Pre-70s ORtg/DRtg officially posted on b-r now.

[Doctor MJ]

Here's the article:

http://www.sports-reference.com/blog/20 ... 1951-1973/

Our own ElGee gets name checked in it too.

Starting to look at the data now.

[Dipper 13]

In the NBA before 1971, I had to drop the opponent side of the possession formula, making it:

Poss = Tm FGA + 0.4 * Tm FTA - 0.321 * (Tm FGA - Tm FG) + Tm TOV

(The 0.321 part comes from ORB being 30% of TRB; 1.07 * (3/10) = 0.321.)

Exactly how was the total Team TOV number determined here?

[Dipper 13]

This is a solid addition to the site. However it is best to keep in mind that these are just estimates based on fixed percentages which assume every single team in the league rebounded their offensive board and turned the ball over at the exact same rate for well over a decade.

[Chicago76]

This is definitely a step in the right direction. My only real issue with how things were done is that ORB rate was in all likelihood set substantially too low. There are a lot of reasons to believe 73/74 to 75/76 was anomalous with respect to ORB rates, so using these to form estimates for prior years is pretty suspect. This won't change the relative Ortgs or drtgs of teams relative to lg average materially, but it does lead to artificial supression of Ortg. Maybe I'll write a post on this more extensively at some point, but basic reasons to assume ORB rates pre-73/74 are higher.

1) Regression is only as good as the years you use in the sample if looking at team box score data. If you use 73/74, your ORB rate will be anchored around 30.3. If you add additional years until the 3 pt shot, it will move up to 30.8, 31.0, or 31.3, depending upon if you add another 1, 2, or 3 years. This only makes sense if 73/74-75/76 are decent sample years, and they aren't for a few reasons.

2) Age. The league was getting older pre-merger. Age/experience negatively correlates to ORB rate. A typical ORB curve of a "good" player (one getting good minutes consistently for 8 or so years) declines substantially from yr 1 to yr 2 (10.5%) and appreciably every year thereafter (2.5% per year). An older league is a smarter league, so guys aren't wasting energy being overaggressive on the glass and gassing themselves over a full season.

3) Pace. 82 games did a nice ORB rate study. ORB rate is highest off a shot 0-10 sec into the shot clock. 60-61 for example had a scoring opportunity (FGA + 0.4 FTA every 11.7 seconds or so). With pace being so high, there was a lot of transition ORB putback chances before ORBs were recorded.

4) "Shot creation" Ability to create confuses and disorganizes a defense. An easy way to measure this is FGA/0.4 FTA. Basically the inverse of FTA/FGA. There is a very good relationship between FGA/0.4FTA and league ORB rates. League being the key word (vs. team). Team rates vary too much based upon talent, which isn't normally distributed. Example: pace may not be meaningful at the team level. Teams may run because they're athletic (good ORB rates) or teams may run because they aren't athletic enough to break a team down in a half court O (poor ORB rates). On a league basis, the relationship is telling. A high FGA/0.4FTA suggests poor defensive breakdown ability. From the shot clock inception through 71/72, that ratio stayed in the 6 to 7.5 range. 72/73 through 75/76, that ratio was historically very high (8.5 to 9.6). For the three years we have ORB rate data in this period, ORB rates were very low (below 31). The FGA/0.4FTA ratio dropped 76/77 and stayed below 8.3 until 2001/02. ORB rates were always between 31.8-33.4 until 94/95, which is when the 3 pt shot really took off (10% or more of total FGAs were 3s for the first time and the line was moved in).

5) Historical context. The league really matured in the mid-60s with a talent influx. There was little expansion to date and younger guys (who typically have higher ORB rates) were finding it tougher to break in by that time. League pace was declining, and shot creation was down. The game was getting more labored. Then the ABA came along. It took them a few years, but they started winning the draft war. In the mean time, they did get some excellent creators in guys like Roger Brown, Connie Hawkins, and they wooed guys like Barry and Cunningham. By 72, the league was at near parity and players were segregating along stylistic lines. The ABA was more creative wing and dirty work big heavy while the NBA had more technicians (inside and outside). The FGA/0.4FTA ratios of both spiked at this point of divergence, perhaps indicating that you need an optimal mix of all four types to get the foul draw rates up. Curiously, the ORB rates of the ABA stayed up, while the NBA did not. Fewer threes might have had something to do with this.

FWIW, a very basic regression ORB rate to FGA/0.4xFTA (NBA only) suggests an ORB rate of around 30 for 72/73, 32.8-34.3 going further back to 57/58 and 34.5 to 35.5 for the first three years of the shot clock.