A Statistical Analysis of Shot Success in Basketball

[Jaleng22]

Basketball is a sport where every shot matters, and understanding the factors that influence success is critical for optimizing performance. Using play-by-play data from the 2019-2020 NBA season, this study analyzed over 539,000 rows of data to evaluate how shot distance and game quarter impact the likelihood of a successful shot. The findings provide a data-driven perspective on shot success in professional basketball.

The dataset used for this analysis contained detailed information on shot attempts, including variables such as shot distance, shot outcome ("make" or "miss"), game quarter, shot type, and game time. Key observations included:

• Shot Distance: Close-range attempts, such as layups and dunks, had significantly higher success rates compared to long-range shots like three-pointers. The average shot distance was 14.05 feet, with most shots occurring within 25 feet.

• Game Quarter: Shot attempts were evenly distributed across quarters, with no significant changes in shooting patterns. However, fourth-quarter shots showed a slight decline in success, likely reflecting increased game pressure or fatigue.

Three statistical models were employed to predict shot success:

1. Logistic Regression: This model identified a clear negative relationship between shot distance and shot success. Each additional foot of shot distance decreased the odds of success, but the model achieved low accuracy (40.02%) due to its inability to handle non-linear patterns.

2. Decision Tree: This model improved accuracy to 63.27%, identifying a critical threshold at 3 feet, where close-range shots had a 68.8% success rate compared to 38.6% for longer shots.

3. Random Forest: Matching the decision tree in accuracy (63.27%), the random forest model demonstrated the importance of shot distance while accounting for more complex relationships.

• Distance Determines Success: Shot distance emerged as the most influential factor. Shots within 3 feet of the basket had the highest success rates, while attempts from beyond 30 feet were significantly less likely to result in a "make".

• Minimal Impact of Game Quarter: The game quarter had little effect on shot outcomes. Success rates were consistent across quarters, except for slight increases in overtime periods.

Findings
Key metrics for shot distance included:

• Mean Distance: 14.05 feet
• Standard Deviation: 10.85 feet
• Success Rate for Shots <3 feet: 68.8%
• Success Rate for Shots ≥3 feet: 38.6%

These findings have practical implications:

• Shot Selection: Prioritizing high-percentage shots, particularly within 3 feet of the basket, can significantly improve team scoring efficiency.

• Coaching Strategies: Coaches can leverage these insights to design plays that increase close-range opportunities while minimizing low-percentage attempts.

While the study revealed critical patterns, it did not account for variables such as player fatigue, defensive pressure, or skill level, which could also influence shot success. Future studies incorporating these factors would provide a more comprehensive understanding of shooting performance.

This analysis highlights the pivotal role of shot distance in basketball performance. Close-range shots consistently yield higher success rates, emphasizing their value in strategic planning. While game quarter showed minimal impact, other contextual factors may hold greater influence. The findings underscore the importance of data-driven decision-making in basketball, offering actionable insights for players, coaches, and analysts.

[Crow]

539,000 shot attempts? That has to be several seasons.


Would take ball tracking data, but I'd want to see study of best shot trajectory for shots from 3-6 feet. Is more arc and / or backboard better than not?

[Mike G]

Did AI write this article?

[Crow]

It is likely but I try to leave room for it to prove otherwise.

If there is no reply, the chances of it being AI go to near 100%.

[DSMok1]

Yeah, I'll likely delete the thread and user in a few days if it is as it seems.

[Crow]

An AI user may try to pass that product off as independent research. And / or may be playing with generation and placement approaches. And looking for detection.

Could use it as a tool and
be transparent about it.

As I mentioned initially, the post appears to have a major flaw with the size and timespan of the dataset. If there is error there, might be more errors.

Most of it sounded ok but nothing new.

If AI is to become really useful, it will have to be implemented very carefully and well or get a generation or several better.

[Jaleng22]

I personally collected the data, performed the analysis, and developed the findings. AI assisted in structuring the article but that’s about it. If this isn’t allowed on the platform I understand. I can remove my post.

[Jaleng22]

539,000 shot attempts? That has to be several seasons.


Would take ball tracking data, but I'd want to see study of best shot trajectory for shots from 3-6 feet. Is more arc and / or backboard better than not?

I’m looking for more contextual data so that I can include it in research that I plan to do in the future

[Jaleng22]

Also it’s 539,000 rows of data not shot attempts. 35 columns

[Crow]

Thanks for the replies / clarifications. Welcome.

Your thread is fine now.

We have gotten some weak AI stuff recently (possibly as pretext for spam) and a stream of outright commercial spam.


Other recent study showed change in shot behavior at end of quarters.

Have you looked at impact on average of court angle from basket at similar distances?

Any study of impact of shot sequencing on results?

Are you planning larger publication of the findings? Interested in providing access to the database?

Willing to identify the 35 columns?

Have any questions for discussion or identified next research steps?

Have access to older or newer seasons and interest in tracking changes?

Other work to mention or interests to discuss?

See anything here recently that interests you?

[Mike G]

Prioritizing high-percentage shots, particularly within 3 feet of the basket, can significantly improve team scoring efficiency.

This has been known since day 1 of basketball. However, there are limitations and other factors:
- Knowing the shooting% in an area doesn't tell you the efficiency, since turnovers are a big factor. Relatively fewer TO occur attempting a longer distance shot.
- A made shot beyond 22 feet is worth 50% more than one at closer range: Hitting 38% of your 3's is like 57% of 2s.
- Late in the shot clock, you have to take what the defense allows. This is true at any time, but especially then.
- Less important but still significant: A low% shot means a higher chance of an offensive rebound. This somewhat mitigates the shooting% difference.
- Driving the ball inside raises the chance of an offensive foul. It's a turnover AND your players get in foul trouble.
- Many or even most of high% shots near the basket are enabled by the threat of a 3-pointer. Without that threat, unless you have a prime Shaq or Jokic, you don't get much inside.

[Jaleng22]

Prioritizing high-percentage shots, particularly within 3 feet of the basket, can significantly improve team scoring efficiency.

This has been known since day 1 of basketball. However, there are limitations and other factors:
- Knowing the shooting% in an area doesn't tell you the efficiency, since turnovers are a big factor. Relatively fewer TO occur attempting a longer distance shot.
- A made shot beyond 22 feet is worth 50% more than one at closer range: Hitting 38% of your 3's is like 57% of 2s.
- Late in the shot clock, you have to take what the defense allows. This is true at any time, but especially then.
- Less important but still significant: A low% shot means a higher chance of an offensive rebound. This somewhat mitigates the shooting% difference.
- Driving the ball inside raises the chance of an offensive foul. It's a turnover AND your players get in foul trouble.
- Many or even most of high% shots near the basket are enabled by the threat of a 3-pointer. Without that threat, unless you have a prime Shaq or Jokic, you don't get much inside.

Hi Mike G,

I’m just emphasizing if you focus on those closer shots then you’ll have a higher chance of making your shot. We’re talking an 88.8 percent chance. Also, I’m not sure turnovers are a factor in shooting. Yes 3 pointers are worth more but longer distance shots are harder to make. To do a deeper dive I’ll work on getting more data like fatigue, defense proximity those types of things.

[Jaleng22]

Also, players are shooting more 3’s in today’s game. I’m not saying don’t shoot 3’s but players are shooting 10 3s a game individually. If it’s harder to make longer distance shots then It takes away from the efficiency if those shots will more likely result in a miss. Look at Jayson tatum, lamelo ball, Anthony Edward’s. Shooting that many 3s per game hurts players efficiency and shooting percentages when they more than likely result in a miss. 3s are good for the game but it’s not necessarily translating to winning basketball for teams. I’ve seen wemby shoot upwards of 10 3s in a game. Get easy buckets and make the game easier.

[TeemoTeejay]

Basketball is a sport where every shot matters, and understanding the factors that influence success is critical for optimizing performance. Using play-by-play data from the 2019-2020 NBA season, this study analyzed over 539,000 rows of data to evaluate how shot distance and game quarter impact the likelihood of a successful shot. The findings provide a data-driven perspective on shot success in professional basketball.

The dataset used for this analysis contained detailed information on shot attempts, including variables such as shot distance, shot outcome ("make" or "miss"), game quarter, shot type, and game time. Key observations included:

• Shot Distance: Close-range attempts, such as layups and dunks, had significantly higher success rates compared to long-range shots like three-pointers. The average shot distance was 14.05 feet, with most shots occurring within 25 feet.

• Game Quarter: Shot attempts were evenly distributed across quarters, with no significant changes in shooting patterns. However, fourth-quarter shots showed a slight decline in success, likely reflecting increased game pressure or fatigue.

Three statistical models were employed to predict shot success:

1. Logistic Regression: This model identified a clear negative relationship between shot distance and shot success. Each additional foot of shot distance decreased the odds of success, but the model achieved low accuracy (40.02%) due to its inability to handle non-linear patterns.

2. Decision Tree: This model improved accuracy to 63.27%, identifying a critical threshold at 3 feet, where close-range shots had a 68.8% success rate compared to 38.6% for longer shots.

3. Random Forest: Matching the decision tree in accuracy (63.27%), the random forest model demonstrated the importance of shot distance while accounting for more complex relationships.

• Distance Determines Success: Shot distance emerged as the most influential factor. Shots within 3 feet of the basket had the highest success rates, while attempts from beyond 30 feet were significantly less likely to result in a "make".

• Minimal Impact of Game Quarter: The game quarter had little effect on shot outcomes. Success rates were consistent across quarters, except for slight increases in overtime periods.

Findings
Key metrics for shot distance included:

• Mean Distance: 14.05 feet
• Standard Deviation: 10.85 feet
• Success Rate for Shots <3 feet: 68.8%
• Success Rate for Shots ≥3 feet: 38.6%

These findings have practical implications:

• Shot Selection: Prioritizing high-percentage shots, particularly within 3 feet of the basket, can significantly improve team scoring efficiency.

• Coaching Strategies: Coaches can leverage these insights to design plays that increase close-range opportunities while minimizing low-percentage attempts.

While the study revealed critical patterns, it did not account for variables such as player fatigue, defensive pressure, or skill level, which could also influence shot success. Future studies incorporating these factors would provide a more comprehensive understanding of shooting performance.

This analysis highlights the pivotal role of shot distance in basketball performance. Close-range shots consistently yield higher success rates, emphasizing their value in strategic planning. While game quarter showed minimal impact, other contextual factors may hold greater influence. The findings underscore the importance of data-driven decision-making in basketball, offering actionable insights for players, coaches, and analysts.

So, not tryna be a dick lol, but so was the analysis basically:

“Shots at the rim go in more often than shots far away?” Or am I misunderstanding something

[Jaleng22]

Basketball is a sport where every shot matters, and understanding the factors that influence success is critical for optimizing performance. Using play-by-play data from the 2019-2020 NBA season, this study analyzed over 539,000 rows of data to evaluate how shot distance and game quarter impact the likelihood of a successful shot. The findings provide a data-driven perspective on shot success in professional basketball.

The dataset used for this analysis contained detailed information on shot attempts, including variables such as shot distance, shot outcome ("make" or "miss"), game quarter, shot type, and game time. Key observations included:

• Shot Distance: Close-range attempts, such as layups and dunks, had significantly higher success rates compared to long-range shots like three-pointers. The average shot distance was 14.05 feet, with most shots occurring within 25 feet.

• Game Quarter: Shot attempts were evenly distributed across quarters, with no significant changes in shooting patterns. However, fourth-quarter shots showed a slight decline in success, likely reflecting increased game pressure or fatigue.

Three statistical models were employed to predict shot success:

1. Logistic Regression: This model identified a clear negative relationship between shot distance and shot success. Each additional foot of shot distance decreased the odds of success, but the model achieved low accuracy (40.02%) due to its inability to handle non-linear patterns.

2. Decision Tree: This model improved accuracy to 63.27%, identifying a critical threshold at 3 feet, where close-range shots had a 68.8% success rate compared to 38.6% for longer shots.

3. Random Forest: Matching the decision tree in accuracy (63.27%), the random forest model demonstrated the importance of shot distance while accounting for more complex relationships.

• Distance Determines Success: Shot distance emerged as the most influential factor. Shots within 3 feet of the basket had the highest success rates, while attempts from beyond 30 feet were significantly less likely to result in a "make".

• Minimal Impact of Game Quarter: The game quarter had little effect on shot outcomes. Success rates were consistent across quarters, except for slight increases in overtime periods.

Findings
Key metrics for shot distance included:

• Mean Distance: 14.05 feet
• Standard Deviation: 10.85 feet
• Success Rate for Shots <3 feet: 68.8%
• Success Rate for Shots ≥3 feet: 38.6%

These findings have practical implications:

• Shot Selection: Prioritizing high-percentage shots, particularly within 3 feet of the basket, can significantly improve team scoring efficiency.

• Coaching Strategies: Coaches can leverage these insights to design plays that increase close-range opportunities while minimizing low-percentage attempts.

While the study revealed critical patterns, it did not account for variables such as player fatigue, defensive pressure, or skill level, which could also influence shot success. Future studies incorporating these factors would provide a more comprehensive understanding of shooting performance.

This analysis highlights the pivotal role of shot distance in basketball performance. Close-range shots consistently yield higher success rates, emphasizing their value in strategic planning. While game quarter showed minimal impact, other contextual factors may hold greater influence. The findings underscore the importance of data-driven decision-making in basketball, offering actionable insights for players, coaches, and analysts.

So, not tryna be a dick lol, but so was the analysis basically:

“Shots at the rim go in more often than shots far away?” Or am I misunderstanding something

Yes, that's the main takeaway—closer shots, like those at the rim, go in more often than long-range ones. I know it's not groundbreaking, but the analysis emphasizes that easy buckets are the way to go when developing strategies to maximize scoring efficiency.