Author Archives: Devin Dignam

DeMar DeRozan is killing the Raptors

Posted on by
3

Recently, the perennially unproductive “offensive weapon” of the Toronto Raptors — Andrea Bargnani — went down with another injury. Although he’s having the best season of his career to date, Bargnani is still a below average player, and there is little evidence to suggest that he makes his teammates better. Shouldn’t that mean that the Raptors will be playing better without him?

Theoretically, but unfortunately, Toronto seems to generate unproductive players like the Hydra grows heads; when one is eradicated, another couple materialize. Exibit A is the Raptors’ Wins Produced for the 2011-12 season to date (taken from Patrick Minton’s wonderful The NBA Geek):

As noted above, Bargnani is now producing in the positive range, and will be out for the next couple of weeks. But the Raptors also added four new unproductive players during the offseason: Rasual Butler, Gary Forbes, Anthony Carter, and Jamaal Magloire. While Butler and Forbes figured to be unproductive even before the season started, Carter and Magloire projected to be close to average (at almost 37 and 34, respectively, age may have taken its toll on these players). Another unproductive player — Linas Kleiza — has returned from injury, and he is barely producing in the positive range. Still on the roster are Jerryd Bayless — who has never been productive — and Leandro Barbosa, who has not been productive since 2008-09 season.

On top of all that, DeMar DeRozan — who undeservedly made the All-Star Sophomore team last year and is regarded as one of the team’s promising young players — has reached a new low in productivity. At this point in the season, DeRozan is playing much worse than he did last year; as a matter of fact, DeRozan’s productivity has declined during each of his three years in the league:

DeRozan's career stats.

*Note: average SG/SF taken from 2011-12 season; wins pro-rated for an 82 game season

As you can see, DeRozan has really struggled this year. Every single statistical category is either worse off than last year or shows no change. His shooting percentages are particularly painful — especially given that he has not reduced his shot attempts — but DeRozan is also below average when it comes to rebounds, steals, turnovers, assists, and fouls. The most positive part of his game right now is his ability to get to the free throw line, and even there he is doing worse than he did last season.

All told, excluding the injured (and so far improved!) Bargnani, that leaves the Raptors with eight unproductive players. And already these eight players have cost Toronto about two wins. This number grows to about 7.5 wins if we replaced them with average players. This is a shame, because Toronto actually has some good players on its roster. Well, at least three: Jose Calderon, Amir Johnson, and Ed Davis.

There are unfortunately two problems with Toronto’s “big three”. The first is that Calderon is very often underrated. Calderon’s lone below average season was his rookie year. Since then he has never failed to produce fewer than five wins in a season, and at his best he was a top five player in the league. That hasn’t stopped the Toronto front office from almost pulling the trigger on moving him or kept fans from considering him overpriced. As the only good player pulling significant minutes on the Raptors it is not a good place to be with him so close to the trading block.

The problem with Johnson and Davis is they play the same position. This puts the Raptors in a bind as when one of their top players in on the floor it pretty much comes with the expense of another being on the bench. With so many bad players guaranteed to be on the floor at once it hurts to have one of the few bright spots of the team seated.

The Raptors will not be a good team with so few good players and so many bad players. It’s also unlikely that things will turn around for the Raptors if they remain willing to shop their good players and insist on holding onto their bad players. In fact, the Raptors’ problems are so plentiful that blaming a single player or fixing the team is not a simple task. If we wanted to point at the biggest culprit though, as it currently stands DeRozan is killing the Raptors. And for that the front office seems ready to pay him even more money.

-Devin (with a little help from Dre)

 

Did Toronto beat Phoenix because of the Bargnani effect?

Posted on by
21
via http://blog.raptors.com

Recently I read Tom Liston’s article over on Raptors Republic. While he completes the article with two comments that I agree with — that Gary Forbes should see more playing time and that DeMar DeRozan takes too many long jumpers — at the very top of the article he also repeated a tired old cliché:

The Raptors’ offence certainly suffers without their main weapon.

For those of you who haven’t paid much attention to the NBA’s lone Canadian team since Bosh (or even Vince Carter) left town, the “main weapon” Liston is referring to here is Andrea Bargnani, who has missed the past six games due to injury. The Raptors started the season with a 4-5 record, and when Bargnani went out, the team was sporting a 4-7 record. Without Bargnani, Toronto has gone 0-6. To some, this is a clear sign that the Raptors are playing worse without Bargnani. The fact that Bargnani just returned, scored over 30 and helped Toronto end its winless streak against Phoenix cements it for others.

To Liston’s credit, he doesn’t simply cite the win-loss record to support his claim; instead, after four games (and yes, he mentions the small sample size as well), Liston uses the following numbers:

Table 1: Selected Toronto Raptors team statistics with and without Bargnani after game #15

W/ Barngani W/O Bargnani Change
FG% 43.9% 39.7% -4.2%
3P% 32.1% 25.9% -6.2%
eFG% 47.7% 42.0% -5.7%
AST 21.3 16.3 -5.0
PTS 87.8 77.8 -10.0

In the four games without Bargnani, the team’s field goal shooting efficiency, assists, and point scored declined, suggesting that Bargnani has some sort of “Melo Effect” on his teammates. According to this theory, his teammates’ shooting, passing, and scoring suffer without him in the game to “draw attention” from defenders. As of today (January 23rd), Toronto has now played six games without Bargnani. Perhaps the numbers have changed?

Table 2: Toronto Raptors team statistics with and without Bargnani after game #17

W/ Bargnani W/O Bargnani Change
FGM 33.4 30.2 -3.2
FGA 76 78.3 2.3
FG% 43.9% 38.7% -5.2%
3PM 5.7 4.5 -1.2
3PA 17.7 14.7 -3.0
3P% 32.3% 38.7% 6.4%
FTM 15.4 16.2 0.8
FTA 19.3 22.8 3.5
FT% 79.7% 64.4% -15.3%
ORB 9.4 11.2 1.8
DRB 32.5 29.7 -2.8
TRB 41.9 40.8 -1.1
AST 21.3 17.8 -3.5
STL 6.4 6.7 0.3
BLK 5.3 4.5 -0.8
TO 16.1 14.8 -1.3
PF 24 25.5 1.5
PTS 87.8 81 -6.8

A few things have changed, but the story remains largely the same. While the team’s three-point percentage has increased significantly, the decrease in three-point attempts, decreased field-goal percentage, and increase in field-goal attempts still add up to poorer team shooting in Bargnani’s absence. Without Bargnani taking free throws, the team’s free-throw percentage has also plummeted, but this is simply because Bargnani’s free-throw shooting is no longer there to prop up the averages of his teammates. It would be hard to argue that Bargnani makes his teammates better at making free-throws, now wouldn’t it?

Beyond this, the team is also doing worse with respect to defensive rebounds, assists, blocks, and points. The team is doing slightly better with respect to offensive rebounds, steals, and turnovers. But basically, it’s pretty clear that the team has performed worse in the games without Bargnani.

But why is this? If the reason is due to Bargnani’s absence, we’d expect the rest of the team to be performing worse than they do when Bargnani is playing. The above table doesn’t address this, so we’ll have to come up with something else to check this out. A simple method is to take Bargnani’s numbers and add them into the numbers his teammates accumulated without him. Then we control for minutes played — I multiplied the team totals by 85.3% and then added in the stats Bargnani would’ve accumulated in six games (given his averages). Finally, we compare this to the numbers his team posted while he was playing. If the totals achieved without Bargnani are smaller than the ones achieved with Bargnani playing, then the team has done worse without him.

Table 3: Comparing the Raptors with a real and an imaginary Bargnani

W/ Bargnani W/O Bargs, Bargs added Change
FGM 33.4 33.8 0.4
FGA 76 83.4 7.4
FG% 43.9% 40.6% -3.3%
3PM 5.7 4.8 -0.9
3PA 17.7 16.1 -1.6
3P% 32.3% 30.1% -2.2%
FTM 15.4 18.9 3.5
FTA 19.3 25.6 6.3
FT% 79.7% 73.8% -5.9%
ORB 9.4 10.3 0.9
DRB 32.5 31.1 -1.4
TRB 41.9 41.4 -0.5
AST 21.3 17.1 -4.2
STL 6.4 6.0 -0.4
BLK 5.3 4.5 -0.8
TO 16.1 14.7 -1.4
PF 24.0 23.4 -0.6
PTS 87.8 91.4 3.6

Perhaps surprisingly, the numbers seem to suggest that Bargnani’s presence is having an effect. Without him, the team did worse with respect to all aspects of shooting (although free-throw attempts were up significantly), defensive rebounds, assists, steals, and blocks. On the positive side, the team did grab more offensive rebounds and had fewer fouls and turnovers, but that isn’t enough to make up for the negatives.So that’s it then: when Bargnani plays, he makes it easier for his teammates. It says it right there in the data!

Not so fast. Sure, it appears that Bargnani’s teammates performed worse without him. But why? Is it because of Bargnani, or perhaps something else? What about controlling for the quality of the opponents?

Table 4: Toronto Raptors opponent strength before and after Bargnani injury

W/ Barngani W/O Bargnani
Opp W% 46.3% 65.3%
Opp Exp W% 45.7% 65.3%
Opp Ortg 101.2 104.4
Opp Drtg 102.6 99.6
Opp Diff -1.4 4.8

As we can see, there is a far more obvious answer as to why the team performed more poorly without Bargnani: they faced better opponents. It’s a lot harder to maintain your performance level when the competition gets tougher. Accordingly, the Raptors’ stats suffered without Bargnani because they faced better teams than they did before Barngani was injured.

Does this completely rule out the existence of a “Bargnani effect”? No. But it certainly doesn’t support it. And this isn’t all that surprising, as the numbers from last year’s Carmelo Anthony trade don’t support the existence of such an effect either. As a matter of fact, given the data, it seems more likely to me that the Raptors actually performed better than we would’ve expected them to, given the quality of their opponents. And this is likely because — year after year — Andrea Bargnani is one of the least productive players in the league.

Of course his triumphant return last night against Phoenix may very well signal the possibility of a Bargnani effect. When we look at Phoenix’s record prior to playing Toronto, their team stats were:

  • 6-10 record
  • expected record 7-9
  • 101.4 Ortg
  • 103.7 Drtg

So the real Bargnani effect for this season appears to be that when he is healthy, the Raptors play weaker opponents, and when he is injured, they play more difficult ones. Of course, with only a small fraction of the season completed, we’ll have to wait and see if this trend holds up :)

- Devin

Is Kyrie Irving a better rookie than Chris Paul or LeBron James?

Posted on by
4
He’s no Chris Paul

Over at TrueHoop today, Justin Havens asked the following question: Irving better than Paul, LeBron as rookie?

Justin answered his own question by comparing each player’s rookie PER. Among all rookies who played a minimum of 2000 minutes, Irving’s PER is the eighth-best mark in the three-point era. Chris Paul’s rookie PER ranks 6th and LeBron’s 26th; therefore, after 13 games, it looks like rookie Irving is better than rookie James, but not as good as rookie Chris Paul.

The problem with PER is that, while it captures the perception of productivity, it doesn’t actually explain productivity very well. PER is heavily driven by scoring totals and is good for predicting All-Star selections, All-NBA teams, and MVP votes — but not wins. In order to answer the question of who was the better player as a rookie, we should use Wins Produced instead.

Player Age MP PER WP48 Wins*
Chris Paul 20 2808 22.1 0.253 14.8
Kyrie Irving 19 368 21.8 0.123 5.9
LeBron James 19 3122 18.3 0.069 4.5

* Wins for Irving extrapolated for an 82 game season

As we can see, LeBron’s rookie season doesn’t compare to Paul’s and Irving’s, and that is true whether we use PER or Wins Produced. But Paul and Irving’s rookie seasons have very similar PERs and very different WP48s; according to PER, Paul and Irving had similar seasons; according to Wins Produced, Paul was roughly 2.5 times better than average, whereas Irving has only been about 1.2 times better than average. Wins Produced says that Paul’s rookie season was about twice as productive as Irving’s rookie season to date. Does that seem a little far-fetched? Well, let’s ignore Wins Produced for a moment and let the numbers speak for themselves:
Stat*  Chris Paul Kyrie Irving Average* PG
PTS 21.5 30.0 19.5
FGA 16.2 23.3 16.5
FTA 7.9 6.7 4.5
TS% 54.6 57.1 52.9
TRB 6.8 6.0 4.8
STL 3.0 1.4 2.0
TOV 3.1 5.9 3.7
Net Poss 6.7 1.6 3.1
AST 10.4 8.9 8.3
BLK 0.1 1.0 0.3
PF 3.7 4.4 3.3

* Stats are per 48 minutes where applicable; average PG numbers from the 2011-12 season

Irving is certainly scoring well — he’s taking almost seven more field goal attempts and more than two more free throw attempts than the average point guard. Not only that, he’s also making these shots at a very good rate. But another important part of basketball is retaining possession of the ball, and it’s something that Irving has struggled with so far this season. Irving’s net possessions (calculated by adding rebounds and steals and subtracting turnovers) are about half the mark of an average point guard. This is mostly driven by his large number of turnovers. Irving is also below average with respect to steals and fouls.

Chris Paul, on the other hand, was also a good scorer — although not as good as Irving has been so far this season. But Paul’s main strength was his ability to gain possessions; his net possessions were more than double that of an average player and more than four times better than Irving’s have been so far. By giving his team extra possessions, Paul gave his team additional opportunities to score points, and as it turns out, this is very valuable. Paul was also above average in every other statistical category other than blocks and fouls; in particular, he was very good at getting assists and drawing fouls.

That is not to say that Irving is a bad player. Irving looked very productive coming out of College, and as I said before, Irving has been about 1.2 times as productive as an average NBA point guard. Not many rookies perform better than average, so this is a pretty good achievement. As well, young NBA players can improve drastically across their first few years in the league, so Irving could still improve on his production in the near future. But let’s not get out of control and start comparing him to Chris Paul just yet.

- Devin

Why Spencer Hawes could be the next LeBron James

Posted on by
1

Prior to this season, Spencer Hawes had not been a very productive player. He was below average in most statistical categories. And yet — just this past weekend — I added Spencer Hawes to my Fantasy Team. Why? To begin to answer this question, let’s first take a look at his career Wins Produced numbers:

Year Age Minutes WP48 Wins
2007-08 19 931 -0.047 -0.9
2008-09 20 2259 -0.041 -1.9
2009-10 21 1904 0.003 0.1
2010-11 22 1718 0.011 0.4
2011-12 23 133 0.388 1.1

Hawes has been shockingly and uncharacteristically productive so far this season. As a matter of fact, Hawes has produced more wins in his first 133 minutes this season than he did in all of his previous minutes combined. Given his career history, is this small sample size a sign of things to come, or should we expect significant regression to the mean in the coming weeks?

Truthfully, I must say that it’s more likely that Hawes will regress back down to his more typical production levels. But there is definitely a chance — and it’s hard to say how big of a chance — that Hawes has actually improved. This is because Hawes is still only 23 years old, and young players can improve significantly in their first few years in the league. Three recent examples of this are LeBron James, Kevin Durant, and Derrick Rose:

Year Age Minutes WP48 Wins
2003-04 19 3112 0.069 4.5
2004-05 20 3388 0.228 16.1
2005-06 21 3361 0.211 14.8
2006-07 22 3190 0.196 13.0
2007-08 23 3027 0.257 16.2
Year Age Minutes WP48 Wins
2007-08 19 2768 0.009 0.6
2008-09 20 2885 0.126 7.6
2009-10 21 3239 0.219 14.8
2010-11 22 3038 0.155 9.8
2011-12 23 221 0.260 1.2
Year Age Minutes WP48 Wins
2008-09 20 3000 0.087 5.4
2009-10 21 2871 0.090 5.4
2010-11 22 3026 0.161 10.2
2011-12 23 174 0.243 0.9

Each of these players had/has improved significantly by age 23. Also of note is that each of these players played heavy minutes leading up to age 23. While Hawes’ jump would seem out of place amongst these other players’ jumps, Hawes played significantly fewer minutes than the other players. In fact, Hawes played just over half of the minutes that LeBron James did, 57% of the minutes that Durant played, and 77% of the minutes that Derrick Rose played (and in one more season than Rose). It’s not out of the question that Hawes needed more on court experience before he was ready to become more productive.

Of course, there are also examples of players who never become productive. Jason Collins, Darko Milicic, and Andrea Bargnani are examples of such players:

Year Age Minutes WP48 Wins
2001-02 23 1407 0.085 2.5
2002-03 24 1900 0.076 3.0
2003-04 25 2220 0.077 3.5
2004-05 26 2542 0.046 2.5

Darko:

Year Age Minutes WP48 Wins
2003-04 18 159 -0.216 -0.7
2004-05 19 254 -0.196 -1.0
2005-06 20 767 0.063 1.0
2006-07 21 1913 0.078 3.1
2007-08 22 1663 -0.033 -1.2
2008-09 23 1034 0.060 1.3
2009-10 24 685 0.015 0.2
2010-11 25 1686 -0.052 -1.8
2011-12 26 98 0.087 0.2
Bargnani:
Year Age Minutes WP48 Wins
2006-07 21 1629 -0.013 -0.4
2007-08 22 1861 -0.096 -3.7
2008-09 23 2453 0.002 0.1
2009-10 24 2799 -0.007 -0.4
2010-11 25 2353 -0.126 -6.2
2011-12 26 182 0.011 0.0

Unfortunately, this type of player is far more common than the previous group. And while there are a few cases of players improving after the age of 26, this doesn’t happen very often; if Hawes is ever going to step up his play, it will almost certainly be sometime during the next three seasons. I added him to my fantasy team on the off chance that his early play is a sign that it happens this year.

- Devin

Toronto will be terrible.

Posted on by
1

The following post uses the Wins Produced metric. During the long offseason Dave Berri took a little time to tweak it a little. You can find all of the numbers for last decade or so here, as well as a handy tutorial!

It’s opening day as I write this, so now’s a good time to get my Raptors prediction out there. Numbers first, with an explanation to follow:

Player Pos. Pred. WP48 Pred. Mins Pred. Wins Lockout Adj.
Ed Davis PF 0.190 2321 9.19 7.39
Jose Calderon PG 0.168 2033 7.13 5.74
Amir Johnson PF 0.196 1615 6.59 5.30
Aaron Gray C 0.123 700 1.80 1.45
DeMar DeRozan SG 0.031 2800 1.80 1.45
Jamaal Magloire C 0.156 423 1.37 1.11
James Johnson SF 0.032 1000 0.66 0.53
Gary Forbes SF 0.027 1000 0.56 0.45
Anthony Carter PG 0.083 300 0.52 0.42
Leandro Barbosa SG 0.010 1136 0.24 0.19
Jerryd Bayless PG 0.005 1603 0.16 0.13
Rasual Butler SF -0.004 1100 -0.08 -0.07
Linas Kleiza SF -0.034 836 -0.59 -0.47
Solomon Alabi C -0.189 300 -1.18 -0.95
Andrea Bargnani C -0.072 2513 -3.77 -3.04
Total 0.060 19680 24.40 19.64

Last year’s Raptors finished with 22 wins, and if this year’s team played a full 82 game season, I’d put them at 24-25 wins. As this season will only be 66 games, I have to adjust that total accordingly, and I get about 19-20 wins.

Individually, the team should be led by three players — Ed Davis, Jose Calderon, and Amir Johnson — who should account for the vast majority of the team’s wins. I expect all other players on the team’s roster to account for only about one win.

This season’s new acquisitions — Aaron Gray, Jamaal Magloire, Gary Forbes, Anthony Carter, and Rasual Butler — will only add about three wins. Gray and Forbes are the most promising, but probably won’t see very much playing time; Magloire and Carter are formerly decent veterans who are old, have been declining for years, and could drop off a cliff at any moment; and Butler should end up being the least productive of the bunch (older and never productive). Thankfully, each of these newly acquired players is cheap and on a one-year deal, so the long term ramifications of these signings will be minimal.

As far as departed players go, Sonny Weems will not be missed. After a promising 2009-10 campaign, Weems regressed significantly last season. It’s a good thing he’s now stuck overseas. The other Raptor stuck overseas is Joey Dorsey. Dorsey was quite productive in limited minutes last season — as he has been for his entire career — so it’s a shame he won’t be back. The other two players the Raptors should miss are Julian Wright and Reggie Evans. Wright was also productive in limited minutes last season, but with the arrival of James Johnson, there was very little chance of him returning. Hopefully he gets another chance with an NBA team in the future. Evans was excellent last season, but he was limited to about 800 minutes due to injuries. While he is a productive player, he is older, injury prone, and plays the same position as two of the team’s three best players. I’m glad he ended up on a team that’s likely to be playoff-bound and will have a use for him.

(Brief aside: it’s also a good thing that Chris Kaman is no longer on the Clippers, because that might have been awkward)

Yes, the Raptors were relatively stacked at the power forward position last season. As I noted above, two of the team’s “modest three” play this position. If I arrange the numbers by position, we can see the impact this has even more clearly:

Pos. Pred. WP48 Pred. Mins Pred. Wins Lockout Adj.
C -0.022 3936 -1.78 -1.43
PF 0.192 3936 15.78 12.70
SF 0.007 3936 0.56 0.45
SG 0.025 3936 2.04 1.64
PG 0.095 3936 7.81 6.29
Total 0.060 19680 24.40 19.64

Power forward will continue to be the team’s strength. The point guard position will be roughly average — a relative strong point for Toronto. Shooting guard, small forward, and centre will all be around a WP48 of zero; there will be a small number of wins produced at SG, but most of these wins will be offset by the losses created by the team’s centres (specifically the team’s least productive player).

I will conclude the post with my standard disclaimer for making pre-season predictions: I do not pretend to have exceptional forecasting abilities. My prediction method involves a three year weighted average of both WP48 and minutes played, with adjustments made to minutes to make the minutes add up properly. I can’t predict injuries, trades, or playing time, nor can I predict players who have uncharacteristic seasons. But I’m sure that you knew that already.

- Devin