PDA

View Full Version : Feature Request for Future Robot Update


stu
5th October 2009, 04:45.02 PM
Currently in the Robot2 race filters, there is a way to seperate 'fast' dirt from 'wet' dirt.

After thinking about Saturday's turf races at Belmont, it would be nice to seperate out 'firm/good' turf from 'soft/yielding/dead' turf course. I would suspect that they produce two different pace profiles.

Stu

Gollygeewiz
5th October 2009, 07:49.00 PM
I think what you are mentioning is available under Race Filters -
Fast Dirt
Turf
Wet Dirt

but there is no such b/out for turf... might be useful !

km
5th October 2009, 08:24.10 PM
thanks Stu, i'm considering some NO Filters on the Negatives screen

NO Wet Turf
No Races taken off the turf

the data is not sufficient with the variety of turf course situations to merit separation by condition in the main section, and more importantly separation by rail position in the Robot to add a number of new filters just for unusual grass conditions, sample sizes would be very small for such an elaborate filter addition

You can use the HX4 to get it all though

Paladin
6th October 2009, 07:01.45 AM
Some more thoughts:
For the "No" Filters, how about No Distance Switch and No Surface Switch.
And for the "Range" Filters, how about [ Beaten Lengths = X ]

motorhead
11th October 2009, 12:44.45 AM
hello all

three ideas for future update

1 - longest losing streak (could possibly give an idicator of what to expect long term)

2 - adding QP% to the robot. add it to the range filters and to learn more.

3 - number of first time starters added to range screen and learn more.

km
14th October 2009, 01:28.46 PM
Thanks Paladin and Motorhead for the suggestions. I'll test them and add to Robot2 if they show promise. Check another thread in the Systems and Angles forum on the Surface switch.

BillN
14th October 2009, 03:32.29 PM
Hi Ken, I'd like to see check boxes for simple drops and rises in claiming price. Some trainers do better in one direction and a horse rising in class, although they win less frequently than the droppers they generally pay better. I'd love to be able to parse those horses out. :D

Rise in class would be great too. :D

Bill

Mike Brown
21st October 2009, 09:54.51 AM
Ken, ever given any thought to adding the capability of using Morning line odds and K line odds at the same time to create a Spot Play.

Thx
Mike

km
22nd October 2009, 12:47.57 AM
Thanks for the suggestions - I print them and put them into a future projects folder. I'm making a small update to the Robot for the Nov newsletter (after Br Cup), then we'll go forward from there for new stuff in 2010.

Huguenot
24th October 2009, 02:51.26 PM
I've found "AP Top" to be a negative in many instances.
Any thought to adding this as a negative?

km
24th October 2009, 07:58.42 PM
Andy and everyone, I have requests for virtually every filter to be reversed as a Negative. I'm unable to do this because the Robot would bloat beyond reason with hundreds of additional filters such as "NO A/P Top" or "NO Bid But Hung" or even "No Razor Sharp" ,etc., etc.

A/P Top is a small sub group for most spot plays and the horses that lose (bounce) off it are most likely those that Won or finished ITM, so you could try eliminating those to increase the ROI. Low odds are also common with horses that just had a "Top" - try testing with higher odds ranges and look at Learn More with the various odds spreads to find the 'sweet spot' with the play.

My experience with bad results on a single small filter sample is that they overlap in a larger sub group that is causing most of the ROI depletion. Low Odds is usually the likely culprit. But BAD CLASS or poor (K) below K=5 is also a common demoninator.

Be a detective, try running a separate test on that bad boy (A/P Top) and see where it is getting killed or has poor I.V. and ROI well below its normal Play All results.