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:

0.344 — Odubel Herrera batting against Wily Peralta
0.332 — Daniel Murphy batting against Jon Moscot
0.328 — Carlos Gonzalez batting against Luis Perdomo
0.324 — Aledmys Diaz batting against Jake Peavy
0.324 — Andres Blanco batting against Wily Peralta
0.323 — Jimmy Paredes batting against Wily Peralta
0.323 — Cameron Rupp batting against Wily Peralta
0.321 — DJ LeMahieu batting against Luis Perdomo
0.321 — Stephen Piscotty batting against Jake Peavy
0.319 — Charlie Blackmon batting against Luis Perdomo

Wily Peralta and Jake Peavy look like the pitchers of the day for allowing hits. Three of Peavy’s last four starts were good, however. Peralta consistently gives up more hits than innings pitched.

Now the NN picks:

0.344, 0.742 — Odubel Herrera batting against Wily Peralta.
0.290, 0.739 — Jose Altuve batting against Sonny Gray.
0.332, 0.738 — Daniel Murphy batting against Jon Moscot.
0.273, 0.737 — Miguel Cabrera batting against Jose Quintana.
0.278, 0.729 — Ian Kinsler batting against Jose Quintana.
0.323, 0.726 — Jimmy Paredes batting against Wily Peralta.
0.305, 0.725 — Nick Hundley batting against Luis Perdomo.
0.286, 0.725 — Robinson Cano batting against Derek Holland.
0.321, 0.724 — Stephen Piscotty batting against Jake Peavy.
0.297, 0.723 — Yunel Escobar batting against Gerrit Cole.

The NN agrees with log5 on Odubel Herrera owning the best probability of getting a hit today. The net is also not as down on Jake Peavy, and a little more down on Jose Quintana.