Database Handicapping Software- JCapper

JCapper Message Board

          JCapper 101
                      -- How do you organize your UDM's?

Home Register
Log In
By How do you organize your UDM's?
Charli125
6/25/2009
7:39:39 PM
I've only been using the software for about 3 months, so this is my first time posting. Go easy on me!

I am running into an issue where I have too many UDM's. I have made them very specific, all geared to a single track, and usually to a single surface and distance. Obviously this results in too many UDM's to keep track of. I'd like to simplify a bit. I'd also like the ability to be able to plug in the pp's from a track I don't usually play, and be able to come to some conclusions.

I'm wondering how you guys have handled this issue. Do you create a playlist that is created from all of the pp's for a year? I'm not sure if you would get any good data out of that, and I don't feel like spending the time it would take if will just spit out garbage. GIGO, you know.

Thanks guys, I appreciate any advice.

Reply
jeff
6/25/2009
8:54:10 PM
http://www.jcapper.com/messageboard/avatars/06252009HOLR7.JPG

The above link is a screenshot of my Live Play Module (in ML/Manual mode) for today's 7th at HOL.

I'll start by pointing out that there is a framework to my UDMs:

1. BUSINESS UDMs - these are similar in makeup to the type of UDM I described in the Finding An Edge help doc. I only carry a handful of these - usually between 5-10 UDMs only depending on what tracks are running and the time of the season. Each UDM that I have of this type is generally what I refer to as a "tight" model UDM - meaning that it is very tight when it comes to horse selection. Any horse picked by a tight model UDM undergoes a screening process of at least a few dozen factor constraints. Usually, but not always, horses picked by tight model UDMs have very few holes in their records. I should also point out that I require any horse I play to be flagged by one of my Business UDMs. Using the mark chars field in the UDM Wizard I've color coded my Business UDMs to display in bright green. The first thing I look for on an HTML Report or the Live Play Module is green color.

2. Layering UDMs - these (for me) are usually based on factors that aren't necessarily part of my UPR... for instance: riders, trainers, track profile... ok track profile sometimes IS part of my UPR. But the idea behind them is to tell me something at a glance that (hopefully) leads to an informed decision about a horse. My layering UDMs are what I refer to as "loose" model UDMs. IOW they are somewhat loose when it comes to picking horses. For example, the UDM named layer_EarlyDR shown on the screen capture is track profile based. It's based on strong CPace, track code, and distance. Nothing else. Its presence tells me that the horse selected not only has strong CPace relative to the rest of the field but that strong CPace is a really good commodity to have for the surface-distance of today's race. I won't bet horses selected by layering UDMs on their own. I use them to help me form a mental picture of each horse. I color code my layering UDMs in the following manner: light green means something worth noting - maybe a small positive, light blue = something a little stronger - maybe a hot rider or trainer, bright blue means a layering effect that's even stronger, and grey text for negative expectation UDMs and/or question marks... such as horses making a route or turf debut.

What I'm trying to achieve through all of this is to have the software present me with a single screen where I can make an educated play or pass decision.

And that play or pass decision usually comes down to getting the following question answered:

Q. Given the UDM mix converging on the individual horse, is ExpectedValueBettorsToteProb strong enough for me to bet this horse profitably?

IOW, if I were to run 100,000 similar horses through the Data Window what would my results look like?

I have a few rules...

First I need to see bright green. If there's positive layering to go with it - so much the better.

No bright green, no bet.

Bright green with positive layering... THAT gets my attention every time.

Bright green with positive layering and a question mark or two? Believe it or not sometimes that's even better. Horses with obvious question marks in their records often make the public look elsewhere.

How many business UDMs does a player really need?

I may currently have about 8 of them... but if you want my advice that may be about 4-5 too many for most players.

In the end you want to be working with enough good ideas to provide you enough solid plays. "Enough" means different things to different players.

Try to stay within your comfort zone.

One of the most successful players I know in terms of roi makes an average of about 5 bets per week.

That's too little action for me.

My conmfort zone is that I want between 8-15 potential plays a day downloading 5-7 tracks a day.

And believe it or not I purposely deactivate business UDMs if I am getting too many plays.

And to answer your question - YES - I do run my UDMs against samples of all tracks everywhere... even if they are track specific when I play live... I want to see decent performance elsewhere... it let's me know that my UDMs are based on solid ideas.

One more point... Someone else may have an entirely different take on UDM framework/approach to the game. And that framework, although different from my own, might be every bit - and even more so - valid than my own.



Sorry for being so long winded...

Hope that helps,


-jp

.



~Edited by: jeff  on:  6/25/2009  at:  8:54:10 PM~

Reply
JustRalph
6/26/2009
1:15:11 PM
you organize your UDM's?

wow....I have a hundred and they are in no way organized.....

Reply
Charli125
6/26/2009
3:51:51 PM
That's exactly what I was looking for. I would guess that my UDMs are all layered UDMs as you described. I haven't gotten far enough along to build solid business UDMs. Thanks for the input, that will send me down the right path.

Reply
PT
6/26/2009
8:42:35 PM

--quote:
"you organize your UDM's?

wow....I have a hundred and they are in no way organized.....
"
--end quote


Funny that you mention that...as I've been getting comfortable with the program, I was heading down that same path...and with only a few tracks. I figured I better clean things up when I was finding two or three UDM's landing on different horses in the same race. Then I'd need a UDM to choose between my UDM's!

-Paul

Reply
busseb
6/27/2009
6:26:44 PM
For those of me who isn't as sophisticated as most
of the others who post here, I just use the first
letter to designate what type of UDM it is:

T-NameOfUDM a trainer UDM colored yellow
J-NameOfUDM a rider UDM colored yellow
x-NameOfUdm a negative expectancy UDM colored pink

I only play MNR and WOX, so:
W-NameOfUdm a Woodbine positive expectancy UDM
colored green
M-NameOFUDM a Mountaineer positive expectancy UDM
colored green

**a secondary consideration is positive
expectancy UDMs with win rate >35% are
colored hot pink for either track

The color scheme is basically yellow=information,
green is a bet, hot pink is a WOW bet and light pink
is a no bet.

I also start UDMs with an "_" if I want it to be at
the top of the list - used for multiple track UDMs.
The "_" UDMs are still color coded as to green,
light pink or hot pink.

You can even break it down further by using G for
grass,S for Sprints, R for Routes and whatever other
letters you want and even numbers for distances such
as "6-6.5f-Bris3" where the UDM is built around 6
and 6.5 furlong races where the top consideration is
BRIS3. Then you could also do a
"T-R-G-Steve_Amussen" for a Steve Asmussen trained horse running a Turf Route and color it as to
whether it is positive or negative or informational.
Add in Class Desrciptor also and you get
"G-MSW-S-NameofUDM" for a Maiden Special Weight
sprint race on the grass.

Your only limit is your imagination!

Not fancy, but it works for me.

ElPaso



Reply
busseb
6/27/2009
7:11:18 PM
One more thing that took me a year to figure out:
Is a UDM getting better or worse?

Over time, you will have roi's go up and roi's go
down and for some UDMs, they will never perform like
you had then set up and eventually, you just delete
them and start over. Of the UDMs I create, about 1/2
of them get deleted and the rest are undergoing an
ongoing improvement. I constantly check UDMs during
the racing day to see how certain factors interact.

So, how do you know if the UDM is performing? Put
the win%, roi and ending balance in the betting
window. "w .33 roi 1.55 eb $166"

For instance, if I have a trainer UDM that shows a
for good win % for both grass and dirt, if the horse
running today is a CmpdLte rank 1, it's a natural
for turf, but how does that work on dirt horses.
When I check to see how the UDM works for CmpdLte on
the dirt, I also confirm my overall stats -win %,
roi and eb. If my 3 percentages are going down, then
I do some data windows inquiries and modify the UDM
do bring it back up - I refer to this as tightening
the UDM.

Next time you run the UDM through the data window,
compare the 3 numbers and you will know if the UDM
is performing to expectations. If the win % is going
down, tighten it up. You could even change the color
to blue to show the UDM is under test because it may
not hold a positive roi. As in my previous post, I
use hot pink for UDMs that show over a 35% win rate
and that are solid and tested over time. If the
win % is holding or going up, put a note after the
eb that you know the UDM is solid or just use a
new/diferent color and bet it with confidence .

ElPaso

Reply
JackB
6/28/2009
10:11:59 PM
No need to delete those UDM's you put all that time. Just turn them off.

Reply
Reply

Copyright © 2018 JCapper Software              back to the JCapper Message Board              www.JCapper.com