Page 1 of 2 12 LastLast
Results 1 to 10 of 14
  1. #1

    Default MyCalc Poll separation seems arbitrary?

    Either I'm missing something - very probable - or Mycalc doesn't allow equal rankings. For example, in a poll of Place percentage, with 4 horses equal on 50%, Excel (and logic?) gives them equal ranking. Mycalc gives them 4 different ranks.

    Can someone explain the logic behind this? And exactly how are they separated? At first I thought it was by Tab no., but that's not the case.

    Is it possible to work around the issue or have I just got it all wrong???

    Steve W

  2. #2
    Join Date
    Feb 2004
    Location
    New Norfolk, Tas, Australia.
    Posts
    526

    Default

    Steve

    I think there is a built in formula in BS that applies something like the AVB1 rating whenever an equal result arises?

    Mike.

  3. #3
    Join Date
    Feb 2004
    Location
    New Norfolk, Tas, Australia.
    Posts
    526

    Default

    Although I cannot find a way to totally eliminate the occurrences of dual selections or only select the highest rated without relying on the BS built in CALCRANKMAX to separate them here are a few ways to almost eliminate the dual qualifiers.

    Code:
    0.REPNAME=any,CALCMIN=1,CALCMAX=99999,CALCRANKMIN=1,CALCRANKMAX=1,CALCNAME=any
    1.LOCATION$"SMBAWXsmbaw" 
    
    MYCALC = IF(CPLCP<101,CPLCP*AVB1,0)
    
    SYSTEM RESULTS FOR: MY RANKINGS_ANYRANKING 01/08/2018-31/08/2018
                     WIN      PLACE    QUINELLA    EXACTA    TRIFECTA   FIRST FOUR
    Races Bet:    1297      1286       1297       1295        1297         1253
    Races Won:     255       592        364        186         137           98
    S.R./Race:      19.7%     46.0%      28.1%      14.4%       10.6%         7.8%
    Outlay($):    1297.00   1286.00    6588.00    6577.00    27112.00     82542.00
    Return   :     991.10   1053.80    4123.50    3851.30    15173.20     47303.40
    $ Profit :    -305.90   -232.20   -2464.50   -2725.70   -11938.80    -35238.60
    % P.O.T. :     -23.6%    -18.1%     -37.4%     -41.4%      -44.0%       -42.7%
    ********************************************
    
    MYCALC = IF(CPLCP<101,CPLCP*H:TNO*H:BP,0)
    
    0.REPNAME=any,CALCMIN=1,CALCMAX=99999,CALCRANKMIN=1,CALCRANKMAX=1,CALCNAME=any
    1.LOCATION$"SMBAWXsmbaw" 
     
    SYSTEM RESULTS FOR: MY RANKINGS_ANYRANKING 01/08/2018-31/08/2018
                     WIN      PLACE    QUINELLA    EXACTA    TRIFECTA   FIRST FOUR
    Races Bet:    1287      1276       1287       1285        1287         1244
    Races Won:     153       401        224        117          88           66
    S.R./Race:      11.9%     31.4%      17.4%       9.1%        6.8%         5.3%
    Outlay($):    1287.00   1276.00    6857.00    6846.00    30102.00    100362.00
    Return   :     808.30    918.50    3549.60    3814.30    12418.10     53175.40
    $ Profit :    -478.70   -357.50   -3307.40   -3031.70   -17683.90    -47186.60
    % P.O.T. :     -37.2%    -28.0%     -48.2%     -44.3%      -58.7%       -47.0%
    *******************************************
    
    MYCALC = IF(CPLCP<101,CPLCP,0)
    
    SYSTEM RESULTS FOR: MY RANKINGS_ANYRANKING 01/08/2018-31/08/2018
    Created by Bet Selector at 14:34:48 on 14/09/2018 based on Tattsbet TAB prices
                     WIN      PLACE    QUINELLA    EXACTA    TRIFECTA   FIRST FOUR
    Races Bet:    1297      1286       1297       1295        1297         1253
    Races Won:     246       593        365        176         131           96
    S.R./Race:      19.0%     46.1%      28.1%      13.6%       10.1%         7.7%
    Outlay($):    1297.00   1286.00    6598.00    6587.00    27212.00     83202.00
    Return   :     964.70   1062.60    4738.70    4416.20    15180.00     65902.40
    $ Profit :    -332.30   -223.40   -1859.30   -2170.80   -12032.00    -17299.60
    % P.O.T. :     -25.6%    -17.4%     -28.2%     -33.0%      -44.2%       -20.8%
    ********************************************
    
    IF(AVB1V>1,AVB1V,0)
    
    SYSTEM RESULTS FOR: MY RANKINGS_ANYRANKING 01/08/2018-31/08/2018
    Created by Bet Selector at 15:29:51 on 14/09/2018 based on Tattsbet TAB prices
                     WIN      PLACE    QUINELLA    EXACTA    TRIFECTA   FIRST FOUR
    Races Bet:    1305      1293       1305       1303        1305         1260
    Races Won:     291       645        393        225         162          119
    S.R./Race:      22.3%     49.9%      30.1%      17.3%       12.4%         9.4%
    Outlay($):    1305.00   1293.00    6499.00    6489.00    25994.00     75762.00
    Return   :    1093.60   1130.50    5430.00    6161.50    15993.60     64965.80
    $ Profit :    -211.40   -162.50   -1069.00    -327.50   -10000.40    -10796.20
    % P.O.T. :     -16.2%    -12.6%     -16.4%      -5.0%      -38.5%       -14.3%
    Last edited by mikemcbain; 14-09-18 at 04:47 PM.

  4. #4

    Default

    Thanks Mike. I'm going to have play with that before I pretend to understand it. Today's race day, tomorrow's drinking day. If you can, check in later next week. I'm sure to have some questions.

    Regards,

    Steve

  5. #5
    Join Date
    Jul 2015
    Location
    Adelaide
    Posts
    48

    Default

    Steve

    Can you provide an example where the results aren't ranked by TAB number ?
    All of the tests I've done rank them this way.

    E.G. For today's Sydney races, there are 6 horses with a career place % of exactly 50%.

    MyCalc called placefifty :

    IF(Within(CPLCP,50,50),100,0)

    Expert system rules

    0.CALCMIN=100,CALCNAME=placefifty
    1.LOCATION$"S"

    This correctly selects the six horses. If you add the condition CALCRANKMAX=1, the horse returned is the lowest TAB number, rank 2 gives the next TAB number etc etc

    Thanks
    Brian

  6. #6

    Default

    Thanks Brian. Tab number was my first impression, so I may have been mistaken when I thought I'd found an aberration. My concern here is that (using your example and assigning them Tab nos 1-6) Horse no 6 could be could be ranked a clear 1st in the 3 other polls of a 4 poll system, yet because of the arbitrary application of Tab nos, not the selection.

    Probably unrealistic of me to expect Neale to change his algorithms But is there a way around it? I'll have a look at Mike's solution this week and see if that helps.

    Regards,

    Steve

  7. #7

    Default

    Mike, If I understand correctly, you're suggesting several ways to split a poll points element (cplcp) using other factors. I think BS already does that using Tab numbers, but I don't want them separated. If 2 or more horses have the same percentage, then I want them to have the same PPRANK. Starting to look like it's not possible.

    regards,

    Steve

  8. #8
    Join Date
    Jul 2015
    Location
    Adelaide
    Posts
    48

    Default

    Steve

    In my last post, the six horses are not all in the same race - there's 1 in race 3, 2 in race 7 and 3 in race 9.
    So they are given the same rank (or value) by applying the arbitrary value of 100 if they meet the criteria of having a place percentage of exactly 50%.
    Testing for the value of 100 correctly selects the 6 horses in the 3 races.
    Is this not what you're looking for ?

    Cheers
    Brian

  9. #9

    Default

    Hi Brian,

    The following table shows the Place Percentage Rankings for Rosehill R5 25/8/18. It is the race which has the aberration I spoke of. You will see there are 4 horses with 50% placings. Excel ranks them all equally, but Bet Selector artificially separates them - in most races by Tab number, however in this case horse no. 5, Plague Stone gives the lie to that.

    My ambition is to have BS replicate the Excel rankings. Excel gives them all the lowest available ranking (from 5,4,3,2), but I would be happy with them all having 5, or even the average (3.5). My point is that equal results should result in an equal ranking.

    The question is can this be achieved and if so, how?

    Thanks for your help,

    Steve

    Track Race TabNo Horse Plc% Excel Bet Sel
    R'hill 5 1 Performer 50.00 2 5
    R'hill 5 2 Paquirri 80.00 7 7
    R'hill 5 3 Jonker 66.67 6 6
    R'hill 5 4 Dio D'Oro 83.33 8 8
    R'hill 5 5 Plague Stone 50.00 2 2
    R'hill 5 6 Graff 100.00 9 9
    R'hill 5 7 Tchaikovsky 50.00 2 4
    R'hill 5 8 Neutrality 50.00 2 3
    R'hill 5 9 Ringerdingding 20.00 1 1

  10. #10
    Join Date
    Jul 2015
    Location
    Adelaide
    Posts
    48

    Default

    Steve

    Try as I might, I cannot get BS to rank equal selections by any other means than TAB number, unless the underlying calculation produces a different ranking order.

    So I'm assuming that the rules of your poll are changing the rankings.

    Neale's help file says "MyCalc rankings assume your MyCalc returns positive numbers with the best horse having the highest number." It doesn't offer any insight in to how equal rankings are treated.

    I've run the calculation 'placefifty' shown above on the Race 5 you mention with the result
    Code:
    25/08/2018 (SAT)     ROSEHILL         (ROS)       WIN   PLC    QN    EX     TT      FF 
    SR 5 # 1 1L0x Performer             100.00  3RD        2.00 
    SR 5 # 5 5x14 Plague Stone          100.00                  
    SR 5 # 7 310x Tchaikovsky           100.00                  
    SR 5 # 8 1x4x Neutrality            100.00  2ND        1.70  12.0
    This is the only way I can get all four horses - but even though they have the same MyCalc score of 100, BS still ranks them 1-4 by TAB number - and this can be verified from the CSV file from the Sytem Analysis.

    Cheers
    Brian

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •