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:

0.339 — Xander Bogaerts batting against Tyler Duffey
0.334 — Daniel Murphy batting against Jeremy Hellickson
0.315 — Josh Harrison batting against Michael Wacha
0.315 — Dustin Pedroia batting against Tyler Duffey
0.314 — Carlos Beltran batting against Mike Pelfrey
0.309 — Starlin Castro batting against Mike Pelfrey
0.309 — David Ortiz batting against Tyler Duffey
0.308 — Jacoby Ellsbury batting against Mike Pelfrey
0.308 — Martin Prado batting against Patrick Corbin
0.306 — Marcell Ozuna batting against Patrick Corbin
0.306 — Mookie Betts batting against Tyler Duffey
0.306 — Ryan Braun batting against Matt Harvey

Log5 is down on Mike Pelfrey, who allowed 78 hits in 58 2/3 innings so far this season against a heating up Yankees offense. It’s also down on Tyler Duffey, who allowed 57 hits in 47 innings against a great Red Sox offense.

Here is the NN list:

0.285, 0.754 — Jose Altuve batting against Matt Andriese.
0.334, 0.746 — Daniel Murphy batting against Jeremy Hellickson.
0.339, 0.744 — Xander Bogaerts batting against Tyler Duffey.
0.301, 0.741 — Danny Valencia batting against Anthony DeSclafani.
0.263, 0.740 — Miguel Cabrera batting against CC Sabathia.
0.315, 0.738 — Josh Harrison batting against Michael Wacha.
0.269, 0.733 — Ian Kinsler batting against CC Sabathia.
0.245, 0.733 — Ben Revere batting against Jeremy Hellickson.
0.308, 0.731 — Martin Prado batting against Patrick Corbin.
0.289, 0.730 — David Peralta batting against Justin Nicolino.

The Log5 method doesn’t like the Detroit starter, but the NN doesn’t like his opponent, CC Sabathia. The consensus picks are Bogaerts, Murphy, and Harrison.

As always there’s a good chance that your pick does not get a hit.