Astros ALDS a Review of the xwOBA Matchup Model- How did we do?
There was some questioning of my xwOBA matchup prediction model in the ALDS matchup guide on social media and even some in the media. There was a significant amount of disrespect pushed at me.
NO ONE is more clear about reviewing whether the things I say are true or turn out to be incorrect. If I am wrong, unlike others, I search for why.
Here is how the Astros hitters ACTUALLY did in the THREE game series compared to what my model SAID they would do from an xwOBA basis. I weighted the stats based on the IPs of the Seattle pitchers. I suppose if I wanted to I could calculate this on every PA (batter, pitcher, pitch, prediction).
You can see from the table above who in the THREE game series over performed and underperformed their predicted xwOBA. This is too small a sample size to really do this, but I wanted to give you the data because ACTUALLY does a stunningly good job of predicting the series performance of the team. Here are a few of the exceptions.
Hot AS expected
Yordan Alvarez- I wrote "Alvarez is by far the one Astros batter this system projects to rake- surprise!" Yordan lived up to expectations and was the MVP in my book.
Kyle Tucker- Tucker was expected to be really good and he was.
Hotter than expected
Yuli Gurriel- Yuli had a great series! I wrote about a theory I had after game 1 here. You will note Yuli had an off day before every game in the ALDS.
Alex Bregman- Breggy was vintage Breggy. He was expected to be near average in the ALDS and was great.
Jeremy Pena- When the Astros scored runs in the ALDS, Pena seemed to be in the middle of it. Pena may need to stay at #2 until the day Altuve retires.
Chas McCormick- Chas was expected to have a subpar series and he hit at an average level vs. a good pitching staff.
Colder than expected
Jose Altuve- Jose wasn't expected to rake this series and he was just terrible. Altuve will bounce back.
Trey Mancini- Mancini is in a funk. I hope he finds something in the days off.
Christian Vaźquez- Vazquez never started, was always substituted in vs. the Mariners bullpen, and never got in a groove. I am not sure what it will take to allow Vazquez to start when the metrics show he is a better option. Don't tell me about defense until you have read everything I have written on our catchers.
Martin Maldonado- Maldly was expected to be bad and was worse at the plate.
The truth is folks is the sample sizes for all of these players is too small to expect convergence with any prediction. The fact that the team actual wOBA is only 0.014 worse than the 0.311 weighted average prediction is encouraging.
Working on ALCS data now.
The statistics referenced here come from Fangraphs.com and baseballsavant.com
Media members including bloggers and podcasters- BEFORE you share this content you must get permission at email@example.com. You must also encourage folks to subscribe to LarryTheGM.com when you do.
Tell me what you think. Subscribers, PLEASE leave your comments here or in the chat.
If you liked this article or any of the others here, you really should subscribe to LarryTheGM.com (Below the header and above the articles). You will get notified of new content and you can use the chat feature.