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:

0.369 — Daniel Murphy batting against Adam Morgan
0.332 — Jose Altuve batting against Robbie Ray
0.322 — Xander Bogaerts batting against Mike Wright
0.321 — Eduardo Nunez batting against Sean Manaea
0.317 — Marcell Ozuna batting against Jonathon Niese
0.316 — Martin Prado batting against Jonathon Niese
0.311 — Carlos Gonzalez batting against John Lamb
0.310 — DJ LeMahieu batting against John Lamb
0.309 — Victor Martinez batting against Matt Shoemaker
0.304 — Gerardo Parra batting against John Lamb

I’ll note that John Lamb’s inclusion on the list is based on allowing an extremely high BABIP and HR/9 IP over not too many innings.

The NN produces this top ten:

0.332, 0.763 — Jose Altuve batting against Robbie Ray.
0.369, 0.755 — Daniel Murphy batting against Adam Morgan.
0.322, 0.744 — Xander Bogaerts batting against Mike Wright.
0.302, 0.741 — Josh Harrison batting against Adam Conley.
0.269, 0.739 — Buster Posey batting against Williams Perez.
0.260, 0.738 — Ben Revere batting against Adam Morgan.
0.321, 0.734 — Eduardo Nunez batting against Sean Manaea.
0.292, 0.733 — Miguel Cabrera batting against Matt Shoemaker.
0.316, 0.732 — Martin Prado batting against Jonathon Niese.
0.310, 0.731 — DJ LeMahieu batting against John Lamb.

The two systems seem to be in agreement today on the top three picks, if not the exact order.

On a side note, there are a number of very hot hit streaks in progress right now. With the exception of Alcides Escobar, everyone with a streak of at least eight games is tearing the cover off the ball. Bogaerts is at 24 games, Carlos Gonzalez at nine games with an OBP over .500 and a slugging percentage over 1.000, and Daniel Murphy’s streak has him flirting with .400. These aren’t 1 for 4 streaks.