Using the Data Window in sql mode, one of the tests that I subject my UDMs to is to break the data out by RGN.
RGN is a randomly generated number between 1 and 100 assigned to each starter by the Database Builder at the point in time when the StarterHistory table is populated.
Below are two data samples that illustrate what I am talking about.
The first data sample is a UDM from its development data set:
query start: 3/8/2013 12:23:03 PM
query end: 3/8/2013 12:23:53 PM
elapsed time: 50 seconds
`
Data Window Settings:
Connected to: C:\JCapper\exe\JCapper2.mdb
999 Divisor Odds Cap: None
Betting Instructions: EU 0.85
`
UDM: Layer4_JacksOrBettor --(Tagged as: Date Last Modified: 05-21-0212)
`
SQL: SELECT * FROM STARTERHISTORY
WHERE (UPR >= 75
AND (RANKUPR=1...
`
AND AGEOFHORSE >= 3
AND STARTSLIFETIME >= 4
AND RACEVOLATILITY >= 30
AND FIELDSIZE >= 6
AND DIST >= 880
AND PACEINDEX < 90
AND MLOR3 >= 1.65
AND MLOR2 >= .9
AND RANKMLINE BETWEEN 2 AND 9
AND MLINE >= 4.5)
AND [DATE] >= #01-21-2011#
AND [DATE] <= #05-20-2012#
ORDER BY [DATE], TRACK, RACE
`
`
Data Summary Win Place Show
-----------------------------------------------------
Mutuel Totals 5877.10 4468.80 4086.60
Bet -4172.00 -4172.00 -4172.00
-----------------------------------------------------
P/L 1705.10 296.80 -85.40
`
Wins 410 697 959
Plays 2086 2086 2086
PCT .1965 .3341 .4597
`
ROI 1.4087 1.0711 0.9795
Avg Mut 14.33 6.41 4.26
`
`
By: RGN
`
>=Min < Max P/L Bet Roi Wins Plays Pct Impact
--------------------------------------------------------------------------------------
-999.00 5.00 90.70 162.00 1.5599 21 81 .2593 1.3191
5.00 10.00 178.80 244.00 1.7328 31 122 .2541 1.2928
10.00 15.00 143.30 232.00 1.6177 25 116 .2155 1.0965
15.00 20.00 89.80 174.00 1.5161 18 87 .2069 1.0526
20.00 25.00 -128.80 246.00 0.4764 10 123 .0813 0.4136
25.00 30.00 58.10 216.00 1.2690 21 108 .1944 0.9893
30.00 35.00 20.40 210.00 1.0971 15 105 .1429 0.7268
35.00 40.00 -21.50 188.00 0.8856 12 94 .1277 0.6495
40.00 45.00 207.80 172.00 2.2081 21 86 .2442 1.2424
45.00 50.00 -5.30 208.00 0.9745 17 104 .1635 0.8317
50.00 55.00 231.10 206.00 2.1218 24 103 .2330 1.1855
55.00 60.00 131.90 202.00 1.6530 25 101 .2475 1.2594
60.00 65.00 164.70 210.00 1.7843 25 105 .2381 1.2114
65.00 70.00 168.50 228.00 1.7390 24 114 .2105 1.0711
70.00 75.00 10.50 228.00 1.0461 18 114 .1579 0.8033
75.00 80.00 43.40 198.00 1.2192 18 99 .1818 0.9251
80.00 85.00 66.80 228.00 1.2930 22 114 .1930 0.9819
85.00 90.00 137.40 170.00 1.8082 22 85 .2588 1.3168
90.00 95.00 10.40 186.00 1.0559 16 93 .1720 0.8753
95.00 999999.00 107.10 264.00 1.4057 25 132 .1894 0.9636
Note that in the above data sample, the RGN data is broken out in 20 separate rows - each row having an increment of 5 points each. Also note that 17 of the 20 rows (or 85%) are individually profitable.
Attempting to relate that to months - if live play were restricted to only those plays generated by the single UDM used in the above sample AND IF THE UDM'S ACTUAL PERFORMANCE DURING LIVE PLAY REMAINED ON PAR WITH THE PERFORANCE SEEN IN THE DEVELOPMENT SAMPLE, ADMITTEDLY A TALL ORDER:
Each month would have an (approximate) 85% chance of profitability. Put another way: The player's expectation would be an impressive 10.2 profitable months out of each 12 (something I have not personally been able to achieve.)
Here is the same UDM run through the Data Window broken out by RGN... except this time, the time period of the sample is that after I began using the UDM to generate live plays:
query start: 3/8/2013 12:38:39 PM
query end: 3/8/2013 12:39:00 PM
elapsed time: 21 seconds
Data Window Settings:
Connected to: C:\JCapper\exe\JCapper2.mdb
999 Divisor Odds Cap: None
Betting Instructions: EU 0.85
UDM: Layer4_JacksOrBettor --(Tagged as: Date Last Modified: 05-21-0212)
SQL: SELECT * FROM STARTERHISTORY
WHERE (UPR >= 75
AND (RANKUPR=1...
`
AND AGEOFHORSE >= 3
AND STARTSLIFETIME >= 4
AND RACEVOLATILITY >= 30
AND FIELDSIZE >= 6
AND DIST >= 880
AND PACEINDEX < 90
AND MLOR3 >= 1.65
AND MLOR2 >= .9
AND RANKMLINE BETWEEN 2 AND 9
AND MLINE >= 4.5)
AND [DATE] >= #05-21-2012#
ORDER BY [DATE], TRACK, RACE
`
`
Data Summary Win Place Show
-----------------------------------------------------
Mutuel Totals 2761.50 2253.80 2073.30
Bet -2394.00 -2394.00 -2394.00
-----------------------------------------------------
P/L 367.50 -140.20 -320.70
`
Wins 221 381 523
Plays 1197 1197 1197
PCT .1846 .3183 .4369
`
ROI 1.1535 0.9414 0.8660
Avg Mut 12.50 5.92 3.96
`
`
By: RGN
`
>=Min < Max P/L Bet Roi Wins Plays Pct Impact
--------------------------------------------------------------------------------------
-999.00 5.00 -38.10 74.00 0.4851 5 37 .1351 0.7319
5.00 10.00 0.70 150.00 1.0047 12 75 .1600 0.8666
10.00 15.00 2.60 120.00 1.0217 10 60 .1667 0.9027
15.00 20.00 76.30 110.00 1.6936 13 55 .2364 1.2802
20.00 25.00 6.00 118.00 1.0508 14 59 .2373 1.2852
25.00 30.00 76.70 128.00 1.5992 15 64 .2344 1.2694
30.00 35.00 3.30 96.00 1.0344 10 48 .2083 1.1284
35.00 40.00 -37.00 112.00 0.6696 9 56 .1607 0.8705
40.00 45.00 70.10 138.00 1.5080 14 69 .2029 1.0990
45.00 50.00 79.00 120.00 1.6583 13 60 .2167 1.1735
50.00 55.00 3.80 128.00 1.0297 10 64 .1563 0.8463
55.00 60.00 -1.60 126.00 0.9873 11 63 .1746 0.9457
60.00 65.00 24.20 128.00 1.1891 10 64 .1563 0.8463
65.00 70.00 -2.80 134.00 0.9791 11 67 .1642 0.8892
70.00 75.00 -13.10 118.00 0.8890 7 59 .1186 0.6426
75.00 80.00 0.40 136.00 1.0029 11 68 .1618 0.8762
80.00 85.00 64.60 98.00 1.6592 13 49 .2653 1.4370
85.00 90.00 49.50 140.00 1.3536 16 70 .2286 1.2380
90.00 95.00 18.70 88.00 1.2125 8 44 .1818 0.9848
95.00 999999.00 -15.80 132.00 0.8803 9 66 .1364 0.7386
Note that in the above data sample, the RGN data is broken out in 20 separate rows - each row having an increment of 5 points each. Also note that 14 of the 20 rows (or 70%) are individually profitable.
Attempting to relate that to months - if live play were restricted to only those plays generated by the single UDM used in the above sample - which this time reflects actual performance seen during live play:
Each month would have an (approximate) 70% chance of profitability. Put another way: The player's expectation would be (approximately) 8.4 profitable months out of each 12.
Based on personal experience using my own UDMs, 8.4 months of profitable play out of each 12 is realistically achievable.
Ideally I would love to be profitable EVERY month - but the reality is I have never been able to achieve that.
~Edited by: jeff on: 3/8/2013 at: 3:07:20 PM~
| 