On his last couple podcasts, Bill Simmons wonders why home court advantage doesn’t matter anymore. He listed several possible reasons, which were mostly moronic. Bill, if something impacts both the home and visiting team equally, it has no explanatory power. Obviously. Let’s look at this a bit more rigorously than he did.
Is it true?
First, is the assumption true? Is the home court advantage diminishing?
In the regular season, the home court team has gone from winning ~60%+ up to the 1990s to ~55% today. Vegas bakes in 2.5-3 points for the home team, it used to be 3-4.5.
What about the playoffs? This is harder to figure because home court is given to the better team. But here again we’ve gone from ~65% to ~58%, and Vegas bakes in 2.5-3 points, though it varies heavily on context.
Conclusion: Yes, the assumption is true.
Do we know why?
Since the phenomena is true, the next step is to see if we already know why it happened. The answer is yes. It’s been studied and known for decades that the home court advantage comes primarily from unconscious referee bias. They don’t mean to, they don’t want to, but referees are human and react to the fans yelling. One of the ways we can see this is how home court advantage completely disappeared in the bubble.
If home court advantage comes from referee bias, has it changed? Yes it has, and that is the answer. Refereeing has become increasingly professionalized. Modern refereeing is heavily scrutinized, graded by micro-cameras, subjected to centralized NBA replay center, play-by-play reports of the last two minutes, increasing social attention, etc. The days of cowboy referees are long over. For all the jokes about Scott Foster, individual referees just don’t impact a game like they used to.
There you go, that’s the answer.
Other factors:
Three-point shooting: The rise of three point shooting increases scores for both teams, but more importantly increases the variance in the outcomes. Increased variance eats away at the home court advantage. Think of an exaggerated case – let’s say at the end of each game, each team takes one shot. If the shot goes in, they get 100 points. In this imagined scenario, the variance would be enormous, swamping everything else that goes on in the game. (It’s analogous to Quidditch, where the optimum strategy is to ignore everything except getting the Snitch. Everyone on the team should be a seeker.) Note that in the last five years of playoffs the Boston Celtics (who shoot 3-pointers aggressively) have a win-loss percentage 10 points higher when playing on the road than at home.
Easier travel: Every team has chartered flights, most have team planes built especially for their comfort. They stay in nicer hotels. The NBA has reduced the number of back to back games. All of this means that the visitors are in much better shape than they used to be. This also applies to the bubble year control case, so there’s some conflating variables there.
It’s an AI fight, Gemini vs ChatGPT! Actually you make a very convincing case for 3-point shooting being a bigger factor than reffing. I may have to fold! At least I’ll say they’re very even.
3-point shooting: While the systemic “cowboy” referee bias was significantly curbed by these reforms, the data shows that the macro-decline of HCA did not move in direct lockstep with referee accountability milestones. The continuous, linear decline of HCA down to historic lows near 52% to 54% in the late 2010s and 2020s maps cleanly onto the league-wide explosion in three-point volume (which climbed from an average of ~18 attempts per game in 2010 to over 35 attempts per game by 2020).
A related factor I was thinking about but didn’t post: Unconscious referee bias manifests most on subjective, high-contact plays—inside play. Not as much for outside shooting. So even outside the variability of higher points for those shots, they are inherently more immune to referee bias.
This is a pretty interesting history. It leaves out the growth of social media. And analytics – teams actively scout referees just like their opponents.
Introduction of Instant Replay 2002
The NBA introduces limited instant replay review, restricted exclusively to end-of-quarter buzzer-beaters. This marks the first time on-floor official judgments can be explicitly overturned by video.
The Donaghy Scandal & The Pedowitz Report 2007 – 2008
Following the federal investigation into referee Tim Donaghy for betting on games, the NBA commissions the independent Pedowitz Report. This triggers a total overhaul of the officiating department, leading to the creation of data-driven tracking, stricter performance grading, and the establishment of a formal referee operations command structure.
Launch of the Secaucus Replay Center 2014
The NBA opens a centralized, high-definition replay network in Secaucus, New Jersey. Instead of on-court referees viewing isolated court-side monitors alone, decisions are fed to and processed by a dedicated remote operations team to standardize rule application.
Last Two Minute (L2M) Reports Launched 2015
The league begins publicly releasing play-by-play assessments of all calls and material non-calls in the final two minutes of close games (within 3 points). This introduces public and media scrutiny directly to individual official error rates.
Introduction of the Coach’s Challenge 2019
The league introduces a formal mechanism allowing head coaches to trigger a replay review of specific fouls, out-of-bounds calls, or violations, removing absolute finality from the on-floor whistle during active gameplay.
Follow up to my last comment:
I again used ChatGPT to compare the NBA to NHL game 7s.
According to ChatGPT, the home team winning percentage in NHL game 7s was 60-65% from 1940-60, but has remained in the 55-60% range ever since. So there basically has been no change in the last 65 years in this metric, while in the NBA the home team win percentage has cratered.
Presumably travel for NHL players has improved. Reffing quality probably too. Yet no impact on the winning trend.
I think this leads us to the 3 point percentage as the main scapegoat for this trend in the NBA. If the NHL had a 3-goal rule, that if you scored on a slap shot from 80 feet it was worth 3 goals, then the randomness introduced could winnow away the 55-60% down to 50%ish.
A different question is why the NBA home team game 7 winning percentage was once so high (85%) while it’s so much lower in the history of the NHL. You could say there are a lot less goals in hockey and that in itself causes randomness. But that hasn’t stopped the NHL from having as many or more dynasties than in the NBA. The Canadiens won 5 straight cups in the 50s and 4 straight in the 70s. The Islsnders won 4 straight in the 80s. The Gretzky Oilers teams won 5 in 7 years. Meanwhile the NBA has only the 50s-60s Celtics winning more than 3 in a row. If hockey was that random, why do we see so many dynasties with the best team winning over time? Food for thought.
I just asked ChatGPT about the trend in home team winning percentage in game 7s. Here’s what it said:
In the 20th century, the home team won 87% of game 7s. Including 17 in a row from 1951-1967 and 20 in a row from 1984-1994.
In the 2000s the winning percentage dropped to 72%, in the 2010s it dropped to 65%, and in the 2020s its all the way down to 47%! Small sample sizes, yes, but not that small, as the playoffs have expanded over the years to more series and more games.
If we believe the suggested explanations, then we would have to believe that refs and travel didn’t have any improvement at all from 1950-2000, but then it had a giant improvement in the past 25 years. But I would argue that travel improved considerably from 1950-2000, and reffing probably too.
So that leaves three point shooting. Given that the winning percentage decline among home teams is inversely proportional to the explosion of three point shooting in the modern NBA, I think this is the answer. For better or worse, in many games the winner is simply determined by who shot well from 3. The Celtics are notorious for this. Shoot above 40% and they are practically unbeatable. But shoot under 30% and they lose more than they win.
One other factor that would affect a regular season decline in home team winning percentage (but not affect playoffs) is tanking and load management and the overall cheapening of the regular season. Teams just don’t give as much of a shit about the regular season anymore and it’s far more important to be rested and healthy going into the playoffs than achieve a high seed. Someone needs to tell Joe Mazz this because running Tatum into the ground upon his return so the Cs could be a 2 seed was very suspect strategy when Tatum’s overuse led him to missing game 7 vs the Sixers. The decline in importance in home court advantage in the playoffs will only cheapen the regular season even more.