Fair enough Dave but remember I was only doing a check on Peter Mair's statement above about the chances of the favourites form his Top Of The Class Form Analysts for Flemington Races 2 and 7 :-
On the good news side, coupling clear favourites with speed maps in races 2 and 7 over 1400 m, suggests the favourites will win both -- and the double will pay about $5. ]
. But here from the computer's Race Integrity Checker software calculations they were both quite wrong. :-
VF140817 FL72306F * WMSPD ORD TNO [ 14 1 13 3 6 8 4 5 2 12 11 10 7 ] * ] BMVRS/W [ ]*SPL*
VF140817 FL72306F * WMSPD RTG ORD [ 1018 959 956 952 947 942 936 933 921 913 905 893 864 ] IMPORTANT
VF140817 FL72306F * RQOD WMSPD RO [ 3 9.6 10.4 11.7 14 16.5 20 22.7 35.5 48.7 65.1 87.4 201.3 ] IN WMSPD ORDER. $PRICE/1 CALCULATED RQD ODDS ON WT MODIFIED SPEED RATINGS *
Just think of the first " [ " left bracket in the WMSPD, Weight Modified Speed line to be The Winning Post , and that no 14 has the highest Speed rating of 101.8 and No 1 has the second highest Speed Rating of 95.9 which says No 14 ought to beat No 1 by (101.8 minus 95.9 ) = by 5.9 lengths.
Now no 14 had its best speed at KENSINGTON over 1300 metres, ( KEN Equalisation not correct ) so perhaps overlook that best chance and think of No 1 as the more likely winner and then look at any horse with a speed rating within 3.0 lengths of 95.9 which would be 92.9 and that would inclinclude down to 921 or 92.1. IE [ 1 13 3 6 8 4 5 2 ] All within 3.0 lengths of the rating for No 1, have a chance to WIN and you would still consider No 14 as some chance, but only for a place.
The Speed Ratings have been equalised by computer over many racetracks and distances allowing one to compare race ratings over different tracks and distances. Sometimes it is right?.
The computer then calculates chances of winning ( using Don Scott's method in his book Winning ) and consequently a Required Price for that horse to be a VALUE Bet. This is shown in the Line * RQOD WMSPD RO [ 3 /1 9.6/1 10.4/1 etc... and sometimes that is useful.
The System was never designed to Just Pick The Winner, but in effect mainly to show Just how hard it was to pick a single selection as an iron clad winner.
The computer having generated about 140 lines from three independent systems attempts to develop a First Four Bet starting with likely Roving Bankers:-
* HORSE NOS [ ]%VAL HILOW. !SELN $20 # 0 NOW! @$0. LR7DS=[ ] TrJkC[ 1 4 5 6 7 8 11 13 ]SPL
*BOX7F4 KSN[ 1 3 2 5 8 10 12 13 14 ]BASE DPLGWP[ 2 3 5 7 8 12 ]*LRWD= 13 1 12 3 2 ]
*DUBL ROV BNKR [ 1 3 2 5 8 ] X [SDup 1 2 3 12 13 <<GW1+ [ 3 8 14 13 5 ]* T4FV[ 1 3 2 13 ]. {<RSE NOS [ >}
*TRPL ROV BNKR [ 1 3 2 5 8 1 3 0] X [ *I/CCL= 1 13 3 6 8 4 ]*MVRvBnk<< SPGM 3 5 <[TLMV 12 2 7 5 3 ? 13 +
T3VIC$ 1 13 12 13 ]%VAL HILOW*. KEEP TRIPLS. * SPLRESIDS** [ 1 2 3 5 7 11 12 13 ] CRUSHD[ 1 12 2 7 11 ]
*IN TX2[ # 3 =13 ] # 1 =12 ] # 13 =11 ] # 2 =10 ] # 5 =9 ] # 12 =8 ] # 8 =7 ] # 7 =5 ] # 11 =3 ] <TRY RVBs
*RvBnkS { IG 3 8 13 } { IL 1 3 13 } { GL 3 13 } { MI 3 } { MG 3 5 } { ML 2 3 12 }*KEY[ TYPE IN
However the computer is still trying to improve its selecting. From this it would eventually select Roving bankers of 3 and 13 and probably include 1 5 12 8. IE [ 3 13 RVB ] x [ 1 5 8 12 2 then maybe 7 11 ] as residuals. Not sure I would take that multi with such skinny Pools, nor believe it can blindly make that type of bet, resulting in a profitable WIN.
Not even close. Top 7 KSN Line boxed F4, would have returned for $8.40 a per 1% outlay, a return of $70.