Over the past week, I finished a baseball model I have been working on to try to quantify different angles I look at.
Using Pinnacle lines, it has produced 151 plays (50 ML, 50 RL, 51 totals) yielding +45u. I understand that with extra vig regular books charge that is realistically more like ~35-37u. Before everyone collectively rolls their eyes, I PMed someone who posts in here my plays yesterday before they started and they went 21-5 +19u (the first five days averaged about +5u per day). If they want to verify, that's up to them. I have only played a few for real cash.
1. I know I'm not that good to beat Vegas that badly
2. The sample size is still EXTREMELY small
3. The correlation between plays is very high. I always take the RL along with the ML (the RL results have actually produced better results than the ML). For example, it had Oakland ML, Oakland RL, and Oakland o7.5 yesterday. That will produce some big swings.
4. It sucks sometimes like projecting overs in all 4 Bal/KC games that went 1-3 and weren't close.
5. Variance will kill me sooner or later. Most big favorites have won so far. It only takes a few of those losses to kill results.
I played these for 1/4u (favorites to win 1/4u, dogs each 1/4u) each on 5d overnight lines:
Angels -205, -1.5 +100, o7.5 -122
Royals -163, -1.5 +125, o8.5 -110
Tigers -101, -1.5 +155, o7.5 -105
Brewers +122, -1.5 +185 at bovada today, no total play
Nats -169, -1.5 +135, o6.5 -105
I won't be surprised when they go 2-12. It has to happen sooner or later.
*edit: yeah, it produces a ####load of plays each night so far.