Beat the Streak Picks

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). This post discusses an NN that includes the ballpark. I recently update the models, and the results of those tests are here.

For 2017, I am just going to publish the Log5 hit averages and the NN probabilities with parks factored in. I am keeping track of the results here.

First, the Log5 Method picks:

0.283 — J.T. Realmuto batting against Bartolo Colon
0.279 — Daniel Murphy batting against Lance Lynn
0.277 — Gerardo Parra batting against Jered Weaver
0.276 — Devon Travis batting against Wily Peralta
0.275 — DJ LeMahieu batting against Jered Weaver
0.273 — Kevin Pillar batting against Wily Peralta
0.273 — Brandon Drury batting against Jeff Samardzija
0.271 — Yadier Molina batting against Gio Gonzalez
0.271 — Josh Donaldson batting against Wily Peralta
0.271 — Nolan Arenado batting against Jered Weaver

Might this be the day Daniel Murphy gets knocked out of the top spot? Miami plays in a tough park for offense. The Rockies against Jered Weaver in Colorado should do well.

The NN with Park sees the top ten this way:

0.279, 0.722 — Daniel Murphy batting against Lance Lynn.
0.275, 0.709 — DJ LeMahieu batting against Jered Weaver.
0.283, 0.708 — J.T. Realmuto batting against Bartolo Colon.
0.277, 0.699 — Gerardo Parra batting against Jered Weaver.
0.266, 0.697 — Yunel Escobar batting against Cole Hamels.
0.268, 0.696 — Charles Blackmon batting against Jered Weaver.
0.276, 0.694 — Devon Travis batting against Wily Peralta.
0.271, 0.693 — Nolan Arenado batting against Jered Weaver.
0.266, 0.690 — Dee Gordon batting against Bartolo Colon.
0.262, 0.688 — Starling Marte batting against Rookie Davis.
0.273, 0.688 — Brandon Drury batting against Jeff Samardzija.
0.271, 0.688 — Yadier Molina batting against Gio Gonzalez.

That’s six days in a row the NN picked Murphy as most likely to get a hit in a days game. Murphy currently owns a seven game hitting streak, making the model look really good. Note that Murphy is also the consensus pick, as his average position is 1.5, while Realmuto’s average position is 2.0.

Of course, in playing beat the streak, do we really want to stick with the hot player? Hitting streaks end. If you look at a batter game as a random event, it should be independent of the past. Murphy’s chance of getting a hit today should have little to do with the length of his streak, it should have to do with the pitching match-up, the park, and the health of the hitter. Murphy is a career .353 hitter against Lance Lynn, 6 for 17. Realmuto is 9 for 21 against Colon, .429.

No matter which batter you choose, there is a better than 1 in 4 chance they won’t get a hit. Good luck!

from baseballmusings.com http://ift.tt/2pot3OQ