The Grizzlies Make a Move
The Grizzlies in 2005-06 were led in Wins Produced by four players: Pau Gasol, Mike Miller, Shane Battier, and Eddie Jones. In the off-season, Battier was traded for Stromile Swift and Rudy Gay, two players who one would not expect to offer the production of Battier. Gasol was also injured while playing for Spain during the World Championship, and hence missed the start of the 2006-07 season. And apparently Eddie Jones also started the season injured. Given the loss of three of its four best players, the Grizzlies quickly sank to the bottom of the NBA.
Over the past few weeks, though, Gasol has returned from injury and is producing at the same pace he offered last year. Jones, who started the year playing very badly (again, probably due to injury), has posted very good numbers in January. His per-minute Win Score this month has been 0.255, a mark well above average for a shooting guard.
For Jones to play well is no surprise. In twelve NBA season he has produced 105 wins and posted a Wins Produced per 48 minute [WP48] of 0.167. Yes, he did play very badly in the first couple of months of the season. But after missing a few games at the start of January he came on in the latter half of the month with numbers quite similar to what Jones had offered throughout his career.
Clearly with Jones and Gasol healthy and producing, the Grizzlies were posed to start winning a bit more frequently. Yesterday, though, the Grizzlies decided to waive Jones. Why would the Grizzlies suddenly decide to waive a player who appears to be healthy and ready to help this team win games?
Losing to Win
Let’s go back to last April when the Memphis Grizzlies faced the LA Clippers. The loser of this game knew it would draw the Denver Nuggets in the first round of the NBA playoffs. The winner would take on the Dallas Mavericks. Although there is evidence neither team played their very best players as much as they could, Memphis managed to play a bit better and won the game. Their reward was a sound thrashing by the Mavericks in the first round. The Clippers defeated the Nuggets and performed well enough in the second round to be considered potential title contenders this year (a forecast that one should not have made after the team signed Tim Thomas, but that’s a story for another day).
The waiving of Eddie Jones suggests that the Grizzlies have learned the lesson taught last April. Sometimes losing today is better for your team tomorrow.
Why would the Grizzlies want to lose now?
The key is the NBA draft. A few days ago a reader of The Wages of Wins Journal – Erich Doerr – sent me his analysis of the college performance of players expected to be available in the 2007 draft. The analysis suggested – as many have already guessed –that there are some extremely talented players available in next year’s NBA draft. To have the best shot at this talent, though, you need to lose today.
To see the talent available to the NBA’s losers, consider the following table, where the per-minute Win Score of fourteen college players ESPN thinks might be lottery choices is reported.
Table One: The Performance of Potential Lottery Choices in 2006-07
The per-minute Win Score is calculated with data from this season (which I downloaded from ESPN.com today). These results were compared to the college performance – at each position — of players taken in the first fifteen choices of each draft since 1994 (specifically, college performance is the per-minute Win Score for the last year each player played in college).
As one can see, Kevin Durant is currently posting numbers that are 0.134 above the per-minute college numbers posted by small forwards from the past thirteen drafts. To be more specific, Durant is currently posting numbers that are better than any small forward taken in the top 15 slots since 1994. Second on the list if Greg Oden, followed by Al Horford and Joakim Noah. Each of these players is offering much more than the average player at their respective position.
The analysis provided in Table One is hardly necessary for people paying attention to the NBA draft. Most analysts expect each of these players to help their future NBA employer win games. The only problem is that you can’t get these players if you win today.
The Predictive Power of College Performance
It should be emphasized that college performance does not correlate perfectly to professional productivity. Since 1994, there is a 0.69 correlation between a rookie’s per minute Win Score in the NBA and what that rookie did his last year in college (assuming the rookie played at least 500 minute his rookie season).
What this means is that it’s likely that Durant, Oden, Horford, and Noah – based on what we have seen so far in 2006-07 – will be productive NBA players in 2007-08. But that’s not guaranteed. Likewise, players at the bottom of the above list – Thaddeus Young, Spencer Hawes, and Darrell Arthur – are not guaranteed to be unproductive NBA players. Still, one might ask how players who are supposed to be “good” in the NBA could play relatively poorly at the college level.
Losing to Win, Again
Would NBA teams reduce their chances to win today to land a player like Durant or Oden?
As mentioned in The Wages of Wins and last May in this forum, one of the best recent articles written in the field of sports economics was “Losing to Win: Tournament Incentives in the National Basketball Association.” by Beck Taylor and Justin Trogdon. This paper, which appeared in The Journal of Labor Economics in 2002, offered evidence that before the institution of the draft lottery, non-playoff bound NBA teams actually played to lose towards the end of each season. And when the lottery introduced weights to give the worst teams a better shot at the best talent, there was again evidence that bad teams lost more than you would expect even bad teams to do.
So when I see the Grizzlies cut Eddie Jones, and the Sixers cut Chris Webber, I immediately suspect that the focus of these teams has shifted to the next draft lottery. Sure, maybe there are other reasons for these players to be cut. But the leading article of faith in economics is that people respond to their incentives. And when we see the quality of players available in the 2007 draft, bad teams today have a clear incentive to enhance their “badness” this season.
– DJ
Wages of Wins Journal 
Jason
January 31, 2007
Though bumping up your lottery odds may have something to do with decisions to dump Jones and Webber, there are other economic incentives to consider that also favor these moves. In Webber’s case, his contract was bought out. Per the league’s CBA, it is his buyout amount, prorated over the remaining life of the contract rather than the contract value that counts against both the salary cap and, probably as important, the luxury tax.
In Webber’s case, the 6ers were on book to pay him more than $20million this year and next. The buyout amount was for a bit less than that. With the tax threshold going up, Webber’s salary next year reflecting the lower (albeit still almost $20 million) the 6ers could be under the lux tax threshold next season. In addition to avoiding the dollar for dollar penalty on salaries in excess of the threshold, they’ll also receive a share of the tax revenues, split equally among the non-taxed teams. Depending on the number of teams taxed, this can be anywhere from $5 to $25 million. The can’t get under this year, but next year is a different story. The weight of Webber’s salary next year was not only the salary, but potentially losing out on a share of the tax. Because of this, to keep him to play poorly, they’d be in for much more than they’re in for letting him go.
Jones may have also been a lux tax casualty. He apparently agreed to a buyout as well. The Grizz were very close to the ~65 million lux tax threshold this year. While I believe they were under it, they could not have acquired even marginal minimum salary players without going over. Again, the penalty for going over is more than just the penalty in dollars paid, but also missing out on the revenue sharing, thus making the decision about either keeping Jones and paying him to be a limited player or cutting him and paying him, but also improving the odds of getting back some of that salary by not being over the lux tax.
If the reality is that neither Webber nor Jones were going to help their teams gain substantial revenue through winning this year, and I suspect that this is true even if Jones did return to being a productive player, it would seem that mitigating the real dollar cost would be a wise move, especially if it has the added value of potentially landing a better player for the future.
Katie Gold
January 31, 2007
John Hollinger at ESPN.com wrote a very intriguing article on this topic, “To Tank or Not To Tank? That is the question.” Hollinger makes a compelling argument that there is a “sweet spot” between 53 and 62 losses where it makes sense for a team to lose games on purpose. However, this strategy does not seem to make sense for the three teams that need the top pick the most: Philly, Memphis, and Portland, because they are already in the sweet spot and wouldn’t gain much.
Its a really interesting article, and it would be great to get your perspective on his analysis (Insider subscription required):
http://insider.espn.go.com/nba/insider/columns/story?columnist=hollinger_john&id=2747955
dberri
January 31, 2007
Jason,
I think you are correct that the salary cap played a role in both decisions. I did find it odd, though, that the decision to cut Jones came right after the he appeared to be playing well. Clearly the Grizzlies have to be worried about playing so well that the both miss the playoffs and miss out on the top talent.
Katie,
I saw Hollinger’s take on this. I thought what he said was interesting, although I wish he would have noted the work of Taylor and Trogdon. I think the big story is not so much whether this strategy works, but that there is evidence that teams have in the past not tried to win as much as possible down the stretch to secure a better draft pick.
Different Jason
February 1, 2007
Grizzlies need to be treated differently than other teams, as it is easier to sell a bare team in the NBA than it is to sell a mediocre team with their payroll at the cap. Any potential buyer would much rather have a situation similar to the Charlotte bobcats (barely at the league minimum) for a couple years with the belief that it is easier to build a team from scratch than toboth maintain a team and get it to rise to the next level. When people look at the Bobcats low payroll, they don’t assume that the Bobcats play to lose, rather that they feel better off building their own talent up and froming a foundation instead of paying large contracts for players like Brian Cardinal.
noah body
February 1, 2007
I have long been interested in the numbers of the draft lottery, but have never seen anyone break down the “best” spot to have going into the lottery. The worst team in the league only has a 25% chance of winning the lottery, which means it has a 75% chance of moving down. And, of course, a 0% chance of moving up. I do not like those odds.
The No. 2 team has a 20% chance of moving up, an 18.91% chance of staying even, and a 61.09% chance of moving down.
The No. 3 team has a 31.54% chance of moving up, a 15.7% chance of staying at No. 3 and a 50.76% chance of moving down. And the further you go down, the smaller the odds are of moving up or down in the draft.
Do you see my point? Statistically, what is the best place to have in the lottery to maximize your chance of getting a good pick while minimizing your chance of getting screwed? I am guessing it is either No. 4 or No. 5, but my poor statistics skills are not good enough to figure it out definitively.
Jake
February 2, 2007
Noah,
There is a little known caveat in the lottery process. The lottery balls are only drawn for the first 3 draft spots, this means that no NBA team can fall more than 3 spots from their original position. So, using your numbersm the worst team would have a 25% chance of 1st, 75% chance of picking between 2 and 4, and 0% chance of moving below 4.
While what you say is true, that the worst team has 0% chance to move up, that does not mean they have a worse spot than a the 2nd team, or any other team.
For instance, even if the 5th team were to have a 50% chance of moving up, they would still have a 50% chance of selecting 5th or worse.
Especially in this draft, where it appears there are 4 players who are clearly the best of the group (i’m including the two florida kids), the best spot is the #1 lottery pick where you know you will be able to select at least one of those 4.
– check out the wikipedia artivle on the NBA draft lottery for more complete info.
http://en.wikipedia.org/wiki/NBA_Draft_Lottery
Jon
February 2, 2007
In the case of the Sixers, they are actually a better team without Webber. While there isn’t enough data to support this just yet, the fact that the Sixers are 5-3 when starting Steven Hunter just shows that they are playing more like a team. 2 of those wins were against the preseason cream of the East – the Heat and Cavs. Playing out a .625 winning percentage over the final 43 games would give them 37 on the season. Following your prediction of 35 wins after the Iverson trade, cutting Webber would yield 2 additional wins.