pinnalce matchups for the belmont 06/10/2006

[formula_2002]

a slighly different selection process.
This will also change in about 1 to 2 weeks

horse line
201 BLUEGRASS CAT -157 *play
202 BOB AND JOHN 140

205 BLUEGRASS CAT -110 play
206 SUNRIVER -106

207 BOB AND JOHN -150 play
208 JAZIL 133

213 DEPUTY GLITTERS -141
214 SACRED LIGHT 124 play

215 DEPUTY GLITTERS 119
216 JAZIL -135 play

219 HEMINGWAYS KEY 103 play
220 OH SO AWESOME -119

225 SACRED LIGHT 187
226 STEPPENWOLFER -210 *play

[ryesteve]

a slighly different selection process.
This will also change in about 1 to 2 weeks
Details?

BTW, I think my head would explode if I tried to track 7 different matchup bets in the same race :D

[formula_2002]

Details?

BTW, I think my head would explode if I tried to track 7 different matchup bets in the same race :D

I'm using more factors, Prime fig, bris, speed, pace,turn time....about 15 or so,plus, some of Jeff P's figures. It's my "buckshot" or perhaps I'll call it "the quantum method" ;)

[ryesteve]

Are you using different weights for different types of races? These factors would have different importance depending on distance and surface.

[formula_2002]

Are you using different weights for different types of races? These factors would have different importance depending on distance and surface.

I'll be runing the data through a multinominal logit formula (just to see what all that means to me).

I dont have as much data as I would like, about 250,000 horses. I could get more data from my "Always" files, about another 250,000 horses.

You may or may not get more accurate results by breaking down things by class, distant, surface, track condition, etc etc..atleast not more meaningful results.The public is just too sharp, but I may be able to beat a fixed odds platform.
Thats why I like a very early line from Pinnacle.. Anything after that could be arbitrage +.

[ryesteve]

I didn't mean you need to break it down that finely. What I meant was, if you run all your races through a model, early speed is going to look good, because most races in your database will be dirt sprints. However, as you may have noticed, most races offered as matchups are turf routes. A model that favors early speed is going to give you a ton of false overlays when you apply it to turf routes.

[formula_2002]

I'm not including these results in my daily scoring..
Too new...

[formula_2002]

I'm not including these results in my daily scoring..
Too new...
New but good.

5 wins in 7 plays, $274 profit, $962 played