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:
0.334 — Eduardo Nunez batting against Jered Weaver
0.326 — Yunel Escobar batting against Ricky Nolasco
0.322 — Ryan Braun batting against Matt Cain
0.316 — Victor Martinez batting against James Shields
0.313 — Jonathan Lucroy batting against Matt Cain
0.311 — Ian Kinsler batting against James Shields
0.306 — Danny Valencia batting against Cesar Ramos
0.300 — Mike Trout batting against Ricky Nolasco
0.300 — Miguel Cabrera batting against James Shields
0.298 — Cameron Maybin batting against James Shields
While the Log5 method does not like Ricky Nolasco and James Shields, Eduardo Nunez against Jered Weaver is by far the most likely combination to produce a hit. The Twins shortstop sports a high batting average with few walks and few strikeouts. So his BABIP is high and he puts the ball in play often.
Here’s the NN list:
0.326, 0.748 — Yunel Escobar batting against Ricky Nolasco.
0.334, 0.735 — Eduardo Nunez batting against Jered Weaver.
0.300, 0.732 — Miguel Cabrera batting against James Shields.
0.283, 0.732 — Daniel Murphy batting against Kyle Hendricks.
0.268, 0.731 — Buster Posey batting against Chase Anderson.
0.306, 0.728 — Danny Valencia batting against Cesar Ramos.
0.263, 0.727 — David Peralta batting against Michael Bolsinger.
0.311, 0.726 — Ian Kinsler batting against James Shields. Parameters:
0.278, 0.723 — Francisco Lindor batting against Edinson Volquez.
0.300, 0.722 — Mike Trout batting against Ricky Nolasco.
The NN flips one and two, but they seem to agree that Nunez and Escobar are the best bets. Miguel Cabrera is always a wise choice as well. As always, note that even with the best picks, there is about a 25% chance that the player will not get a hit.
from baseballmusings.com http://ift.tt/24LgBGp
No comments:
Post a Comment