Below, I describe 5 attributes that I think encapsulate the qualities I look for in a defensive player: Mobility: Will cover ground well in order to stay in front of defenders, close gaps to prevent penetration, close out on shooters, get back in transition, rebound outside of space, get to loose balls Length: Possesses good height/arm length/leaping ability to contest shots, to help snare loose balls, secure rebounds Strength: Will not be moved easily by post players or rebounders Intelligence: Understands the game plan and is mentally alert. Sees the floor well. Makes good decisions, communicates well with teammates to help them recognize situations and make good decisions. Versatility: Has the necessary physical tools to defend different types of players and cover against different types of plays. Gives coaches flexibility in cross-matching or switching. The first 3 attributes describe how effectively a player is able to make use of his physical tools. Some may view the last attribute as redundant, as the versatility of the player depends on his physical abilities. But I feel that players that possess a combination of physical tools that make them versatile on defense afford their team an extra advantage, so I wish to give extra credit for this. How much weight we should put on each of these attribute is very much subjective. I propose the following: DRTG = (1/6)*( MOB + LEN + STR + 2*INT + VER) where DRTG, MOB, LEN, STR, etc. will have a value between 0 and 10. I'll rate our 9 rotation players, again subjectively, for each of these categories. I also take into account which positions they are asked to defend (e.g. Kyle Lowry and and of course Chuck Hayes are asked to defend out of position quite a lot, which will help their mobility and versatility rating but bring down their length/strength rating). 8-10: excellent 6-8: good 4-6: average 2-4: below average 0-2: very poor So, below is what I came up with. The final ratings make sense to me for the most part. I was initially bothered that David Andersen ended up rating about the same as Aaron Brooks. But then I know that research has been done (Rosenbaum) that indicates bigs tend to contribute more on defense than smalls. So, while in absolute terms Andersen may be about the same, relative to their respective positions and their defensive roles I think Brooks likely hurts the team less. Pos player MOB LEN STR INT VER DRTG G/F Battier 7.5 8.0 5.0 10.0 8.0 8.1 C Hayes 9.0 1.5 8.0 9.5 8.5 7.7 G/F Ariza 8.0 8.5 4.0 5.0 8.5 6.5 G Lowry 8.5 3.0 6.0 6.0 4.5 5.7 F/C Landry 7.0 5.0 6.0 4.0 5.5 5.3 G/F Budinger 7.0 7.5 4.0 4.0 4.5 5.2 F/C Scola 4.5 4.0 4.0 6.0 5.0 4.9 PG Brooks 8.0 2.5 3.0 5.0 1.0 4.1 C Andersen 3.0 4.5 3.0 5.0 4.0 4.1 As sort of a sanity check, I took a weighted average of these player's final ratings (based on minutes played) and it came to 5.8, which is a little above average. According to basketball-geeks schedule adjusted power rankings, the Rockets are rated 12th in defense right now. So that at least seems to check out ok. Where do you guys feel I might have overrated or underrated players? I'd like to refine it and reach some type of "subjective consensus" on how effective each player is on defense.
I don't want to chime in first and i do like your setup, but i have a few issues with the ratings. I will wait and come back after other have chimed in :grin:
In before the people who say Ariza isn't a good defender. I've always felt this way but didn't know there was research done on the topic.
I wont comment on the validity of your system but I will say if I looked at your list of defensive ability, I would say Landry does not enough with what he has and Brooks does more with what little he has.
Some more thoughts on qualities we desire in a good defensive player that I jotted down yesterday while watching the game. Not sure if my five attributes listed above really covers everything. Even if subjective, I'd like it to be as comprehensive as possible. want players who can defend their man in 1-on-1 situations; the less help required the better good individual defense happens before one's assignment receives the ball. with suitable help-side defense, a defender may apply extra pressure on his man before he receives the ball to either keep him from a desired spot on the floor or deny him the ball. invariably, offensive actions will necessitate help defense to some degree. want players to quickly recognize these situations and provide suitable help defense. this takes many forms: double teaming an offenisve player to protect against a mismatch or induce an offensive mistake getting in the proper position (spacing is important) to take away driving lanes and give teammates a chance to recover to their man. apply strong-side pressure to take away easy options for the offense, e.g. extra backside pressure to force a ball reversal and take time off the clock. closing out a possession by securing the rebound after a miss is essential. things that we want players to do to contribute to this: boxing out early to carve rebounding space around the basket effective pursuit of the basketball entails proper timing, suitable agility, and length. effective transition defense: players needs to run back consistently matchup quickly when facing a numbers disadvantage, try to force a miss, force a turnover, or (last resort) take a foul. sloppy offensive execution / spacing will create transition opportunities for the opposition. Bad shots and bad turnovers should be avoided.
I've got a few gripes about the system: 1. Why are you weighing Intelligence twice as much as other factors? (and shouldn't this really be called Basketball IQ?) This just seems like a way to boost Battier's stats. I think Intelligence is a good factor in making a great defensive player, but I don't think it should be weighed twice as much as other factors - they all should be equal parts in the formula, IMO. 2. Why is Battier an 8 on the length factor? Length isn't and shouldn't be about just height -- or else you'd automatically be punishing shorter players and making them not as effective defenders. http://www.draftexpress.com/profile/Shane-Battier-1619/ Battier should score below 5 in this category, IMO. He is 6'8", but he does NOT have good length, which is essentially what the entire category is. 3. I'm not really sure how strength factors in. Or why you think Lowry is only almost half of what Battier/Ariza/Hayes are. Because he can't guard centers? So you're penalizing guards for not being able to guard centers? I rarely, if ever, see Lowry get pushed around. I'm not sure strength should really be a factor, or if it is, the ratings need to not take into account height and take into account their strength relative to their position. (i.e., Lowry should score high; I suspect you did this in the Mobility department, as there's no way Chuck's more mobile than Shane or Ariza head up, but for his position, he's very mobile.)
I think the weight that you've given to each of those factors seems to be fairly arbitrary. In the NBA, I would argue that length and mobility are more important than strength. I think that intelligence is an important factor to include, but to say that it is twice as important as those other factors isn't necessarily accurate. There are a lot of excellent college level defenders who get by on mostly intelligence, but those players can't guard anyone in the NBA because they lack the physical tools. I also think that you might make a place for effort and energy. Guys like Amare, Eddy Curry, among others have the physical tools to make an impact on the defensive end, but they don't due to a lack of effort. You might be including this in intelligence, but they are really two different things. I also think that the ratings you've given to players is still biased, even if you're trying to be as fair as possible. There are measurements and physical tests that can be used to measure some of these things, but it looks like you're assigning numbers based on what you see or a player's reputation. For example, you rate Battier as having excellent length, but he's pretty average in size for a small forward.
My version (all based on current season performances): -----------MOB----LEN----STR----INT----VER----DRTG Hayes-------10------1-----10-----10------10------8.5 Battier-------7------7------6------10------8-------8 Lowry--------8------3------6------9-------8------6.8 -- Ariza---------8------9------5------4-------7------6.2 Scola-------6--------3------5------6-------3------4.8 Budinger----5--------6------4------5-------3------4.7 Landry------7-------3------5------5--------2-----4.5 Brooks------8--------2------2------4-------2------3.7 Andersen----5-------5------2------2--------2-------3 One extraneous factor will be the fact that the bench players may seem like they play better defense than they really do because they're usually playing against worse players rather than the studs the starters defend. So I actually accounted for that and imagined the defensive ability of each player against the entire set of players at the positions each player plays at and his relative defensive effectiveness to the entire set of possible opponents. Also, just for kicks: -----------------MOB----LEN----STR----INT----VER----DRTG Yao--------------2-------9-------9------6-------2-------5.7 Dwight Howard---10------10------8------7-------9-------8.5 Lebron-----------10------10------9------8-------8-------8.8 Kobe-------------8-------8-------9------9-------7-------8.3 Wade------------9-------7-------9------8-------7-------8 T-mac(current)---3-------10------3------2-------3-------3.8 T-mac(prime)----8--------10------6------8-------10------8.3 Another thing I've discovered is that you may want to weigh the different attributes differently for different positions. Like LEN and STR would probably be more important for Centers than Point Guards and should probably count for 2x for Centers while INT counts for just 1x while for wingmen INT might continue to count for 2x while for PG it might be 2x for MOB but not LEN and STR, etc.
(edited for bad math) My version (all based on current season performances): -----------MOB----LEN----STR----INT----VER----DRTG Hayes-------10------1-----10-----10------10------8.5 Battier-------7------7------6------10------8-------8 Lowry--------8------3------6------9-------8------7.2 Ariza---------8------9------5------4-------7------6.2 Scola-------6--------3------5------6-------3------4.8 Budinger----5--------6------4------5-------3------4.7 Landry------7-------3------5------5--------2-----4.5 Brooks------8--------2------2------4-------2------3.7 Andersen----5-------5------2------2--------2-------3 One extraneous factor will be the fact that the bench players may seem like they play better defense than they really do because they're usually playing against worse players rather than the studs the starters defend. So I actually accounted for that and imagined the defensive ability of each player against the entire set of players at the positions each player plays at and his relative defensive effectiveness to the entire set of possible opponents. Also, just for kicks: -----------------MOB----LEN----STR----INT----VER----DRTG Yao--------------2-------9-------9------6-------2-------5.7 Dwight Howard---10------10------8------7-------9-------8.5 Lebron-----------10------10------9------8-------8-------8.8 Kobe-------------8-------8-------9------9-------7-------8.3 Wade------------9-------7-------9------8-------7-------8 T-mac(current)---3-------10------3------2-------3-------3.8 T-mac(prime)----8--------10------6------8-------10------8.3 Another thing I've discovered is that you may want to weigh the different attributes differently for different positions. Like LEN and STR would probably be more important for Centers than Point Guards and should probably count for 2x for Centers while INT counts for just 1x while for wingmen INT might continue to count for 2x while for PG it might be 2x for MOB but not LEN and STR, etc.
I think there is quite a lot that goes into "intelligence" or "IQ". Knowing the gameplan, knowing the other team's gameplan, making smart decisions, helping teammates to make smart decisions. You can have all the physical tools, but if your IQ is poor (i.e. you don't see the floor well, don't know the team's gameplan, don't make good decisions) I think its very hard to be a really good defender. According to my proposed system, if you rate poorly by IQ (INT=2), then the best defensive rating you can hope for is 7.3. That seems fair to me. First of all, Battier is guarding SGs as often as he guards SFs, so he does have above average size/length on that basis. Further, I think Battier uses what length he possesses as well as any player in the league. According to hoopdata, amongst rotation swing-men Battier ranks 7th in blocks/min (only Donte Greene, Jamario Moon, Kirilenko, Wade, Marion, and Gerald Wallace are ahead of him). I don't rate Lowry very highly by strength for two reasons. (1) he guards SGs about as often as he guards PGs. (2) I don't think strength factors in that heavily for guards anyway, so I'd rather compress his rating and have it close to 5 than put it at 7 or 8.
Looks like a good rating system to me. I was going to nitpick versatility for Ariza and Hayes but after thinking about it they do cover pretty well on switches. Good Post!
I considered that, but even that gets problematic because teams may employ different defensive schemes based on the abilities of their players. Varejao may be just as an effective a defender as Yao even if his LEN and STR are far less, because the way CLE plays defense with Varejao as a center makes use of his strengths (his mobility), while the way HOU plays defense makes use of Yao's strengths (his size).
I considered putting effort/energy in a separate category, but I figured it is only relevant so far as it allows a player to make use of his mobility/length/strength. So I factored that into my subjective ratings for those attributes -- especially mobility and strength. A player might, in theory, have great quickness/strength, but if he's so lazy that he lets players routinely get around him or push him around on the boards then I can't rate him highly in those categories. If you or anyone else is willing to suggest alternate ratings for the players, I'd like to read them.
There definitely should be a hustle/energy category. although I agree that intelligence isn't 2x more valuable than other areas, it is more important than most individual categories. i'd rather Battier be smart than have longer arms. Propensity for fouls may equate somehow as well although not much.
Fouling a lot would either fall under poor technique, which would fall under IQ in my view, or suggests a physical deficiency in mobility, length, strength. Similarly, hustle/energy can be subsumed by the mobility/length/strength attributes in my opinion. Having great physical tools but providing no hustle/energy means very little. To make any positive plays with your physical gifts, you need to expend some energy.
I think the 2nd paragraph admits that hustle/energy needs to be a separate category. And great job with this. This can be something that always evolves.
Not sure I agree. Suppose I made it a separate category, and there was a player who had great physical gifts (excellent in mobility, length, strength), but was a 0 in hustle/energy. It seems to me that if he provides no hustle/energy than he's not going to make use of his physical advantages. To me, it makes more sense to just reduce his MOB, LEN, STR ratings, rather than create a separate category for hustle/energy.
