There’s a more detailed breakdown of how the process works [here](https://www.reddit.com/r/baseball/comments/1chqa4f/comment/l24290o/?utm_source=share&utm_medium=web3x&utm_name=web3xcss&utm_term=1&utm_content=share_button). In short, xwOBA and xwOBA allowed (xwOBAA) are converted into runs scored and allowed using the [Weighted Runs Created formula.](https://library.fangraphs.com/offense/wrc/) Those numbers are then fed into the pythagorean expectation to come up with an expected winning percentage. For each game, the two teams’ winning percentages can be compared with the [log5 method](https://tht.fangraphs.com/tools-game-and-series-win-probabilities/), giving the probability that the home team wins. Then I essentially flip a weighted coin to determine who wins and repeat that process for each game 10,000 times, adding the simulated wins to the actual wins to get the season win totals. I also track how many times a team wins their division, earns a bye, or makes the wildcard and divide those counts by 10,000 to get the playoff probabilities.
Please let me know if you have any questions or criticisms. Thanks to FanGraphs, Baseball-Reference, and Baseball Savant for all the data used in this project.
hypnoticus103
Wow this shit hates the Brewers.
Boooooo
Sliiiiime
We’re expected to win fewer games with more of our starters?
3 Comments
This is a continuation of a series of posts I’ve made using hitting and pitching xwOBA to predict win totals and playoff probabilities. You can see a summary of how accurate they were [here](https://www.reddit.com/r/baseball/comments/16yfk71/comment/k38ah5x/?utm_source=share&utm_medium=web2x&context=3) (tldr: in the same ballpark as the FanGraphs playoff odds). The numbers from last month can be found [here](https://www.reddit.com/r/baseball/comments/1d60mh8/xwoba_projected_standings_and_playoff_odds/).
There’s a more detailed breakdown of how the process works [here](https://www.reddit.com/r/baseball/comments/1chqa4f/comment/l24290o/?utm_source=share&utm_medium=web3x&utm_name=web3xcss&utm_term=1&utm_content=share_button). In short, xwOBA and xwOBA allowed (xwOBAA) are converted into runs scored and allowed using the [Weighted Runs Created formula.](https://library.fangraphs.com/offense/wrc/) Those numbers are then fed into the pythagorean expectation to come up with an expected winning percentage. For each game, the two teams’ winning percentages can be compared with the [log5 method](https://tht.fangraphs.com/tools-game-and-series-win-probabilities/), giving the probability that the home team wins. Then I essentially flip a weighted coin to determine who wins and repeat that process for each game 10,000 times, adding the simulated wins to the actual wins to get the season win totals. I also track how many times a team wins their division, earns a bye, or makes the wildcard and divide those counts by 10,000 to get the playoff probabilities.
Please let me know if you have any questions or criticisms. Thanks to FanGraphs, Baseball-Reference, and Baseball Savant for all the data used in this project.
Wow this shit hates the Brewers.
Boooooo
We’re expected to win fewer games with more of our starters?