Mortal Locks

[km]

I ran a test of horses with odds <= 0.5; those that would pay $3.00 or less to Win.

This would include the following on the tote: 1/9, 1/5, 2/5, 1/2 (max)

Purse > $10,000; there were about 1600 plays in the last 365-days

62% of them won, ROI = 0.87 (not bad), 88% finished ITM (1-2-3)

Without question the HF was the key positive and had impressive results.
Plays = 468
Win = 74%
ROI = 1.02
ITM = 95% (wow)

Flat bet profits with the most obvious of all favorites.

But beating these horses can produce major scores for those that bet against them. At these odds they command 80% or more of the win pool, and certainly that much or more in the exotics. Everybody is using them as a singleton in the pickX. If they lose, the booing will be big and so will the payoffs.

What factors beats them? Easy answer = anything below K=1.

K= 2-7 with odds <= 0.5
Plays = 277
Win = 42%
ROI = 0.43
ITM = 78%

An amazing loss rate for such obvious chalk, 58% of them lose if NOT K=1.

[njcurveball]

Ken,

Thanks for posting this. I think the key for software users to increase their advantage in todays overpopulated by software game is to view a race in terms of favorite in addition to the usual factors (track, class, distance, surface, etc.)

Perhaps a future addition to one of the exports could include Favorite ML odds? Or perhaps a flag for whether the race contains an XF or HF favorite? I use the flag for the horse currently and then go back and add it to the race in my database.

I dont have the large amount of races that most people have in their database, but you certainly can improve an ROI and win percentage if you leave out races with heavy favorites.

You have taken it one step further and given us valuable information on when to try to go against these horses. One more thing I have to look at now, so sleep is canceled for the next 2 weeks. :-)

thanks again,
Jim

[km]

Thanks Jim.

The robot has a Race Filters feature that removes HF and XF races. Lots of people use that one at the tournaments with GET MY PLAYS to skip those races. The Post-Time report also has them clearly marked.

Export should not be a problem to pull that information using these fields and determine if the race contains a horse that you don't want to bet against.

HX4
Field 146 nXFAV
(2 = HF, 1= XF)

Field 19 nMLO
Field 96 rMLO
(1 = MLO fav)

Maybe Rick, Donnie or Mike can help here. How would you filter out a race that has either an XF, HF, K110+ in it?

Query for the heavy MLO fav that is NOT ranked K=1 should be easy task.

rMLO=1
nMLO <= 2
rK <> 1

[Donnie]

Simple....build one query that identifies these races, then use that query in a query with the races you wish to bet. When you put a query into a equi-join, the only horses (races) that then come back are horses (races) where both sides of the join are equal. Your first query would identify races that held horses as XF or HF but NOT rK=1....then start a new query and join HX4 to the first query...in the show tables box you have 3 tabs...one for tables, one for saved queries, one for both. Draw from your saved queries that first one that holds only races where there is an XF that is also non-rK=1. Keep the join an equi-join.

[njcurveball]

Donnie / Ken,

Thanks for the response. Since I am only exporting the data into an Oracle database, I will have to keep populating it after the fact. This is to allow me to do queries on hundreds of races, not just the current day. That was why I was hoping for a field like the NCHAOS one, that is a race factor and not a horse factor.

It is easy to do this, just time consuming.

Thanks again for your help.
Jim

[km]

Jim, send me an email and i'll substitute something in HX4 with the items you need so you can speed up your Oracle query.

[hurrikane]

JIm
I use mysql,
i use a sum by race on xfav getting the races equal to 0. This gives you all the races that have no xfav and exludes the ones that do.
if you want to restrict it to only xfav that is k1. use rk =1 sum of xfav = 0

Interesting note on Xfav. Last race at the LRL tourney, cole norman horse won at 14-1 against an xFAV last race of the day.
If you looked at the xFAV horse in fig 2 there is no way you could make this horse a stand out and I was pretty stong on him not winning.
If you are playing a tourney and your last races are xFAVs you have to look at them anyway. MHO.

Not sure now you can pull out the weak xFAV rk=1 except to test a lot of factors that might overcome them. Testing by track may help. nK < 110? Maybe Ken could test that one

[njcurveball]

Ken/Hurricane,

Thanks for the info and replies. I think most people here concentrate on getting plays for the day. I am working on trends and patterns for the entire database. And with Oracle I could sit down and write complex queries.

I guess after doing it all day, it gets a little cumbersome to do that amount of "grunt work" at night as well. So I usually "take it easy" at home and flatten the tables, rather than go nuts.

I am almost there with Ken's 1000 race challenge. I can segment the data and get there, but in a reasonable query I still have 4 or 5 winners in the 1,000 to remove. I doubt I ever will since the game is flesh and blood and nothing is 100%.

I also have worked on a high percentage query that gets 96% in the money for more than 50 races. Funny thing is that it shows a higher profit for win and place. With the place win% around 88%.

These are just "cerebral" things as most of my handicapping is done on a race by race basis, checking past performances, velocity numbers, energy percentages, trainer/jockey stats, post position/track bias, etc.

All of us with a computer usually have the same dream. Produce a report that has all winners on it. :)

The great thing about this game are the tournaments where we definitely have a BIG advantage over the masses.

thanks,
Jim