JCapper Message Board
Email questions - Player framework given current state of the game
|| Email questions - Player framework given current state of the game
|From the email inbox...|
"Q. Do you find your process has become more and more complex over the years or do you find you have simplified things as you have learned more? By that I mean is it more important to continually improve your UPR to increase win percentage a quarter of a point or to rely on your number and patiently wait for the 5/2 horse going off at 4/1 or the 6/1 horse going off at 10/1.--end quote
In a recent post you mentioned that you look mostly at your SQL UDM Plays Report during live play. What this tells me is that you are not overthinking things to beat the individual race as much as trusting the math of your UDM Plays and making the play/pass decision based on the odds. Am I reading that right?
Do you have a difficult time going against the horse that is your UPR Rank 1, to say play your UPR Rank 3 or 4 when there are favorable odds vs. your odds line and you know the lower ranked UPR horse is being under bet?"
Let's start with:
"Do you find your process has become more and more complex over the years?"
Parts of my process have become more and more complex. That's mostly the result of me being me. Other parts of my game have become simpler.
I'll dive into areas of my game that have become more and more complex first. Then later, I'll come back to the parts that have become simpler.
I'm always striving to become a better player. One of the many areas where I do that is I'm always making incremental improvements to my UPR and UserFactors - sometimes in creative ways.
I need to emphasize that (just like UDMs) I've learned (the hard way) how important it is to validate (even small) changes to UPR and UserFactors going forward in time using fresh races before working said changes into live play.
It's one thing to either work out the math suggesting an improvement to a UserFactor and/or to "improve" a UserFactor using a data sample made up of the same recent past races other players or syndicates are data-mining in hopes of improving their models.
In my mind "improvements" to UPR, UserFactors, or UDMs don't really exist until after you've analyzed performance going forward in time over a large enough sample of fresh races.
In the UDM Wizard whenever I create a new UDM I'll give that UDM an inactive status and a "tag date" that represents the day after the last race date included in the development sample. From there I'll let the UDM sit for the next 45-90 days or so (depending) before coming back to it.
When I do come back to it I'll test from the "tag date" forward. I'll only activate the UDM and use it for live play if performance (win rate and roi) from the "tag date" forward is: a. an improvement over what I'm already doing and b. in line with results from the development sample. FYI, I handle changes to UPR and UserFactors the same way.
Let's move on to:
"Is it more important to continually improve your UPR to increase win percentage a quarter of a point or to rely on your number and patiently wait for the 5/2 horse going off at 4/1 or the 6/1 horse going off at 10/1?"
In my opinion THIS is one of if not the most important area of the game. To my way of thinking, the ability to sit on your hands and wait for a true Plus EV (positive expected value) situation - and pull the trigger only when a Plus EV situation is staring you in the face - THAT is the crux of the game.
If you presently lack the ability to do that - or think your game could stand some improvement in that area - my recommendation would be to work on that first before a. risking real money at the windows and/or b. spending real time developing UPR or UserFactors.
More to come...
~Edited by: jeff on: 9/10/2014 at: 4:09:22 PM~
"In a recent post you mentioned that you look mostly at your SQL UDM Plays Report during live play. What this tells me is that you are not overthinking things to beat the individual race as much as trusting the math of your UDM Plays and making the play/pass decision based on the odds. Am I reading that right?"
Absolutely correct - and this is one area of my game that has gotten (exponentially) simpler over the years.
The reason 98% of my attention goes to the SQL UDM Plays Report on race day?
The primary piece of information displayed on the SQL UDM Plays Report?
Horses flagged by business UDMs. Or put another way, a list of the most compelling horses for the day in military post time order.
Once I have scratches and changes for a race with the SQL UDM Plays Report generated:
The handicapping part is done.
This frees me up for the more important stuff:
• Mathematical estimation of edge (both postitive and negative.)
• Play or pass decision making - including which pool(s) to use.
• Bet sizing (based in part) on mathematical estimation of edge.
• Ticket structure and wager submission.
• Trip notes and record keeping/wager history.
All of the above are important to me.
Even if horses in my trip notes show up next time at some track I never play - the act of making trip notes while watching today's races serves a purpose. It keeps me immersed in the game. If there's a bias (inside path/outside path... early/late... what have you) I'll spot it.
And while I very rarely use that info to come up with new plays for today's card - more often than not track profile or bias info gleaned while a card is underway proves to be of help with plays already identified that I am likely to make later on that same card.
For example, knowing that a horse on the SQL UDM Plays Report in R8 fits (or doesn't fit) today's track profile enables me to make a more accurate estimation of edge. In turn that results in better play or pass decision making and bet sizing.
Obviously, I can't do any of that if I have to spend the entire race day bogged down with "handicapping."
More to come...
~Edited by: jeff on: 9/10/2014 at: 4:35:33 PM~
"Do you have a difficult time going against the horse that is your UPR Rank 1, to say play your UPR Rank 3 or 4 when there are favorable odds vs. your odds line and you know the lower ranked UPR horse is being under bet?"
One of the things I look for are horses I know are overbet.
Even better if an overbet horse scores/ranks well based on my UPR.
I say that because I've studied both the factor makeup of my UPR and results produced by that factor mix. I should have a clear uderstanding of conditions that produce both good results and bad results for my UPR horses.
Also, not all of the horses flagged on my SQL UDM Plays report are top UPR. (Actually, only about 60% of them are.)
If I'm keeping my head in the game on race day (instead of handicapping) I should also know what's happening or what the conditions are at the tracks I am playing.
From that, I should be able to recognize when a top UPR horse is overbet. I should especially be able to recognize the rare situation when a top UPR horse is overbet to the point of being ridiculous.
I'll work through the most glaring situational example I saw today...
Today at ALB in R8 my top UPR horse (#3 PLEA AGREEMENT) was 1/5 in the odds facing up to the gate. My estimated UPR strike price was somewhere around 2.6 to one. (So the horse was way underbet.)
The horse was also flagged by a couple of negative UDMs that I have... the first is based on LAMBERT CASEY T as an underperforming rider on early speed horses this year... the second negative UDM is based on a combo of excessive RaceStrength on MLine favorites that often leads to overbet horses. (A glance at the html report told me to insist on getting my strike price before pulling the trigger because of the negative UDMs.)
Looking at the PPs now (but not earlier today) I see the horse has speed figs that tower over the rest of the field.
One thing to keep in mind is that almost all claimers have their share of physical problems and are hardly win machines as a result.
This horse is a 6 year old and was eligible to be claimed out of this race for $15k.
Digging a little deeper, a look at the PPs shows he was coming off a pair of stakes wins in his two most recent starts.
Now, why in the world would you drop your stakes horse into a race where somebody could claim him from you for $15k?
In hindsight, this situation reeks of the words "suspicious class drop!" Oddly enough there was a horse claimed out of this race but it wasn't the overbet favorite.
As it turns out, another horse in this same race (#4 CALL ME HOSS) was flagged by a UDM I have based on FormConsensus. My estimated UPR strike price here was about 4.7 to one - which is about where the horse was facing up to the gate.
Going off on a tangent for a moment... Even if you don't have my UPR, this second horse was arguably the second best horse in the race based on QRating. Also, the horse's saddle cloth number was auto-highlighted at the bottom of the html report because of FormConsensus. (The point being made here is the horse was get-able as a logical alternative using the html report if you happened to think the favortite wais suspect.)
In this case (based on the SQL UDM Plays Report and the HTML Report without digging into the PPs) I had no qualms whatsoever about pulling the trigger on the second horse.
I'll admit I had no special confidence the horse would win.
What I did have confidence in though was this:
a. The favorite was overbet for reasons that make sense to me.
b. I was getting my strike price on the horse I was betting.
That's really about all you can ask for.
~Edited by: jeff on: 9/10/2014 at: 9:06:48 PM~
This is great. Thank you.
|Keep those e-mails coming !|
"Jeff, I've been meaning to ask you this for a while but haven't gotten around to it. There was a pre-loaded UPR factor that came with my Platinum program, but I don't have any idea what it's made up of. Can you let me know?--end quote
Love the program!"
My reply - short answer:
The pre loaded UPR entries are the same as the pre loaded entries for JPR.
Under the DEFAULT GroupName I used the ImpactValues Table Wizard to create a simple power rating based on the following factors:
Running Reports to see the entries:
If you launch the User System Definitions Screen you can see the name of the GroupName you have persisted that controls your UPR. (Write it down. Don't change it.)
After writing the GroupName down close down x-out of the User System Definitions Screen.
From there you can go into UPR Tools and run a report that shows you all of the entries for that or any GroupName:
1. Launch the Impact Values Table Wizard (or UPR Tools) Interface.
2. Click the Reports button.
3. On the Reports Screen, select the UPR GroupName from the drop down you want to run a report for - hint: use the same GroupName that you wrote down above when you were in the User Sys Defs Screen.
4. Check the Detail Report box.
5. Click the Run Report button.
6. Scroll through the report to see all of the entries for that GroupName.
Alternately, copy the report to your Windows ClipBoard and paste it into Word or app of your choice to make it easier to read.
~Edited by: jeff on: 9/20/2014 at: 12:35:35 PM~