Unabated - Modeling for Beginners

Podcast Release Date:
Podcast Link:
hagrinBPR Review Date: 2024-04-12 06:42:24
Rating: 2.00

Notes:
It's an intro pod, but we didn't even turn the engine on and leave the driveway. Here's what we covered for modeling - excel, python, fangraphs, BRef, timers for scrapers, Monte Carlo, Bayes, outlier events. That's it. That's the pod. Like, we said the names and didn't get down to even a little detail.

However, all is not lost. It's just the pod's format that caused this. Just go to the guy's Youtube "Wagered on Tilt" and if you're a newbie trying to just get exposed to some basic terms, this is going to be a good start. In fact, I think the pod, instead of the haphazard Qs that were asked, should have just gone right through his Youtube and that would have made for a far better intro pod.

2:30 - Jack is doing a lot of "baseball betting watching" looking at his outs looking at what they are offering, doing a lot of scouting. "It's too early - you don't know what you don't know."

4:45 - T introduction. Does a lot of Youtube content (I am unfamiliar). Started mostly Excel, doing some Python now. (At the end of the pod, I went to the Youtube and it's perfectly fine for people nearer the intro level.)

7:15 - "If you get stuck in a bad habit and you just copy your code, you will get stuck." (Not only this, but you start incurring "technical debt".)

9:45 - "Where do you start to model baseball?" First fives because it's an easier problem to solve. For data - Fangraphs, Baseball Reference the big two. Scrapers have a timer in them so as to not get banned by the site for scraping. (I know this is an intro pod but geez.)

Then there's a q about early vs midseason which turns into I stop betting in July or if the model is taking a beating and then picks up back around playoffs. Not sure what that has to do with the question asked or the topic of the pod but okay. Followed by top-down vs building your own projections and there's not much new here.

19:15 - "Have you found that different models work better for different parts of baseball?" - this is a good question by the host back on topic, but the answer is just listing the variables included inside of each respective model. Host smartly follows up looking for some clarity on actual modeling techniques.

22:00 - "A lot of people use averages, I prefer to use medians." Think we're about done here.