Some Insight Into Individual Defense - 3019
Jul 18, 2015 11:59:03 GMT -6
Heynong Man, MJ, and 2 more like this
Post by 20s Navidad on Jul 18, 2015 11:59:03 GMT -6
One of the problems that has long plagued the league is that it is difficult to judge how well an individual player plays defensively. All we really have to go by is their defensive letter grade and their steal and block numbers. Everyone realizes that there is more to defense than steals and blocks, but because that is all we have available, it is often how we compare players' defense. Another way to compare is their letter grade, but as we have long suspected, and eric's work has shown us recently, letter grades can be misleading. Especially in the case of defensive letter grade due to inexplicable positional rating caps where a player's overall defensive rating stops increasing after a certain point for each attribute (even if the attribute keeps going up). We know 95 perimeter defense is better than 80, but at times the overall defense grade does not reflect that.
An alternate method that many GMs use to attempt to judge their players defensive abilities is to look at their box scores and see how their opposition performs offensively. If I want to judge Mike Conley's defense, I can look at the 76ers box scores and see how the PGs he faces off against perform. This is a very effective way of determining how good of defense Conley plays, however it is tedious and anecdotal if you are only looking at certain box scores.
Thanks to the proliferation of box score data collection (hat tip to eric for collecting the 3019 box score data), we can now look at all the box score data together in an easy manner. I have decided to total up the box score results for each team's opposition in 3019 to take a look at how players really perform on the defensive end of the floor. This article summarizes the results.
Before we get into it, let's address what the data is telling us and known flaws in the data. The data comes from the box scores that we are all familiar with. We know that players are only listed as one position in the box score even if they get minutes at multiple positions during the game. The position listed is the position we are using for these purposes. So a potential flaw is Antonio McDyess plays 26 minutes against the Nets, 16 at PF and 10 at C. The box score lists him as PF 26 minutes. All of his production is going to count against the PF defense for the Nets, even though for 10 of those minutes he was being defended by a C. This could cause some small discrepancies, but over the course of an entire season, the data is still very useful and should tell us what we want to know.
Another thing to consider is that this is entire game data so it is not only starter's defense that is being measured (although they play a large majority of the minutes). So when I am looking at the Thunder's PG defense, I am looking at 38 minutes of Aaron Harrison's defense and 10 minutes of Frank Johnson's defense. With that said, let's take a look.
Looking at the PG numbers yields a few surprises. Chris Duhon earns another feather in his undervalued cap as he lead the Kings to the best defense against PGs in TMBSL in 3019. Scoonie Murray has been thought to be a valuable PG due mainly to his passing and defensive abilities and last year the Hornets were one of two teams that held opposing PGs to a true shooting percentage of less than 50%. Across the league, PGs scored on average 25 points per game on .535 TS%. Keep in mind this is the PG positiion for the entire game, not just the starter. Dutch Burch and his low defensive grade performed well for the Clippers as they came in third. This goes to show us that simply looking at letter grade does not tell the full story at all when it comes to defense.
The SG results are shocking. The 76ers were by far (a full 2% over 2nd place) the best at defending opposing SGs in 3019. Their starting SG? Pee Wee Kirkland. Pee Wee had a B- defensive grade and has long been touted as a guy who can fill it up on offense but is a slouch on the defensive end of the floor. Well, he was doing something right last season. This data goes to show us once again that simply looking at defensive grade can be misleading and is not a direct indicator into how their opponent will perform. Overall, SGs had the highest TS% in TMBSL in 3019 at 55.2%. The SG position averaged 24.5 points per game. Clyde Drexler and his mega contract in Charlotte played below average SG defense. Quentin Richardson shows that he earned his most improved player award by playing stellar defense as well.
One of the things I was curious about with the "big men" who get moved to SF is how that affects them defensively. It looks like they can still be quite successful defensively. The Clippers and Bullets both played former big men at SF and they came in at 1-2 in the rankings. Al Horford and the Warriors ranked 6th. This goes to show that at least some of the big men who get shifted down to SF are still quite effective on the defensive side of the court as well. This puts me more firmly in favor of the new position rules that soup is enforcing. Across the league, SFs averaged 22 points per game on .538 true shooting percentage. Another interesting thing I noted was that the Champion Thunder had below average defense from their SF spot.
Unlike SG, the top of the PF rankings are no surprise at all. It is pretty widely recognized that Jermaine O'Neal and Shareef Abdur-Rahim are the best two defensive PFs in TMBSL and the rankings back that up. Some may be slightly surprised that the Nets rank ahead of the Thunder, but they are close and I liked the Nets backup bigs better than the Thunder last year. Kenyon Martin played strong defense for the Warriors which helps to make up for his subpar offense and the Kings split of Catfish/Myles Turner rated highly as well. On average, PFs scored 23.8 points per game on a true shooting percentage of .514.
So much for that Hasheem Thabeet defensive big man theory, eh? Everyone was well aware that he was very inefficient on offense, but most assumed that the big man at least was a plus on the defensive end of the floor. Well, in 3019, the Hawks were the 3rd worst team in TMBSL at defending the center position and he was playing a large majority of their minutes there. At least he can rebound. The Clippers reign supreme once again. The rest of the top ten is not that surprising either - Camby/Catfish for the Kings, Chet Noe for the Bucks, Vlade and the Thunder, David Robinson and the Sixers, Dwight for the Hornets, Ben Wallace in Brooklyn, Harrell in Phoenix, Oden in Minnesota and Okafor on the Warriors. Overall, centers are the least efficient scorers in TMBSL (which only confirms what we were already aware of). They averaged 19.4 points on a true shooting percentage of .490.
I hope you guys find this information enjoyable and beneficial when evaluating players as it took a long time to compile. I will be interested to see if the same players consistently land atop these rankings or if it is more of a year by year thing so I can keep doing this if people enjoy it.
An alternate method that many GMs use to attempt to judge their players defensive abilities is to look at their box scores and see how their opposition performs offensively. If I want to judge Mike Conley's defense, I can look at the 76ers box scores and see how the PGs he faces off against perform. This is a very effective way of determining how good of defense Conley plays, however it is tedious and anecdotal if you are only looking at certain box scores.
Thanks to the proliferation of box score data collection (hat tip to eric for collecting the 3019 box score data), we can now look at all the box score data together in an easy manner. I have decided to total up the box score results for each team's opposition in 3019 to take a look at how players really perform on the defensive end of the floor. This article summarizes the results.
Before we get into it, let's address what the data is telling us and known flaws in the data. The data comes from the box scores that we are all familiar with. We know that players are only listed as one position in the box score even if they get minutes at multiple positions during the game. The position listed is the position we are using for these purposes. So a potential flaw is Antonio McDyess plays 26 minutes against the Nets, 16 at PF and 10 at C. The box score lists him as PF 26 minutes. All of his production is going to count against the PF defense for the Nets, even though for 10 of those minutes he was being defended by a C. This could cause some small discrepancies, but over the course of an entire season, the data is still very useful and should tell us what we want to know.
Another thing to consider is that this is entire game data so it is not only starter's defense that is being measured (although they play a large majority of the minutes). So when I am looking at the Thunder's PG defense, I am looking at 38 minutes of Aaron Harrison's defense and 10 minutes of Frank Johnson's defense. With that said, let's take a look.
Looking at the PG numbers yields a few surprises. Chris Duhon earns another feather in his undervalued cap as he lead the Kings to the best defense against PGs in TMBSL in 3019. Scoonie Murray has been thought to be a valuable PG due mainly to his passing and defensive abilities and last year the Hornets were one of two teams that held opposing PGs to a true shooting percentage of less than 50%. Across the league, PGs scored on average 25 points per game on .535 TS%. Keep in mind this is the PG positiion for the entire game, not just the starter. Dutch Burch and his low defensive grade performed well for the Clippers as they came in third. This goes to show us that simply looking at letter grade does not tell the full story at all when it comes to defense.
The SG results are shocking. The 76ers were by far (a full 2% over 2nd place) the best at defending opposing SGs in 3019. Their starting SG? Pee Wee Kirkland. Pee Wee had a B- defensive grade and has long been touted as a guy who can fill it up on offense but is a slouch on the defensive end of the floor. Well, he was doing something right last season. This data goes to show us once again that simply looking at defensive grade can be misleading and is not a direct indicator into how their opponent will perform. Overall, SGs had the highest TS% in TMBSL in 3019 at 55.2%. The SG position averaged 24.5 points per game. Clyde Drexler and his mega contract in Charlotte played below average SG defense. Quentin Richardson shows that he earned his most improved player award by playing stellar defense as well.
One of the things I was curious about with the "big men" who get moved to SF is how that affects them defensively. It looks like they can still be quite successful defensively. The Clippers and Bullets both played former big men at SF and they came in at 1-2 in the rankings. Al Horford and the Warriors ranked 6th. This goes to show that at least some of the big men who get shifted down to SF are still quite effective on the defensive side of the court as well. This puts me more firmly in favor of the new position rules that soup is enforcing. Across the league, SFs averaged 22 points per game on .538 true shooting percentage. Another interesting thing I noted was that the Champion Thunder had below average defense from their SF spot.
Unlike SG, the top of the PF rankings are no surprise at all. It is pretty widely recognized that Jermaine O'Neal and Shareef Abdur-Rahim are the best two defensive PFs in TMBSL and the rankings back that up. Some may be slightly surprised that the Nets rank ahead of the Thunder, but they are close and I liked the Nets backup bigs better than the Thunder last year. Kenyon Martin played strong defense for the Warriors which helps to make up for his subpar offense and the Kings split of Catfish/Myles Turner rated highly as well. On average, PFs scored 23.8 points per game on a true shooting percentage of .514.
So much for that Hasheem Thabeet defensive big man theory, eh? Everyone was well aware that he was very inefficient on offense, but most assumed that the big man at least was a plus on the defensive end of the floor. Well, in 3019, the Hawks were the 3rd worst team in TMBSL at defending the center position and he was playing a large majority of their minutes there. At least he can rebound. The Clippers reign supreme once again. The rest of the top ten is not that surprising either - Camby/Catfish for the Kings, Chet Noe for the Bucks, Vlade and the Thunder, David Robinson and the Sixers, Dwight for the Hornets, Ben Wallace in Brooklyn, Harrell in Phoenix, Oden in Minnesota and Okafor on the Warriors. Overall, centers are the least efficient scorers in TMBSL (which only confirms what we were already aware of). They averaged 19.4 points on a true shooting percentage of .490.
I hope you guys find this information enjoyable and beneficial when evaluating players as it took a long time to compile. I will be interested to see if the same players consistently land atop these rankings or if it is more of a year by year thing so I can keep doing this if people enjoy it.
