Great analysis of the Hawks improvement this season

[AtLaS]

The Atlanta Hawks took a big leap forward this year. Coming in everyone thought Cleveland, Boston, Orlando would walk over the East. But from the beginning Atlanta has been just as good. This is a surprise considering they had a very similar team last year.

So what's the change? Let's take a look:

08-09 Hawks: 10th offensively, 12th defensively - 47 wins. Pythagorean - 46 wins. SRS/point differential - 1.71 (10th overall)

09-10 Hawks: 3rd offensively, 12th defensively - 26-14. On track for 53 wins. Pythagorean - 55 wins. SRS/point differential - 5.35 (4th overall)

They are identical defensively, the difference is all offense. A jump from 10th to 3rd is huge for a team's outlook with their expected W/L, as shown by them jumping from 46 wins to 55 wins. They went from a weak +1.71 in SRS to a very very good +5.35. In comparison LA is +6.75, Cleveland is +5.94, Boston is +5.38, Orlando is +4.18. Defeating LA without homecourt advantage is a longshot, but this indicates Atlanta has an excellent ECF or Finals chance

What happened? Conventional logic says Jamal Crawford's 6MOY worthy addition and Josh Smith and Al Horford's offensive improvement have pushed them over the top. So let's see if the stats back that up:

2008-2009

eFG% - .504 - 9th

TOV% - .125 - 8th

Orb% - .260 - 19th

FT/FGA - .238 - 14th

2009-2010

eFG% - .506 - 11th

TOV% - .114 - 1st (2nd highest is .121, 3rd is .124)

Orb% - .281 - 9th

FT/FGA - .212 - 25th

Surprisingly the Hawks' scoring efficiency hasn't gone anywhere. They are slightly more efficient from the field, but getting much less help from the FT line. That likely means they've been worse overall

The real leap has been in TOs and Offensive Rebounds. They are by far the best TO team in the league right now. The difference between them and the 2nd place team is the same as between 2nd and 10th. Looking at the stats the most noticeable difference is Johnson and Horford improved their TO rate by 2% and Flip to Crawford is a 2% jump. Furthermore getting rid of Law's minutes is good, Bibby and Crawford are 4% less turnover prone than him. Finally the bench rotation of Zaza, Joe Smith, Morris has been 2-3% better protecting the ball this year

As for the offensive rebounding, similar story. Horford and Smith are up 2%, the Johnson and Crawford combo are 2% higher than Johnson/Flip, and Joe Smith has been giving them another Zaza level offensive rebounder off the bench.

So there you have it. The Hawks offense is almost the exact same as last year's, but by having outstanding TOV% and vastly improving their offensive rebounding, they are now a fringe title contender. Basketball is a game of inches and incrimental improvement. Getting a handful of extra possesions a game can change an entire season. The Hawks are doing themselves a huge favor by protecting the ball so well

We've always complained about turnovers but the stats don't back it up. Even in the past we weren't a big turnover team. But, our improvement in ball security as well as Smith's reluctance to shoot has given us a big boost. Hopefully we keep it up.

[Admin]

Interesting stuff there, where did it come from?

[jerrywest]

there was a Hollinger insider a few weeks ago where he illustrated how Denver/Phoenix/Atlanta became the 3 best offensive teams in 3 different ways. Hawks by the way of least turnovers.

[RedDawg#8]

I agree that we have done better turnover wise this season, but the past week or so we have finished games above our average in turnovers so we can still do better I think. Its probably the reason games have been too close for comfort lately.

[jerrywest]

We had more energy last year. The talent/depth/familiarity has overcome the energy drop.

[Guest]

Furthermore getting rid of Law's minutes is good, Bibby and Crawford are 4% less turnover prone than him.

Hey whatever happened to all the posters that dogged Woody for not playing Acie? I mean, I thought Acie was supposed to be a star..

What is interesting is that TOV% just increases the number of successful chances you have to score. It has nothing to do with FG%, but they really should. A team could have a lower FG% but have a higher probability to score. Why is that? Well if you turn the ball over more often, you don't actually shoot so it doesn't affect FG%. But turning the ball over doesn't allow you to shoot. So if we just simplify everything and say a team only shoots 2 pointers and doesn't shoot FT it'll be easier to see:

Team A: .500 FG% .100 TOV% -> this team will shoot the ball 90% of the time and make 50% of the shots, so every possession they should expect 2*(.9)*(.5) = .9 points

Team B: .520 FG% .150 TOV% -> this team shoots better, but only shoots the ball 85% of the time, so every possession they should expect 2*(.85)*(.52)= .884 points

You can play around with the numbers, but in the end having a lower TOV% will increase your expected value of points per possessions.

But then offensive rebounding is a much simpler concept. You increase the number of possessions you have. So you could have a lower expected value of points per possessions, but if you have more possessions then you can increase the number of Total points.

I like this post on the Hawks, good find Atlas.

[AHF]

You can play around with the numbers, but in the end having a lower TOV% will increase your expected value of points per possessions.

But then offensive rebounding is a much simpler concept. You increase the number of possessions you have. So you could have a lower expected value of points per possessions, but if you have more possessions then you can increase the number of Total points.

Maybe you could generate some kind of true scoring metric that marries TS% with the number of possessions a team utilizes (factoring in both TOs and offensive rebounds).

This was a good find and good read.

[AHF]

Hey whatever happened to all the posters that dogged Woody for not playing Acie? I mean, I thought Acie was supposed to be a star..

I dogged Woody for this but not because I thought Acie was a star. I dogged him because I thought we needed to find out what we had in Acie and either bring in another young PG to develop ASAP or to get Acie ready to play if we couldn't resign Bibby or Bibby's age/injuries caught up to him. Kudos to Woody for recognizing that Law was not going to cut it without needing the additional playing time. If he has accurately seen that in Teague as well, then we should likewise limit his minutes, trade him for value, and get a replacement in who can be at least a limited part of the rotation ASAP.

[macdaddy]

I don't think anyone was dogging him because we though Acie was a star. We were dogging him because DNPs assured that he and the rest of the bench would be totally lost if we got to the playoffs and had key players injured. Glad that didnt' happen.

[exodus]

The lack of turnovers is actually a side benefit of their weak offensive scheme. When JJ or Crawford dribbles between their legs 5 times an put up a contested jumper it is unlikely that they will turn it over. A lot of those shots get missed and some of those become offensive rebounds. I am hesitant to give the hawks much credit for this because i know that they should move the ball more on offense even though that would likely result in more turnovers.

[Guest]

Maybe you could generate some kind of true scoring metric that marries TS% with the number of possessions a team utilizes (factoring in both TOs and offensive rebounds).

This was a good find and good read.

I do enjoy thinking in terms of expected values for points per possession and number of possessions. Maybe thats the nerdy statistics side of me. But one thing I do not like is when people just throw statistics together and then slap their name onto a metric (ex. Christopher Reina over at RealGM). Most statistics just don't have much behind them in terms of effort put behind creating the metric and effort of explaining why its important. So unless I put a lot of effort into thinking about a metric concerning utilization of possessions, I wouldn't want to actually have a metric. I think we get too cluttered with these Excel Wizards and Basketball-Reference.com.

I dogged Woody for this but not because I thought Acie was a star. I dogged him because I thought we needed to find out what we had in Acie and either bring in another young PG to develop ASAP or to get Acie ready to play if we couldn't resign Bibby or Bibby's age/injuries caught up to him. Kudos to Woody for recognizing that Law was not going to cut it without needing the additional playing time. If he has accurately seen that in Teague as well, then we should likewise limit his minutes, trade him for value, and get a replacement in who can be at least a limited part of the rotation ASAP.

I guess my main point wasn't about how Woody used Acie. My main point was that Acie Law was statistically one of the worst players in the NBA. It just always frustrated me when posters would claim Acie as a budding star based on nothing more than some sentimental reasoning.

[AHF]

I guess my main point wasn't about how Woody used Acie. My main point was that Acie Law was statistically one of the worst players in the NBA. It just always frustrated me when posters would claim Acie as a budding star based on nothing more than some sentimental reasoning.

I agree he never showed himself to be a budding star, but it is expected that some posters embody the term "fanatic" in their support of a player. My point was that we had a very strong need for a contigency if Bibby left and for a long-term successor and that we should fish or cut bait with Law.

[AHF]

I do enjoy thinking in terms of expected values for points per possession and number of possessions. Maybe thats the nerdy statistics side of me. But one thing I do not like is when people just throw statistics together and then slap their name onto a metric (ex. Christopher Reina over at RealGM). Most statistics just don't have much behind them in terms of effort put behind creating the metric and effort of explaining why its important. So unless I put a lot of effort into thinking about a metric concerning utilization of possessions, I wouldn't want to actually have a metric. I think we get too cluttered with these Excel Wizards and Basketball-Reference.com.

I agree with this 100%. You can't arbitrarily mash numbers together. However, I would think there is real potential to do something that layers the scoring efficiency numbers of a team with the ability of a team to generate possessions in which they can score. If you could adequately describe the creation/loss of possessions, that would seem to give you a way to describe more than the efficiency of an offense per possession resulting in a shot (TS%) and get closer to efficiency per possession (i.e., no longer limiting it to possessions resulting in a shot). It may not be able to be done, but it is interesting to think about.

Edited January 22, 2010 by AHF

[TheNorthCydeRises]

The lack of turnovers is actually a side benefit of their weak offensive scheme. When JJ or Crawford dribbles between their legs 5 times an put up a contested jumper it is unlikely that they will turn it over. A lot of those shots get missed and some of those become offensive rebounds. I am hesitant to give the hawks much credit for this because i know that they should move the ball more on offense even though that would likely result in more turnovers.

In other words, the offense isn't pretty, but it turns out to be efficient, because the scheme is low-risk.

Kind of like watching Navy's offense in football. You know the formation, you know what they're going to run. And you know the 2 people who will get most of the carries. Question is . . . can you stop it? Most teams that they play . . . can't. So when that offenseis clicking on all cylinders, it's a beautiful thing to watch.

LOL . . you probaly won't agree with the analogy though.

[exodus]

In other words, the offense isn't pretty, but it turns out to be efficient, because the scheme is low-risk.

Kind of like watching Navy's offense in football. You know the formation, you know what they're going to run. And you know the 2 people who will get most of the carries. Question is . . . can you stop it? Most teams that they play . . . can't. So when that offenseis clicking on all cylinders, it's a beautiful thing to watch.

LOL . . you probaly won't agree with the analogy though.

But good teams can, especially in a 7 game series. They are playing one of the teams tonight that they might face in the playoffs. Not surprisingly the Hawks scored 83 pts the last time we played them.

[Crank]

The lack of turnovers is actually a side benefit of their weak offensive scheme. When JJ or Crawford dribbles between their legs 5 times an put up a contested jumper it is unlikely that they will turn it over. A lot of those shots get missed and some of those become offensive rebounds. I am hesitant to give the hawks much credit for this because i know that they should move the ball more on offense even though that would likely result in more turnovers.

We are 7th in the league in APG as a team . The past 3 games we had 29 ,27 ,25 we move the ball just fine most of the time but we must do a better job executing when our shots are not falling.

[exodus]

We are 7th in the league in APG as a team . The past 3 games we had 29 ,27 ,25 we move the ball just fine most of the time but we must do a better job executing when our shots are not falling.

It only takes one pass to get an assist. You can get a lot of assists by doing iso's all game long and just passing when a double comes. The problem is that running isos doesn't make the defense move and good defenses can exploit that.

Also the Hawks get a lot of assists on the fast break. Last i checked they were 3rd in the league in fast break pts.