Does Winning A Specific NBA Quarter Matter?

By: Matty Cheng

In the 2018 playoffs, the Golden State Warriors have typically dominated the third quarter, especially with Stephen Curry detonating, but is domination in the third quarter more important than doing so in the second quarter? Or does a first quarter explosion demoralize opponents early and is the most important quarter to win? Do home teams jump out to leads in the 1st quarter when the home crowd is likely most active? Numberfire did some analysis on NBA quarters for 2014-15 games, but I would like to examine a larger sample size.

Scraping the NBA boxscores since the 1950 from basketball reference, we can see some patterns from the 61,390 games.

As seen in the table above, the home team wins the first quarter (i.e. scores more points than the away team in the 1st quarter) 57.12% of the time, the highest win percentage of any single quarter. The data also shows a clear home court advantage as home teams win 61.91% of the games.

Next, we can look at the percentage that a team wins a game conditional on winning a specific quarter:

Note: Win% is conditional on winning that specific quarter (example: the first data point 74.96% means that when the home team wins the first quarter, it wins the game 74.96% of the time.)

Since the home win percentage in column1 are all almost 15% higher than the normal home win percentage of 61.91% found in the previous table, we can conclude that winning a specific quarter increases the probability of winning a game. However, the conditional win percentages for each quarter are marginally different, suggesting that no quarter is much more significant than another. The first and third quarter have slightly higher win percentages, but only by 1-2%. Running a statistical proportions test, we find that though the differences are small, they are statistically significant:

REGULAR SEASON VS. PLAYOFFS

We can also examine whether home teams have a greater advantage in the playoffs:

Home teams have a better advantage in the playoffs that is extremely statistically significant when running a statistical proportions test. This could be due to playing at home being more important or that the higher seeds get more home games in a seven game series. Oddly, the home teams win the third quarter and OT less than they do in the regular season.

Finally, we can compare the percentage that a team wins a game conditional on winning a specific quarter in the regular season versus the playoffs:

The conditional win percentage increases by 3-4% for the home team in the playoffs, but decreases by 4-5% for the away team in the playoffs. This means that home court is more important in the playoffs as winning a specific quarter means less in winning the game for the away team. Interestingly, when the away team wins the second or fourth quarter, they win the game less than 50% of the time. However this is still higher than the away team’s normal game win percentage of 34.31%. Once again, all of these percentages are extremely statistically significant when running a statistical proportions test.

CONCLUSION
Home court matters and it is even more important in the playoffs. Winning a specific quarter matters as it increases win probability, but no quarter is extremely more important than another as the first and third quarter are only increase win probability by 1-2% compared to the second and fourth quarter. The results are consistent over all time periods since 1950 as the 1950s, 1960s, 1970s, etc all show extremely similar patterns. One implication of the small but statistically significant increases in win probability for the first and third quarter is that perhaps star players should rest at the beginning of the second and fourth quarters instead of resting at the end of first and third quarters.

If you have any questions for Matthew about this article, please feel free to reach out to him at chengm@college.harvard.edu

About the author

harvardsports

View all posts

Leave a Reply

Your email address will not be published. Required fields are marked *