With all the irrepressible superlatives the draft season brings, a few dirty words make their way into our collective lexicon. Tweener. Below-the-rim player. Combo guard. The last of these has killed many a draft stock price, sending fine college guards into the late first round purgatory generally filled with white power forwards, 5'11 point guards, and 18-year-old Frenchmen no one but the Spurs have heard of.
This year, several serious prospects have been cursed with the Combo Guard label, a nametag which could send lottery talents to the 20s, first round players to the second. But what the hell does Combo Guard even mean? Could front-office draft analysts even pick a Combo Guard out of a real NBA line-up?
In NBA analysis, few even care about the classification of players beyond the arcane positional taxonomy or measures of quality. A player is a good point guard, or a crap power forward, or an average swingman. Every metaphor weighs almost solely on final justification (or renouncement) of a player's aptitude at basketball in total, not in the vital aspects of basketball. Why? Few players (LeBron, for one) can excel at everything, so why do we measure players by their ability to do everything?
Even in deeper consideration of the point guard position, quality assessment lords over assessment. John Hollinger has a metric named Pure Point Rating. By design, it is an improvement on the contemptible assist-to-turnover ratio. In short, Hollinger's PPR discounts assists by a third, and adjusts for minutes and team tempo. In the end, you get a list of good point guards with particular emphasis on the top assistmen. Quality is the central component -- Paul, Calderon and Nash topped the list, in that order.
But quality is not the same as purity. Hollinger has an affinity for naming his metrics to suggest improvement over the standard fare (True Shooting percentage, for example), so perhaps he means PPR as "a point guard rating freed of the oppressive shackles of overweighted assists and slow teams." But PPR certainly does not measure a point guard's purity. Would anyone argue Paul as a more pure PG than Nash or Calderon, or Jason Kidd? Of course not, by our common definition of purity (one who looks to pass first, score second).
So how do we aptly measure purity? By taking quality out of the equation. We do not judge here, we only describe. Let's take measureable point guard activities on one side (that'd be assists) and measureable impure PG activities on the other (that'd be shots, which can be boiled into FGAs and FTAs). We'll adjust FTAs to account for the reality of 2-shot fouls ('FTAs x 0.44' is the standard here). We'll find the league average for point guards and adjust accordingly to get a nice round scale. And, of course, we'll stick a catchy name on it and slap it into a shiny graphic. Thus, The Purity Scale.
photo
It's a 0-100 scale, with higher numbers telling of a higher level of Point Guard Purity. Above, 0-100 moves left to right. Leandro Barbosa (not really a PG) sits at 23 and Jason Kidd tops out at 97 (!). The scale includes only those PGs who a) played at least 800 minutes in 2007-08, and b) averaged at least 30 minutes per game. In the list here, I have added those PGs who averaged 20 minutes per game, but still topped 800 total minutes.
Again, this is not a ranking of quality. This tells you only who are the most pure (and impure) PGs in the league. Unsurprisingly, Kidd, Nash, Calderon are quite pure. Monta Ellis? Jason Terry? A.I.? Not so much.
Based on the groupings we see on our Scale, there are a lot of incredible "combo guards" in our league. Baron Davis. Reigning Finals MVP Tony Parker (who has roughly three shooting possessions for every assist he makes). Chauncey Billups. Of course, some of these can be attributed to the offense system these players use. Famously, San Antonio basketball results in few assists (though I'll note that Avery Johnson averaged 7+ assists in 31 minutes in SA's first title season, playing in the exact same offense, versus 6 assts in 33 minutes for Parker this year; Parker is less pure than AJ, certainly).
All four teams still competing boast impure PGs, combo guards who shoot at least twice as often as they earn an assist. Two (Billups, Parker) are considered among the best in the league; the others (Rondo, Fisher) are considered solid at worst. Yet we curse every prospect without the vision of Kidd (his team lost in five in the first round) or Nash (ditto), relegating these lepers to the bargain bin.
Chris Paul stormed the universe this year. Deron Williams was also brilliant. These matters seemingly have the greater Chicago area set on Derrick Rose, lauding THE RISE OF THE PURE POINT GUARD. And while these fellows are more pure than most, they aren't holy angels of unselfishness and good intentions. Paul was a 20 ppg scorer, for goodness sake. And even more, Rose's freshman totals would place him with Marbury on our scale!
Our perspective on what makes a point guard great is seriously warped, and I blame it all on the false heralding of the assist as a game-changer and of purity as the singular path to point guard greatness. Because we believe assists to be of utmost import, and because pure point guards are more valued than scorers, we consider PGs who get lots of assists to be pure and thus, the best. They supposedly raise the game of their teammates. They make everything offense easier. They lead, muzzled or not, because they pass. It's malarkey (and I offer Jason Kidd as proof).
This map shows a team's offensive rating (points per 100 possessions) on the x axis and a team's rate of assists per made field goals on the y axis. Good offensive teams to the right, bad ones to the left. High assist teams up top, low assist teams on the bottom. Based on conventional wisdom, you'd expect to see all the good offenses in the top right (with lots of assists) or the bad guys in the bottom left (with few assists), right?
Wrong.
photo
There's no measureable relationship between a team's rate of assists and a team's offensive performance. (For the statgeeks, the correlation for the past four seasons is -0.011.) There's only a tiny correlation between ast/FGM and effective FG% (0.061). It's the same for ast/FGM and two-point FG% (0.065). You can be good at offense and a good passing team (Utah, Phoenix), but you can also be bad at offense and a good passing team (New Jersey, the Clippers). You can be good at offense without racking up assists (Golden State, Orlando). The two concepts, on a leaguewide basis, are not dependent on each other.
This is not to say assists should be ignored. Many systems require ball movement. Quite honestly most personnel in the league (including every player on the 2005-08 Phoenix Suns not surnamed Nash or Diaw) needs help from a good pass to perform at peak efficiency. But as described in a thorough Wall Street Journal Numbers Guy blog post by Carl Bialik, the assist's measureable value is in serious question. Quite possibly, no assist is created equal, and our rash cookie-cutter generalizations only cloud the value of the pass, discounting the truly vital (an alley-oop, a backdoor bounce for an easy layup) to the level of the coincidental (the shuffle pass which leads to single-clutch fall-away from 15).
But despite the veil over the value of the pass, we demand purity. Some of our best point guards (Parker, Billups) are not pure in any sense, but we shun the combo guard. On a league level, assists having nothing to do with offensive production, but we dismiss those who can't drop eight dimes a game. This is all very absurd, as four conference finalists are showing us right now. It must stop.
And my dear Combo Guards, please do not try to become something you're not. Jerryd and Russell and O.J. and even you Derrick: Screw your reputation. Fire away, brothers. Fire away.
