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).

First, the Log5 Method picks:

0.313 — Eric Hosmer batting against Mike Wright
0.307 — Lorenzo Cain batting against Mike Wright
0.301 — Ben Zobrist batting against Adam C Morgan
0.301 — Jose Altuve batting against Colby Lewis
0.298 — Francisco Lindor batting against James Paxton
0.297 — Paulo Orlando batting against Mike Wright
0.291 — Victor Martinez batting against J.A. Happ
0.290 — Matthew Szczur batting against Adam Morgan
0.288 — Steve Pearce batting against Robbie Ray
0.288 — Evan Longoria batting against Robbie Ray

Looks like Mike is Mr. Wright for hitters Monday night.

The NN produces this list:

0.301, 0.762 — Jose Altuve batting against Colby Lewis.
0.307, 0.740 — Lorenzo Cain batting against Mike Wright.
0.287, 0.735 — Josh Harrison batting against Steven Matz.
0.269, 0.734 — Miguel Cabrera batting against J.A. Happ.
0.298, 0.731 — Francisco Lindor batting against James Paxton.
0.283, 0.730 — James Loney batting against Jonathon Niese.
0.313, 0.726 — Eric Hosmer batting against Mike Wright.
0.276, 0.725 — Ian Kinsler batting against J.A. Happ.
0.251, 0.722 — Nick Hundley batting against Michael Bolsinger.
0.241, 0.721 — Ender Inciarte batting against Christian Friedrich.

I find the James Loney pick interesting. Niese owns the highest hit average (hits/PA) allowed, among the starters in this list. Loney is off to a good start with the Mets, and long term he is the kind of high average, low OBP players who should have a high hit average. That’s what I like about the NN, it finds patterns you might not otherwise notice.

That said, the consensus picks appear to be Lorenzo Cain, Jose Altuve, and Francisco Lindor. As always, even someone with a .74 probability of getting a hit in a game takes an 0 for fairly often. Good luck!