My post from a few days ago was relatively well-received at reddit, and one of the readers there said that he’d like to see the same graph for some of the all-time great point guards.

So with a little help from Basketball-Reference.com, I compiled a list of (all?) the Hall of Fame point guards: Oscar Robertson, Lenny Wilkens, Bob Cousy, Jerry West, John Stockton, Isiah Thomas, Magic Johnson, Dennis Johnson, Tiny Archibald, Calvin Murphy, Pete Maravich and Walt Frazier. Unfortunately, the NBA didn’t start keeping track of turnovers until the 1977-78 season, so there’s no assist-to-turnover data for the first four (Robertson, Wilkens, Cousy, West) and the data for Archibald, Murphy, Maravich and Frazier is incomplete, so I could only use their post-1977 numbers.

I also compiled a list of the top non-HOF point guards who are both retired and still active: Jason Kidd, Mark Jackson, Steve Nash, Gary Payton, Rod Strickland, Maurice Cheeks, Terry Porter, Tim Hardaway, Andre Miller, Muggsy Bogues, Kevin Johnson, Derek Harper, Stephon Marbury (yes, Stephon Marbury), John Lucas, Norm Nixon, Mookie Blaylock, Sam Cassell, Avery Johnson, Baron Davis, Nick Van Exel, Allen Iverson, Chauncey Billups and Mike Bibby. All of these players have at least 5,400 career assists, which seemed to be the cutoff for players I was interested in using for this study.

Lastly, I added seven of the top current point guards who have yet to break the 5,400-assist barrier: Tony Parker, Stephen Curry, Russell Westbrook, Rajon Rondo, Derrick Rose, Deron Williams and of course, Chris Paul.

I first tackled this subject two years ago, and settled on the shot-to-assist ratio to determine whether a player is “pass-first” or “shoot-first.” The higher the number, the more of a “shoot-first” player he is. To determine whether or not a player is “turnover-prone,” I calculated each player’s assist-to-turnover ratio. The higher the number, the better the player is at taking care of the ball, relative to what he’s asked to do as a playmaker for his team. The graph takes a gentle downward slope because assists are part of both calculations. (Note: While I do like FGA/A as the criteria for shoot-first/pass-first, I am not completely sold on A/TO as the criteria for turnover-prone. Perhaps (A+FGA)/TO would show shoot-first guards in a better light? Maybe I’ll try that next year.)

Here’s how the graph turned out. Click on it to see a bigger version.

Pass-first/shoot-first goes left to right, while takes care of the ball/turnover-prone sits on the vertical axis.

A few random thoughts:

– This is not typically an indictment of a player’s game. Most offenses call for a playmaking point guard, while others do not. For example, Sam Cassell was definitely a scoring point guard, but he was a great fit in the Houston offense alongside Hakeem Olajuwon, who was very adept at drawing the double-team and kicking the ball out to the open shooter. Cassell’s role on that team was to hit open shots, not run the pick-and-roll over and over in an attempt to set up other shooters (like John Stockton).

– Murphy, Frazier, Iverson and Maravich were so far to the right on the shoot-first scale that if had I shown their plot points, the rest of the graph would have been unreadable. Their shot-to-assist ratios ranged from 3.40 (Frazier) to 4.75 (Murphy). In the modern era, Iverson is probably the quintessential shoot-first point guard, even though he did average 6.2 assists per game.

– Speaking of quintessential, it’s no surprise that John Stockton and Chris Paul are in the top-left quadrant, but Muggsy Bogues’ presence is a little surprising. He couldn’t score very well (career 7.7 ppg), but his assist-to-turnover ratio (4.69) was easily the best of the players in the study.

– Other players in the top-left quadrant like Jason Kidd, Steve Nash, Mark Jackson, Maurice Cheeks, Rajohn Rondo and Avery Johnson are your traditional pass-first point guards. Magic Johnson also qualifies, but he was special because he could drop 35 points on you if necessary.

– Just because a player is not a traditional pass-first point guard does not mean that he can’t have great success. Isiah Thomas, Sam Cassell, Tony Parker and Dennis Johnson all have multiple NBA titles on their resumes, while Chauncey Billups has one of his own. This is good news for Rose, Westbrook and Curry, who are not traditional pass-first point guards.

We hear it all the time. NBA analysts call one point guard “pass-first” and another “shoot-first.” Or they say one guy is “turnover-prone” while another “takes care of the ball.” But really, what makes a player a “pass-first” point guard? How carefree must he be with the ball to be considered “turnover-prone”?

I first tackled this subject two years ago, and settled on the shot-to-assist ratio to determine whether a player is “pass-first” or “shoot-first.” The higher the number, the more of a “shoot-first” player he is. To determine whether or not a player is “turnover-prone,” I calculated each player’s assist-to-turnover ratio. The higher the number, the better the player is at taking care of the ball.

I narrowed the list of players to 33, one for each team plus a few extra for teams like Cleveland, Sacramento and Denver, who have a couple of players manning the position. I also added eight prospects (indicated in green) just to see where a few of the younger guys land. Here’s the graph — it’s small, but if you click it, you’ll get to a bigger version:

So the pass-first/shoot-first aspect goes left to right, and the turnover-prone players will be towards the bottom, while the guys that take really good care of the ball will be up top. Players indicated with a blue diamond are in the Top 10 in this group in Efficiency Per Minute. I set the axis for each category at the average of the 33 players in question, so 1.97 for FGA-to-assist and 2.70 assist-to-turnover.

Two years ago when I conducted this study, seven of the top 10 EPM performers were in the top left quadrant (pass-first, takes care of the ball). This year, only five of the top 11 (I included both Rondo and Calderon, since they tied for #11) are in that quadrant. This is due to the emergence of three shoot-first, (fairly) turnover-prone guards who are emerging as stars: Derrick Rose, Russell Westbrook and Stephen Curry.

A few takeaways:

Ty Lawson looks like a star. He is #13 in this group in EPM, so if he can continue to post big numbers in bigger minutes he’s poised to join the Top 10 sometime soon.

Kyle Lowry isn’t too far behind him. It’s not clear if Houston has decided that Lowry is their point guard of the future, but he is #16 in EPM and is well above average in assist-to-turnover ratio.

Brandon Jennings can still become a Top 10 point guard. Two years ago, I would have said that Jennings would have to drastically change his game to join the elite group of point guards, but with the way Rose, Westbrook and Curry have emerged, he can continue to be himself. One thing I’ve noticed with his game is that he has a tough time finishing around the hoop when he gets bumped, which is something that Westbrook and Rose excel at. Jennings needs to master the floater — that way, he can use his quickness to get into the lane pretty much any time he wants, and then get his shot off without taking a big hit.

Roddy Beaubois is an interesting prospect, but it’s doubtful he’s going to be a quality starter in this league… unless he changes his game to become more of a distributor. He doesn’t need to turn into Jason Kidd, but right now he’s an outlier so he needs to take better care of the ball and create a few more opportunities for teammates. You may notice that Tyreke Evans is very close to Beaubois, but he really needs to play shooting guard.

If John Wall can take care of the ball, he’ll be right there with Deron Williams. For a rookie, Wall’s numbers look great. Sure, he’s a little turnover prone, but no worse that Westbrook, Billups, Collison or Curry. He’s a playmaker, so once he cuts back on the turnovers, his assist-to-turnover ratio really has a chance to skyrocket. I wouldn’t be surprised if he’s up in the area of Maynor and Lowry with a blue diamond next to his name in a couple of seasons.

We hear it all the time. NBA analysts call one point guard “pass-first” and another “shoot-first.” Or they say one guy is “turnover-prone” while another “takes care of the ball.” But really, what makes a player a “pass-first” point guard? How carefree must he be with the ball to be considered “turnover-prone”?

Taking an analytical approach to these questions, I decided to bust out an Excel spreadsheet and try to come up with some answers. Below you’ll see a graph that attempts to classify the top point guards in the league. But first, a little background…

I chose to categorize each player based on two stats. First, to determine if he’s “shoot-first” or “pass-first,” I calculated the shot-to-assist ratio for each player. The bigger the number, the more of a “shoot-first” mentality the player has. Second, to determine whether or not a player is “turnover-prone,” I calculated each player’s assist-to-turnover ratio. I thought about using turnovers per 48 minutes, but I like the idea of including assists so that playmakers are rewarded for the positive as well as the negative. Next, I calculated each player’s Efficiency Per Minute (EPM) to see if there is any correlation between these other statistics and the overall efficiency of the player in question.

The last step was to plot each player based on the two ratios. The graph is a little small, but if you click it, it will take you to a larger version.

Each player is indicated by a separate plot point with his name next to it. (Duh.) The blue diamonds represent the top 10 point guards in terms of EPM. One point guard was chosen per team. Usually, I went with the player that got the most starts at the position.

The further the player is to the right on the (horizontal) x-axis the higher his shot-to-assist ratio and the more of a “shoot-first” mentality he has. Conversely, the further the player is to the left, the more of a “pass-first” point guard he is.

On the (vertical) y-axis, the further the player is towards the top, the better job he does of taking care of the ball (in comparison to the number of assists he posts). Players towards the bottom of the axis have poor assist-to-turnover ratios.

Each axis is set at the average of the 30 players included in the study. The shot-to-assist ratio average is 2.12, while the average assist-to-turnover ratio is 2.66.

A few interesting things to note…

As a whole, the data set takes on a downward slope. Intuitively, this makes sense. “Shoot-first” point guards are generally that way because that is their strength – scoring the ball. They are generally not as adept at being in the playmaker role, so they won’t have a good assist-to-turnover ratio. Conversely, players that are “pass-first” are that way because playmaking is their specialty. They handle the ball well and have good vision.

The sweet spot is in the second (top-left) quadrant. Seven of the top EPM performers at the position – Jose Calderon, Jason Kidd, Chris Paul, Rajon Rondo, Deron Williams, Steve Nash and Chauncey Billups – all have a better-than-average assist-to-turnover ratio and a “pass-first” menality. The other three top 10 EPM performers – Jameer Nelson, Tony Parker and Devin Harris – are reasonably close to that top-left quadrant. The 10 point guards in the top-left quadrant play for teams that had a combined record of 471-349 (.574) in the regular season and seven of those teams made the playoffs. These facts are telling.

Jameer Nelson and Tony Parker aren’t in the ideal quadrant, but they are a good fit for their teams. Think about it – Nelson and Parker play with two of the best big men in the game in Dwight Howard and Tim Duncan. Point guards that play with a great big man will find that assists are a little tougher to come by. Typically, when someone feeds Howard or Duncan, they take their time in the post before they make their move, so the player that passed them the ball doesn’t necessarily get the assist. Plus, anyone can feed the post – it doesn’t have to be the point guard. Think about those seven players in the top-left quadrant for a moment. For the most part, they play with jump shooters (and are therefore more likely to rack up assists).

The Magic are built in the Hakeem Olajuwon-era Rockets mold. They have a dominant big man and the idea is to surround him with great jumpshooters to keep defenses honest. Conversely, Parker oftentimes takes advantage of the amount of attention that defenses pay to Duncan to get to the rim. When he attacks the hole, he’s trying to score and will only pass if the defense forces him to.

Finally, it’s no coincidence that Nelson and Parker are two of the most efficient scorers in this study. In terms of FG%, Parker is first and Nelson is tied for third. Both players shoot better than 50% from the field, so it makes sense that they’d have a “shoot-first” mentality.

Ramon Sessions is a keeper. Luke Ridnour got more starts than Sessions, but I like Sessions more and the Bucks have to decide whether or not to sign the up-and-coming point guard to an extension this summer. Here’s my pitch: He has the #11 EPM (just behind Chauncey Billups) and is on the verge of becoming a very good point guard. His Player Efficiency Rating (PER) of 17.65 is #13 amongst qualified PGs. Throw in the fact that the NBA is becoming increasingly dependent on good point guard play, and the 23-year-old looks like a great (and relatively cheap) prospect. GM John Hammond has to find a way to keep this kid.

Baron Davis isn’t a “shoot-first” point guard?!? Apparently not. His career shot-to-assist ratio is 2.05, so even when we consider his entire body of work, he’s still leans “pass-first.” This is underlined by his 2008-09 campaign, where he was even more “pass-first.”

Derrick Rose turns the ball over too much. I hadn’t yet looked at his numbers, but I could tell this about Rose just watching the Bulls/Celtics series. Every so often he throws a pass that might have made it in college, but isn’t going to make it in the NBA. Most rookie point guards struggle with this, so I’d assume that Rose will start to work his way into that top-left quadrant (or at least the top-right quadrant) in the next year or two. In fact, his assist-to-turnover ratio rose from 2.52 before the All-Star break to 2.56 after. I think he’ll eventually settle in somewhere near Nelson and Parker. He seems like more of a scorer than a pure playmaker, partly do to his incredible ability to finish at the rim.

Where would a healthy Gilbert Arenas fit in? Remember him? Agent Zero has only played 15 games in the last two years, so it’s not really fair to judge him based on those numbers. To get an idea of where he’d land, I calculated his ratios for the 2006-07 season. You’ll find him bunched with Jamal Crawford, Mo Williams, Randy Foye and Aaron Brooks in the bottom-right quadrant, which makes sense because those are all “shoot-first” point guards.

Is Mo Williams’ position on the chart due to his playing with LeBron? After plotting Arenas, it made me wonder about Mo. He was one of the top assist men in the league in his last year in Milwaukee. How does the 2007-08 Mo Williams compare with the 2008-09 version? It turns out that the ’07 version of Mo Williams is a lot like the ’08 Devin Harris, which makes sense considering that both players are very capable scorers and playmakers, but at heart are “shoot-first.”

CONCLUSION

I think the big thing to take away from this is that it’s generally better to have a “pass-first” point guard who takes care of the ball. However, if you have a dominant big man (or a superstar wing like LeBron or Kobe), a “shoot-first” point guard can be just as effective, especially if he can shoot it like Nelson or Williams, or get to the rim like Parker does.

Is there anything that I missed?