MikeDC
Member
- Joined
- May 24, 2007
- Messages
- 5,643
- Likes
- 16
- Points
- 38
I posted the whole article in the Bulls forum, but I thought you guys might be interested in the discussion as well
http://sportstwo.com/threads/162385-NBA-Draft-Measurements-Drafting-by-Numbers
I've been doing a little bit of numbers work on draft measurement numbers now that we've got several years to work with to see how numbers seem to play out in actual NBA action.
Obviously, the measures taken at the NBA draft combine for size and athletic talent aren't everything, but they are something. So what do they count for? Which numbers are important and why? These are the sorts of questions I'll try to answer.
The data
The analysis I did is extremely simple in its basic scope. Since I wanted to do this by today, I haven't checked a lot, but the numbers look about right. I looked at the career minutes per game of everyone I could get measurement and athletic statistics for. Then, I divided them into two groups. Guys who average >20mpg and guys who average <= 20mpg. By position. Then I took the average measures of players, by position, in each category.
That gives us a table like this for players who average more than 20mpg:
For players who average 20mpg and under, we get:
That is, you can read this as saying I have measures for 50 PGs who play at least 20mpg, and their average sprint time was 3.17. For the 98 guys who don't play much, or at all, the average time measured as 3.21. So the average time is faster, but not by very much.
To get an idea of how much it matters, you need to know how dispersed the sample is from the average. To do this, I took the standard deviation of the good (high minute) players.
This means, for example, that if you look at those 21 Centers who play at least 20mpg, about 70% of them will fall within + or - 17.4 lbs of the 255.5lb average.
Findings
To get an idea of what to do next, look at the average weight of a low minute center. It's 251lbs, and that's well within the range, and not much different from a high minute center. So we should probably conclude that looking at the weight of a center prospect probably isn't going to tell us much about his success level.
That doesn't mean it's meaningless. When evaluating a draft prospect, we probably still want to know if a guy is Keith Closs or ahem... DeMarcus Cousins. On average, it's better to be bigger, I think, but it ought to be muscular bigger, not fat bigger.
To get a really basic idea of what matters and what doesn't, I've compared the difference between the high and low minute averages in each category to the standard deviation. The higher the number you see in this table, the bigger the difference is between guys who play a lot in the NBA and guys who don't.
In plain English, this means that, for example, the average high minute center has 2.9 inches of flat footed leaping ability that the average low minute center doesn't. The standard deviation for is only 2.58 inches, so this is pretty solid evidence that this is a meaningful measure. The better centers can usually jump noticeably better than the worse ones.
What doesn't matter
Weight and bench press values don't seem to matter very much. Like I said, other things being equal, you still want in shape, athletic players, but I'd probably put more emphasis on other factors when evaluating.
What matters
Jumping ability (NSV and MaxV), height, wingspan, and sprinting ability seem to be important across the board. This makes pretty obvious sense when you consider basketball is more or less about being long, running and jumping.
When you start to break down things by position, it starts to make even more sense.
Good Centers are separated distinguished by less good ones based on their ability to athletic ability, not on their size. Guys who can jump, run the floor, and are fairly agile are going to be more successful. Big oafs are not. Often the center decision comes down to a guy who looks like a center but can't move, vs. a guy who can move but might be undersized. Probably you wan the guy who can move.
Good Power Forwards tend to be bigger than less good ones, and better jumpers. Again, this sounds like common sense when you think about it. The scarcity involved in getting a PF isn't in getting a guy who can move quickly, it's in getting a big guy who can move quickly and jump.
Good Small Forwards might be the toughest to distinguish out of all the positions. Looking at the findings above, it's hard to tell a high minute player from a low minute player based on his size or athleticism. In some cases, the low minute ones look better than the high minute ones on paper.
That shouldn't tell us size and athleticism doesn't matter. It's still better to be big and a great athlete. What's going on here is that, in many ways, the SF is the utility man on the basketball floor. He's got to do some inside work and some outside work. And there's a pretty good supply of 6'7 guys around, so what's going to really distinguish the good ones from the bad ones is their skillset. So when judging SFs, look for a guy who has the proper athletic tools, but emphasize their basketball skills if they're in the right athletic range.
Good Shooting Guards are typically bigger and better athletes across the board than the low minute SGs. Being able to run the floor, cut, out jump, and outreach your opponent turns out to be pretty important for them. This makes sense if you think of their role, typically, as perimeter shooters, defenders, and one-on-one options. Obviously you need the skills to go along with it, but a slow SG who can't jump is probably going to be much easier to defend, and have a harder time defending other SGs.
Good Point Guards typically look a bit like good SGs in that athleticism is important, but there's less emphasis on physical size. The emphasis here is on speed and agility. A PG needs to be able to get to where he wants to go on the court. While it's nice to be big and strong too, most PGs are typically good because they get to where they need to go.
http://sportstwo.com/threads/162385-NBA-Draft-Measurements-Drafting-by-Numbers
I've been doing a little bit of numbers work on draft measurement numbers now that we've got several years to work with to see how numbers seem to play out in actual NBA action.
Obviously, the measures taken at the NBA draft combine for size and athletic talent aren't everything, but they are something. So what do they count for? Which numbers are important and why? These are the sorts of questions I'll try to answer.
The data
The analysis I did is extremely simple in its basic scope. Since I wanted to do this by today, I haven't checked a lot, but the numbers look about right. I looked at the career minutes per game of everyone I could get measurement and athletic statistics for. Then, I divided them into two groups. Guys who average >20mpg and guys who average <= 20mpg. By position. Then I took the average measures of players, by position, in each category.
That gives us a table like this for players who average more than 20mpg:
Code:
pos weight height wspan sreach NSV MaxV Bench Agil Sprint sample
C 255.5 82.7 87.0 110.1 29.75 33.00 14.55 11.65 3.36 21
PF 240.7 79.9 85.4 107.5 29.80 33.76 12.86 11.54 3.28 31
SF 216.6 78.3 83.5 104.6 30.17 35.50 9.09 11.48 3.26 40
SG 204.3 76.1 80.8 101.4 31.46 36.36 9.77 11.10 3.19 43
PG 187.2 72.8 77.2 97.1 30.67 36.09 8.62 11.07 3.17 50For players who average 20mpg and under, we get:
Code:
pos weight height wspan sreach NSV MaxV Bench Agil Sprint sample
C 251.6 82.5 86.9 110.1 26.84 30.37 11.72 12.10 3.45 116
PF 237.2 79.5 84.7 106.5 29.01 33.22 13.84 11.55 3.30 113
SF 217.7 78.3 82.8 104.6 29.59 34.36 11.82 11.33 3.26 59
SG 201.9 75.3 80.2 100.7 29.80 35.40 11.03 11.27 3.23 98
PG 185.3 72.7 76.8 96.9 29.16 34.70 9.78 11.14 3.21 98That is, you can read this as saying I have measures for 50 PGs who play at least 20mpg, and their average sprint time was 3.17. For the 98 guys who don't play much, or at all, the average time measured as 3.21. So the average time is faster, but not by very much.
To get an idea of how much it matters, you need to know how dispersed the sample is from the average. To do this, I took the standard deviation of the good (high minute) players.
Code:
pos weight height wspan sreach NSV MaxV Bench Agil Sprint sample
C 17.4 1.9 2.2 2.3 2.58 2.50 4.03 0.59 0.13 21
PF 19.6 1.3 1.8 2.0 2.65 3.21 4.54 0.53 0.11 31
SF 13.3 1.1 2.1 2.7 3.15 3.07 4.50 0.43 0.10 40
SG 14.1 1.6 2.3 2.2 1.92 2.91 5.22 0.49 0.10 43
PG 13.9 1.9 2.9 3.9 3.03 3.82 4.79 0.39 0.09 50This means, for example, that if you look at those 21 Centers who play at least 20mpg, about 70% of them will fall within + or - 17.4 lbs of the 255.5lb average.
Findings
To get an idea of what to do next, look at the average weight of a low minute center. It's 251lbs, and that's well within the range, and not much different from a high minute center. So we should probably conclude that looking at the weight of a center prospect probably isn't going to tell us much about his success level.
That doesn't mean it's meaningless. When evaluating a draft prospect, we probably still want to know if a guy is Keith Closs or ahem... DeMarcus Cousins. On average, it's better to be bigger, I think, but it ought to be muscular bigger, not fat bigger.
To get a really basic idea of what matters and what doesn't, I've compared the difference between the high and low minute averages in each category to the standard deviation. The higher the number you see in this table, the bigger the difference is between guys who play a lot in the NBA and guys who don't.
Code:
pos weight height wspan sreach NSV MaxV Bench Agil Sprint
C 22% 8% 2% -2% 113% 106% 70% -77% -67%
PF 18% 33% 40% 50% 30% 17% -21% -1% -20%
SF -9% -5% 32% 1% 18% 37% -61% 36% -7%
SG 17% 54% 24% 30% 87% 33% -24% -34% -46%
PG 13% 0% 13% 4% 50% 36% -24% -18% -47%In plain English, this means that, for example, the average high minute center has 2.9 inches of flat footed leaping ability that the average low minute center doesn't. The standard deviation for is only 2.58 inches, so this is pretty solid evidence that this is a meaningful measure. The better centers can usually jump noticeably better than the worse ones.
What doesn't matter
Weight and bench press values don't seem to matter very much. Like I said, other things being equal, you still want in shape, athletic players, but I'd probably put more emphasis on other factors when evaluating.
What matters
Jumping ability (NSV and MaxV), height, wingspan, and sprinting ability seem to be important across the board. This makes pretty obvious sense when you consider basketball is more or less about being long, running and jumping.
When you start to break down things by position, it starts to make even more sense.
Good Centers are separated distinguished by less good ones based on their ability to athletic ability, not on their size. Guys who can jump, run the floor, and are fairly agile are going to be more successful. Big oafs are not. Often the center decision comes down to a guy who looks like a center but can't move, vs. a guy who can move but might be undersized. Probably you wan the guy who can move.
Good Power Forwards tend to be bigger than less good ones, and better jumpers. Again, this sounds like common sense when you think about it. The scarcity involved in getting a PF isn't in getting a guy who can move quickly, it's in getting a big guy who can move quickly and jump.
Good Small Forwards might be the toughest to distinguish out of all the positions. Looking at the findings above, it's hard to tell a high minute player from a low minute player based on his size or athleticism. In some cases, the low minute ones look better than the high minute ones on paper.
That shouldn't tell us size and athleticism doesn't matter. It's still better to be big and a great athlete. What's going on here is that, in many ways, the SF is the utility man on the basketball floor. He's got to do some inside work and some outside work. And there's a pretty good supply of 6'7 guys around, so what's going to really distinguish the good ones from the bad ones is their skillset. So when judging SFs, look for a guy who has the proper athletic tools, but emphasize their basketball skills if they're in the right athletic range.
Good Shooting Guards are typically bigger and better athletes across the board than the low minute SGs. Being able to run the floor, cut, out jump, and outreach your opponent turns out to be pretty important for them. This makes sense if you think of their role, typically, as perimeter shooters, defenders, and one-on-one options. Obviously you need the skills to go along with it, but a slow SG who can't jump is probably going to be much easier to defend, and have a harder time defending other SGs.
Good Point Guards typically look a bit like good SGs in that athleticism is important, but there's less emphasis on physical size. The emphasis here is on speed and agility. A PG needs to be able to get to where he wants to go on the court. While it's nice to be big and strong too, most PGs are typically good because they get to where they need to go.
I did a similar bit of number crunching a few years back on another chat site. I agree with your conclusions... my one suggestion is that you convert the measurements into feet and inches rather then just straight inches as it's what we're more used to.