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 updated 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. I added a graph that gives a visual representation of the probability and success each day. The sheet also includes a table that summarizes the length of positive and negative streaks.
First, the Log5 Method picks:
0.320 — Jose Altuve batting against Marcus Stroman
0.320 — Corey Dickerson batting against Rick Porcello
0.307 — Daniel Murphy batting against Julio Teheran
0.303 — T.J. Rivera batting against Adam Wainwright
0.303 — Wilmer Flores batting against Adam Wainwright
0.302 — Jean Segura batting against Chris Smith
0.299 — Xander Bogaerts batting against Alex Cobb
0.299 — Dee Gordon batting against Jeff Samardzija
0.298 — J.T. Realmuto batting against Jeff Samardzija
0.297 — Mallex Smith batting against Rick Porcello
Altuve and Dickerson tie for the top spot. Altuve is red hot, with multiple hits in seven of his last ten games, including four games in a row with three hits. Dickerson is 8 for 25 against Porcello, a .320 BA. Note that Xander Bogaerts still has a tender hand after a HBP the other night.
Here is how the NN with Park ranks the players:
0.320, 0.760 — Jose Altuve batting against Marcus Stroman.
0.307, 0.757 — Daniel Murphy batting against Julio Teheran.
0.302, 0.748 — Jean Segura batting against Chris Smith.
0.279, 0.732 — DJ LeMahieu batting against Jose Quintana.
0.299, 0.730 — Xander Bogaerts batting against Alex Cobb.
0.299, 0.729 — Dee Gordon batting against Jeff Samardzija.
0.279, 0.729 — Charlie Blackmon batting against Jose Quintana.
0.320, 0.729 — Corey Dickerson batting against Rick Porcello.
0.298, 0.726 — J.T. Realmuto batting against Jeff Samardzija.
0.279, 0.721 — Jose Ramirez batting against Justin Verlander.
0.286, 0.721 — David Peralta batting against Luis Castillo.
The NN agrees with Altuve as number one, with Daniel Murphy not far behind.
As always, your best pick will fail to get a hit about 25% of the time.
Here is the daily list of active streaks of plate appearances without a hit, with pitchers eliminated:
Good luck!
I wonder if “being due” is quantifiable. Like one BTS strategy that is employed is to pick a good hitter who went hitless the day before. Usually good hitters don’t go 2 straight games without getting a hit. I was wondering if a stat exists that gives us hit% of .300+ hitters that have gone hitless the game before. Any idea how I could go about figuring that out besides manually sifting through all the games of prospective hitters that I am interested in?
Earl_Squire » The question is, are the games independent or dependent? That’s the whole idea behind measuring hot hands. A good hitter will have about a 75% chance of getting a hit in any particular game. So we need to answer the question, is p(hit in game | no hit last game) = p(hit in game)? If so, the games are independent. The big problem is that for good hitters the sample size of games with no hit in the last game is going to be small, so it’s going to be difficult to get a meaningful sample.
Also, no Justin Turner on either list today? Seemed like he had a good matchup today. Interesting
I see. That makes a lot of sense. Jim Leyland used to say that momentum is only as good as your next days starting pitcher. Hit streaks and slumps exist where batters might be feeling good and seeing the ball well or not at all, but at the end of the day the previous game should have no bearing on the following one.