definitely different than with most boxscore metrics because team shot defense is included in WS/48, in a broad way.
Some other metric would be needed to capture counterpart shot defense, if one wanted to. Adjusted +/- tries to capture it alongside overall team defense. Counterpart WS/48 might be desirable but not available. Counterpart PER and Wins Produced might be less desirable but are available.
Kevin Martin had the best WS/48 of any player with a 1000+ minutes played a WS/48 above .130 and a Defensive rating worse than 108. It is a short list of only 12 guys (mostly starters, all near or over 2000 minutes). Only Houston, Denver and New York had more than one such guy and New York actually had 3 post-trade (all of "the big 3"). Nash had the worst Defensive Rating in the group.
How did one year RAPM evaluate these 12 guys overall? 8 positive, 4 negative. Nash getting the highest RAPM at +3.9, Amare the lowest at -0.5. RAPM actually had Nash estimated at a slight positive on defensive RAPM. Amare was estimated at the 4th worst defensive RAPM in 2011 and that more than wiped out his offensive impact and pulled him down more on RAPM than on WS/48.
So ultimately WS/48 in not the only metric to check and rely on but it seemed interesting / useful to check around with it.
All but 4 played for playoff teams, for an average of 2.2 per playoff team. The 8 teams who got out of the first round had 21 for an average of 2.6. The 4 conference finalists had 11 for an average of 2.75. The Lakers and Thunder had 4, the Heat 3.
If you don't have a really super superstar who gives large edge over other #1 players then the strength of your 2nd, 3rd and 4th players is the next main way to distinguish yourself.
The Bulls had 2 over .200 and 5 just under .150 (but over .140) and 2 more over .100. Influenced by the great shot defense and the shared credit for it.
Non-playoff teams averaged just 2.1 guys over .100 with 1000+ minutes played. Influenced by only 1 non-playoff team with an above average defense. On non-playoffs almost everyone should be available in trade discussions if any upgrade option can be found.
The playoff teams averaged 5.8. (The Bulls had 9.)
The view with WS/48 is definitely different than with most boxscore metrics because team shot defense is included in WS/48, in a broad way.
Some other metric would be needed to capture counterpart shot defense, if one wanted to. Adjusted +/- tries to capture it alongside overall team defense. Counterpart WS/48 might be desirable but not available. Counterpart PER and Wins Produced might be less desirable but are available.
Kevin Martin had the best WS/48 of any player with a 1000+ minutes played a WS/48 above .130 and a Defensive rating worse than 108. It is a short list of only 12 guys (mostly starters, all near or over 2000 minutes). Only Houston, Denver and New York had more than one such guy and New York actually had 3 post-trade (all of "the big 3"). Nash had the worst Defensive Rating in the group.
How did one year RAPM evaluate these 12 guys overall? 8 positive, 4 negative. Nash getting the highest RAPM at +3.9, Amare the lowest at -0.5. RAPM actually had Nash estimated at a slight positive on defensive RAPM. Amare was estimated at the 4th worst defensive RAPM in 2011 and that more than wiped out his offensive impact and pulled him down more on RAPM than on WS/48.
So ultimately WS/48 in not the only metric to check and rely on but it seemed interesting / useful to check around with it.
Re: Exploring with WinShares per 48 minutes
The average age of the WS leader on the championship team, since 1947, is right at 28 years.
The '90s were especially an old man's league, with only Duncan in '99 (age 22) being under 27 and leading a team to a title.
In these 65 seasons, 33 different players have led a titling team in regular season WS.
The average age at first doing it is 27.1 years. The oldest first-timer was Havlicek in '74, age 33. At age 32, Hayes in '78 and Dirk this year.
Besides Duncan, Magic also was team leader at age 22, in '82.
%W is (leader) WS/tm (pythagorean wins)Surprising (to me) team WS leaders: Sharman, Howell (2), Maxwell, Laimbeer (2). And Kobe (not).
The '90s were especially an old man's league, with only Duncan in '99 (age 22) being under 27 and leading a team to a title.
In these 65 seasons, 33 different players have led a titling team in regular season WS.
The average age at first doing it is 27.1 years. The oldest first-timer was Havlicek in '74, age 33. At age 32, Hayes in '78 and Dirk this year.
Besides Duncan, Magic also was team leader at age 22, in '82.
%W is (leader) WS/tm (pythagorean wins)
Code: Select all
Win Shares leaders of champs pyth yr Champ leader age WS tm %W
yr Champ leader age WS tm %W 1979 Sea Sikma 23 8.8 48 .183
2011 Dal Nowitzki 32 11.1 53 .209 1978 Was Hayes 32 8.3 43 .193
2010 LA Gasol 29 11.0 54 .204 1977 Por Walton 24 10.2 55 .185
2009 LA Gasol 28 13.9 61 .228 1976 Bos Cowens 27 10.7 47 .228
2008 Bos Garnett 31 12.9 67 .193 1975 GS Barry 30 12.7 50 .254
2007 SA Duncan 30 13.0 64 .203 1974 Bos Havlicek 33 9.7 51 .190
2006 Mia Wade 24 14.4 52 .277 1973 NY Frazier 27 13.0 59 .220
2005 SA Duncan 28 11.2 63 .178 1972 LA Wilt 35 15.8 67 .236
2004 Det Billups 27 11.3 59 .192 1971 Mil Kareem 23 22.3 67 .333
2003 SA Duncan 26 16.5 57 .289 1970 NY Frazier 24 15.0 62 .242
2002 LA Shaq 29 13.2 60 .220 1969 Bos Howell 32 11.3 55 .205
2001 LA Shaq 28 14.9 51 .292 1968 Bos Howell 31 10.1 51 .198
2000 LA Shaq 27 18.6 64 .291 1967 Phl Wilt 30 21.9 61 .359
1999 SA Duncan 22 8.7 39 .223 1966 Bos Russell 31 11.7 52 .225
1998 Chi Jordan 34 15.8 61 .259 1965 Bos Russell 30 16.9 60 .282
1997 Chi Jordan 33 18.3 68 .269 1964 Bos Russell 29 17.3 59 .293
1996 Chi Jordan 32 20.4 70 .291 1963 Bos Russell 28 13.5 56 .241
1995 Hou Olajuwon 32 10.7 47 .228 1962 Bos Russell 27 15.5 60 .258
1994 Hou Olajuwon 31 14.3 53 .270 1961 Bos Russell 26 13.0 52 .250
1993 Chi Jordan 29 17.2 58 .297 1960 Bos Russell 25 13.8 54 .256
1992 Chi Jordan 28 17.7 66 .268 1959 Bos Russell 24 12.9 50 .258
1991 Chi Jordan 27 20.3 63 .322 1958 StL Pettit 25 11.0 39 .282
1990 Det Laimbeer 32 10.1 57 .177 1957 Bos Sharman 30 10.4 44 .236
1989 Det Laimbeer 31 9.0 56 .161 1956 PhW Johnston 26 13.9 46 .302
1988 LA Magic 28 10.9 56 .195 1955 Syr Schayes 26 12.0 40 .300
1987 LA Magic 27 15.9 62 .256 1954 MnL Mikan 29 12.7 45 .282
1986 Bos Bird 29 15.8 63 .251 1953 MnL Mikan 28 14.6 52 .281
1985 LA Magic 25 12.7 58 .219 1952 MnL Mikan 27 14.4 49 .294
1984 Bos Bird 27 13.6 57 .239 1951 Roc Risen 26 9.3 42 .221
1983 Phl Moses 27 15.1 60 .252 1950 MnL Mikan 25 21.1 55 .384
1982 LA Magic 22 12.9 53 .243 1949 MnL Mikan 24 20.9 47 .445
1981 Bos Maxwell 25 11.0 56 .196 1948 Bal Jeannette 30 8.2 33 .248
1980 LA Kareem 32 14.8 55 .269 1947 PhW Fulks 25 16.3 40 .408
Re: Exploring with WinShares per 48 minutes
The age of the WS leader on the championship team was less than 26 in only 12% of cases during the last 32 years. Will Kevin Durant or Derrick Rose be another exception in the next 2 years? With the proposed CBA changes how many of their current teammates will still be around in 3+ years? Will title contention windows become shorter? Will dynasties be less likely? Will "re-loading" be easier favored teams / top markets, not being as locked into stars past their peak contention years (which appear right now to be later than their peak individual performance years)?
The last 5 champs averaged being $22 million over the cap. If you can't overspend the cap at all or by much or by much without the true cost after the luxury tax going up a lot more than in the past does quality Coaching and Analytics and General Manager decision-making become more important / more valuable / more sought & paid for?
The WS leader was 29 or older in almost 50% of the last 32 years. Nearly all these guys were on their third contract already but in the future if contract lengths are shortened by a year or two it will move even further to being all of them and it will move that decision point up further ahead of the average age of title success. How many will stick with their original team all that way, thru 2 chances to move elsewhere, especially if it is even harder to keep key elements of the crew they had during their 2nd contract with them? Will there be more star player alliances and player alliances hatched before the end of their current contracts? Do shorten contracts tend to put more power in the hands of players & agents to determine the competitive landscape? How thoroughly have the league and the owners and particular owners thought thru all the implications of the proposed changes to the CBA? Are they really fully aware of and willing to accept all that comes with it?
A hard or harder cap would seem likely to favor locations with player desired weather, amenities and big market endorsement & business opportunities. The big market may lose some of their edge but if they still have edge in other areas, the results of player movement and competitive outcomes might not shift that much or at least not as much as if these other factors didn't exist.
The last 5 champs averaged being $22 million over the cap. If you can't overspend the cap at all or by much or by much without the true cost after the luxury tax going up a lot more than in the past does quality Coaching and Analytics and General Manager decision-making become more important / more valuable / more sought & paid for?
The WS leader was 29 or older in almost 50% of the last 32 years. Nearly all these guys were on their third contract already but in the future if contract lengths are shortened by a year or two it will move even further to being all of them and it will move that decision point up further ahead of the average age of title success. How many will stick with their original team all that way, thru 2 chances to move elsewhere, especially if it is even harder to keep key elements of the crew they had during their 2nd contract with them? Will there be more star player alliances and player alliances hatched before the end of their current contracts? Do shorten contracts tend to put more power in the hands of players & agents to determine the competitive landscape? How thoroughly have the league and the owners and particular owners thought thru all the implications of the proposed changes to the CBA? Are they really fully aware of and willing to accept all that comes with it?
A hard or harder cap would seem likely to favor locations with player desired weather, amenities and big market endorsement & business opportunities. The big market may lose some of their edge but if they still have edge in other areas, the results of player movement and competitive outcomes might not shift that much or at least not as much as if these other factors didn't exist.
Re: Exploring with WinShares per 48 minutes
AYC's post from the previous version of the forum:
Crow, you inspired me. I looked at 75 players and came up with 44 with a WS/48 of .190 or better in their "prime". I defined prime as a player's best stretch of consecutive seasons with a WS/48 above his career average (minimum of 5 consecutive seasons. I excluded players with under 500 career games, and also players who entered the league before the shot clock era). Here it is:
1987-98, 830g, .281 WS/48, M.Jordan
1971-81, 853g, .267 WS/48, K.A-Jabbar
1960-68, 706g, .267 WS/48, W.Chamberlain
1990-01, 845g, .258 WS/48, D.Robinson
2005-11, 548g, .248 WS/48, L.James
1989-01, 1028g,.235 WS/48, K.Malone
1994-05, 801g, .234 WS/48, S.O'Neal
1961-68, 604g, .233 WS/48, O.Robertson
1964-72, 623g, .232 WS/48, J.West
1981-91, 797g, .231 WS/48, M.Johnson
1988-97, 713g, .231 WS/48, C.Barkley
2001-11, 864g, .226 WS/48, D.Nowitzki
1999-10, 895g, .222 WS/48, T.Duncan
1988-01, 1094g,.220 WS/48, J.Stockton
2002-11, 747g, .220 WS/48, K.Garnett
2004-11, 564g, .220 WS/48, M.Ginobili
1980-88, 711g, .217 WS/48, L.Bird
1956-64, 670g, .216 WS/48, B.Pettit
2003-10, 613g, .211 WS/48, C.Billups
1958-65, 601g, .210 WS/48, B.Russell
1981-86, 461g, .210 WS/48, S.Moncrief
2005-10, 379g, .210 WS/48, A.Stoudamire
2007-11, 403g, .209 WS/48, D.Howard
2005-11, 486g, .203 WS/48, D.Wade
1980-88, 611g, .202 WS/48, A.Dantley
2000-09, 748g, .201 WS/48, K.Bryant
1984-91, 598g, .200 WS/48, K.McHale
1973-78, 429g, .200 WS/48, B.Lanier
1969-75, 550g, .199 WS/48, W.Frazier
1960-66, 529g, .199 WS/48, S.Jones
2005-11, 489g, .199 WS/48, P.Gasol
1988-95, 583g, .198 WS/48, C.Drexler
1977-84, 624g, .197 WS/48, J.Erving
1974-78, 385g, .197 WS/48, B.McAdoo
1979-87, 700g, .196 WS/48, M.Malone
1996-00, 319g, .196 WS/48, A.Mourning
1959-63, 341g, .195 WS/48, E.Baylor
1986-95, 746g, .194 WS/48, H.Olajuwon
1989-94, 387g, .194 WS/48, M.Price
1990-98, 726g, .193 WS/48, R.Miller
1961-69, 712g, .193 WS/48, B.Howell
1958-63, 450g, .192 WS/48, C.Hagan
1967-75, 719g, .191 WS/48, C.Walker
1968-74, 416g, .190 WS/48, W.Reed
Crow, you inspired me. I looked at 75 players and came up with 44 with a WS/48 of .190 or better in their "prime". I defined prime as a player's best stretch of consecutive seasons with a WS/48 above his career average (minimum of 5 consecutive seasons. I excluded players with under 500 career games, and also players who entered the league before the shot clock era). Here it is:
1987-98, 830g, .281 WS/48, M.Jordan
1971-81, 853g, .267 WS/48, K.A-Jabbar
1960-68, 706g, .267 WS/48, W.Chamberlain
1990-01, 845g, .258 WS/48, D.Robinson
2005-11, 548g, .248 WS/48, L.James
1989-01, 1028g,.235 WS/48, K.Malone
1994-05, 801g, .234 WS/48, S.O'Neal
1961-68, 604g, .233 WS/48, O.Robertson
1964-72, 623g, .232 WS/48, J.West
1981-91, 797g, .231 WS/48, M.Johnson
1988-97, 713g, .231 WS/48, C.Barkley
2001-11, 864g, .226 WS/48, D.Nowitzki
1999-10, 895g, .222 WS/48, T.Duncan
1988-01, 1094g,.220 WS/48, J.Stockton
2002-11, 747g, .220 WS/48, K.Garnett
2004-11, 564g, .220 WS/48, M.Ginobili
1980-88, 711g, .217 WS/48, L.Bird
1956-64, 670g, .216 WS/48, B.Pettit
2003-10, 613g, .211 WS/48, C.Billups
1958-65, 601g, .210 WS/48, B.Russell
1981-86, 461g, .210 WS/48, S.Moncrief
2005-10, 379g, .210 WS/48, A.Stoudamire
2007-11, 403g, .209 WS/48, D.Howard
2005-11, 486g, .203 WS/48, D.Wade
1980-88, 611g, .202 WS/48, A.Dantley
2000-09, 748g, .201 WS/48, K.Bryant
1984-91, 598g, .200 WS/48, K.McHale
1973-78, 429g, .200 WS/48, B.Lanier
1969-75, 550g, .199 WS/48, W.Frazier
1960-66, 529g, .199 WS/48, S.Jones
2005-11, 489g, .199 WS/48, P.Gasol
1988-95, 583g, .198 WS/48, C.Drexler
1977-84, 624g, .197 WS/48, J.Erving
1974-78, 385g, .197 WS/48, B.McAdoo
1979-87, 700g, .196 WS/48, M.Malone
1996-00, 319g, .196 WS/48, A.Mourning
1959-63, 341g, .195 WS/48, E.Baylor
1986-95, 746g, .194 WS/48, H.Olajuwon
1989-94, 387g, .194 WS/48, M.Price
1990-98, 726g, .193 WS/48, R.Miller
1961-69, 712g, .193 WS/48, B.Howell
1958-63, 450g, .192 WS/48, C.Hagan
1967-75, 719g, .191 WS/48, C.Walker
1968-74, 416g, .190 WS/48, W.Reed
Re: Exploring with WinShares per 48 minutes
How well does Win Shares predict (or retrodict) the Sixth Man of the Year award?
Starting from this page -- http://bkref.com/tiny/y7uVE
-- most Win Shares since 1982 by players starting no more than half their games; I just sorted by season, jotted down their votes for 6th man, ranked by votes received.
Sixth man award started in 1983, and votes are known since 1985.
The top 100 WS for non-starters range from 11.1 down to 6.4.
In a 30-year interval, that's 3.3 per year. A few times there is only one qualifying player; he won the award each time.
When the winner of the 6moy did not have 6.4 WS, I had to look him up.Egregiously out of order stats are noted by increasing numbers of # over the column.
Most of the ##+ head scratchers are guards, esp. after the first few years.
Sometimes an out of order winner may be favored by having started fewer games.
Missing games doesn't seem to be a penalizer. Lots of "make up calls".
Starting from this page -- http://bkref.com/tiny/y7uVE
-- most Win Shares since 1982 by players starting no more than half their games; I just sorted by season, jotted down their votes for 6th man, ranked by votes received.
Sixth man award started in 1983, and votes are known since 1985.
The top 100 WS for non-starters range from 11.1 down to 6.4.
In a 30-year interval, that's 3.3 per year. A few times there is only one qualifying player; he won the award each time.
When the winner of the 6moy did not have 6.4 WS, I had to look him up.
Code: Select all
year 1998 votes Tm G GS Min WS WS/48
. 1982 votes Tm G GS Min WS WS/48 Manning 57 PHO 70 11 1794 6.5 .174
Ruland WSB 82 0 2214 7.8 .170 T Murray 1 WAS 82 12 2227 6.5 .139
McHale BOS 82 33 2332 7.5 .155 A Davis 0 IND 82 12 2191 7.4 .162
. 1983 # 1999
B Jones (w) Phl 74 0 1749 6.4 .175 Armstrong 54 ORL 50 15 1502 6.4 .205
Woodson KCK 81 3 2426 7.8 .153 2000 #
McHale BOS 82 13 2345 7.5 .154 R Rogers 104 PHO 82 7 2286 7.4 .154
. 1984 McGrady 3 TOR 79 34 2462 6.6 .129
McHale (w) BOS 82 10 2577 10.5 .196 Je Williams 2 DET 82 1 2102 7.6 .174
Cooper LAL 82 9 2387 6.5 .130 2001 ## ###
. 1985 McKie 57 Phl 76 33 2394 5.9 .118
McHale 57 BOS 79 31 2653 11.0 .199 Best 5 IND 77 21 2457 7.2 .140
. 1986 ## ## S Smith 4 POR 81 36 2542 8.2 .155
Walton 32 Bos 80 2 1546 5.0 .157 K Thomas 0 NYK 77 29 2125 8.2 .185
Pierce 11 MIL 81 8 2147 8.3 .186 A Daniels 0 SAS 79 23 2060 6.6 .153
. 1987 2002 ## ##
Pierce 41 MIL 79 31 2505 9.4 .180 Williamson 56 Det 78 7 1701 5.1 .144
Stockton 0 UTA 82 2 1858 6.7 .174 B Jackson 30 Sac 81 3 1750 5.4 .148
. 1988 Horry 1 LAL 81 24 2140 6.9 .154
Tarpley 67 DAL 81 9 2307 7.8 .163 J Barry 0 DET 82 6 1985 7.5 .180
Bailey 13 UTA 82 10 2804 6.8 .116 Kirilenko 0 UTA 82 40 2151 7.3 .164
Levingston 0 ATL 82 32 2135 7.0 .157 Pollard 0 SAC 80 29 1881 6.8 .175
Rodman 0 DET 82 32 2147 6.8 .153 2003 ### ###
. ## # B Jackson 52 Sac 59 26 1676 5.5 .157
. 1989 votes Tm G GS Min WS WS/48 Redd 33 MIL 82 14 2316 8.0 .167
E Johnson 33 PHO 70 7 2043 6.5 .152 Kirilenko 10 UTA 80 11 2213 9.2 .199
Bailey 26 UTA 82 3 2777 6.5 .113 Bradley 0 DAL 81 39 1731 7.4 .204
Rodman 17 DET 82 8 2208 8.1 .175 2004 #
J Williams 4 WSB 82 1 2413 6.9 .136 Jamison 43 DAL 82 2 2376 9.0 .183
HR Williams 1 CLE 82 10 2125 7.7 .173 Ginobili 33 SAS 77 38 2260 9.1 .194
Pierce 1 MIL 75 4 2078 6.7 .155 Cardinal 0 GSW 76 11 1634 7.2 .212
. 1990 ### ## 2005 ##### ####
Pierce 77 Mil 59 0 1709 5.8 .163 B Gordon 88 Chi 82 3 2002 3.5 .084
Schrempf 8 IND 78 18 2573 8.1 .151 Boykins 11 Den 82 5 2162 6.0 .133
McHale 0 BOS 82 25 2722 11.1 .195 Szczerbiak 1 MIN 81 37 2558 7.3 .136
HR Williams 0 CLE 82 29 2776 8.7 .150 Chandler 0 CHI 80 10 2189 8.3 .183
. 1991 # Okur 0 UTA 82 25 2304 7.0 .147
Schrempf 38 IND 82 3 2632 9.0 .165 A Daniels 0 SEA 75 2 2026 6.8 .160
Majerle 37 PHO 77 7 2281 7.2 .152 2006
McHale 8 BOS 68 10 2067 7.9 .182 M Miller 88 MEM 74 9 2268 7.4 .157
Pierce 6 TOT 78 2 2167 7.9 .174 2007 ## ###
Schayes 0 MIL 82 38 2228 6.8 .147 Barbosa 101 PHO 80 18 2613 8.2 .151
. 1992 Ginobili 18 SAS 75 36 2060 10.6 .246
Schrempf 54 IND 80 4 2605 9.8 .181 Maggette 0 LAC 75 31 2291 7.0 .146
Majerle 19 PHO 82 15 2853 9.8 .165 Lee 0 NYK 58 12 1731 6.9 .191
HR Williams 0 CLE 80 12 2432 9.2 .181 B Barry 0 SAS 75 28 1631 6.8 .201
Pinckney 0 BOS 81 36 1917 7.5 .187 Varejao 0 CLE 81 6 1932 6.6 .164
Kemp 0 SEA 64 23 1808 6.7 .177 2008
# # Ginobili 123 SAS 74 23 2299 11.1 .232
. 1993 votes Tm G GS Min WS WS/48 Terry 0 DAL 82 34 2579 8.6 .160
C Robinson 89 POR 82 12 2575 6.8 .127 Childress 0 ATL 76 0 2274 7.6 .160
Mason 3 NYK 81 0 2482 7.4 .143 Lee 0 NYK 81 29 2356 7.4 .151
Strickland 0 POR 78 35 2474 7.4 .143 Scola 0 HOU 82 39 2024 6.5 .154
. 1994 ## ## 2009 # #
Curry 46 ChH 82 0 2173 5.0 .110 Terry 111 DAL 74 11 2491 7.3 .140
McMillan 37 Sea 73 8 1887 6.0 .154 Odom 2 LAL 78 32 2316 6.9 .143
Gilliam 2 NJN 82 5 1969 6.8 .166 Millsap 0 UTA 76 38 2290 7.7 .162
. 1995 # 2010 ## ###
Mason 47 NYK 77 11 2496 8.6 .166 Crawford 110 ATL 79 0 2460 7.3 .143
Gilliam 1 NJN 82 30 2472 6.6 .128 Terry 6 DAL 77 12 2540 6.5 .123
Perkins 0 SEA 82 37 2356 8.5 .173 Ginobili 2 SAS 75 22 2150 9.7 .216
. 1996 Varejao 2 CLE 76 7 2166 8.1 .179
Kukoc 45 CHI 81 20 2103 10.1 .231 Odom 1 LAL 82 38 2585 7.7 .143
Sabonis 24 POR 73 21 1735 8.4 .233 Millsap 1 UTA 82 8 2277 7.2 .151
Kerr 1 CHI 82 0 1919 8.3 .208 Landry 0 TOT 80 29 2469 7.7 .150
Perkins 0 SEA 82 20 2169 6.9 .153 Redick 0 ORL 82 9 1808 6.5 .173
Legler 0 WSB 77 0 1775 6.8 .184 2011
. 1997 ## #### Odom 96 LAL 82 35 2639 10.1 .184
Starks 84 NYK 77 1 2042 5.8 .136 Harden 0 OKC 82 5 2189 7.1 .156
Kukoc 7 CHI 57 15 1610 6.9 .204 Lawson 0 DEN 80 31 2103 6.7 .153
Perkins 4 SEA 81 4 1976 7.1 .173 R Anderson 0 ORL 64 14 1424 6.4 .217
Kerr 0 CHI 82 0 1861 7.5 .192
Most of the ##+ head scratchers are guards, esp. after the first few years.
Sometimes an out of order winner may be favored by having started fewer games.
Missing games doesn't seem to be a penalizer. Lots of "make up calls".
ESPN rankings and WinShares for Past Champs and the Thunder
Previously posted at DailyThunder.com:
So on the broad-based, subjective ESPN NBA ranking project, 8 teams had more top 50 guys than the Thunder. All 3 of the other conference finalist were top 5 on this while the Thunder were in a 4 way tie for 9th (to 12th).
The Thunder in a 2 way tie for 5th for most guys in the top 100. Dallas was number 1 here. Is that the Mav’s and Thunder’s basic strategy? Seems like it. Also seems like the less popular strategy (at least this season) as only 1 of the conference finalists met this criteria.
I guess you could win either way but it probably would be possible to use some objective criteria and compare top 50 vs top 100 strategies over the last 5-10-20-or more seasons and see which won titles more often.
The west in definitely tougher than the east. Maybe the best strategy varies by conference. Again, could be researched.
Only 7 of the 25 guys in the top 5s by position made the conference finals. Dallas was the only one with just 1 instead of 2. And he was the very oldest such star. Another tidbit of championships most often going to grizzled vets not young guns.
Do Harden and / or Ibaka make it to the top 50 in the next season or two? If so then their strategy changes complexion and moves to the path that was more successful this season at least to the level of conference finalist.
In the last 12 seasons every champ has had 5-6 guys in the top 100 on Winshares except once for the Shaq-Kobe Lakers when tehy had just 4. The 2010-11 Thunder had just 4 (Westbrook, Harden, Durant, Ibaka). I’d say they need to get to 5. It is somewhat unlikely to be from Collison or Thabo getting better, though it could be if the team defense were to go back to the way it performed with Adams in 2009-10. It could also come from Jackson or a new guy via trade or free agency or a future draft.
In the last 12 seasons 9 of the champ have at least 3 guys in the top 50 on Winshares. Once the 2005-6 Heat won with just one top 50 guy but 5 in the top 100. Twice teams won the title with just 2 top 50 guys, this season with Dallas and once with the Lakers when they had 4 in the top 100.
Seems like to be fully covered you’d want at least 3 in the top 50 and at least 5 in the top 100. Of the 4 conference finalists last season only Chicago had that. 9 of the last 12 champs met both criteria.
The 2010-11 Thunder were one short on each. According to Winshares.
ESPN must not have counted Perkins for the Thunder. That would give them 5 in the ESPN top 100. If Harden or Ibaka advance to top 50 then they will meet both ESPN criteria. The WinShares count could still be different. But the concept is similar.
On “WinShares per 48 minutes” last season Perkins was well outside that top 100 at 274th. His best season was 2007-08 when he was 50th on this per minute metric and 70th of overall Winshares. If he could get to one or both of those levels that would help or help a real lot.
Green was 180th on WinShares last season. Not good enough for what they needed from him, what they envisioned for him, what they originally traded for. But they finally realized the shortfall.
They eventually freed themselves of the salary commitment originally tied to Allen but there were probably other better ways to do it (in my opinion at the time of that original deal and certainly in retrospect) and "worst case scenario" it would have been over this upcoming season. But Allen is still productive. If fact he was 17th best on WinShares (in Boston) last season.
The larger, more important perspective though is that 9 of the last 12 champs met both the top 50 and top 100 WinShares criteria.
So on the broad-based, subjective ESPN NBA ranking project, 8 teams had more top 50 guys than the Thunder. All 3 of the other conference finalist were top 5 on this while the Thunder were in a 4 way tie for 9th (to 12th).
The Thunder in a 2 way tie for 5th for most guys in the top 100. Dallas was number 1 here. Is that the Mav’s and Thunder’s basic strategy? Seems like it. Also seems like the less popular strategy (at least this season) as only 1 of the conference finalists met this criteria.
I guess you could win either way but it probably would be possible to use some objective criteria and compare top 50 vs top 100 strategies over the last 5-10-20-or more seasons and see which won titles more often.
The west in definitely tougher than the east. Maybe the best strategy varies by conference. Again, could be researched.
Only 7 of the 25 guys in the top 5s by position made the conference finals. Dallas was the only one with just 1 instead of 2. And he was the very oldest such star. Another tidbit of championships most often going to grizzled vets not young guns.
Do Harden and / or Ibaka make it to the top 50 in the next season or two? If so then their strategy changes complexion and moves to the path that was more successful this season at least to the level of conference finalist.
In the last 12 seasons every champ has had 5-6 guys in the top 100 on Winshares except once for the Shaq-Kobe Lakers when tehy had just 4. The 2010-11 Thunder had just 4 (Westbrook, Harden, Durant, Ibaka). I’d say they need to get to 5. It is somewhat unlikely to be from Collison or Thabo getting better, though it could be if the team defense were to go back to the way it performed with Adams in 2009-10. It could also come from Jackson or a new guy via trade or free agency or a future draft.
In the last 12 seasons 9 of the champ have at least 3 guys in the top 50 on Winshares. Once the 2005-6 Heat won with just one top 50 guy but 5 in the top 100. Twice teams won the title with just 2 top 50 guys, this season with Dallas and once with the Lakers when they had 4 in the top 100.
Seems like to be fully covered you’d want at least 3 in the top 50 and at least 5 in the top 100. Of the 4 conference finalists last season only Chicago had that. 9 of the last 12 champs met both criteria.
The 2010-11 Thunder were one short on each. According to Winshares.
ESPN must not have counted Perkins for the Thunder. That would give them 5 in the ESPN top 100. If Harden or Ibaka advance to top 50 then they will meet both ESPN criteria. The WinShares count could still be different. But the concept is similar.
On “WinShares per 48 minutes” last season Perkins was well outside that top 100 at 274th. His best season was 2007-08 when he was 50th on this per minute metric and 70th of overall Winshares. If he could get to one or both of those levels that would help or help a real lot.
Green was 180th on WinShares last season. Not good enough for what they needed from him, what they envisioned for him, what they originally traded for. But they finally realized the shortfall.
They eventually freed themselves of the salary commitment originally tied to Allen but there were probably other better ways to do it (in my opinion at the time of that original deal and certainly in retrospect) and "worst case scenario" it would have been over this upcoming season. But Allen is still productive. If fact he was 17th best on WinShares (in Boston) last season.
The larger, more important perspective though is that 9 of the last 12 champs met both the top 50 and top 100 WinShares criteria.
Re: Exploring with WinShares per 48 minutes
Mike doesn't look like all the data you intended to display made the transfer to the forum.
Win Shares and Win Shares per 48 minutes only called the winner in 6 of the 15 cases displayed, if I am reading it right.
The Voters are probably not thinking much about player impact on team shot defense in their 6th man award voting while it is crudely estimated in WinShares.
Win Shares and Win Shares per 48 minutes only called the winner in 6 of the 15 cases displayed, if I am reading it right.
The Voters are probably not thinking much about player impact on team shot defense in their 6th man award voting while it is crudely estimated in WinShares.
Re: Exploring with WinShares per 48 minutes
There should be 30 years covered, 28 with a 6moy award. Do you see 2 sets of columns, the first 16 years on the left and the latest 14 on the right?
If your screen is too narrow, try copying into Notepad or something. That ought to contain it.
Sometimes WS and WS/48 aren't the same guy.
Regarding the seeming bias favoring guards, here are median WS/48 rates of players with 1000 to 2000 minutes in the last 5 seasons.
SS# is 'sample size', the number at each position, in b-r.com's newly expanded position description.
Centers (and C-F) are some 55% "better than" guards, not counting elite players nor scrubs.
If your screen is too narrow, try copying into Notepad or something. That ought to contain it.
Sometimes WS and WS/48 aren't the same guy.
Regarding the seeming bias favoring guards, here are median WS/48 rates of players with 1000 to 2000 minutes in the last 5 seasons.
SS# is 'sample size', the number at each position, in b-r.com's newly expanded position description.
Code: Select all
pos SS# median
G 245 .073
G-F 38 .077
F-G 10 .085
F 234 .095
F-C 31 .100
C-F 5 .116
C 76 .112Re: Exploring with WinShares per 48 minutes
Weird. The table didn't look exactly the same to me as it did a few hours ago. Still cropped severely but not quite as bad. (Then next time I log in it was back to the way it was before.) Must be related to my old browser that can't be updated on this system. I can see the full table in notepad though. Thanks for the tip (again, as this probably had this happen before).
Looking at all 30 seasons, Win Shares and WS/48 each still agreed with the voters' choice less than half the time.
Looking at all 30 seasons, Win Shares and WS/48 each still agreed with the voters' choice less than half the time.
Re: Exploring with WinShares per 48 minutes
Sorted into bins of minutes per game, in the last 10 seasons, median WS/48 by position, alongside the number of player-seasons.The latest center who averaged 40 minutes is Shaq in 1999-00.
There isn't any mpg interval in which C's number half as many as F or G. Mostly it's 1/3 or less.
If many F's are playing out of position at C, that may hurt their numbers relative to C; but why are guards so relatively low?
Code: Select all
. minutes per game
pos 36+ ws/48 33-36 ws/48 30-33 ws/48 27-30 ws/48 24-27 ws/48
C 15 .184 31 .145 46 .130 50 .116 45 .111
F 204 .147 149 .115 157 .104 136 .098 134 .098
G 202 .124 145 .114 134 .095 141 .084 154 .076
. minutes per game
p 21-24 ws/48 18-21 ws/48 15-18 ws/48 12-15 ws/48 9-12 ws/48
C 53 .095 52 .090 71 .084 78 .080 73 .038
F 161 .096 162 .089 175 .083 203 .067 211 .054
G 168 .068 175 .064 173 .055 166 .037 159 .033
There isn't any mpg interval in which C's number half as many as F or G. Mostly it's 1/3 or less.
If many F's are playing out of position at C, that may hurt their numbers relative to C; but why are guards so relatively low?
Re: Exploring with WinShares per 48 minutes
Guards have the highest usage, yet the lowest average TS% and highest TO%. And lowest rebounding. Affected by how roles are given out and performed. (Whether big man can handle / should get more usage is a question for another time.)
There seems to be a mini-run of recent champs being tall. I think most or all of the last 6 title winners have been top 10 on average roster height (not weighted by minutes). It wasn't that way in the decade before.
Looking at B-R data it appears that this season's conference finalists were all top 7-8 on minutes played by players listed as F-C, C-F and C, though there are guys listed as Fs who did play some center based on play by play estimates at 82 games so these numbers are not precise. Based on average WS per 48 for the positions that might be giving them some advantage against teams playing smaller, especially if they are having more minutes played by players listed as guards.
Using 6-9 or taller as the definition of a big man instead, it appears that at least 3 of the 4 conference finalists (and 6-7 of those who got out of the first round) were probably top 12 in minutes given to them (traded players could affect that and I didn't feel like sorting that out). Dallas was not one of them strictly, but with Marion's reach, they might effectively qualify.
There seems to be a mini-run of recent champs being tall. I think most or all of the last 6 title winners have been top 10 on average roster height (not weighted by minutes). It wasn't that way in the decade before.
Looking at B-R data it appears that this season's conference finalists were all top 7-8 on minutes played by players listed as F-C, C-F and C, though there are guys listed as Fs who did play some center based on play by play estimates at 82 games so these numbers are not precise. Based on average WS per 48 for the positions that might be giving them some advantage against teams playing smaller, especially if they are having more minutes played by players listed as guards.
Using 6-9 or taller as the definition of a big man instead, it appears that at least 3 of the 4 conference finalists (and 6-7 of those who got out of the first round) were probably top 12 in minutes given to them (traded players could affect that and I didn't feel like sorting that out). Dallas was not one of them strictly, but with Marion's reach, they might effectively qualify.
Re: Exploring with WinShares per 48 minutes
Using eWins and some different position designations (from dougstats), I get a different breakdown:PF are, presumably, F who also rebound; PG are G who assist a lot.
Still, guards play 43% of all minutes, centers just 17%. That's 3% of NBA lineups = 2F + 3G
Maybe (probably) bigs are just more important generally, and PF who don't score much are asked to play C; leaving the PF category as by definition players who do a lot: Griffin, Gasol, Love, Aldridge, Zach, Dirk, Amare, Bosh, Garnett, Smoove, DWest, Millsap, Odom, Scola, Brand, Boozer, DLee, Humphries, ...
These positional disparities may suggest eWins still doesn't weigh assists quite enough. And WS even more so.
When I remove Dwight from consideration, C's average .099 eW/48. Further, shift Duncan back to PF, and they're .096
If Gasol and Amare are among the C, it's PF .117 - .109 C .
Code: Select all
pos eW/48 %Min
C .103 .170
PF .120 .196
SF .087 .205
SG .085 .208
PG .103 .221
Still, guards play 43% of all minutes, centers just 17%. That's 3% of NBA lineups = 2F + 3G
Maybe (probably) bigs are just more important generally, and PF who don't score much are asked to play C; leaving the PF category as by definition players who do a lot: Griffin, Gasol, Love, Aldridge, Zach, Dirk, Amare, Bosh, Garnett, Smoove, DWest, Millsap, Odom, Scola, Brand, Boozer, DLee, Humphries, ...
These positional disparities may suggest eWins still doesn't weigh assists quite enough. And WS even more so.
When I remove Dwight from consideration, C's average .099 eW/48. Further, shift Duncan back to PF, and they're .096
If Gasol and Amare are among the C, it's PF .117 - .109 C .
Re: Exploring with WinShares per 48 minutes
From the old forum location:
Mathketball (October 20)
Crow, you mention above that several of the .200+ guys have tailed off in the playoffs and ultimately their teams didn't win the championship. I bet it would be pretty interesting to compare regular season vs. post season numbers for WS/48. It might give added insight what guys show up for the big game and what guys don't.
Mathketball (October 20)
Crow, you mention above that several of the .200+ guys have tailed off in the playoffs and ultimately their teams didn't win the championship. I bet it would be pretty interesting to compare regular season vs. post season numbers for WS/48. It might give added insight what guys show up for the big game and what guys don't.
Re: Exploring with WinShares per 48 minutes
I did some year to year account of regular season to playoff changes for the every top teams but I'll look at it again more broadly based on the suggestion.
In the last 12 years there have been an average of 8 players over 2000 minutes and also over .200 WS/48. In the playoffs the average drops to 6. That is not dramatic / probably expected given tougher competition.
Here are some of the regular seasons over .200 vs playoffs by prominent player
Name RS P
R Allen 1 2
Anthony 0 1
Billups 5 4
Bryant 6 2
B Davis 0 2
Duncan 9 4
Durant 1 1
Garnett 5 1
Gasol 3 3
Ginobili 4 3
Howard 3 2
James 7 3
Kidd 0 1
K Malone 2 0
Ming 2 2
Nash 4 0
Nowitski 8 6
S O'Neal 5 5
Paul 3 2
Wade 4 3
Westbrook 0 1
Pierce 2 1
D Robinson 3 2
Rose 1 0
Stockton 2 2
Stoudemire 3 2
Of course these are cumulative totals for all seasons within the period and not all are matched pairs.
The most regular seasons over .200 were by Dirk, TimmyD and LeBron. The most such performance in the playoffs were by Nowitski, O'Neal, Duncan and Billups.
The biggest over-performer in the playoffs compared to regular season was B. Davis, but that could perhaps also be explained as under-performance in the regular season compared to his salary and expectations.
The biggest negative differentials between regular season and playoffs were by Duncan, Bryant, Garnett, James and Nash, with Nash they only one in this group to never score .200+ in the playoffs (probably due to team defense). Being on teams that missed the playoff deprived James of 1 opportunity to repeat such a performance in the playoffs and Garnett of 2 but Duncan, Bryant and Nash don't have those excuses and end up with the biggest negative differentials. James tended to only miss by a little on WS/48 in the playoffs. With Duncan it was sometimes close, sometimes not. With Garnett, Bryant and Nash it was usually not that close to a very wide miss. Even with his WS/48 playoff misses Bryant has 5 titles though. The clear candidate for least WS/48 clutchness in the playoffs is Nash.
In the last 12 years there have been an average of 8 players over 2000 minutes and also over .200 WS/48. In the playoffs the average drops to 6. That is not dramatic / probably expected given tougher competition.
Here are some of the regular seasons over .200 vs playoffs by prominent player
Name RS P
R Allen 1 2
Anthony 0 1
Billups 5 4
Bryant 6 2
B Davis 0 2
Duncan 9 4
Durant 1 1
Garnett 5 1
Gasol 3 3
Ginobili 4 3
Howard 3 2
James 7 3
Kidd 0 1
K Malone 2 0
Ming 2 2
Nash 4 0
Nowitski 8 6
S O'Neal 5 5
Paul 3 2
Wade 4 3
Westbrook 0 1
Pierce 2 1
D Robinson 3 2
Rose 1 0
Stockton 2 2
Stoudemire 3 2
Of course these are cumulative totals for all seasons within the period and not all are matched pairs.
The most regular seasons over .200 were by Dirk, TimmyD and LeBron. The most such performance in the playoffs were by Nowitski, O'Neal, Duncan and Billups.
The biggest over-performer in the playoffs compared to regular season was B. Davis, but that could perhaps also be explained as under-performance in the regular season compared to his salary and expectations.
The biggest negative differentials between regular season and playoffs were by Duncan, Bryant, Garnett, James and Nash, with Nash they only one in this group to never score .200+ in the playoffs (probably due to team defense). Being on teams that missed the playoff deprived James of 1 opportunity to repeat such a performance in the playoffs and Garnett of 2 but Duncan, Bryant and Nash don't have those excuses and end up with the biggest negative differentials. James tended to only miss by a little on WS/48 in the playoffs. With Duncan it was sometimes close, sometimes not. With Garnett, Bryant and Nash it was usually not that close to a very wide miss. Even with his WS/48 playoff misses Bryant has 5 titles though. The clear candidate for least WS/48 clutchness in the playoffs is Nash.