Meeting 1 : November 6th 2018 - QMIND-Team/Sabermetrics GitHub Wiki
Problem Definition
- Predict the game.
- Mix between a classification problem and a regression.
- Mostly pandas, scikit-learn, py-baseball (give me stats),
- Lots of past models
- New stats every week
- 80% accuracy the outcome of a game
Planning
Things to learn:
- Pandas
- Scikit-learn
- Sql (manipulating databases)
- Tutorial mike posted (most naive model) (second part is predicting who will be in the hall of fame)-
- Research on sabr
- Starting to understand baseball stats (lots of abbreviations)
Delegating Tasks
Vedant
- Tutorial Walkthrough
- Sabr stat
- OPS/OBP
Will
- Tutorial Walkthrough
- Sabr stat
- DIPS (which dips are the best dips)
Eric
- Tutorial Walkthrough
- Sabr stat
- BABIP (batting average on balls in play)
Mike
- Break down data collection into tasks
- SIERA