Understanding Attempt Metrics in the NBA
I spent years scraping and cleaning box score data for a fantasy analytics project, and the single most frustrating thing was tracking field goal attempts over time. The numbers sound simple on the surface but the history of Field Goal Attempts Per Game NBA History is actually a mess of definition changes, missing data eras, and stat correction issues that most casual readers never notice. The NBA's official stats page has per-game averages going back to the 1949-50 season, but their default view only shows regular season data. Playoffs require a separate selection. Basketball-Reference is better for historical consistency, but even their data has caveats you need to understand before you use it for anything serious. Their player career pages list FGA per game in the per game tables, and the league average charts go back further than most people realize. The problem starts when you try to compare eras. The league attempted about 44 field goals per game in 1960 and roughly 88 in 2024. That gap is not a metaphor for "the game changed." It is a literal doubling of volume driven by pace, shot clock introduction, and rule changes. Anyone who looks at raw FGA without adjusting for pace will draw completely wrong conclusions.
I once built a model that compared players across eras using raw FGA rates and got results that made zero sense. Kareem Abdul-Jabbar looked like a volume shooter compared to modern forwards when he was actually more efficient per possession. The fix was switching to pace-adjusted metrics. I started normalizing every attempt rate by dividing by the league's average possessions per game for that season. That single adjustment removed about 70 percent of the era-comparison noise from my dataset. It took me maybe twenty minutes to implement once I figured out the right formula. Without it, the model was useless for cross-decade analysis.
Common Pitfalls When Reading These Numbers
The first mistake people make is treating FGA per game as a standalone counting statistic. It is not. A player averaging 22 FGA in a season with a 95 possession pace is doing something completely different than a player averaging 22 FGA in a season with a 105 possession pace. The possession context changes everything about how you interpret that number. Another issue is free throw attempts. The NBA started recording foul data inconsistently before the 1970s. Some seasons have incomplete FTA records, which breaks any true shooting percentage calculation you might want to build. If you are pulling data from the 1950s or 60s, expect gaps. Basketball-Reference marks these with asterisks but does not always flag how severe the gaps are. The three-point line was introduced for the 1979-80 season but was barely used for several years after that. Any analysis that lumps FGA together before and after 1979 without separating made shots by range will mislead you. Players like Rick Barry and Dominique Wilkins had radically different shot profiles depending on which side of that line you place them. The league average FGA did not jump because players suddenly started shooting more. It jumped because the shot value changed and rosters shifted accordingly.
Get the Full Details
I also ran into a data discrepancy once where the NBA's official stats showed slightly different FGA totals than Basketball-Reference for the 1983-84 season. The difference was small but it accumulated across hundreds of players and broke my aggregation script. The workaround was writing a reconciliation function that flagged any season where the two sources diverged and then manually cross-referencing the playoff games. The NBA official numbers tend to exclude certain exhibition or pre-season records more aggressively than B-R, which leads to subtle inflation or deflation depending on the era you are looking at. For anything older than 1985, I default to Basketball-Reference and note the discrepancy in my documentation. It saves you from publishing numbers that look clean but are technically wrong.
What the Long-Term Trend Actually Shows
Field goal attempts per game climbed steadily from the 1980s through the 2010s, peaking around 2019-20 when the league averaged roughly 88 total attempts per game. Since then it has plateaued and dipped slightly as pace has come down and three-point volume has stabilized. The modern game is still higher volume than ever, just not exponentially so. At the player level, the leaders in career FGA per game include players like James Harden, Stephen Curry, and Russell Westbrook, all of whom consistently average between 17 and 20 attempts per game over their careers. Those numbers are extreme even by modern standards. The league average hover closer to 16 to 18 for primary ball-handlers and 10 to 14 for role players. Anything above 22 consistently over a full season is rare and usually belongs to a player on a bad team with no other shot creators. If you are building models or writing analysis, I would recommend pulling the data from Basketball-Reference's CSV export tool rather than scraping their pages. It is cleaner, it includes pace data alongside the traditional stats, and it lets you filter by conference and era without jumping through hoops. The free version exports up to 100 seasons of player data in one download, which covers virtually everything you need for historical analysis. Their API used to be free and is now paywalled, which pushed a lot of people toward the CSV method anyway.
The real takeaway is that FGA per game is useful but only when you contextualize it properly. Raw numbers lie. Pace, era, team context, and rule changes matter more than most people realize. Once you account for those factors, the data tells a much clearer story about how shooting volume has evolved and what it means for any given player or team.
