Kyle Stevens Email & Phone Number

Pittsburgh, PA, US

East Amherst, New York, United States

• Utilizing various data mining techniques to create and maintain a unique dataset containing over 100,000 results of NBA player prop bets at FanDuel Sportsbook.
• Analyzing and exploiting inefficiencies in “under” bets in the NBA player prop market through the use of pivot tables, querying and data visualization in Excel, SQL, and Tableau.
• Implementing random forest machine learning model in Python to project DraftKings Points (DKP) for NBA/MLB players.
• Betting ROI of 200% during NBA regular season 2020/21.
• DFS ROI of 1200% during NBA and MLB regular seasons.
• Providing findings and projections at a standard monthly rate.

Durham, New Hampshire, United States

• Used various clustering and statistical modeling techniques to enhance pre-existing Batter vs Pitcher predictive models for the New York Mets Analytics Department.
• Produced heat maps in Tableau for various pitch types to determine that certain pitch profiles have more success when thrown in certain locations.
• Ex: Concluded that Sliders must be located below the strike zone and away from the batter to induce more swinging strikes (75% of slider-induced swinging strikes occurred under these parameters).

Durham, New Hampshire, United States

• Implemented neural network model (RSquare = 0.988) and multiple data visualization techniques to evaluate predictors of player efficiency in Python.
• Determined that True Shooting % is the most significant predictor of Player Efficiency Rating (PER) and that shooting more 3’s can have an inverse effect on player efficiency.
• Introduced further implications within contract negotiations and prospect scouting/drafting.

Getzville, New York, United States

• Analyzed, revised, and queried to pull accurate historical delinquency data for all consumer products.
• Queried performance analysis of individual collectors based on various factors over a certain period of time.
• Assisted in creating and formatting reports for Drury Deck Project that presents an overview of all consumer data for upper-level management.
• Reconciled consumer delinquency data with risk management delinquency data to establish more precise predictive consumer modeling.