By |
MLOR3 |
bettheoverlay 9/25/2008 10:53:39 AM | I've always been confused on this factor. The glossary states that it is driven by JPRMLOddsProb. I thought I read somewhere that if your odds line is based on your UPR then MLOR3 is driven by your UPRMLOddsProb. Is that correct?
|
jeff 9/26/2008 12:51:02 AM | In the Data Window: MLOR3 = ML / ((1 / JPRMLProb) - 1) MLOR5 = ML / ((1 / UPRMLProb) - 1)
...where ((1 / Prob) - 1) represents an odds line derived from the probability estimate.
On the System Definitions Screen: If you use the program default JPR ML OL as your odds line type you will see JPRMLProb on your HTML Report. If you have the Enable Overlay Highlighting box checked, the Overlay Highlighting Cutoffs specified on the Sys Defs Screen will be driven by MLOR3.
Once you've spent time developing UPR to the point to where you are seeing UPRMLProb outperform JPRMLProb in the Data Window you may want to select User Def OL as your odds line type on the Sys Defs Screen.
When User Def OL is the selected odds line type you will see UPRMLProb on the HTML Report. Overlay Highlighting will be driven the cutoffs you have defined on the Sys Defs Screen and MLOR5.
BettorsToteProb: When User Def OL is your selected odds line type on the Sys Defs Screen, BettorsToteProb in the Data Window and in the Live Play Module is based on current odds and UPR.
When JPR ML OL is your selected odds line type on the Sys Defs Screen, BettorsToteProb in the Data Window and in the Live Play Module is based on current odds and JPR.
MLOR3 in UDMs: If you add factor constraints for MLOR3 to a UDM, both the Profile Marker and Data Window will evaluate MLOR3 (based on JPR) against the UDM Definition. The odds line type selected on the Sys Defs Screen has no effect on this. When it comes to UDMs MLOR3 is always based on JPR.
I THINK that covers it... whew!
-jp
.
~Edited by: jeff on: 9/26/2008 at: 12:51:02 AM~
|
Hdcper 9/26/2008 1:20:17 AM | Jeff,
In the next upgrade, can you add MLOR5 to both the data window (as a factor)for testing and in the Udm Wizard so it can be added to a Udm (didn't find it in the profile table either)?
Thanks,
Bill
|
bettheoverlay 9/26/2008 1:32:06 AM | I have been using my UPR for an odds line for a year so I understand the system definition part. But I've never seen MLOR5. Sounds like a great factor for UPR users, where is it? I also don't see a glossary definition for OddsRatio.
Edit:I see Hcapper beat me to the MLOR5 question.
~Edited by: bettheoverlay on: 9/26/2008 at: 1:32:06 AM~
|
jeff 9/26/2008 2:01:25 AM | --quote:"In the next upgrade, can you add MLOR5 to both the data window (as a factor)for testing and in the Udm Wizard so it can be added to a Udm (didn't find it in the profile table either)?" --end quote
Data Window? Yes.
UDM Wizard? No.
The UDM Wizard's memory stack is maxed out. The only way I can add anything new to the factors it handles is to add the new factor to the factor XML. Not worth the effort at this point. Working on implementing a 3rd profile table. Realistic ETA towards the end of this year.
If after seeing MLOR5 in the Data Window you want to enforce cutoffs for it in UDMs - your best bet would be through a combination of Overlay Highlighting (with User Def OL selected on the Sys Defs Screen) and the BettingInstructions field... something like the following keywords would do the trick: need yellow.
-jp
.
|
Hdcper 10/11/2008 1:47:28 AM | Hey Jeff,
Dropped you an email a few days back about MLOR5 in the data window and correction of Rdebut. Was wondering about the time frame to get them finalized.
Thanks,
Bill
|
jeff 10/11/2008 12:25:45 PM | Bill, working on it.
I have MLOR5 working in a version of the Data Window as I write this.
But I want to set expectations from the beginning. MLOR5 is going to be Data Window only.
I have no plans to add it to the UDM Wizard. My own UPR easily outperforms JPR. But I'm not seeing real benefit for MLOR5 roi-wise in my own Data Window queries when comparing it side by side against MLOR3.
A morning line opinion that differs substantially from the program's opinion is what makes MLOR3 effective in the first place. Substantial differences are be picked up by both MLOR3 and MLOR5 in the great majority of cases.
Route Debut is another matter. Yes, you've discovered and reported a bug. Through human error I've managed to break the route debut preset filters at some point in the past when implementing a change to the program.
Stepping through the code that handles route debut horses a few days ago I isolated one chunk of code that is causing those preset filters to fail.
My concern is that there are LOTS of places in JCapper where those filters are referenced. I have to step through a LOT of code one line at a time - testing the RouteDebut True/False attribute of every horse - making sure that each block of code in the program that relies on a RouteDebut True/False attribute is receiving the correct True/False information.
I WILL get there as soon as humanly (humanely) possible. I used the term humanely because stepping through that much code one line at a time looking for something that's broken when you don't know exactly what's broken is tedious. It wears on a guy.
To be able to specifically say when it'll be ready isn't possible.
What I can say is that as soon as I know that it's right - that's when I'll publish it as a bug fix in a program update.
-jp
.
|
Hdcper 10/11/2008 9:13:05 PM | Works for me,
Thanks,
Bill
|
jeff 10/14/2008 4:11:12 PM | Got it. The bug fix will be part of (a really tiny part of) the next program update, most likely Thursday of this week.
Believe me when I say this:
You guys are really going to like the new toys.
-jp
~Edited by: jeff on: 10/14/2008 at: 4:11:12 PM~
|