For my final project, I used diving scores from various meets for potential women diving recruits for CSU-P. The intent would be to see which girls have progressed the best and create a model to predict who will have the best results in their collegiate diving career.
I broke down each individual diver with their one meter and three meter scores. I created a scatter plot with a linear regression line in the y=mx+b format to create a simple regression model. Second, I create an exponential model for both 1 meter and 3 meter and compared to the data set as the diver may be progressively getting better instead of a more linear and normal increase in their scores at each meet. It will take some considering to see which model will work best depending on the data set provided for each individual diver. Finally, I did the “meet” squared to create a quadratic multiple regression analysis that includes one dependent variable and three independent variables. This gives me three ways to analyze each diver. Depending on the data set and the output from each model, I will use one of the three models to create the best prediction for their performance next year as a CSU-P athlete.
After figuring out which model will best fit the data set for each given diver, I can compare and contrast each diver to see which diver will be the best performing (at least in theory).
As I will be the new head diving coach for CSU-P, I plan to use the tools provided from stats to create a more successful program. The model described above will be the starting point, at least for recruiting. Once the divers are with CSU-P, I plan on using some simple descriptive statistics to help quantify the divers progress. The goal will be this:
- Score each dive individually and create a data set for each individual dive.
- Reduce the Standard Deviation to as low of a number as possible – which means the dive is more consistent
- Try to continually raise the scores of each dive
- Show the diver their progress with trending analysis
- Show the diver which dives are strong and which dives are weak
- Show the diver which dives are consistent and which dives are inconsistent
- Create different models to best pick which dives a diver should compete to maximize their overall score
I strongly feel that using the statistics provided above to quantify the performance and progress of a diver will be extremely beneficial and I predict that it will yield better and more consistent performances compared to arbitrarily diving everyday without a goal or measuring progress.
Best Regards,
Nick O’Quin