How Much is Home Court Actually Worth?

[Bowtothebill23]

With this being our 30th season, I decided to take a stab at answering the age-old question of how much is home court actually worth in points. It's clear that playing at home translates to a better chance to win, but how much exactly is it worth? And if you were a betting man in sim league, how much influence on the point spread should playing at home have?

METHODOLOGY

There are plenty of ways to calculate this. I did it very simply: using wins and losses from every team who's played at home over the course of each of the 29 seasons. Now this isn't the most accurate data source possible. That would be using point differential or (if you can really mine data: point differential with respect to teams' records, injuries, etc.). But that might be too complicated to pull from our standings site, and, with 34,645 regular season games played coming into this season, I deemed the sample size large enough to be able to calculate essentially the exact same number.

I then used 3 different ways to calculate the winning percentage of home teams.
1) First, I simply took the total number of home wins and divided it by total number of games played.
2) I used a linear system where the 1990 season is worth 1 point, 91 worth 2... 2018 worth 29 points and used that to calculate a winning percentage
3) I used a backward system where I made the 1990 season worth 5 points, 1991 worth 4... 1993 worth 2, and everything else worth 1.

So you might ask why I used the 3rd system. Surely, if we're trying to calculate how much home court is worth today, last season should be worth more than 29 years ago. But this takes into account one of the inherent weaknesses of the win/loss method: team skill is not accounted for. Take, for example, an extreme scenario: the 2001 Hawks, who went 72-10 with a point differential of 19.7, play against the 2000 Supersonics, who went 5-77 with a point differential of -22.6, 100 times with 50 times at each team's home stadium. Without injuries, the Hawks probably go 50-0 at home and something like 48-2, 49-1, or 50-0 on the road. That doesn't really teach us anything about home court advantage because, looking at numbers, there is essentially no such thing as home court advantage since home teams win 50% of the time. So as there is less parity in the league with more teams tanking and more super teams, home court advantage shrinks with this system of wins/losses. So this system gives more weight to the first few years of the league, when everyone had just gone through the creation draft and there was more parity than we've ever had. I gave 5 points for the first year and progressively decreased until it got to 1 point which was the year after all the first rounders from the creation draft hit free agency.

Next, I used the converted winning percentage to wins in a season by simply multiplying the percentage by 82 games. After that, I subtracted that number from 41 to find the number of games above .500 home teams are in a given season. Lastly, one positive point in point differential is worth approximately 2.7 wins. So we divide the number of games over .500 by 2.7 to find Home Court Advantage.

RESULTS

Using our 3 methods, we get winning percentages of home teams of
1) 61.49%
2) 61.32%
3) 61.82%

Converting those to wins over a season, we get:
1) 50.42
2) 50.28
3) 50.69

And converting that into point differential, we get:
1) 3.49
2) 3.44
3) 3.59

Interestingly, the highest winning percentage of any season in our league's history is, unsurprisingly, the first season in 1990 with a Win % of 64.17%. The lowest season? Just last year in 2018 with a Win % of 58.05%. While that could theoretically be a trend worth looking into, taking a closer look, we find that home teams have NOT been steadily winning less. Instead, this seems to be an anomaly of a season.

With all that said, we can say with some certainty that Home Court in PBSL is worth approximately 3.5 points in a given game. Meaning: home teams, if you win a game by 8 or more, you would've likely won that game on the road. Road teams, if you lose by 6 or fewer, you likely win that game at home.

Hope this information is useful/interesting for y'all.

[digiskunk]

Wow, great idea for an article, and very interesting results. 10/10

[kucoach7]

Very interesting. I was able to control for team winning percentage and opponent winning percentage but didn't look at point differential when I did this (viewtopic.php?f=7&t=4290) although I easily could. Maybe I'll do a follow up article.

[ballsohard]

5 Points!