A Statistical Breakdown of ACC Point Guards (Part I)

WARNING: THIS INCLUDES A LOT OF STATISTICS AND NUMBERS TO MAKE GENERALIZATIONS ABOUT PLAYERS. IF YOU DON’T LIKE DOING THIS THEN I BELIEVE YOU WILL FIND THE FOLLOWING HIGHLY BORING AND/OR CONFUSING.

The conversations in the comment sections over the past few days got me thinking about the point guard play in the ACC. Like everything else related to ACC basketball, the point guard play is very, very good right now. It started out as a glimpse at Bryant Crawford and where he ranks in the ACC, but developed into a pretty full blown statistical breakdown of point guards overall in the ACC at this juncture in the season.

I will provide a TL;DR (too long, didn’t read) in Part III (the Conclusion section) for those of you who are less statistically inclined and prefer to just see the conclusion. I provided the warning initially because I know that we don’t all share the same enthusiasm that I do as far as looking at numbers and how accurate they are in terms of ranking players.

Introduction

In today’s game there are many different versions of point guards, and the charts that are presented in the next two articles will demonstrate that pretty well. On one hand you have your typical "pass-first, manage the game, don’t turn the ball over" point guards (Quentin Snider, John Gillon, London Perrantes). On the other hand you have a newer breed of guard, effectively the combo guards (Dennis Smith Jr, Xavier Rathan-Mayes, Jamel Artis, Bryant Crawford, Ja’Quan Newton).

As would be expected, the first category of guards tend to not jump off the stat sheet at you. This is particularly true in this exercise because the "protect the ball" guards all seem to play in a low-tempo offense (Virginia especially, although Ja'Quan Newton is stuck in a slow Miami offense).

The second category of guards accumulate a lot of stats, usually both positive (points, assists), but also tend to accrue a fair amount of negative stats (turnovers).

Due to the different types of point guards in the conference, it is disingenuous to look at these statistics as the "be-all, end-all". Coaches in the ACC are very intelligent and use their personnel to determine what type of offense is going to be on the court. Or rather, they recruit the type of players that tend to conform to the offense that they like to run.

London Perrantes and Quentin Snider in particular pop up as outliers throughout this. Most people who watch the ACC would put Perrantes in their top tier of ACC guards, but statistically (even pace-adjusted this season), he simply does not fall into that category with other high-scoring guards. Yet I guarantee you there is almost no other guard in the ACC that Tony Bennett would rather have leading his team right now than Perrantes (well, maybe Dennis Smith, Jr., who we will get to later on).

The fit of the point guard with the team matters greatly, and I will be the first to say that statistics don’t always demonstrate who the "best" guard is in any capacity. They usually do a very good job of showing what a player is good and bad at, but the overall value of a player is usually dependent on his role within the team.

Methodology

To determine who the point guard was for each team I simply used Ken Pomeroy’s assessment of who is most likely the team’s point guard. He has put together a formula based on which statistics are likely to be high or low based on position. For example, the point guard is usually (but not always) the player with the highest assist rate on the team. As he says on the website: "position designation is estimated by an algorithm and may not reflect reality."

One questionable "point guard" that will likely be is Grayson Allen as the Duke point guard over Frank Jackson. This makes sense because Frank Jackson is a "shoot-first" guard, while Allen leads the Blue Devils in assists and is a combo-guard. Does Jackson carry the ball up the court after a made basket? Yes, I believe so, but Allen plays more of the facilitator role in that offense, therefore I went with Allen as Duke's point guard.

The only time I broke from this method was in making Jamel Artis the point guard at Pittsburgh. KenPom believes that over the past five games it has been Justice Kithcart, but Artis was certainly the point guard to begin the year. Kithcart only plays 40% of the minutes, and Artis has a higher assist rate. Therefore, Artis is the point guard for this statistical examination

I broke down point guards in two manners: using raw statistical numbers, and using pace adjusted numbers (thanks to Ken Pomeroy and College Basketball Reference). I will do my best to explain any questionable terms, and included footnotes for "pace-adjusted" definitions in Part II, but if at the end there are questions or clarifications needed, please post them in the comment section and I will answer them as they come up.

This is through all games as of January 4th. I wanted to do it after non-conference play but got a little behind with the holidays, so here we are.

After I gathered this data, I used a Borda Count ranking method (the same one that college polls use). If you are first in a category you get 15 points, second gets 14, third gets 13, and so on until the last ranked player in the category gets 1 point. I applied this to each category, added each category up, and thus a ranking system for raw numbers is applied.

Raw Numbers

While raw numbers certainly have their problems (they are very, very pacist against slow teams), a lot of folks like to see those in their glory, so I ran those as part of the breakdown, but to also demonstrate how poor of a job they can do in reflecting reality sans context.

I included what I figured were the most important stats: Points Per Game (PPG),  Rebounds Per Game (RPG), Assists Per Game (APG), Turnovers Per Game (TPG), and Steals Per Game (SPG). I also included the minutes that each player plays per game (MPG), but did not include them in my "ranking" system.

I had to organize the table in some way, and rather than do it alphabetically, I just did it by points per game. Congratulations Jamel Artis, you are first in this table since you lead ACC point guards in PPG.

As the introduction stated, there are many different types of point guards in this conference. Jamel Artis is very clearly a scoring guard who grabs rebounds. He has a score-first mentality, and that fits very well for Pittsburgh, who, other than Michael Young is primarily comprised of role players well under the 20% usage rate mark.

This is the first time that freshman phenom Dennis Smith, Jr. sticks out, and it is for good reason. He got the first triple-double for N.C. State in over a decade last night, and he is averaging 19.6 PPG, 4.1 RPG, and 6.5 APG for the season. This is why he is very likely to be a top 5 pick in the draft, and perhaps No. 1 over Lonzo Ball depending on what team winds up drafting first.

From a Wake standpoint, Bryant Crawford is 6th in the league in scoring, 2nd in Assists, 2nd in steals, and 13th in turnovers. This pretty much plays into the narrative that we have of him—-he fills up the stat sheet, but turnovers can be a problem.

Another theme begins to develop, and that is the tiers of ACC point guards, primarily the bottom with Georgia Tech freshman Justin Moore.

Now that we have seen the raw numbers, let’s break them down into a ranking system for context:

As the table above shows, Dennis Smith, Jr. pretty much dominates everybody else when it comes to the raw numbers. He leads the league in assists and steals per game at the point guard position, is second in points per game, and third in rebounds per game. His downfall is the turnovers (sound familiar Wake fans?). He is an exceptional player, and he fits very well into N.C. State’s offense, and quite frankly any team that would be lucky enough to have him.

What’s that though???? Is that Bryant Crawford tied for second place with The Grim Tripper??? Yes it is! Crawford rides his high assists and steals per game totals (which would both be first if it weren’t for the freak of nature Smith, Jr.) to a tie for second with Allen.

Allen relies on PPG, RPG, and a low TPG mark to grab second place. From there it files down pretty much like the eye test would tell you. Joel Berry is having a very good year at UNC, while Matt Farrell of Notre Dame rounds out the top 5.

Taking a quick look at the bottom there indicates that there is a pretty sharp drop in the final two guys of Shelton Mitchell and Justin Moore. This makes sense from a logical standpoint because Moore is a freshman, while Mitchell is in his first year with Clemson. Continuity and familiarity matter in college basketball (except when it doesn’t in the case of Dennis Smith, Jr.)

I can’t repeat this enough, but this chart is a very good example of why pace-adjusted statistics are far superior to just raw numbers. Perrantes and Snider are certainly not the 12th and 13th ranked point guards in the league, but because they play on a team with slower than average tempo possession within their team’s offense, it produces a very low statistical output from a raw vantage point.

That wraps up Part I for the introduction, methodology, and raw number assessment. Check out Part II for the pace-adjusted numbers, questions about my methodology, and footnotes. Part III holds my conclusions and what I drew from the data available.