November 13, 2007
The Displaced NBA Fan
On Tempo and Game Scores

by Bradford Doolittle

Life as a displaced NBA fan--one living in a town without a team of its own, like my home of Kansas City--hasn't always been as easy as it is now, in this era of myriad media options. We used to have a team here. Now that we just recently opened the doors on a sparkling $276 million arena, we'd like another. That's a story for another day. For now, the reality for a KC/NBA fan is that we have to get our fix from League Pass, the Web and an occasional road trip to catch a game. It's really not so bad.

It used to be harder. When I was growing up in a small Iowa town about three hours north of Kansas City, access to the league came in three ways: radio, The Sporting News and the national Lakers/Sixers/Celtics game of the week on CBS. Once the sun went down, we could pick up the distant Amplitude Modulated waves of ballgames from KC, Chicago, Dallas, Denver and, on a good night, Salt Lake City. My brother and I were transfixed by the scoring exploits of Alex English, Kiki Vandeweghe, Mark Aguirre and Adrian Dantley. When the weather was good, we'd spend afternoons in the driveway trying to emulate them. Heck, even now, one of my great pleasures is to while away a half-hour shooting jump shots in the driveway of my own house, though I don't really have any players in mind when I do so. Something in the imagination has been lost.

Whereas I no longer yearn to be an NBA player (that notion floated away at the onset of puberty), my fascination with the game has never diminished. My first attempt at trying to objectively rate players came way back in those Iowa days. We'd acquired a stat-based tabletop game from a toy store in Omaha, Statis-Pro Basketball. The set of cards we had were based on the 1978-79 season and, thus, were outdated by a few years. After learning the rules, we formed a league and began generating box scores in earnest. We acquired some graphed ledgers made for home accounting purposes and used them to compile our season stats.

Around the same time, I began playing junior high basketball. My coach was also our math teacher and he had a system of rating us where he'd give us, say, two points for a basket, one for a rebound, one for a hustle play, minus one for a blown defensive assignment, etc. I was transfixed. I began to employ a similar system for our Statis-Pro league.

It didn't take long before I wanted to start looking at the real thing. Who was better: Magic or Larry? Moses or Kareem? Dantley, English or Aguirre? The only problem was lack of information. So I borrowed my aunt's typewriter and pounded out a heartfelt screed addressed to the NBA League Office, outlining my idea for "the ultimate basketball stat" and telling them all I needed was the data to make it work. Somehow or another, I found the correct address and sent my letter off to New York.

A month or two later came my response: an official NBA envelope with all the basic statistics from the previous season that are so easily found today with the click of a mouse button, all printed on official league letterhead. There was no personal response to my letter, no grant for my research, no enthusiam for my revolutionary ideas, but that didn't matter. I treated my packet of statistics like a precious work of art.

I don't remember the results of my effort nor the formula that I'd "invented." It probably was something similar to the standard efficiency rating that you still see employed at NBA.com. (This system was popularized by Martin Manley in the late 1980s in his Basketball Heaven books. Martin is pretty much out of the NBA analysis business these days, but he analyzes just about everything else. He's now a colleague of mine at The Kansas City Star, and another displaced NBA fan.) My formula, and my packet of statistics, were lost a long time ago.

Analysis is a lot easier these days, thanks to the availability of information on the Web, with great sites like 82games.com, dougstats.com and basketball-reference.com allowing you to uncover just about any piece of NBA information that you desire. Every night during the season, I code up the box scores of each game and import them into my Excel-based analysis system, called NBAPET. We'll get into the nuts and bolts of NBAPET a lot more as the season goes along.

My passion for the NBA and my ongoing quest to understand the game's underpinnings have not dimished since those early efforts in a country house, surrounded by Iowa corn. For what it's worth, I still play Statis-Pro basketball, even though the Avalon Hill game company quit making it years ago. It's awesome to be able to share my fervor for the professional game with you at Basketball Prospectus.

---

With the season two weeks old, you've probably consumed all the preview content that you care to ingest. I sketched out my team-by-team thoughts and went on record with projected records in my preview at my blog. My finals pick is the same as Kevin Pelton's: the Suns over the Bulls.

Rather than write another preview, I'll instead address my foremost desire for the upcoming season: That the trend towards up-tempo basketball continues to spread across the league.

I suppose that, as someone who came of age in the 1980s, I'm partial to the style of basketball that was played in those years. Teams knew how to execute a fastbreak and typically did so at every available opportunity. The aim was to get the first available open look, which usually came before too much time had burned off the 24-second clock.

In the period since the NBA/ABA merger, scoring peaked in 1984-85 with an average team putting up 110.8 points per game. Even then, the game was already starting to slow. In the 1977-78 season, teams averaged 110.4 possessions per game. By 1985, it was down to 105.5. The average slipped under 100 for the first time in '93-94 and hasn't reached triple digits since then. Scoring has been as low as 91.6, achieved by a bunch of rusty NBA players after their strike truncated the '98-99 campaign.

Why did the pace slow? There's no single answer for that question. Certainly, the addition of the three-point shot had something to do with it. Changing in coaching philosophy, ushered in by the defense-oriented Pistons' championships of 1989 and 1990, probably had an effect. The underdeveloped skills of high-school players who, beginning with Kevin Garnett, invaded the league didn't help matters. The rules governing defense contributed as well, leading to a glut of shot-clock-milking, pick-and-roll offenses that left two or three disinterested offensive players camped out on the three-point line. The Rockets won two titles that way.

Whether or not the controlled pace and emphasis on defense employed so successfully by the Spurs is more effective than a style like that of the Suns is very much an open question. The Spurs have four titles, an effective statement for the defense camp, but that style suits their personnel and was dictated by their good fortune in winning the 1997 draft lottery, the year in which Tim Duncan was turning pro. The Suns have a championship-caliber roster and if not for the unfortunate circumstances during last year's playoff matchup with San Antonio, they could have taken the flag last year. If they had, it might have been a great boon to the cause of up-tempo basketball.

Still, there are signs that things are picking up. Last year's rate of scoring (98.7 points per team per game) was the highest since 1995-96. Possessions (95.1) were the most since 1994-95. The pro game seems to be loosening. Could this be the season the average once again nudges over the symbolically-significant mark of 100 points per game?If it does, it again won't be because of any one reason.

Shooting in the NBA has been improving for a few years now. Two-point field goal percentage dropped from .499 in 1984-85 all the way to a non-strike-year low of .459 in 2003-04. It's gone up the last three seasons, however, reaching .484 last year, the highest mark in 11 years. Three-point shooting has been getting better almost annually since the rule's adoption, save for a trend-skewing blip from 1994-97 when the league brought the line in closer. Last year's .361 mark was the highest-ever rate at the current distance; teams are shooting at a .358 clip in the early going this year.

The fact that possessions per game rose last season was a good sign. When you look at scoring on a per-100-possessions basis, it's higher than it was before the scoring surge of the 1980s. Last season, teams scored at a rate of 103.8 points per 100 possesions. During the first two seasons of the Magic/Bird era, the rates were 102.6 and 102.8, yet scoring was more than 10 points per game lower last season than in 1979-80. So the problem isn't offensive efficiency--it's pace of the game. Let's break out a chart:

Year    P100    Tm    UT     %UT
2007    98.2    30     9    30.0%
2006    96.3    30     7    23.3%
2005    96.6    30     5    16.7%
2004    96.9    30     3    10.0%
2003    99.8    29     0     0.0%
2002    99.2    29     0     0.0%
2001    98.4    29     0     0.0%
2000    99.6    29     0     0.0%
1999    98.7    29     4    13.8%
1998   100.6    29     0     0.0%
1997    98.0    29     0     0.0%
1996    96.3    29     3    10.3%
1995    95.4    29    13    44.8%
1994    94.9    27    20    74.1%
1993    96.6    27    21    77.8%
1992    95.1    27    27   100.0%
1991    95.1    27    26    96.3%
1990    95.4    27    26    96.3%
1989    95.3    27    26    96.3%
1988    95.5    25    25   100.0%
1987    95.1    23    23   100.0%
1986    95.0    23    23   100.0%
1985    95.7    23    23   100.0%
1984    95.2    23    23   100.0%
1983    95.4    23    23   100.0%
1982    98.0    23    23   100.0%
1981    95.8    23    23   100.0%
1980    97.3    23    23   100.0%
1979    97.5    22    22   100.0%
1978    98.9    22    22   100.0%
1977   101.7    22    22   100.0%
1976   103.0    22    21    95.5%

KEY: P100 - the number of possessions required by a team of average offensive efficiency to score 100 points per game; Tm - number of teams in league; UT - number of teams whose possessions per game exceeded Poss100; %UT - percentage of teams playing up-tempo styles.

You'll notice first off that while almost every team used to play at a fast enough tempo to score 100 points per contest, for a while there, no team did. A few teams did manage to crack the 100-point barrier by employing efficient offenses. Still, it was an ugly style of basketball being played in the early years of this decade. A hardcore hoops addict like myself can watch NBA ball no matter what, but it's likely that the slow pace cost the league more than a few fans.

The seven teams that played up-tempo last season were Denver, Golden State, the Lakers, Memphis, Phoenix, Sacramento and Washington. There are nine teams up-tempo so far this season, though with just 95 games in the books, it's way, way too early to jump to any conclusions. The nine include all of last year's teams except for Sacramento. The new up-tempo squads are Utah, Indiana and Seattle.

This is all very encouraging. Teams that get up and down the court are more enjoyable to watch and, combined with the improved shooting around the league, should help to bring back some of the fans the NBA has shed since Michael Jordan retired. But up-tempo teams have to prove that they can win. We're completely objective here, but it would be a great thing for the NBA if the Suns won a championship. Why? Because fans of the fast break have to overcome a startling stat: Since 1995, when scoring first dipped under 100, teams that have played up-tempo ball have won at a rate of just .469. The slower teams (which outnumber the latter about 12 to 1) are at .504.

So go Suns … and Jazz … and Nuggets … and, well, you get the idea.

(Note on possessions per game: It's hard to find two figures on the Web that match when it comes to this statistic. The formula I use is the same as the one at basketball-reference.com, yet my figures rarely match theirs. As I mentioned, I get my raw data from box scores taken from NBA.com. One possible reason for the discrepency is a difference in how turnovers are tabulated--some folks are missing out on team turnovers, like 24-second violations. Mine have those team figures in there. As a result, my possession figures tend to be just a shade higher than some of the others I've seen. There may be other reasons for the difference but if there are, I haven't figured them out. What would unify things is if the NBA would simply keep track of possessions. Note to commish.)

---

The format for this column will evolve. I'll probably start out by listing each week's leaders and laggards in Game Scores for that week. There are various versions of game scores out there, the most prevalent likely being the one used by John Hollinger. You can also look at the NBA's daily "efficiency" recaps as being a form of Game Score. My version is different in that it includes a component for each player's box score counterpart. (I'll post a glossary with my specific explanations and formulas in the near future, for those who want to play along.) It works like this:

Say in a Dallas/Phoenix game, Dirk Nowitzki gets an efficiency score of 34. Efficiency score is based on applying linear weights to the various box score categories. His box score counterpart, say Shawn Marion, gets an efficiency score of 26. However, we know that based on Marion's season statistics and minutes played in that game, his expected efficiency score was actually 29. Thus Dirk held him three points under his typical score. Those three points are added to Dirk's total for a Game Score of 37. If Dirk's typical score and minutes played for that game yielded an expected efficiency of 30, then Marion gets docked the four extra points Dirk scored. Thus Marion gets a Game Score of 22. (This method of counterpart +/- is a key component in my system for rating individual defense. The Game Scores are just an offshoot of that.)

Since a hot shooter or a travel-weary performance can skew Game Scores dramatically, they are really used for entertainment purposes only, not as a rigorous evaluation metric. They are a fun and quick way to look at daily and weekly leaders. We'll use these leaders and laggards to drill down into various issues and trends as the season goes along, and we'll try to stay topical, reacting to league news and player movement as it occurs.

Whew. That's plenty for an intro column, I should think. I'll leave you with the Game Score leaderboard for first two weeks of the NBA season. Yes, Jason Hart has looked as bad as his Game Scores suggest.

Game Score leaders, week one/two:

11/3    Al Harrington       58
11/9    Kevin Garnett       52
11/7    Sam Cassell         52
11/2    Kevin Garnett       49
11/6    Kevin Martin        49
11/10   Leandro Barbosa     48
11/3    Deron Williams      48
11/7    LeBron James        47
11/7    Paul Pierce         45
11/2    Kelenna Azubuike    45
11/9    Kobe Bryant         45
11/6    LeBron James        45
11/9    Tim Duncan          44
11/1    Tracy McGrady       44
11/10   Carmelo Anthony     44
11/9    LaMarcus Aldridge   43
11/7    Steve Nash          43
11/5    Jason Terry         43
11/4    Steve Nash          43
11/7    Hedo Turkoglu       43
11/6    Desmond Mason       43

Game score laggards, week one/two:

11/6    Chris Bosh         -20
11/9    Damon Stoudamire   -16
11/7    Jamaal Tinsley     -15
11/11   Desmond Mason      -14
11/4    Primoz Brezec      -13
11/7    Dwayne Jones       -12
11/10   Keyon Dooling      -12
11/7    Anthony Parker     -12
11/2    Quentin Richardson -11
11/7    Linas Kleiza       -11
11/6    Luke Walton         -8
11/9    Jordan Farmar       -8
11/2    Darrell Armstrong   -8
11/3    Kirk Hinrich        -8
11/4    Melvin Ely          -7
11/4    Jason Hart          -7
11/2    Ryan Bowen          -7
11/1    Jason Hart          -7
11/7    Jason Richardson    -7

Bradford Doolittle is an author of Basketball Prospectus. You can contact Bradford by clicking here or click here to see Bradford's other articles.