Here are the top picks my programs produced for use in Beat the Streak. This post mostly explains the ideas behind the calculations. In addition, this post shows tests on the Neural Network (NN). This post discusses an NN that includes the ballpark. I updated the models, and the results of those tests are here.
For 2018, I am just going to publish the Log5 hit averages and the NN probabilities with parks factored in. I am keeping track of the results here. I added a graph that gives a visual representation of the probability and success each day.
I have been asked to expand the list to the top 25 players for an econometric project.
First, the Log5 Method picks:
0.382 — Jose Altuve batting against Matt Moore.
0.354 — Albert Almora batting against Ivan Nova.
0.337 — Yulieski Gurriel batting against Matt Moore.
0.316 — George Springer batting against Matt Moore.
0.313 — Michael Brantley batting against Artie Lewicki.
0.313 — Kris Bryant batting against Ivan Nova.
0.311 — Kevin Pillar batting against Alex Cobb.
0.309 — Javier Baez batting against Ivan Nova.
0.308 — Tommy La Stella batting against Ivan Nova.
0.307 — Willson Contreras batting against Ivan Nova.
0.307 — Starlin Castro batting against Clayton Richard.
0.305 — Eddie Rosario batting against Nicholas Tropeano.
0.303 — Teoscar Hernandez batting against Alex Cobb.
0.301 — J.T. Realmuto batting against Clayton Richard.
0.301 — Adam Jones batting against Marco Estrada.
0.300 — Christian Yelich batting against Zach Eflin.
0.300 — Ben Zobrist batting against Ivan Nova.
0.299 — Carlos Correa batting against Matt Moore.
0.299 — Brian Anderson batting against Clayton Richard.
0.298 — Jose Martinez batting against Anthony DeSclafani.
0.297 — Yangervis Solarte batting against Alex Cobb.
0.297 — Manny Machado batting against Marco Estrada.
0.297 — Aledmys Diaz batting against Alex Cobb.
0.297 — Dwight Smith batting against Alex Cobb.
0.295 — Alex Bregman batting against Matt Moore.
Both predicted hit averages for Altuve and Almora are astronomical. Opponents are hitting .341 against Moore this season, while Nova allows a high BA and a low OBP. Both batters and both pitchers collect and allow hits at a way above average rate, and those combinations produce a high Log5 result. Altuve is 4 for 12 against Moore, Almora 0 for 3 against Nova.
Here is how the NN with Park ranks the players:
0.382, 0.791 — Jose Altuve batting against Matt Moore.
0.354, 0.756 — Albert Almora batting against Ivan Nova.
0.337, 0.739 — Yulieski Gurriel batting against Matt Moore.
0.313, 0.739 — Michael Brantley batting against Artie Lewicki.
0.288, 0.738 — Jean Segura batting against Nathan Eovaldi.
0.305, 0.730 — Eddie Rosario batting against Nicholas Tropeano.
0.307, 0.725 — Starlin Castro batting against Clayton Richard.
0.298, 0.723 — Jose Martinez batting against Anthony DeSclafani.
0.287, 0.718 — Charlie Blackmon batting against Zachary Godley.
0.301, 0.717 — J.T. Realmuto batting against Clayton Richard.
0.287, 0.716 — DJ LeMahieu batting against Zachary Godley.
0.291, 0.715 — Nolan Arenado batting against Zachary Godley.
0.289, 0.713 — Gerardo Parra batting against Zachary Godley.
0.300, 0.713 — Christian Yelich batting against Zach Eflin.
0.287, 0.711 — Freddie Freeman batting against Ross Stripling.
0.286, 0.711 — Jose Ramirez batting against Artie Lewicki.
0.255, 0.711 — Scooter Gennett batting against Carlos Martinez.
0.280, 0.710 — Jose Abreu batting against Rick Porcello.
0.301, 0.708 — Adam Jones batting against Marco Estrada.
0.264, 0.707 — Matt Kemp batting against Sean Newcomb.
0.299, 0.706 — Brian Anderson batting against Clayton Richard.
0.316, 0.704 — George Springer batting against Matt Moore.
0.274, 0.703 — Corey Dickerson batting against Kyle Hendricks.
0.313, 0.702 — Kris Bryant batting against Ivan Nova.
0.275, 0.701 — Jon Jay batting against Kyle Freeland.
0.309, 0.701 — Javier Baez batting against Ivan Nova.
There is perfect 1-2 agreement between the two systems. Altuve’s .791 probability of a hit is the highest generated this season.
The NN ran 70 days so far this season. Based on the daily predictions it should be right 73.8% of the time, and generated 52 days with hits. It’s been right 82.9% of the time and generated 58 days with hits. It’s current streak is five days, and generated a streak of six, two streaks of seven, two streaks of eight, and a streak of ten. The NN had favorable luck so far this season.
Remember, your best pick will fail about 25% of the time. Good luck!