Breaking Down: Bucks SF Carlos Delfino

Periodically I am going to calculate the five year and the three year Marginal Win Scores for various Milwaukee Bucks players, and I am going to break down their production into each of the three major statistical categories over those time periods.  Then I am going to try to predict how many wins they will produce next season.

The three main categories I referenced will all be based upon a comparison of the player against the counterpart opponents who were on the floor at the same time.  The three main categories of comparison are: (1) Net Effective Scoring (Pts – FGAs – .5FTAs);  (2) Net Possessions Created (Rebs + Stls – TOs); and (3) Net “Helpful” Stats (.5Ass + .5Blks – .5PFs).

After I have broken down the categories, I will determine the player’s 3 and 5 season: (1) Mean and Median Marginal Win Score; and (2) the player’s Marginal Win Score “Ceiling”/”Floor”/”Midground”. 

When I calculate his Ceiling, that is a calculation of what his MWS would be if he produced the best numbers he has produced over the period in each of the three main categories, all in the same season.  The Floor is just the opposite.  The Midground is the dead center between the player’s time period Ceiling and Floor.

Carlos Delfino

Delfino spent the first 3 seasons of his NBA career in Detroit.  There he played mostly shooting guard, and even some point guard.  He did not play a lot of small forward as he did last season in Milwaukee.  He was pretty bad his first two seasons in Detroit (-1.32; -1.24) and then suddenly he was extremely good in his final Detroit season (+2.35).  Delfino then moved to Toronto where he spent most of his time at shooting guard and 10% of his time at small forward.  In Toronto his numbers slipped back, but just below the .500% mark (-0.20).  After that Delfino spent a season in Europe and on his return he was traded to the Bucks, where he posted an excellent MWS48 (+0.96).


Obviously, Delfino has been volatile over the course of his career, but in the past three seasons he has been volatile on the correct side of .500%, which is a good sign for the Bucks.

Delfino is mainly a Possession Creator.  Generally, his Effective Scoring is negative.  Last season with the Bucks was his first positive Scoring season, but it was only slightly positive.  Here are the exact scoring numbers:

Net Five Season Effective Scoring (-3.12; -4.55; -0.21; -2.42; +0.39).
5 season Eff Scoring “Mean”: -1.98  “Median”: -2.42
3 season Eff Scoring “Mean”: -0.75 “Median”: -0.21

As you can see, his effective scoring over the full five seasons has been pretty subpar, and it tilts negative, as the Median is lower than the Mean.  But, it has improved over the last three seasons, with both the Mean and the Median coming much closer to the .500% mark, and with the Median tilt shifting toward positive.

You will also notice that in the two seasons that his scoring was at its best he recorded superb Marginal Win Scores.  You will also notice that in his 4th season, his Toronto season, he was able to record a decent Marginal Win Score despite poor Effective Scoring.

That’s because his Possession Creation has always been good but really picked up in Seasons 2-4.


Here are the five season Possession numbers for Delfino:

5 season Net Possession Creation: (+0.85; +2.87; +5.89; +3.25; +0.83)
5 season Net Poss Creation “Mean”: (+2.74) “Median”: (+2.87)
3 season Net Poss Creation “Mean”: (+3.32) “Median”: (+3.25)

As you can see, he has been a very impressive  Possession Creator.  And, while I don’t show the numbers here, I can say he is good at both rebounding and creating steals, and is generally careful with the ball.  Delfino’s Mean and Median are pretty close together, and each have risen over the 3 season period compared to the 5 season period, which is good for his outloo


Delfino isn’t so hot in the Helpful Act category, except for last season when he really decided to trust the pass.  His marginal assist totals were unusually high.  That was partly due to his passing, partly due to his outstanding ball pressure on the players he defended.  And that was good, because heavy marginal assist totals are about the only way he can stay positive in this area.  Lets face it, he’s not a shot blocker.  And while he isn’t an exorbitant fouler, his offensive style does not lend itself to drawing fouls on his opponents.

Here are his numbers in this category:

5 Season Net “Helpful” Statistics (-0.37; -0.80; -0.98; -1.24; +0.71)
5 Season Net Helpfuls “Mean”: (-0.54); “Median”: (-0.80)
3 Season Net Helpfuls “Mean”: (-0.50); “Median”: (-0.98)

As you can see, like a lot of players, Delfino’s numbers in this area are pretty steady.  Unfortunately they are steady on the wrong side of the zero mark, so they don’t tend to help him create wins above .500%, they tend to slightly drag the effort down.


Okay I’m going to do two things here.  The first is simple.  I first take Delfino’s Marginal Win Score per 5 seasons and per last 3 seasons and get the average for each.  Then I find Delfino’s Weighted MWS in the 5 and 3 season intervals by weighting the seasons according to their percentage of the overall time period.  The corresponding “Personal Winning Percentage” follows each MWS48 calculation.

5 season MWS48 “Mean”: (+0.11) (.521%)
3 season MWS48 “Mean”: (+1.03) (.677%)
5 season Weighted MWS48: (+0.52) (.590%)
3 season Weighted MWS48: (+0.90) (.655%)
MARGINAL WIN SCORE Median and Categorical Median

Alright now what I’m going to give is Delfino’s MWS48 median, or the number right in the middle of his seasonal MWS48s, and then — to try to wash out the influence of any one of the three categories — I give his MWS48 as calculated using the Median from each of the three separate statistical categories.

5 season Median MWS48: (-0.20) (.468%)
3 season Median MWS48: (+0.96) (.665%)
5 season Category Medians MWS48: (-0.17) (.474%)
3 season Category Medians MWS48: (+1.03) (.677%)

As you can see, there’s not much difference between the two.  If you weight your analysis to consider only the past three seasons, as you probably should because he’s played more in that time, then the outlook has to be good for Delfino.  But now let’s consider best and worst season scenarios.


I explained this category above so I will just spill the numbers:

5 season MWS “Ceiling”: (+3.49)
5 season MWS “Floor”: (-2.48)
5 season MWS “Midground”: (+0.50)
3 season MWS “Ceiling”: (+3.49)
3 season MWS “Floor”: (-1.41)
3 season MWS “Midground”: (+1.04)

As you can see, his Ceiling is the same over the two time periods, from which you can deduce that he’s had his best performance in each of the three categories in the last three seasons.  By contrast, his Floor rises over the 3 season analysis, from which you can deduce that at least some of his worst categorical performances came in his first two seasons.  Both positive things for Delfino.


If you look at the numbers, they may be volatile, but they keep coming back to this range:  somewhere between +0.52 and +1.03, tilting toward the latter.

I’m going to go ahead and project Delfino to produce a (+0.95) Marginal Win Score per 48 next season.  I produced this number by averaging each of the MWS48 numbers above plus his 5 and 3 season “Midgrounds”.

As for Game Responsibilities, Delfino will be 28 years old without any major injuries.  Plus, he seemed to cement himself in the starting lineup.  But I don’t know if the Bucks will draft a small forward, which could cut into his playing time.  I will therefore project him as a low to medium minute starter who will play 2400 minutes, which is about 9.9 game responsibilities.

With the aforementioned MWS48 of (+0.95) comes the corresponding winning percentage of .663%, so I project Delfino’s 2010-11 season to go as follows:

2010-11 prediction:

CARLOS DELFINO….6.6 wins and 3.3 losses


2 Responses to “Breaking Down: Bucks SF Carlos Delfino”

  1. Greg Magarian Says:

    Fantastic analysis. I’m still a basketball novice, but I’m pretty well versed in statistical analysis from baseball, so this kind of work really helps me learn. Thanks for doing it.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: