Joined: 12 Dec 2005
Posts: 74
PostPosted: Sat Jan 07, 2006 1:04 pm Post subject: Worst MVP votes Reply with quote
Hello I'm new to this board, so I want to start off with a good discussion. In your opinion what were the worst MVP selections/ votes in NBA history?
Personally the first vote that comes to mind was Allen Iverson's MVP win in 2001. I love AI and I think he's a great player, but he shouldn't have won the MVP award that year over Shaq. Shap was clearly the best player in the game in the 2000-2001 season and he should have won the award a second time. Personally I think that Allen Iverson benifited from the hype he and his team got from locking up the first seed in the East and the belief that he carried his team. However, just because Allen Iverson's year made a great story doesn't change the fact that Shaq was the best player in the known universe.
My other votes will go to Magic Johnson's MVP wins over Michael Jordan in 1989 and 1990. Magic is a legend and his immense reputation were probably the deciding factors in his back to back MVP wins, but in 1989 and 1990 Jordan had arguably the two greatest-non Wilt Chamberlain- seasons in basketball history. I love Magic, but not only should Jordan have won the award in 89 and 90, the votes should have been unanimous. What are your opinions?
Back to top
View user's profile Send private message
Neil Paine
Joined: 13 Oct 2005
Posts: 774
Location: Atlanta, GA
PostPosted: Sat Jan 07, 2006 4:16 pm Post subject: Reply with quote
In reverse chronological order:
Steve Nash, 2005: Led Suns to league's top record, but may not have been the best (or even 2nd-best) player on his own team. Also, as is usually the case with Nash, it was a bad defensive year. Better MVP choices included Tim Duncan (my choice), Shaq, Amare Stoudemire, and Dirk Nowitzki. The league's best player was Kevin Garnett, but his team underperformed enough to eliminate him from contention.
Iverson, 2001
Charles Barkley, 1993: Not that Sir Charles had a bad year, mind you, but having anyone but Michael Jordan as MVP that year was totally inexcusable.
Magic Johnson, 1990 and 1989: See above, 1993. Jordan was the league's best player both years, and by a pretty wide margin.
I'm not sure about 1981... Dr. J was having a good year, but I think it was pretty obvious that Kareem was the league's best player again. I'll call it a toss-up, if only because the Sixers won 62 games.
Bill Walton, 1978: Sure, he was good... when he played. He only suited up for 58 games, the fewest of any MVP winner ever, and was the only winner to average under 20 ppg until Nash came along. There were plenty of other viable choices, too: David Thompson, George Gervin, and Kareem (sensing a pattern here?). Walton's selection also pretty much invalidated the award...
... that is, assuming that Dave Cowens, 1973 hadn't already. Cowens is easily the worst player ever to win the MVP, but sportswriters needed someone to credit for the Celtics' insane 68-win season. The only other choice on Boston probably would have been John Havlicek, but had he won he would have also been the worst MVP ever. The Celtics were a good team that year, with a lot of depth but no single superstar, so it was hard to find one guy to credit for the big season. Kareem Abdul-Jabbar was the real MVP this year, hands down, though Tiny Archibald's NBA-leading PPG and APG performance is a sentimental favorite. But honestly, how many MVPs did Kareem get screwed out of? Geez...
Willis Reed, 1970: With all due respect to Willis for hobbling out onto the court in Game 7, either Wilt Chamberlain, Jerry West, or Kareem would have been far better MVPs this season. In yet another case of a guy not even being the best player on his team (Clyde Frazier had a better overall year), sportswriters needed somebody to credit for the Knicks' season. While Reed's defense was probably a key to the Knicks' performance, Wilt did the same things while providing more offense.
Wes Unseld, 1969: Unseld had a very good rookie year for the Bullets that year, and 18 rebounds per game is always impressive, but come on now... MVP? I think not. Earl Monroe was a better player for Baltimore in '69, and he was not even close to MVP status. Again, West or Chamberlain would have been the best picks as most valuable, or even their L.A. teammate Elgin Baylor. Too bad Cincinnati was a bad defensive team that year, or Royals Oscar Robertson and Jerry Lucas might have had a shot.
Bill Russell, 1962: Russell was great, one of the best ever, but speaking of the "best ever": Wilt Chamberlain's 1962 may have been the best ever individual season in the history of the game. The Big Dipper led the NBA in at least 10 statistical categories that season, including, of course, 50.4 (!) points per game (100 in one game). Sure, Russell's defense was dominating (and we can only conjecture that Wilt's wasn't, as Philly was last in points allowed/game, yielding almost 4 more points per game than average), but could it ever possibly offset Wilt's offenive contribution? Even with that last-place D, the Warriors finished at 49-31, eclipsed 10,000 points scored on the year, and took Russell's Celts to 7 games in the East finals before finally bowing out. Say it with me: Chamberlain deserved the MVP in 1962.
Bill Russell, 1961: Elgin Baylor had a better year. Plain and simple. Those who say that Russell's defense could offset any other player's offense should note that Baylor's Lakers were tied with the Celtics for the league's top defense, with Baylor playing 3133 minutes... if he were a significantly worse defensive player than Russell, you'd think it would bear out in the numbers. As for offense, Baylor was the NBA's best, even better than Chamberlain: nearly 35 points per game, nearly 20 rebounds, and 5 assists, plus the league's best PER. Give him the award!
Finally, Bob Cousy in 1957. It was arguably the best season in Bob Petit's considerable career, but the voters didn't seem to care about his 25 points and 15 rebounds per game, giving the award to the Nash-esque Cooz. 7.5 assists per game (2.4 more than anyone else) and an NBA title (over Petit's Hawks) are impressive, but Petit was easily the league's most productive player, on a team with few other options. Meanwhile, Cousy racked up his assists (basically his only claim to the award) next to Bill Sharman, Tommy Heinsohn, and Russell.
So, those are my undeserving MVPs. There are probably more out there, though; I tried to give the voters huge benefits of the doubt, but some omissions were too glaring to ignore.
Back to top
View user's profile Send private message Visit poster's website
Mike G
Joined: 14 Jan 2005
Posts: 3630
Location: Hendersonville, NC
PostPosted: Sun Jan 08, 2006 9:09 am Post subject: Reply with quote
Hey, I think I may actually agree with 100% of the opinion so far in this thread. Consider these comments to be additions and not criticisms:
1998: Jordan was named MVP at the beginning of the season. Shaq (again) missed 20 games. Robinson was almost as great but missed 9 games. Therefore, Karl should've had 3 in a row.
1993: Hakeem was a close 2nd and played every game (MJ missed 4).
1981: Far from Kareem's worst snub, and Julius was due
1979: Moses' time hadn't actually arrived. Kareem still averaged almost 4 blocks, and 5 assists!
1978: Wylie, you don't say who should've won, rather than Bill. Kareem missed 20 games himself. Lanier missed 19. My vote: Artis Gilmore!
1973/69: Who was 'worse': Cowens or Unseld. Wes' year would be pretty typical of his career, yet he was never again a major vote-getter. This was just the umpteenth slap at Wilt, and his 'me-first' attitude (real or perceived). Nobody likes Goliath.
In 1970, I like West and Billy C.
1961: Yeah, this was Baylor's best shot. He never led the league in anything, and he never won a ring.
1957: I rank 9 guys above Cousy, and they are all 'bigs'. These 9 split 50 votes, while Cooz took 31. (The next-best G was running-mate Sharman).
This is still a predictable factor in MVP voting: Guard bias. More specifically, little-man bias. No one loves Goliath, and everyone loves the underdog. Most voters are closer to Iverson's size than to Shaq's.
From Cousy to Nash, I think it comes down to voter perception of what brings more 'value' to the league, as an enhanced commodity. Do spectators want to watch Wilt/Kareem/Shaq overwhelming people? Or do they like well-oiled-machine/Team basketball? -- whether orchestrated by 'team-first' Centers (Russell/Unseld/Cowens/Walton) or by point guards.
Makeup calls are also a distinct element. Maybe Magic 'had to' win 3 MVP's to match Bird. Cousy, Oscar, Erving, Barkley, Olajuwon, Robinson 'got theirs'; not in seasons that were necessarily their best.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
Jim Raynor
Joined: 11 Dec 2005
Posts: 65
Location: New Jersey
PostPosted: Sun Jan 08, 2006 7:09 pm Post subject: Reply with quote
This isn't a selection, but the worst MVP vote I can think of was the one P.J. Brown got last season. IIRC, the sportswriter who made that vote admitted that he did that just to bring attention to an underrated player.
Back to top
View user's profile Send private message Send e-mail
jkubatko
Joined: 05 Jan 2005
Posts: 702
Location: Columbus, OH
PostPosted: Mon Jan 09, 2006 12:50 pm Post subject: Reply with quote
On my site I have outlined a method for calculating basketball win shares. Win shares can only be calculated from the 1977-78 season forward. In those 28 years, the MVP has led the league in win shares 17 times (60.71 percent). In the other 11 years, here is the difference between the MVP's win shares and the league leader's win shares (sorted from largest difference to smallest):
Code:
Year Win Shares Leader MVP Diff
1978 David Thompson (39) Bill Walton (24) 15
1989 Michael Jordan (56) Magic Johnson (44) 12
1994 David Robinson (52) Hakeem Olajuwon (43) 9
2001 Shaquille O'Neal (45) Allen Iverson (36) 9
2005 Dirk Nowitzki (44) Steve Nash (36) 8
1990 Michael Jordan (56) Magic Johnson (51) 5
1984 Adrian Dantley (46) Larry Bird (42) 4
1993 Michael Jordan (48) Charles Barkley (44) 4
1997 Michael Jordan (52) Karl Malone (48) 4
1998 Karl Malone (48) Michael Jordan (45) 3
1981 Kareem Abdul-Jabbar (43) Julius Erving (42) 1
For example, in 1978 David Thompson led the league with 39 win shares. The MVP, Bill Walton, had 24 win shares, 15 fewer than Thompson.
_________________
Regards,
Justin Kubatko
Basketball-Reference.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
jkubatko
Joined: 05 Jan 2005
Posts: 702
Location: Columbus, OH
PostPosted: Mon Jan 09, 2006 1:13 pm Post subject: Reply with quote
Another way to look at this is to take the player's PER and convert it into "PER credits." To do this I take a player's PER, divide by 15, and multiply by minutes played. PER can only be accurately calculated since the 1977-78 season in the NBA, but I came up with a way to estimate PER for other seasons (please see this article for more information). Of the 50 NBA MVPs, 27 (54 percent) have led the league in PER credits. Here are the cases where the NBA MVP did not lead the league in PER credits:
Code:
Year PER Credits Leader MVP Diff
1962 Wilt Chamberlain (8313) Bill Russell (4484) 3829
1963 Wilt Chamberlain (8066) Bill Russell (4245) 3821
1961 Wilt Chamberlain (6985) Bill Russell (4182) 2804
2005 Kevin Garnett (5868) Steve Nash (3771) 2097
1973 Tiny Archibald (6184) Dave Cowens (4139) 2046
1965 Wilt Chamberlain (6298) Bill Russell (4513) 1785
1969 Wilt Chamberlain (5355) Wes Unseld (3589) 1766
1978 Artis Gilmore (4808) Bill Walton (3193) 1615
1989 Michael Jordan (6763) Magic Johnson (5186) 1578
1990 Michael Jordan (6654) Magic Johnson (5214) 1440
1957 Bob Pettit (4700) Bob Cousy (3338) 1363
1987 Michael Jordan (6513) Magic Johnson (5233) 1281
1964 Wilt Chamberlain (7780) Oscar Robertson (6558) 1222
2001 Shaquille O'Neal (5893) Allen Iverson (4745) 1148
1993 Michael Jordan (6072) Charles Barkley (4936) 1136
1970 Kareem Abdul-Jabbar (5296) Willis Reed (4180) 1116
1994 David Robinson (6625) Hakeem Olajuwon (5525) 1100
1981 Adrian Dantley (5525) Julius Erving (4803) 721
1958 Dolph Schayes (4705) Bill Russell (4005) 699
1999 Shaquille O'Neal (3473) Karl Malone (3127) 346
1998 Karl Malone (5640) Michael Jordan (5352) 288
2003 Tracy McGrady (5961) Tim Duncan (5710) 251
1984 Adrian Dantley (4902) Larry Bird (4889) 14
This just reinforces how good Davis21wylie2121's list was.
_________________
Regards,
Justin Kubatko
Basketball-Reference.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Neil Paine
Joined: 13 Oct 2005
Posts: 774
Location: Atlanta, GA
PostPosted: Mon Jan 09, 2006 4:54 pm Post subject: Reply with quote
Thanks, Justin. Speaking of Win Shares (their not being able to be calculated prior to 1978, that is), John Hollinger made a kind of off-hand comment in this year's Forecast that I thought was really cool... on page 33 he said, "I have a rule of thumb that for every 2,000 minutes a player plays, each additional point of PER is worth an additional win." Specifically, he used it to compare how many wins Antoine Walker's addition added to Boston's record... Does this mean, though, that (Wins = (PER*(Minutes/2000)))? How do these "PER Wins" compare to Justin's Win Shares (which I think are incredible, btw)? If they do match up (I just thought of this, so I haven't even experimented with this yet), and PER can be calculated for virtually all NBA seasons, we could have a master list of all-time "player-win-type" stat leaders. Any thoughts?
Back to top
View user's profile Send private message Visit poster's website
jkubatko
Joined: 05 Jan 2005
Posts: 702
Location: Columbus, OH
PostPosted: Tue Jan 10, 2006 4:43 pm Post subject: Reply with quote
Davis21wylie2121 wrote:
Thanks, Justin. Speaking of Win Shares (their not being able to be calculated prior to 1978, that is), John Hollinger made a kind of off-hand comment in this year's Forecast that I thought was really cool... on page 33 he said, "I have a rule of thumb that for every 2,000 minutes a player plays, each additional point of PER is worth an additional win." Specifically, he used it to compare how many wins Antoine Walker's addition added to Boston's record... Does this mean, though, that (Wins = (PER*(Minutes/2000)))?
No, I don't think he meant that. I think he meant that if you replace Player A with Player B, and Player B has the better PER, the team will gain (Player B PER - Player A PER) wins for every 2000 minutes Player B plays. I hope that makes sense.
_________________
Regards,
Justin Kubatko
Basketball-Reference.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Mike G
Joined: 14 Jan 2005
Posts: 3630
Location: Hendersonville, NC
PostPosted: Wed Jan 11, 2006 9:59 am Post subject: Reply with quote
jkubatko wrote:
Here are the cases where the NBA MVP did not lead the league in PER credits:
Code:
Year PER Credits Leader MVP Diff
2005 Kevin Garnett (5868) Steve Nash (3771) 2097
1973 Tiny Archibald (6184) Dave Cowens (4139) 2046
1969 Wilt Chamberlain (5355) Wes Unseld (3589) 1766
1978 Artis Gilmore (4808) Bill Walton (3193) 1615
1989 Michael Jordan (6763) Magic Johnson (5186) 1578
1990 Michael Jordan (6654) Magic Johnson (5214) 1440
1987 Michael Jordan (6513) Magic Johnson (5233) 1281
2001 Shaquille O'Neal (5893) Allen Iverson (4745) 1148
1993 Michael Jordan (6072) Charles Barkley (4936) 1136
1970 Kareem Abdul-Jabbar (5296) Willis Reed (4180) 1116
1994 David Robinson (6625) Hakeem Olajuwon (5525) 1100
1981 Adrian Dantley (5525) Julius Erving (4803) 721
1999 Shaquille O'Neal (3473) Karl Malone (3127) 346
1998 Karl Malone (5640) Michael Jordan (5352) 288
2003 Tracy McGrady (5961) Tim Duncan (5710) 251
1984 Adrian Dantley (4902) Larry Bird (4889) 14
This just reinforces how good Davis21wylie2121's list was.
I trimmed Justin's list a bit to eliminate the Wilt/Russell MVP era and before. And I updated my "the real mvp" file for purposes of comparison/corroboration.
Here's my ranking of MVP Credits Disparity.
miss yr Winner mis% Screwed
135 73 Cowens .694 Kareem 442
130 05 Nash .692 Garnett 422
96 69 Unseld .722 Wilt 346
78 89 Magic .835 Jordan 475
71 01 Iverson .831 Shaquille 420
69 90 Magic .851 Jordan 460
55 93 Barkley .874 Olajuwon 442
51 87 Magic .885 Jordan 444
50 70 Reed .861 Kareem 361
48 78 Walton .864 Gilmore 356
41 98 Jordan .904 Karl 424
40 94 Olajuwon .914 Robinson 468
37 79 Moses .909 Kareem 405
16 81 Erving .957 Kareem 385
10 99 Karl .967 Shaquille 316
1 03 Duncan .997 Garnett 416
The miss% shows, for example, that Nash had only 69% of Garnett's 422 'mvp credits', in '05. Nash and Cowens are both ranked #1, either in total or % of displacement.
PER seems to rank high-% scorers (Dantley) higher than versatile players (Erving and Bird in their primes). I've got AD at only 93% of the top value player in '81 and in '84.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
gabefarkas
Joined: 31 Dec 2004
Posts: 1313
Location: Durham, NC
PostPosted: Wed Jan 11, 2006 10:56 am Post subject: Reply with quote
jkubatko wrote:
On my site I have outlined a method for calculating basketball win shares. Win shares can only be calculated from the 1977-78 season forward. In those 28 years, the MVP has led the league in win shares 17 times (60.71 percent). In the other 11 years, here is the difference between the MVP's win shares and the league leader's win shares (sorted from largest difference to smallest):
Code:
Year Win Shares Leader MVP Diff
1978 David Thompson (39) Bill Walton (24) 15
1989 Michael Jordan (56) Magic Johnson (44) 12
1994 David Robinson (52) Hakeem Olajuwon (43) 9
2001 Shaquille O'Neal (45) Allen Iverson (36) 9
2005 Dirk Nowitzki (44) Steve Nash (36) 8
1990 Michael Jordan (56) Magic Johnson (51) 5
1984 Adrian Dantley (46) Larry Bird (42) 4
1993 Michael Jordan (48) Charles Barkley (44) 4
1997 Michael Jordan (52) Karl Malone (48) 4
1998 Karl Malone (48) Michael Jordan (45) 3
1981 Kareem Abdul-Jabbar (43) Julius Erving (42) 1
For example, in 1978 David Thompson led the league with 39 win shares. The MVP, Bill Walton, had 24 win shares, 15 fewer than Thompson.
Justin--
Although this is indeed fascinating, the problem is that the award is called Most Valuable Player, not Most Productive Player...
Back to top
View user's profile Send private message Send e-mail AIM Address
94by50
Joined: 01 Jan 2006
Posts: 499
Location: Phoenix
PostPosted: Wed Jan 11, 2006 11:41 am Post subject: Reply with quote
gabefarkas wrote:
jkubatko wrote:
On my site I have outlined a method for calculating basketball win shares. Win shares can only be calculated from the 1977-78 season forward. In those 28 years, the MVP has led the league in win shares 17 times (60.71 percent). In the other 11 years, here is the difference between the MVP's win shares and the league leader's win shares (sorted from largest difference to smallest):
Code:
Year Win Shares Leader MVP Diff
1978 David Thompson (39) Bill Walton (24) 15
1989 Michael Jordan (56) Magic Johnson (44) 12
1994 David Robinson (52) Hakeem Olajuwon (43) 9
2001 Shaquille O'Neal (45) Allen Iverson (36) 9
2005 Dirk Nowitzki (44) Steve Nash (36) 8
1990 Michael Jordan (56) Magic Johnson (51) 5
1984 Adrian Dantley (46) Larry Bird (42) 4
1993 Michael Jordan (48) Charles Barkley (44) 4
1997 Michael Jordan (52) Karl Malone (48) 4
1998 Karl Malone (48) Michael Jordan (45) 3
1981 Kareem Abdul-Jabbar (43) Julius Erving (42) 1
For example, in 1978 David Thompson led the league with 39 win shares. The MVP, Bill Walton, had 24 win shares, 15 fewer than Thompson.
Justin--
Although this is indeed fascinating, the problem is that the award is called Most Valuable Player, not Most Productive Player...
And therein lies the semantic debate about the meaning of "valuable". I have taken a lot of interest in this thread because the whole idea of the "Most Valuable Player" (and by extension, whether a player should have been the MVP or not) is so variable. Is it just the most productive player? Does a player have to play on a winning team to be truly valuable? Did the player's team improve greatly after adding this player (the dreaded "New Guy" phenomenon)? Did this team play significantly better with this player than without him?
If a player's team doesn't go to the playoffs or otherwise underperforms (like Minnesota or Cleveland last year), the argument as I understand it is that the player's performance isn't truly valuable because the player's team didn't succeed. I disagree. The implication of that theory is that an otherwise legitimate contender for the MVP Award can't actually be the MVP because his teammates weren't good enough. That sounds highly unreasonable and illogical.
I don't advocate selecting MVPs on individual productivity alone, but I'd rather do that than disregard deserving candidates because they had to run with Zydrunas Ilgauskas and Drew Gooden rather than Amare Stoudemire and Shawn Marion.
Back to top
View user's profile Send private message
HoopStudies
Joined: 30 Dec 2004
Posts: 706
Location: Near Philadelphia, PA
PostPosted: Wed Jan 11, 2006 12:15 pm Post subject: Reply with quote
94by50 wrote:
Quote:
....
Justin--
Although this is indeed fascinating, the problem is that the award is called Most Valuable Player, not Most Productive Player...
And therein lies the semantic debate about the meaning of "valuable". I have taken a lot of interest in this thread because the whole idea of the "Most Valuable Player" (and by extension, whether a player should have been the MVP or not) is so variable. Is it just the most productive player? Does a player have to play on a winning team to be truly valuable? Did the player's team improve greatly after adding this player (the dreaded "New Guy" phenomenon)? Did this team play significantly better with this player than without him?
If a player's team doesn't go to the playoffs or otherwise underperforms (like Minnesota or Cleveland last year), the argument as I understand it is that the player's performance isn't truly valuable because the player's team didn't succeed. I disagree. The implication of that theory is that an otherwise legitimate contender for the MVP Award can't actually be the MVP because his teammates weren't good enough. That sounds highly unreasonable and illogical.
We had this debate with Nash's MVP selection, I think.
The most value a team can get is in winning a title (economically and competitively), so I then figure that the player that contributes the most towards his team winning a championship is a good definition of MVP. A player whose performance, as good as it may be, doesn't get his team into the playoffs, is not the most valuable. Yeah, it may be unfair, but the most value is in having a chance to win the title. Since the award is voted on before the playoffs, you have to essentially figure out the odds of winning the title based on regular season -- which is 0 if you don't make the playoffs -- and then parse out credit.
That's a road map for identifying an MVP, I think. It makes it a function of both productivity and team success in a logical way. The odds of winning a title given win-loss record, playoff seed can be found. How you parse out credit is still a bit of personal choice. But I think that method will explain some of the discrepancy noted in this thread between awards and measures of productivity.
The main reason I bring this up is because I think criticism of MVP awards really can make our community sound bad. When Hollinger (who I'll pick on because he was most public, not because he was unique) chose Garnett for MVP last year, it was a case of just looking at his PER value. The main criticism of stat guys is that all they look at is numbers -- they're not looking at the whole picture. And selecting Garnett last year was a clear case of John doing just that because it was simple. Stats are a simplification of basketball. Making a straight decision based on a simplification of basketball is a strong and legit critique of what John did. No one else on ESPN made that selection and John looked like the geeky stat guy who can't get a date because his calculator doesn't tell him how (uh, not speaking from experience, really). It's fine that productivity measures showed that Garnett was most productive, but, yawn, that isn't the question. Most valuable vs most productive. Let's elevate the discussion to Most Valuable, not just most productive. And maybe we'll get a few more hot dates...
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top
View user's profile Send private message Visit poster's website
jkubatko
Joined: 05 Jan 2005
Posts: 702
Location: Columbus, OH
PostPosted: Wed Jan 11, 2006 12:43 pm Post subject: Reply with quote
Perhaps I misled some people, but I am not advocating the use of a statistical system for selecting the MVP. Anthony started this thread by asking about the worst MVP selections in NBA history. One way to get a starting point to answer that question is to find players who statistically did not "look like" an MVP. I actually think the NBA MVP voters have done a pretty good job over the years, much better than the MLB MVP voters.
Valuable can mean different things to different people. Dean, have you talked to John about why he chose Garnett? If not, your implication that he was lazy (you said he made the "simple" choice) is unfair.
Oh, and some of us are married with two beautiful children, so we don't need to worry about getting dates.
_________________
Regards,
Justin Kubatko
Basketball-Reference.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
kjb
Joined: 03 Jan 2005
Posts: 865
Location: Washington, DC
PostPosted: Wed Jan 11, 2006 12:52 pm Post subject: Reply with quote
jkubatko wrote:
Oh, and some of us are married with two beautiful children, so we don't need to worry about getting dates. Smile
Just like a stat guy to break out how many kids he has.
Umm...I have 3.
Wink
Back to top
View user's profile Send private message AIM Address Yahoo Messenger
bchaikin
Joined: 27 Jan 2005
Posts: 690
Location: cleveland, ohio
PostPosted: Wed Jan 11, 2006 1:02 pm Post subject: Reply with quote
Let's elevate the discussion to Most Valuable, not just most productive...
ok - last season in 04-05 shawn marion was the most valuable player on the suns, followed very closely by amare stoudemire. this season in 05-06 it's shawn marion again whose most valuable in phoenix....
And maybe we'll get a few more hot dates...
sorry - already married (but my calculator's real hot)...
The main reason I bring this up is because I think criticism of MVP awards really can make our community sound bad....
i think a sound discussion of why or why not a player should have been named MVP is fruitful in bringing out to the public the different methodologies for determining just this. the fact that the steve nash selection has spurred such debate is good reason for looking at the process of selection more closely, in particular via stats analysis...
where many years earlier the choice was made with less stats available and possibly those who did the voting not seeing each of the players on a regular basis, today there is so much available in terms of stats, video, and stats analysis such as this community has developed to make a qualified choice. debating the MVP for example brings all these into play and gives one more tools to make an educated choice...
i can't imagine, for example, that there isn't an nba draft war room where discussions just like this occur all the time - one scout's idea of why one player is good or bad versus another's, or video evidence versus statistical evidence...
i'm not saying the method for choosing the MVP should be changed, but only that we as a community should in fact debate the choice...
Stats are a simplification of basketball....
stats are a vindication of basketball....
HoopStudies
Joined: 30 Dec 2004
Posts: 706
Location: Near Philadelphia, PA
PostPosted: Wed Jan 11, 2006 1:02 pm Post subject: Reply with quote
jkubatko wrote:
Perhaps I misled some people, but I am not advocating the use of a statistical system for selecting the MVP. Anthony started this thread by asking about the worst MVP selections in NBA history. One way to get a starting point to answer that question is to find players who statistically did not "look like" an MVP. I actually think the NBA MVP voters have done a pretty good job over the years, much better than the MLB MVP voters.
Valuable can mean different things to different people. Dean, have you talked to John about why he chose Garnett? If not, your implication that he was lazy (you said he made the "simple" choice) is unfair.
You're right. I didn't talk to him. But I saw nothing in his writing about his selection that suggested anything other than a choice based on PER (which, I realize, sounds a bit like an argument being made regarding a Supreme Court nominee). He seemed to be making the productivity argument and nothing else. I'd be glad to be wrong, but the point withstands the example.
You yourself have said several times that you don't advocate basic productivity as the way to choose an MVP. I just remembered vividly John's standing out last year (because he is so public) as making the simple statistical argument rather than the more complex argument that really gets at value.
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top
View user's profile Send private message Visit poster's website
schtevie
Joined: 18 Apr 2005
Posts: 413
PostPosted: Wed Jan 11, 2006 1:59 pm Post subject: Reply with quote
Of course, the fundamental problem is one of definition: what constitutes value? Given that the voting precedes the Finals, presumably the notion of value supercedes success in the playoffs. Personally, I have always preferred an alternative specification of the question of MVP, where the question is, if the league were starting anew, who would be the number one draft pick? This is the ultimate measure of value, I think, as it incorporates not just some absolute value of skill, but relates to the relative scarcity of talent, across all notional categories. As such, Hollinger's answer is not geeky at all, as long as it is properly explained. And not much explanation is needed, I should think, as it is conventional wisdom that championships are not won without adequate "supporting cast".
Back to top
View user's profile Send private message
Kevin Pelton
Site Admin
Joined: 30 Dec 2004
Posts: 979
Location: Seattle
PostPosted: Wed Jan 11, 2006 2:09 pm Post subject: Reply with quote
schtevie wrote:
Personally, I have always preferred an alternative specification of the question of MVP, where the question is, if the league were starting anew, who would be the number one draft pick?
Unless you specified that you were drafting only for that given season, wouldn't that enter age into the discussion? I don't think LeBron was the most valuable player in the NBA in 2004-05, but in terms of future value he was a clear-cut #1.
With regards to the Garnett pick in particular, don't you at least have to address the fact had such a mediocre Roland Rating (+3.6)? That seems pretty damning if we're talking value.
Back to top
View user's profile Send private message Send e-mail Visit poster's website
94by50
Joined: 01 Jan 2006
Posts: 499
Location: Phoenix
PostPosted: Wed Jan 11, 2006 2:10 pm Post subject: Reply with quote
HoopStudies wrote:
The main criticism of stat guys is that all they look at is numbers -- they're not looking at the whole picture... It's fine that productivity measures showed that Garnett was most productive, but, yawn, that isn't the question.
I'll give you that... certainly we don't want to be known as "basketball accountants" or something like that, slavishly following the bottom line without question. I definitely don't wish to be thought of like that.
At the same time, I keep bristling at media types, talking heads, and fans who overrate or underrate athletes because they either don't win championships or haven't won one yet. Alex Rodriguez and Peyton Manning are two of the most prominent examples. They're both Hall-of-Fame caliber athletes, but they get questioned by the media and crushed by the general public because they're somehow personally responsible for never winning championships.
I do agree with Justin that compared to MLB MVP selections (not to mention their Hall of Fame selections), the NBA MVP voting has been excellent. I just thought it was silly that Nash might have been the third-best player on his team last year and still got an MVP out of it. Throw in the fact that MVP voting has been biased towards players who either lead or are very near their team lead in scoring, and I couldn't see why Nash was the designated candidate from the Suns, other than that he was the "new guy". And personally, I would have voted for Stoudemire last year.
Back to top
View user's profile Send private message
HoopStudies
Joined: 30 Dec 2004
Posts: 706
Location: Near Philadelphia, PA
PostPosted: Wed Jan 11, 2006 2:42 pm Post subject: Reply with quote
schtevie wrote:
Of course, the fundamental problem is one of definition: what constitutes value? Given that the voting precedes the Finals, presumably the notion of value supercedes success in the playoffs. Personally, I have always preferred an alternative specification of the question of MVP, where the question is, if the league were starting anew, who would be the number one draft pick? This is the ultimate measure of value, I think, as it incorporates not just some absolute value of skill, but relates to the relative scarcity of talent, across all notional categories. As such, Hollinger's answer is not geeky at all, as long as it is properly explained. And not much explanation is needed, I should think, as it is conventional wisdom that championships are not won without adequate "supporting cast".
Ah, conventional wisdom.
If all we do is spout conventional wisdom, we aren't useful either.
John's point for Garnett as MVP was anti-conventional wisdom and I respect that as an important aspect of what we have to do. I'm just pointing out that we have to spell out good reasons behind what decisions we make, not just a number. The rationale for a number needs spelling out and it appeared to me that John's primary (or only) justification for Garnett was the rationale behind PER. And that rationale is reasonable for productivity, but not the only story being fairly touted in the MVP race and I don't think he addressed the other stories.
We're fighting a HUGE uphill battle to get this work used, despite gains that have been made. LARGE, HUGE uphill in a raging snowstorm both ways (my dad would be proud). We have to be really smart in how we make arguments. John's public forum makes him the primary voice outsiders hear. If he looks like the geek by choosing someone ignoring other (more difficult to quantify but reasonable) ways, that reflects on everyone here.
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top
View user's profile Send private message Visit poster's website
Ben
Joined: 13 Jan 2005
Posts: 266
Location: Iowa City
PostPosted: Wed Jan 11, 2006 3:06 pm Post subject: Reply with quote
If the only valid measure of value is contributing to chances of winning the championship, then the NBA should simply vote after the championship. Obviously, there would be a huge correlation between Finals MVP and MVP then.
As long as it's a regular season award, I prefer most productive season as the measure (Garnett didn't pick his teammates) and presumably Hollinger does as well. This doesn't seem to be a conflict that can be settled by evidence, but the argument that it makes one seem geeky to choose one measure of value is uniquely unpersuasive.
Back to top
View user's profile Send private message
jkubatko
Joined: 05 Jan 2005
Posts: 702
Location: Columbus, OH
PostPosted: Wed Jan 11, 2006 3:46 pm Post subject: Reply with quote
HoopStudies wrote:
You yourself have said several times that you don't advocate basic productivity as the way to choose an MVP.
That's right. I think statistics can (and should) be used to get a reasonable list of candidates. If one candidate stands head-and-shoulders above the rest statistically -- like Kevin Garnett in 2004 -- then I think the choice is rather clear-cut. Last year was a different story, as there were a number of good candidates. My finalists would have been Dirk Nowitzki, Amare Stoudemire, and Garnett, although good cases can be made for Shawn Marion, Tim Duncan, and LeBron James. These guys were fairly close in terms of production, so at this point I would use things like team success, etc. to separate them. My first cut would be Garnett, since his team didn't make the playoffs. The choice between Nowitzki and Stoudemire is tough. In the APBRmetric board poll I voted for Nowitzki, so I'll stick with that choice now, although I could be convinced to change to Stoudemire. You could not convince me that Garnett was the league's MVP last year.
_________________
Regards,
Justin Kubatko
Basketball-Reference.com
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Neil Paine
Joined: 13 Oct 2005
Posts: 774
Location: Atlanta, GA
PostPosted: Wed Jan 11, 2006 3:58 pm Post subject: Reply with quote
All of this is why I tried to temper the advanced numbers (Win Shares, PER, TS%, etc.) with conventional measures in my list. It's stupid to go back and look at MVP selections from 40 years ago, and evaluate them based on only one statistical measure (one that wouldn't even be invented until decades later, for that matter), and then say "well, according to my Super Index Value Rating (insert stat-name here), Wilt Chamberlain was the real 1962 MVP... how I pity those poor, stupid 1962 sportswriters!" Wilt Chamberlain was the true 1962 MVP for a variety of reasons, statistical and otherwise, not simply because he led the earth in "Points Created per 48 minutes," or what have you. So whoever said we shouldn't boil it down to one number to evaluate players is totally right. But I trust that John didn't do that, either, when he said that KG was MVP, because he's more responsible than that. He probably picked KG because he thought that KG was the best player in the League... which he was. So John was simply using another definition of "Most Valuable". Which brings us to...
Quote:
Of course, the fundamental problem is one of definition: what constitutes value?
Yeah, that. I think that it's up to either the league itself, or the sportswriters who vote, to make a decision: are we talking about the most productive (or "best") player in the NBA... or are we talking the literal "Most Valuable" player? Because both are fairly easy to narrow down, if you think about it.
"Value", according to The American Heritage Dictionary, is defined: "An amount, as of goods, services, or money, considered to be a fair and suitable equivalent for something else; a fair price or return." We also have means of quantifying (at least in the ballpark) a player's production, or the quality of his "services", to his team. So, if we're talking productivity, we look at the PER, the Win Shares, the Points/40 min., and we watch the games... voila! We have a pretty clear picture: about a handful of guys who could really be the NBA's best player. If we're talking literal "value," that's easy too. Salary information is pretty accessable nowadays, so just divide a player's production by the amount of money that his team has deemed "a fair and suitable equivalent" for that production. Shouldn't the answer be the literal "Most Valuable Player"? Nash cost the Suns $8,750,000 last year; Stoudemire cost only $2,032,200. Even if you (and this being generous to Nash) call them equal production-wise, Stoudemire was way more valuable by this measure. I think the voters are confusing multiple definitions of "value," but in order for the debate to make sense, there must be a uniform definition.
Davis21wylie2121 now steps down off soapbox.
Back to top
View user's profile Send private message Visit poster's website
HoopStudies
Joined: 30 Dec 2004
Posts: 706
Location: Near Philadelphia, PA
PostPosted: Wed Jan 11, 2006 4:13 pm Post subject: Reply with quote
Ben wrote:
If the only valid measure of value is contributing to chances of winning the championship, then the NBA should simply vote after the championship. Obviously, there would be a huge correlation between Finals MVP and MVP then.
As long as it's a regular season award, I prefer most productive season as the measure (Garnett didn't pick his teammates) and presumably Hollinger does as well. This doesn't seem to be a conflict that can be settled by evidence, but the argument that it makes one seem geeky to choose one measure of value is uniquely unpersuasive.
The regular season does establish a pretty good set of expectations of who will win the title. It is that set of expectations that eliminated Garnett from a lot of ballots and raised other players on those same ballots. It is a factor and most people would say a valid factor. There has been desire to vote for the regular season awards after the playoffs (I voiced it when I was 19 and thought I could change the world). Failing that change -- and I don't think that change will happen for some good reasons (I ain't 19 anymore) -- people look at the regular season and say that its ability to project post-season success is a valid consideration for MVP.
Regardless, if we make arguments about MVP, those making their case based strictly on individual productivity should both state so and remember that others are looking past that and incorporating more. And those considering more information should not be criticized for doing so without a pretty clear case for why that other information should not be used.
(And some would say that players like Garnett do help pick their teammates, if not directly, then indirectly through the limitations their salary imposes.)
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top
View user's profile Send private message Visit poster's website
Mike G
Joined: 14 Jan 2005
Posts: 3612
Location: Hendersonville, NC
PostPosted: Wed Jan 11, 2006 5:42 pm Post subject: Reply with quote
Some of us have gone to a bit of work to generate overall 'value' estimates, like Win Shares, eWins, etc. Why would it seem that ignoring all this would be "incorporating more".
If a player adds 20 wins to a team that would otherwise win 10 games, or 50 games, how does that change his value?
If the Timberwolves had been in the Eastern Conference last year, they'd have been the 3rd or 4th seed. If Kevin Garnett had been traded to the Atlanta Hawks, they might have become the 3rd or 4th seed; and KG would have been a shoo-in MVP, his team favored to contend. But he wouldn't have been any more 'valuable' than he was in Minnesota.
I think the purpose of this community should be to dispel 'voodoo valuation', rather than to propagate it.
_________________
`
36% of all statistics are wrong
Back to top
View user's profile Send private message Send e-mail
HoopStudies
Joined: 30 Dec 2004
Posts: 706
Location: Near Philadelphia, PA
PostPosted: Wed Jan 11, 2006 6:41 pm Post subject: Reply with quote
Mike G wrote:
Some of us have gone to a bit of work to generate overall 'value' estimates, like Win Shares, eWins, etc. Why would it seem that ignoring all this would be "incorporating more".
Incorporating more meant accounting for increasing the team's chances of winning a title. Part of this is any overall value estimate or productivity estimate. But the part that we've said that these stat methods (most, if not all) ignore is that some players didn't give their teams a chance to win a title as of the end of the season. I definitely did not say to ignore any stat method. I in fact said to use the one you want, but use it in combination with some measure of team odds of winning a championship.
Mike G wrote:
If a player adds 20 wins to a team that would otherwise win 10 games, or 50 games, how does that change his value?
If you want to argue that team success has nothing to do with MVP votes, that's another argument. I'd say it does matter because playoff revenue matters, because winning games in the playoffs matters. If all you want to be is a 20-win adder on a 30 win team your whole career, rather than a 20-win adder on a 70 win team, then you're right -- you can win all the MVPs you want and be happy while remaining a team loser (but I know of no player who considered that optimal). But you're saying that these two players should split MVP votes. And I'd say you're loco. Would you really split your vote between player A on a 30 win team and player B on a 70 win team if they had the same MikeStat?
Mike G wrote:
I think the purpose of this community should be to dispel 'voodoo valuation', rather than to propagate it.
And to do it correctly, incorporate as much info as is reasonable, and not ignore apparently important factors like team success because it's hard to handle.
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers.
Back to top
View user's profile Send private message Visit poster's website
gabefarkas
Joined: 31 Dec 2004
Posts: 1313
Location: Durham, NC
PostPosted: Wed Jan 11, 2006 7:09 pm Post subject: Reply with quote
HoopStudies wrote:
The most value a team can get is in winning a title (economically and competitively), so I then figure that the player that contributes the most towards his team winning a championship is a good definition of MVP. A player whose performance, as good as it may be, doesn't get his team into the playoffs, is not the most valuable. Yeah, it may be unfair, but the most value is in having a chance to win the title. Since the award is voted on before the playoffs, you have to essentially figure out the odds of winning the title based on regular season -- which is 0 if you don't make the playoffs -- and then parse out credit.
Not to split hairs, but what you're describing sounds an awful lot like the Finals MVP, no? Along the lines of what you're saying though, could the (regular) MVP be the guy who contributes the most towards his teams record, however stellar or stinky that record may be?
HoopStudies wrote:
The main reason I bring this up is because I think criticism of MVP awards really can make our community sound bad. When Hollinger (who I'll pick on because he was most public, not because he was unique) chose Garnett for MVP last year, it was a case of just looking at his PER value. The main criticism of stat guys is that all they look at is numbers -- they're not looking at the whole picture. And selecting Garnett last year was a clear case of John doing just that because it was simple. Stats are a simplification of basketball. Making a straight decision based on a simplification of basketball is a strong and legit critique of what John did. No one else on ESPN made that selection and John looked like the geeky stat guy who can't get a date because his calculator doesn't tell him how (uh, not speaking from experience, really). It's fine that productivity measures showed that Garnett was most productive, but, yawn, that isn't the question. Most valuable vs most productive. Let's elevate the discussion to Most Valuable, not just most productive. And maybe we'll get a few more hot dates...
I interpreted it not as the facile choice, but as John standing behind his method (ie, PER). He was saying that by his count (however subjective or objective that may be), KG provided the most value to his team. Perhaps that says that he equates value with productivity, perhaps it says something about a purely objective selection methodology, or perhaps it says something about the method itself.
Back to top
View user's profile Send private message Send e-mail AIM Address
gabefarkas
Joined: 31 Dec 2004
Posts: 1313
Location: Durham, NC
PostPosted: Wed Jan 11, 2006 7:29 pm Post subject: Reply with quote
HoopStudies wrote:
Incorporating more meant accounting for increasing the team's chances of winning a title. Part of this is any overall value estimate or productivity estimate. But the part that we've said that these stat methods (most, if not all) ignore is that some players didn't give their teams a chance to win a title as of the end of the season. I definitely did not say to ignore any stat method. I in fact said to use the one you want, but use it in combination with some measure of team odds of winning a championship.
Let's say Player K increases his team's chances of winning a title from 1-in-100 to 1-in-25, while Player S increases his team's chances of winning a title from 1-in-20 to 1-in-8. Thus, Player K conferred a 400% increase in his team's chances, while Player S conferred only (by comparison) a 250% increase in his team's chances.
So, even though Player K increased his team's chances more, would you still be inclined to favor him for your MVP over Player S? Player K's team would best be described as "sucky" without him and "marginal playoff team" with him, while Player S's team went from "low playoff seed that could possibly be a first-round upsetter" without him to "arguable championship contender" with him.
Dean: I don't mean to come off like I'm just unabatedly challenging you, because I certainly am not. I'm just asking some tough questions in what I believe is the proper forum to ask them, and hoping to spur discussions to a higher level. For some reason, what you say always gets me thinking...
HoopStudies wrote:
If all you want to be is a 20-win adder on a 30 win team your whole career, rather than a 20-win adder on a 70 win team, then you're right -- you can win all the MVPs you want and be happy while remaining a team loser (but I know of no player who considered that optimal). But you're saying that these two players should split MVP votes. And I'd say you're loco. Would you really split your vote between player A on a 30 win team and player B on a 70 win team if they had the same MikeStat?
The latter (20-win adder on a 70-win team) sounds an awful lot like Dennis Rodman...
Back to top
View user's profile Send private message Send e-mail AIM Address
Ben
Joined: 13 Jan 2005
Posts: 266
Location: Iowa City
PostPosted: Wed Jan 11, 2006 7:35 pm Post subject: Reply with quote
HoopStudies wrote:
The regular season does establish a pretty good set of expectations of who will win the title. It is that set of expectations that eliminated Garnett from a lot of ballots and raised other players on those same ballots.
You seem to be implying that most voters vote only on this, but this doesn't square with the history of the voting. Jordan handily won the MVP when his team was tied for the 7th best record in the league. Nobody gave them much of a chance for a title that year. Certainly, Magic was far more deserving by your proposed measure.
As for this season, should Lebron receive no consideration? Currently, Tradesports.com has Cleveland with about a 3% shot at the title. San Antonio has a 1 in 3 chance while Detroit has 1 in 4. It seems like Duncan 's pretty much a lock by this measure - Billups with an outside shot, but Tony Parker and Rip Hamilton far ahead of Lebron.
Back to top
View user's profile Send private message
antcole
Joined: 12 Dec 2005
Posts: 74
PostPosted: Wed Jan 11, 2006 7:40 pm Post subject: Reply with quote
Quote:
It's fine that productivity measures showed that Garnett was most productive, but, yawn, that isn't the question. Most valuable vs most productive. Let's elevate the discussion to Most Valuable, not just most productive. And maybe we'll get a few more hot dates...
I understand where you're coming from Dean. We shouldn't limit our votes for MVP to just stats. The MVP award should focus on the whole picture, particulary how much this player contributed to the team's success. And there are cases where a player can be a great stat guy, but can safely be described as a loser (Yes I'm talking to you, Vince Carter). However, I don't believe the MVP award should necessarily be given to the best player on the team with the best record or a championship contender, because I think that approach is as equally flawed as solely zeroing in on stats.
I don't care how good somebody is, unless that player has a well rounded supporting cast that player is probably not going to be on a championship-quality team, so I don't think it is right to give out an MVP award to a player just because he had a great year on a team with a great to very good supporting cast and overlooking another player who perhaps may have had a far superior year, but played on a dismal or mediocre team (i.e, is anybody honestly going to make the argument that Nash supporting cast was better than KG's supporting class?).
Honestly, when I think of MVP award I ask myself one simple question: who was the best player in basketball this year year? Whoever is the best player in the league in a particular year is the MVP, in my opinion.