• Data loading...

Calculating Win Percentages for the 2020 NBA Season (prior to lockdown)

Calculating Win Percentages for the 2020 NBA Season, so far prior to season continuation.

As happily shown on our MLB 2019 calculating wins article, I decided to put my math skills down to good use, this time with the NBA season. Currently, the NBA season came to a full halt in March due to the worldwide pandemic, but announced its continuation this coming July 30, 2020. In addition to this article, here's a post from NBC Sports asking some players how practice is going in "the bubble" at Orlando, FL.

Okay, so how do we calculate predicted win percentages of NBA teams? If you've read our MLB article, linked above, we said how famous Greek mathematician Pythagoras laid down his equation in finding the area of a right triangle a2 + b2 = c2 known as the Pythagorean Theorem. Although we shall use the same formula, it'll be slightly different. Houston Rockets executive Daryl Morey punched in his research and found that the theorem using the exponent 13.91 offers the most accurate percentage (compared to the exponent of 2 for the MLB). To write out the equation, we get the following:

$${points \ per \ game^{13.91} \over {points \ per \ game^{13.91} + opponent \ points \ per \ game^{13.91}}} = predicted \ wins \ percentage$$

(predicted wins percentage is the estimated percentage of games won often written as a 0.000 decimal format.

points per game 13.91 is the team's average points scored per game, to the 13.91th power.

opponent points per game 13.91 is the team's opposing average points scored/allowed per game, to the 13.91th power.)

Time to get your calculators out and simply plug in the numbers to these variables to see what we get! We're currently based in Los Angeles, so we'll take and test this formula out with the Lakers and the Clippers.

Currently ranked #1 in the Western Conference, prior to the lockdown, the Lakers averaged 114.3 points per game, but their opponents averaged 106.9 points per game. (Lakers have won 49 games and have 14 losses so far.) Plug in the numbers to get the and we get:

$${114.3^{13.91} \over {114.3^{13.91} + 106.9^{13.91}}} = 0.717$$

(Round the final answer to the thousandth place, which is the third number to the right of the decimal point.)

A predicted 72% of games the Lakers have won thus far, which is impressive given their current lineup. Since we're calculating from total games played, only 63 games, we'll find the percentage of wins by dividing the games won to total games played. In other words, the Lakers won 49 games, so out of 63 games played, we get:

$${49 \over {63}} = 0.778$$

(In fraction form, this is 7/9 [reduced]. In other words, Lakers would win 7 out of 9 games.)

Pretty good for a #1 ranked team, pre-lockdown. Based on what we've gotten so far, let's see how much we were off by, known as the error. We just subtract the actual wins percentage to our predicted percentage, and we have:

$${0.778 - 0.717} = 0.061$$

For that, we were about 6% off the predicted percentage. As you calculate these, you may get an answer that's negative (-). A negative percentage means the team won less than that predicted percentage stated. In this case with the Lakers, we got a positive number, meaning they won more games than predicted initially. Simply put: Negative percentage is less games won than originally predicted; Positive percentage means more games won than originally predicted.

Let's go to the other part of town: the Clippers. How are they doing? Let the Pythagorean Theorem tells us:

$${116.2^{13.91} \over {116.2^{13.91} + 109.7^{13.91}}} = 0.690$$

This time, the Clippers played 64 games total and won 44 of them. Thus, we get:

$${44 \over {64}} = 0.688$$

(In reduced fraction form, this is 11/16—out of 16 games, Clippers would win 11 of them.)

To sum it all up, we were off by:

$${0.688 - 0.690} = -0.002$$

That's a very close prediction! We all can thank Daryl Morey for the mathematical research in making this possible. This is as close as we got, so Mr. Morey, if you're reading this, or if your employees sent this article over to you, I personally want to thank you so much!

Because the season came to a stop, just weeks shy from starting the playoffs, a handful of teams never completed their seasons. However, due to the rush in completing the season, teams with losing records are not invited to complete their 82-game schedule. Overall, prior to season continuation, here are the numbers for all 30 teams:

Spreadsheet for calculating win percentages in the 2020 NBA season prior to lockdown.
Feel free to download our whole spreadsheet here. Created and made using LibreOffice Calc.

Looking at the Error column, it's mostly on the positive end as more teams won more than expected. You'd think that an offensive-laden team, like the Golden State Warriors would have a win percentage above 0.500 (50%), but perhaps they themselves, including fans, wouldn't like to even glance at this season's numbers due to injuries, poor performance and team record. Because of that, we're definitely looking at a new champion this year.

The numbers have been crunched, so what? These numbers reflect more on the regular season performances, as it does not state who has the bigger edge in winning the championship. Those predictions require more advanced math, as we continue to expand our research in this amazing field of sports. Without losing focus, we're starting with formulas that aren't too overwhelming, but will dive in to the more rigorous as our research continues!

Now that the season will continue, who do you think will win the championship?

As said previously, for those who are still fearful and disgusted by Math, remember this: Math is like cooking, just follow the recipe. Like any other discipline or hobby, practice makes perfect. You can do this!

This article was sponsored by The Seeds of Math. We WILL NOT provided any secret winning formulas and WILL NOT work with personal, illegal bookies looking to use our research to compose a sports betting model in exchange for money. As of current, we are not affiliated with any sports entity whatsoever and are provided this for extensive fanaticism of sports and informational reasons. In part with adding such techniques to our portfolio, we won't be responsible nor held accountable if such mathematical formulas give the bettor(s) a losing streak. It is encouraged that you gamble responsibly and legally, as any tout or accusations toward our current mathematical research will not be dealt with. Under the KCU Network, we are here to educate and entertain, and will not agree to exchange monetary "juice" for a secret winning math formula, kept under the rugs, to circumvent regulations by the federal government and sports books around the US and the world. Not once do we force our fans and readers to gamble and use any techniques provided, and so any disagreements with this disclaimer is free to detach from our platform, as any legal action taken against us will not be carried further. For more on this topic of discussion, please contact us.

Points Per Game and Opponent Points Per Game stats courtesy of and ESPN. Also, a big thank you to MathJax for the nicely-rendered equations!

Write a Comment