Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.
NBA Injury Prediction
Peter Wen
petwen1027@gmail.com
github.com/rayal34/Predicting_NBA_Injuries
MOTIVATION
● Warriors’ offensive rating with Steph Curry: 119.1
● Warriors’ offensive rating without Steph Curry: 105.3
10...
MOTIVATION
● Increased awareness of player health.
PROCESS
● Collected data by web scraping.
Data
PROCESS
● Used a three week aggregation moving window.
Predict if an injury
will occur in the next
3 games
Aggregate box s...
RESULTS
Winner!
AUC = 0.549
INSIGHTS
● Important features to the model.
INSIGHTS
● Important features to the model.
INSIGHTS
● Important features to the model.
INSIGHTS
● Speed: slower running speed might signal an injury.
● Distance: greater distance ran might also signal an injur...
FUTURE WORK
● Investigate hustle stats.
○ Ex: charges drawn, loose balls recovered.
○ Diving head-first into the stands = ...
Thank you!
Any questions?
Peter Wen
petwen1027@gmail.com
github.com/rayal34/Predicting_NBA_Injuries
NBA Injury Prediction
NBA Injury Prediction
Upcoming SlideShare
Loading in …5
×

NBA Injury Prediction

140 views

Published on

using machine learning to predict the likelihood of an injury in the NBA

Published in: Data & Analytics
  • Be the first to comment

  • Be the first to like this

NBA Injury Prediction

  1. 1. NBA Injury Prediction Peter Wen petwen1027@gmail.com github.com/rayal34/Predicting_NBA_Injuries
  2. 2. MOTIVATION ● Warriors’ offensive rating with Steph Curry: 119.1 ● Warriors’ offensive rating without Steph Curry: 105.3 105.3 = not good!
  3. 3. MOTIVATION ● Increased awareness of player health.
  4. 4. PROCESS ● Collected data by web scraping. Data
  5. 5. PROCESS ● Used a three week aggregation moving window. Predict if an injury will occur in the next 3 games Aggregate box score statistics in the three weeks leading up to the prediction Time
  6. 6. RESULTS Winner! AUC = 0.549
  7. 7. INSIGHTS ● Important features to the model.
  8. 8. INSIGHTS ● Important features to the model.
  9. 9. INSIGHTS ● Important features to the model.
  10. 10. INSIGHTS ● Speed: slower running speed might signal an injury. ● Distance: greater distance ran might also signal an injury.
  11. 11. FUTURE WORK ● Investigate hustle stats. ○ Ex: charges drawn, loose balls recovered. ○ Diving head-first into the stands = higher injury risk!
  12. 12. Thank you! Any questions? Peter Wen petwen1027@gmail.com github.com/rayal34/Predicting_NBA_Injuries

×