Idea for Future Update

[km]

March 9 is the lastest version, fixes a few bugs from March 8 version.

Several users have reported that the "Super Trainer in Race" items are reversed, so i'll fix that soon and await for other bugs before another update and maybe add something new...

Getting ideas from users along the lines that they would like to see something like this =

Won by X Lengths
Lost by X Lengths

the only thing we have along these lines is the @ = lost by 20 lengths or more ("Big Loser" in R2)

These gaps in wins or losses lengths are the basis of many systems and angles. I think it would be interesting too. Examples are the "big win" a horse that wins by 3 or 5 or more lengths. And also setting various parameters on losers such as Fr1 plays that lost by 10 lengths as most of the public will toss these horses because they think they are quitters. We know that is nonsense and is the primary reason why Fr1 used to produce such great profit == big loser in a sprint, moves to a route and wires the field at long odds.

Let me know what you think -

[Huguenot]

Ken,
I think you've answered this before, but I can't remember the answer.

What is the issue with separating out Turf races on softer ground from the "firm" races. If I'm modeling the turf races at Belmont Park, for example, surely the good and soft tracks will run differently than the firm ones.

Also turf related, separating out the inner turf tracks at Belmont and Saratoga, which also have their own idiosyncracies as you know.

I presume there's some software issue here.

[km]

No separation in the Robot Andy, data sampling would be too small and fragmented.

As you know, the "rails out" is a far more common issue with turf racing than wet turf. In fact, most mildly wet turf races ("good") are run with the rails out. Soft or Yielding grass races are taken off the turf unless it is a stakes race in most cases.

Turf courses become entirely different surfaces (and timing setups) when rails are placed at variable distances. GP is the perfect example, with rails out as far as 72 feet - those horses are running on whole different footing than if the rail = 0.

So unless we want to dig into every nuance of turf settings and labels, i don't see the test results as having anything but marginal value because the data would be split up too much.

Our db users might be able to provide you some insight as the com charts have all the rail and turf course conditions in them. But where do you start? Separate by "firm", "good", "soft" and "yeilding" and then separate again by rail placement? That would be the only accurate way to do it, but sample sizes would be counted on one hand in most cases and be unreliable.

I printed a GP Turf Route sample in the newsletter that had a universal outcome for all grass races there. I would rather know that K=8,9 are 0 for 50 on the lead on the turf (regardless of condition) than fragmented sets of data that would show 0 for 7, 0 for 3, 0 for 6, etc.

[BEEGEE]

Ken,
Can you put the spot play number in the heading at the top of the summaries in the original robot?

Is there a reason that the VI numbers are listed by fixed ranges rather than letting us select the values as we do in K, EPR, etc? Perhaps you could consider that capability for Robot 2.

Thanks for the great new additions to the software.
beegee

[chuck]

Along the same lines as what Beegee posted, the learn more shows stats for FC ranges but FC is not included on the range filter screen.

[km]

Thanks guys, we'll get the Vi and FC range (user setting) into the final version of R2.

[Mike T]

Ken,

How about a "No bad Class" option in the negative screen?

[km]

Mike, are you referrring to the older Class (-) from the HTR2 Robot?

I would like to come up with something better, will work on it, thanks for input

[Mike T]

If you find something better, that would be great. If not, I would love the old one put in the Robot 2. I have found that it is a great throw away on many of my spot plays.

[jhilden]

Race 5 tonight @ CT reminded me of a questional play: A horse that never raced today's distance. A bit different than going from sprint to route or vice versa. I always red flag these horses and would like to test it for verification.

[Paladin]

If you find something better, that would be great. If not, I would love the old one put in the Robot 2. I have found that it is a great throw away on many of my spot plays.

I wish I could say the same. I've tried to use it as a "throw away", but Class (-) usually comes up about the same as the other class's or sometimes even better, depending on what other positives the horses have going for them. Also, if the horse is of the lower odds type, it seems that many times it has other things going for it that overcome a (-)Class. And if it is of a higher odds type, then the Class(-) may be a reason it is overlooked and pays big (brings up the ROI). Maybe there is a "sweet spot" to use it. Or maybe something else (as Ken says) could be injected into the robot class section. Wonder if there would be a way to include earning (with a minimum number of starts), added/mixed into an equation with pace/speed of recent races or something to add to the mix.
I'm just thinking off the top of my head with "half baked' ideas. Maybe Ken can completely bake them. He's our chef and has cooked up some tasty dishes into HTR.

[TEJASKIDD]

KM

How about adding a range filter for the KGAP differential.
A K7 in one race is different from a K7 in another. This way we could eliminate the k7's (or any krank) that have too big a differential.
I like to eliminate all KGAPS of -15 or more (except the $/$$ runners).

Also, I'd like the ability to filter races based on the amount of Qfives+
For Example;
Qfives>5
I like to handicap races loaded with speed types. this enables me to quickly identify which races to look at. I already do this in Access, but maybe there are others out there that don't work with access that would like to filter out these type of races.

[km]

Scott (paladin) thanks for that feedback, i agree with what you said based on experience, the Class (-) was a dud, reminds me of the filter for Front Wraps, just too many toss outs without enough benefit. Eliminations need to show consistent improvement in ROI, particularly with the K spread.

Dennis (tejas) thanks for the first suggestion, i have several requests along those lines, i'll work something in to separate the (K) by the gap top to bottom. Some have suggested NO K < 90 which is also a good one.

Your second idea is already in Robot II, check the Race Filters screen and look at the bottom right for the QP6 items (same as QP > 5) there are 4 different ones to use. They will help you test or isolate races with specific Quirin point setups.

[del_dog]

a way to test by post???

example , inside two posts with a k1 and q6 horse??

things along that line???

or in a perfect world a way to test large post switches from race to race....

far outside two posts to inside two posts???

something like that...

sounds more like a db type function , but for us lazy types???

:)

also a box to check to test just the 2 place finisher???
no need for a third that's where the ALL button would be used.

:)


by the way , HTR had an amazing run this last fri-sat-sun or perhaps i just got out of its way.

congrats.

[km]

thank you for the suggestions 'dog

boy i'm with you on the post position change, the inside to outside in SoCal sprints has been one of my best intuitive spot plays for years, but it really is case specific, with dist.surf.field size all being part of the equation. For example, a horse exiting the rail at SA from 5.5f is not an issue, but 6.0f -7.0f in a field of 10 or more almost always results in a shuffle-back for early speed types and they do much better if they draw outside next time.

But that would be tough to test in the Robot without a whole lot of information to develop from current race vs. previous race data in tandem.

[Paladin]

I agree with Ken, there are too many variables (running style, distance) affecting pp change. For example, many years ago one night I had burned the midnight oil over the DRF. Came up with one play for the next day. The horse was going from 6 to 7f, drew the rail, and had Gary Stevens up (again).
The horse was a mid-pack type of runner, but did best when a little closer to the pace.

I lived VERY close to the track (SA), got there early and asked Gary about this race coming up later.
The conversation went something like this:
" Gary this horse you've got coming up, drew the rail, and now you're way back in the 7f chute. Do you think that will hurt his chances by getting shuffled back too far?"
Gary - " No, in fact I think it makes it better."
Me: "How so?"
Gary - " Because the 7f chute is farther away from where you enter the main track. And the speed horses on the outside will wait until they get to the main track before they cross over towards the rail. That gives me more time to get position."

He won the race.

[DanG]

Random thought on Class;

I feel the real value in “classifying” racehorses oddly enough is when they lose. (At least in terms of returns) Many people classify an animal on their wins only; and the real value is an animal running over their head whose form is darkened because of it. The tight rope becomes an issue of form once competition level has been established. Classifying animals off their wins is “usually” just compounding the speed factor that is already in the deep end of the pool.

As you guys said; Ken will cook up something clever…he usually approaches these challenges from an angle most of us would never dream of.

I’ve got it…Day of Week!

True story; sitting in the temporary grandstands Breeders Cup day at Gulfstream years ago behind a few horsemen. One guy asked the other…”Do you have any Saturday horses?” That’s the backstretch code for a stakes horse that normally run on Saturdays.

New Robot filters for weekdays / weekends with a special accelerator for national holidays! :eek: :D

[tomcat]

Alright then, Test my grey horses on turf!

[del_dog]

i have always tested based on weekends being different from weekdays.

there is no doubt the types of races carded are very different.

??

[del_dog]

thank you for the suggestions 'dog

boy i'm with you on the post position change, the inside to outside in SoCal sprints has been one of my best intuitive spot plays for years, but it really is case specific, with dist.surf.field size all being part of the equation. For example, a horse exiting the rail at SA from 5.5f is not an issue, but 6.0f -7.0f in a field of 10 or more almost always results in a shuffle-back for early speed types and they do much better if they draw outside next time.

But that would be tough to test in the Robot without a whole lot of information to develop from current race vs. previous race data in tandem.

yes, i agree lots of noise is "in there".

I was looking at an overall quick way to test using post switch from race to race using the existing variables in the robot.

Was looking for a way to try to discover some tracks distances from a thousand foot view that may be worth digging into.

Thanks for the reply.

[Bob]

Ken,

Not sure how difficult to do but how about a Robot II option to test for individual finish positions.

BobG

[Hovard]

Ken, How about a query by specific age, not just the age requirement of the race. I would like to find 3 in 3u fields, and need an age of horse key. thanks.

[km]

Bob- i think there is sufficient Place and Show information on every report, i don't want to add options for 2nd and 3rd (only) for testing because i would have to redesign each of those reports to account for the individual finish position and re-compute the I.V., High, AvgMut etc., a tremendous amount of extra work and text space for marginal advantage.

If you were referrring to finish in last race, there is already a box for 1st and 2nd place finisher last out and now the Negative screen adds "Did NOT Win" and "Did NOT Finish ITM". I see no value in having individual finish for '5th' , '6th' , '7th' in last start etc. separately.

Hovard - Full age listing is in the LEARN MORE report now in R2. I'll consider a filter for it if there is space later.

thank you both for the input and suggestions

[Bob]

Ken,

Thanks for letting me know and for considering it.

I wanted to do some superfecta analysis. However, there are bigger fish to fry and, as for me, I always prefer a full plate of goodies! :D

Regards,
Bob G

P.S. I just now remembered that MaxVel has 4th place stats...

[Huguenot]

Well, as long as we're asking .....;)

HTR is a great program as it is of course, but I'm feeling a bit greedy.

On the XFIG screen, any of these doable, realistic or even wanted?

1. Replace purse with another more valuable measure. Purses have lost their worth long ago with the onset of state-bred races and racinos. I would favor either SOR or 1st fraction.


2. Have an option (an Easter Egg perhaps) providing more flexibility on synthetic track races. It would be helpful to have the ability to view synthetic races with just turf races, for example instead of having to opt for either JUST synth or ALL surfaces, including dirt. Similarly in handicapping turf races, having the ability to mix just synth and turf without including dirt.

[smallchange]

Ken,

Is it possible to change the Minimum Lifetime Starts to a range such as 0 to 999?

I have some ideas about non FTS but I want lightly raced horses.

Terry

[km]

Hug' this is R2 forum, i'll put up a suggestion box for HTR2 after the Ky Derby and reply to your suggestions then. You always have good ideas and most of them make it into the program - so thank you :cool:

smallchange' not a bad idea, i can change that one to a range between 00 - 99 life starts. Consider it done for the next update.

[Huguenot]

In my defense, we don't go on daylight savings time until Saturday so I've been kinda disconbobulated for about three weeks.

:o

[Bob]

Ken,

The first Robot prints the date range for the robot run...any way that Robot II can do the same without going to a lot of trouble?

Regards,
Bob G

[km]

Bob, will do - thx

[Bob]

Ken,

Thanks! My memory isn't what it used to be. :(

Bob G

[NC Tony]

What about adding analysis up to fourth place for us Superfecta players?

Just a thought....

NCT

PS That Avatar is Me doing a Frank Stronach imitation at one of his board meetings .........or better yet Kinda like the Beard Situation..............I See Nothing!!!

[km]

thanks for chiming in Tony, glad you are using the Robot

To query 4th place finishers separately, (or 2nd and 3rd place as others have requested) is the task of a database. The Robot exists on its simplicity - it would take quite a bit more overhead to reformat and lay code for individual finishes. Every addition slows it down.

The current Robot reports have columns for 2nd and 3rd finisher % and ROI which i feel is sufficient for its tasking on non-winners right now.

It's becoming apparent that many users want to expand the Robot to the level of a full featured db application with their requests. I appreciate the enthusiasm. But the Robot does not function like a db, it must open every racefile individually and then handicap the race with 100% of the HTR data items being computed. This is different from a db that simply matches the requested query items. The more overhead we place on the Robot, the slower it functions and the more complicated it becomes to beta test for accuracy.

[NC Tony]

Thanks for the reply Ken.

I actually understood the technical response. But yes the data base concept I'm thinking is what most users would spend hours and hours with trying to find that golden kernel.

The only way something like a db could work would be to have the data reside on a server which external users could have access to. I now own a HP Home Server for home use and centralize all of my files (HTR, Photos, Music, Data etc) onto that server. I can "dial up" the data from anywhere I have internet access. These servers are not all that expensive and can hold up to 4 Terabytes of data. Maybe you can develop what could be conidered a premium account whereby users would have to buy into the server service by granting read only access to the data which one can then query. Again just bouncing another idea. That way if certain reports or formats were developed and maintained other less sophisticated users can use the pre-canned querries.

That way those expert access guys don't have a leg up on those of us who don't understand VB or whatever code is used to write access programs.

NCT