Bayesball Picks

Great day today for my model and a strong start to the week. Getting the probable starting pitcher info in the morning is far more reliable than the pulling them the night before. As such, my picks and results should be more consistent and in line with my model’s performance. I have significantly higher favorites won today than the sportsbooks. It seems like I am consistently outperforming the sportsbook anyways, but today I had a really solid difference.

Today’s Stats:

Model favorites won: 7/10
Sportsbook favorites won: 4/10
Picks won: 4/6
Net Units: +2.41u
ROI: 40.17%

Weekly Stats:

Model favorites won: 7/10
Sportsbook favorites won: 4/10
Picks won: 4/6
Net Units: +2.41u
ROI: 40.17%

Absolute disaster day today. I’m going to go ahead with making my picks early morning instead of the night before since the probable starting pitchers were quite inaccurate last night. With the updated probable pitchers in the morning, my model heavily favored the Red Sox, which sucked especially since Nationals was one of my picks with insane EV. It also led the Brewers to be favored over the Marlins, another one of my picks. Unfortunately, I close this week with a loss, but hopefully after getting picks in the morning, my results will be more consistent.

Today’s Stats:

Model favorites won: 7/15
Sportsbook favorites won: 8/15
Picks won: 1/5
Net Units: -2.65u
ROI: -53.00%

Weekly Stats:

Model favorites won: 30/55
Sportsbook favorites won: 28/55
Picks won: 9/20
Net Units: -1.19u
ROI: -5.95%

I made a mistake while fetching the matchups yesterday so some of the starting pitchers were off, which screwed with some of the predictions. Mostly, the predicted teams were the same though, with the exception of the Athletics, which was unfortunately one of my picks.

Today’s Stats:

Model favorites won: 8/15
Sportsbook favorites won: 7/15
Picks won: 2/5
Net Units: -0.7u
ROI: -14%

Weekly Stats:

Model favorites won: 23/40
Sportsbook favorites won: 20/40
Picks won: 8/15
Net Units: +1.46u
ROI: +9.73%

Today was a pretty good day. Brewers and Twins had me sweating until the end though.

Today’s Stats:

Model favorites won: 10/15
Sportsbook favorites won: 9/15
Picks won: 4/5
Net Units: +3.06u, ROI: +61.20%

Weekly Stats:

Model favorites won: 15/25
Sportsbook favorites won: 13/25
Picks won: 6/10
Net Units: +2.16u, ROI: +21.6%

Yankees loss stung quite a bit since it was predicted to have high probability to win.

Model favorites won: 5/10
Sportsbook favorites won: 4/10
Picks won: 2/5
Net Units: -0.90u, ROI: -18%

My Picks

• Miami Marlins ML +100
• New York Yankees ML -135
• Los Angeles Angels ML +110
• Arizona Diamondbacks ML +100
• Seattle Mariners ML -160

Comparison to Sportsbook Odds (Edges and EV)
Odds used are Fliff Odds. Your sportsbook’s odds may differ.

Today’s Favorite Predictions + Probability

Model Stats:
Used 1,141 total games with an 80/20 train-test split
Train accuracy: 67.98%
Test accuracy: 63.32%
Test recall: 85.34%
Test precision: 59.64%