JCapper Message Board

General Discussion
-- Preakness HTML Report Link

Home	Register Log In

Preakness HTML Report Link

jeff
5/19/2012
3:52:02 PM

Here's a link to my html report for the Preakness day card:
http://www.jcapper.com/messageboard/reports/pim05192012.html

-jp

.

jeff
5/19/2012
4:21:05 PM

Some of you have asked me for an opinion in the Preakness itself...

No A UDM horses in that race for me. (Meaning I'll likely sit this one out without making a bet.)

If I have to voice an opinion, I see Bodemeister as the lone speed - with an easier pace scenario in this race to overcome than what he faced in the Derby. However, as the 8/5 ML favorite I see zero value - which dovetails with what the UDMs on the report are telling me. (Pass.)

Good luck to all,

-jp

.

bill2010
5/20/2012
9:46:18 AM

HI --JEFF I WAS LOOKING AT YOUR HTML REPORT -CAN U TELL ME WHAT THE RPL MEANS---THANKS BILL

jeff
5/20/2012
1:30:59 PM

Although I'm using the characters RPL to label it on the report, the factor appearing on the report in that slot is a UserFactor.

RPL stands for repellant. The UserFactor is based on a mixture of low ML, high WinPct for Rider/Trainer, late pace, and high final time based speed figs. Horses ranked highly vs. their fields in this UserFactor tend to attract an inordinate amount of money - and the data says that tendency repels any realistic chance of profiting by betting on such horses.

Thus, the characters RPL (Repellant) instead of UF3.

-jp

.

~Edited by: jeff on: 5/20/2012 at: 1:30:59 PM~

bill2010
5/20/2012
3:37:45 PM

hi jeff thanks---bill

jeff
5/20/2012
4:40:48 PM

I'll try illustrating the point behind creating this type of factor...

The following sample was pulled from a StarterHistory table containing HDW data for all starters 1/1/2011 through yesterday 5/19/2012. The results are for horses ranked 1st in my own UPR that also rank 1st or 2nd in both FormConsensus and EarlyConsensus (F07 and F19 in my sql factor setup respectively.)

The factor combination in the sample... UPR with EarlyConsensus and/or FormConsensus... figures prominently in many of my UDMs. I use it because it produces a decent GENERAL starting point (~30pct win rate and ~.90 flat win bet roi) from which to begin an assault on the pools.

The sample data is further broken out by numeric value for UserFactor3 (F33 in my sql factor setup.)

Note the flat win bet roi through much of the lower area of the sample - where F33 numeric value is less than 150. Also note the roi through much of the upper area of the sample - where F33 numeric value is greater than 250. Those are the sweet spots -although the sweet spot at the upper edge of the sample ( >=250) is mostly a newer development.


query start:         5/20/2012 12:33:34 PM
query end:           5/20/2012 12:34:36 PM
elapsed time:        62 seconds
`
Data Window Settings:
Connected to: C:\JCapper\exe\JCapper2.mdb
999 Divisor  Odds Cap: None
`
SQL:  SELECT * FROM STARTERHISTORY
      WHERE RANKUPR=1
      AND RANKF07 <= 2 
      AND RANKF19 <= 2
`
`
Data Summary         Win     Place      Show
Mutuel Totals   40921.40  40013.90  39271.20
Bet            -43894.00 -43894.00 -43894.00
Gain            -2972.60  -3880.10  -4622.80
`
Wins                6704     10878     13479
Plays              21947     21947     21947
PCT                .3055     .4956     .6142
`
ROI               0.9323    0.9116    0.8947
Avg Mut             6.10      3.68      2.91
`
`
By: SQL-F33
`
  >=Min      < Max      Gain       Bet       Roi   Wins  Plays     Pct   Impact
-999.00       0.00      0.00      0.00    0.0000      0      0   .0000   0.0000
   0.00      25.00     54.70    560.00    1.0977     52    280   .1857   0.6080
  25.00      50.00    -66.40   3070.00    0.9784    357   1535   .2326   0.7614
  50.00      75.00   -304.80   5450.00    0.9441    757   2725   .2778   0.9094
  75.00     100.00   -231.10   6262.00    0.9631    864   3131   .2760   0.9034
 100.00     125.00   -351.30   6522.00    0.9461    969   3261   .2971   0.9728
 125.00     150.00   -270.20   6378.00    0.9576    996   3189   .3123   1.0225
`
 150.00     175.00   -584.50   5386.00    0.8915    850   2693   .3156   1.0333
 175.00     200.00   -440.90   4124.00    0.8931    702   2062   .3404   1.1145
 200.00     225.00   -373.00   2744.00    0.8641    475   1372   .3462   1.1334
 225.00     250.00   -361.60   1664.00    0.7827    284    832   .3413   1.1175
`
 250.00     275.00     29.40    890.00    1.0330    206    445   .4629   1.5155
 275.00     300.00    -18.40    472.00    0.9610    106    236   .4492   1.4704
 300.00     325.00     -9.00    200.00    0.9550     47    100   .4700   1.5386
 325.00     350.00    -32.00     82.00    0.6098     17     41   .4146   1.3574
 350.00     375.00    -16.00     70.00    0.7714     16     35   .4571   1.4966
 375.00     400.00      2.50     20.00    1.1250      6     10   .6000   1.9642
 400.00     425.00      0.00      0.00    0.0000      0      0   .0000   0.0000
 425.00     450.00      0.00      0.00    0.0000      0      0   .0000   0.0000
 450.00  999999.00      0.00      0.00    0.0000      0      0   .0000   0.0000

The objective was to develop a unique factor strongly correlated to how the public bets the races. I think of the lower area of this sample (repellant less than 150) as contenders that repel public money.

This UserFactor isn't a game changer by any stretch of the imagination. But it does provide a unique look at the data - and it does a reasonably good job of predicting which horses the public will like - as well as which horses the public will turn their noses up at.

-jp

.

~Edited by: jeff on: 5/20/2012 at: 4:40:48 PM~

JCapper Message Board

General Discussion -- Preakness HTML Report Link

General Discussion
-- Preakness HTML Report Link