FanPost

Mid-Season Evaluation of Damian Lillard Projections

Earlier this year, around the start of training camp I published a number of fan posts regarding the best comparable NBA players to the Trail Blazers rookies, and used those comparable players to project each players stat line for their first 3 NBA seasons. It is now 41 games into the season, so I thought I would do a recap of those projections. I will do a recap of those projections for Damian Lillard and Meyers Leonard only at this point, simply because the other rookie’s statistical performance is based on very small sample sizes. This post will focus on Damian Lillard, and I will do a separate follow up post on Meyers Leonard. Just for discussion sake I will also make a 3rd post detailing the 10 best compass for a number of the top picks in the 2012 NBA draft, which I found quite revealing.

Here is a link to the Damian Lillard Fan Post (Damian Lillard Comparables), and also the Meyers Leonard Fan Post (Meyers Leonard Comparables).

At the time of these posts my data base of college seasons only included the seasons from 2000-01 through 2011-12. Since then I have expanded the data base to include every NBA player who debuted in the NBA after the 1989-1990 season, excluding players drafted out of high school, players who did not play at an American university (such as players from Europe), and any US college players who did not have complete statistical information for their college careers. I built my database using the data available from Draft Express, Basketball Reference, and College Basketball Reference, and after finalizing the database I found some inconsistencies in some of the data, usually relating to formulas for advanced stats, but a few additional inconsistencies as well. I have "fixed" those inconsistencies now, though I am sure there are still some I haven’t found. Because I have a new expanded database, and I am more confident that all the entries are consistent in their formula calculation, so I decided to redo the comparable analysis for Damian Lillard and Meyers Leonard.

Damian Lillard

Here is the list of the initial 7 best comparable players for Damian Lillard.

Player

Season

Rodney Stuckey

So

George Hill

Sr

Ben Gordon

Jr

Ronnie Price

Sr

Marcus Thornton

Sr

Stephen Curry

Jr

Randy Foye

Sr

This was the projected stat Line for Damian Lillard’s first 3 years.

Season

Shooting Statistics per 36 Minutes

FG

FGA

FG%

3P

3PA

3P%

FT

FTA

FT%

TS%

eFG%

1

5.83

13.54

43.06%

1.47

3.91

37.48%

3.90

4.53

86.0%

56.9%

48.5%

2

6.14

13.63

45.05%

1.64

4.11

39.90%

3.25

3.83

84.9%

58.5%

51.1%

3

6.07

13.73

44.24%

1.59

4.26

37.33%

4.05

4.55

89.0%

58.7%

50.0%

Season

Accumulation Stats per 36 Minutes

ORB

DRB

TRB

AST

STL

BLK

TOV

A/TO

PF

PTS

1

0.84

3.48

4.23

4.07

1.23

0.29

2.33

1.78

3.23

17.83

2

0.63

3.19

3.70

4.58

1.18

0.19

2.12

2.22

3.03

18.04

3

0.65

3.03

3.61

4.52

1.24

0.28

2.30

1.97

2.66

18.62

Season

Usage Stats

ORB%

DRB%

TRB%

AST%

STL%

BLK%

TOV%

USG%

1

3.1%

10.7%

6.9%

18.0%

2.0%

0.6%

13.9%

23.0%

2

2.0%

10.0%

6.0%

20.0%

2.0%

0.0%

13.0%

23.0%

3

2.4%

9.4%

5.8%

20.3%

2.0%

0.6%

13.8%

23.0%

Season

Advanced Measurements

PER

ORtg

DRtg

WS

WS/48

1

15.96

102.6

107.0

2.95

0.067

2

16.87

105.0

108.6

3.75

0.080

3

17.58

106.7

109.1

4.18

0.084

NEW COMPARABLE MODEL

Once I expanded my database I decided to recalculate the list of comparable players to include players prior to 2000-2001, and to expand the number of comparable players from 7 to 10. One of the limitations on the data I have is that there is very little if any physical attribute data for players prior to the 2001 draft. This forced me to find a more inclusive approach to determining the list of best comparable players. For players drafted after 2000 I evaluate both physical and statistical comparability, and for players prior to the 2001 draft I consider only statistical comparability. Whenever possible I try to compare players of the same collegiate seasons, but I am open to comparing players from different classes. Wherever reasonable I try to avoid comparing Freshmen to Seniors and Juniors, but I do compare Sophomores to Freshmen, Sophomores and Juniors.

Here is the list of the new 10 Best Comparable Players to Damian Lillard

Keith McLeod

Sr

J.J. Redick

Sr

Toney Douglas

Sr

Devin Harris

Jr

Eric Maynor

Sr

George Hill

Sr

Jameer Nelson

So

Marcus Thornton

Sr

Delonte West

Jr

Ben Gordon

Jr

From this list of 10 comparable players I generated a new projection for Damian Lillard for his first 3 seasons. The new projection is similar to the old projection, though this new projection does project a little higher production from Lillard.

Below is a raw comparison of Damian Lillard’s statistical performance through 41 games, with the new 10 comparable player first year projection.

Shooting Statistics Per 36 Minutes

Player

FG

FGA

FG%

2 Pt %

3P

3PA

3P%

FT

FTA

FT%

Damian Lillard (Actual)

6.00

14.40

41.7%

46.4%

2.10

6.00

35.0%

2.90

3.40

85.3%

Damian Lillard (Proj)

5.61

13.14

42.7%

45.8%

1.59

4.37

36.4%

3.11

3.73

83.5%

Accumulation Statistics per 36 Minutes

Player

ORB

DRB

TRB

AST

STL

BLK

TOV

PF

PTS

Damian Lillard (Actual)

0.60

2.70

3.30

6.10

1.00

0.30

2.80

1.90

17.10

Damian Lillard (Proj)

0.80

3.15

3.88

4.14

1.24

0.27

2.14

3.78

16.47

Usage Rates

Player

ORB%

DRB%

TRB%

AST%

STL%

BLK%

TOV%

USG%

A/TO

PPR

Damian Lillard (Actual)

1.80

8.90

5.30

29.90

1.40

0.50

15.00

23.70

2.18

3.41

Damian Lillard (Proj)

2.57

9.88

6.19

18.36

1.78

0.55

13.34

22.27

1.98

1.66

Advanced Statistics

Player

PER

TS%

eFG%

ORtg

DRtg

WS

WS/48

EFF/36

FT/FGA

3PA/FGA

Damian Lillard (Actual)

16.50

53.7%

49.4%

107

110

5.83

0.097

16.10

23.6%

41.7%

Damian Lillard (Proj)

15.13

55.3%

48.7%

104

107

5.47

0.091

15.70

28.4%

33.2%

At a cursory glance I am proud to say that the projection was very accurate, especially when you consider that there are 39 different measurements. There are two statistical categories where the model was off a reasonable amount, and those were assists, and 3 point attempts.

The projection indicated 4.14 assists per 36 minutes, and Lillard is nearly 2 assists per 36 minutes higher than that. This wasn’t surprising to me, I was not all that comfortable with the initial projection of 4.14 A/36 Min; I felt his assist rates would exceed that level, but I chose to remain skeptical of Damian Lillard’s ability to pass effectively until he had shown otherwise. One of the results I expected was that if Lillard’s assist increased that it was reasonable to assume that his turnovers would also increase. Overall Lillard’s turnovers increased as well, and his A/TO rates are very similar to his projection.

The Lillard’s 3 point attempts are about 1.4 attempts per 36 minutes higher, but his 3Pt% is almost exactly correct. The difference in points per 36 minutes of 0.63p/36 minutes is due almost exclusively to a slightly higher actual usage rate than his projected usage rate, which is very much team and situation dependent.

Accuracy of Projection Tool

From this I wanted to see how accurate this projection actually is. To do this I used a model similar to what I did to find my initial set of 10 best comparable players. I found the degree of similarity between Damian Lillard’s actual performance and his projected performance. In this analysis I found the degree of similarity for Damian Lillard’s projected rookie season to his actual rookie year to be 95.08% similar.

This 95.08% similarity must be understood in context. Recognize that the statistics I am comparing are of some very basic types;

1.) There are accumulation statistics like Points, Assists, Turnovers which are per 36 minutes, but all of these are on an identical scale, Per 36 Minutes. Because of that, there will naturally be some degree of similarity between players.

2.) There are a number of ratio’s, FGM to FGA, FTA to FTM, Ast to TO etc. There is a natural degree of similarity when using ratios.

3.) There are complex advanced stats that combine a large number of statistics into a single number, such as PER, Eff/36 and WS. Two players can have an identical PER, but have a significant disparity in the various components.

4.) Usage rates are a measure as a % of available. Rebound % is the % of all rebounds available that the player got, which puts all 10 players on the floor into a similar context.

Because of this, a raw degree of similarity of 95.08% may not reveal how accurate the projection is, it is necessary to place this into a context, because all players will have some degree of similarity. How I did this was to calculate this degree of similarity for every player season from the 1990-91 season through the 2011-12 season; a total of 11,690 distinct individual seasons. I then used the same methodology of a player’s performance relative to a range from highest to lowest for each statistical category, to place each player into a percentile of the best overall performance, and found the degree of similarity of all 11,690 players to Damian Lillard so far this season. I then ranked all of those seasons, to understand how accurate the 95.08% is relative to other players. Recognize that I am not comparing any individual’s raw stats to Damian Lillard’s raw stats, but rather relationship to a standard of 0 to 100%. I could not compare individual players raw stats, because I would get division by zero errors, meaning that all players would have an infinite degree of dissimilarity to the target player for that stat category (for instance 3 pointers made, if the target player did not make a 3 point shot).

The median degree of similarity between all the players and Damian Lillard is 84.5%, meaning 5,845 had a degree of similarity of less than 84.5%. The reason why the median is 84.5% and not 50% is because of the 4 reasons detailed above. Overall there were 78 player seasons over the last 23 seasons that were more similar to Damian Lillard’s actual performance than the projection; basically the projection was in the 99.33 percentile of similarity. Based upon all of that I would say the methodology was tremendously accurate at least in the case of Damian Lillard. Damian Lillard though has the overall largest sample size of all rookies so far this season, so it should be more accurate than most. Regardless, the tool did project Damian Lillard’s actual stats extremely accurately.

Just as a point of discussion here are the twenty seasons that are the most statistically similar to Damian Lillard’s actual performance so far.


Player

Similarity

Age

Tm

Season

1

Damian Lillard (Actual)

100.00%

22

POR

2012-13

2

Mike Bibby

96.98%

28

SAC

2006-07

3

Mo Williams

96.80%

27

CLE

2009-10

4

Nick Van Exel

96.64%

24

LAL

1995-96

5

Terry Porter

96.54%

30

POR

1993-94

6

Jason Terry

96.54%

25

ATL

2002-03

7

Jason Terry

96.46%

26

ATL

2003-04

8

Jose Barea

96.43%

26

DAL

2010-11

9

Damon Stoudamire

96.39%

31

POR

2004-05

10

Aaron Brooks

96.33%

25

HOU

2009-10

11

Joe Johnson

96.31%

24

ATL

2005-06

12

Nick Van Exel

96.25%

23

LAL

1994-95

13

Mo Williams

96.07%

28

LAC

2010-11

14

Mike Bibby

96.05%

29

TOT

2007-08

15

Vernon Maxwell

96.00%

27

HOU

1992-93

16

Nick Van Exel

95.99%

31

DAL

2002-03

17

Ricky Davis

95.98%

27

MIN

2006-07

18

Mo Williams

95.98%

23

MIL

2005-06

19

Jerryd Bayless

95.97%

23

TOR

2011-12

20

Tim Hardaway

95.93%

32

MIA

1998-99

21

Joe Johnson

95.88%

26

ATL

2007-08

X
Log In Sign Up

forgot?
Log In Sign Up

Please choose a new SB Nation username and password

As part of the new SB Nation launch, prior users will need to choose a permanent username, along with a new password.

Your username will be used to login to SB Nation going forward.

I already have a Vox Media account!

Verify Vox Media account

Please login to your Vox Media account. This account will be linked to your previously existing Eater account.

Please choose a new SB Nation username and password

As part of the new SB Nation launch, prior MT authors will need to choose a new username and password.

Your username will be used to login to SB Nation going forward.

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

Join Blazer's Edge

You must be a member of Blazer's Edge to participate.

We have our own Community Guidelines at Blazer's Edge. You should read them.

Join Blazer's Edge

You must be a member of Blazer's Edge to participate.

We have our own Community Guidelines at Blazer's Edge. You should read them.

Spinner.vc97ec6e

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9347_tracker