*Handicapping Methods-Spot Play Testing-Software Etc

[OzYank]

Originally Posted by NorCalGreg

.

.

Nice hit - again.....I seem to be choosing the wrong tracks lately for this method...I will never give up trying though!!!!

[NorCalGreg]

Something I and others have learned--Best-Five program can also be excellent for pick-4's, at a reduced cost. It gives you 3 contenders per race--MANY times the top-2 will be enough to win.
2x2x2x2 p-4 ticket is $8. Go through your data files looking for pick-4 possibilities with price potential. If they are all mostly 4-1 and below...use the 2x2 method. If you win it, and it only pays $35...well, you only paid $8 bucks and got 4 X your wager back.

Anyway I like fooling with stats and numbers, and an idea hit me.

What if I ran Best-Five...got the 3 contenders per race--THEN added those to Trainer Moves and deleted (or scratched, if you will ) all the rest of the entries that didn't make the B-5 Contender's List.

So Trainer Moves should have 3 per race to choose a selection.

Sounds plausible in theory--but I'm not optimistic.

Here's the TR Moves CTX Class of the Speed selections using only the Best-Five contenders:



And here is the Top Platinum Power using only the B-5 contenders:
.


This is a live experiment @ Charlestown, which goes off in appx ten minutes.

-NCG

[NorCalGreg]

That little experiment sorta' blew up in my face. The more rules and requirements you add in--the closer to the favorite you get. I only saw one horse worth a wager- GOOD KARMA in the 7th @ 8-1 ML. He was on both lists-went off @ 9-1 - ran a good race, led but faded to grab the 3rd spot.

That was fun

[ReplayRandall]

Originally Posted by NorCalGreg

The more rules and requirements you add in--the closer to the favorite you get.

One of the biggest insights of wagering....Break the rules and be bold, DO what others are fearful to do.

That's how you hit HUGE scores...

[CheckMark]

Originally Posted by ReplayRandall

One of the biggest insights of wagering....Break the rules and be bold, DO what others are fearful to do.

That's how you hit HUGE scores...

Amen to that Randall!

[NorCalGreg]

Exotic Bets does come with a printable tip sheet. Here's Gulfstream today:


.
Good luck,
-NCG

[ric4nier]

Greg:

I just bought pony picker today...that completes the collection, I have them all! I really like the pace view software and how easy they are to use. This forum has been a good tool for how to use the software going thru the 450 pages.

I too learned handicapping from pops & tips john worth books late 80's...I have most Tom Worth's tips software, super cluster, I like the DQBF horses...I use it with pace view.....pace view much easier to use and about 1/4 the price. A lot of DQBF hoses are Platinum best bets horses...


Big Fan.... keep the software coming!
Rick

[NorCalGreg]

Very nice, Rick...thank you so much. I have maybe a dozen or more customers who buy everything I put out for sale--usually on day one. I'm working on my last program...and I hereby promise to send those loyal, wonderful customers that last program--completely free of charge. No idea when it will be completed, but I have some concrete ideas floating around, and a few solid items to include, that have already been tested out. Like everything in life...we'll have to wait and see what happens.

Was looking over Mountaineer and came up with this:

MOUNTAINEER LONGSHOTS WITH A CHANCE

4TH #5 LET'S TWIRL 12-1

5TH #5 BELLA BLESS 15-1

6TH #4 SISTER ANNA 8-1

7TH #1 CHARLIE'S BELLE 6-1

8TH #3 OSMI SLEW 10-1

[NorCalGreg]

Was going through some old files and programs...and found (at least the start of one) a program tentatively titled "Consistent Horse". I played the method on paper, until I was satisfied we had something.

I had grandeurs of the Consistent Horse Method being the basis for a new effective software program. Okay, so I had Steve make up a simple interface...and begin with the "Consistent Horse".

Suddenly, a small problem was apparent: after we got it programmed...we found the method was horrible . It was anything but "Consistent".

Just for giggles here is the CONSISTENT HORSE' selections for LAD this afternoon:





Don't even ask what these ratings include, or how they work--if I have them written down--no clue where they are.

[sharkey11]

not be nosy NCG but how many test samples are you using for your findings sir . i dont use many myself sometimes mabey 40 -50 they say it should be at least 500 or more . hey you rember a guy named jim bayle in vegas here one i got in one of his tests use th best of last 3 races drf speed rating and variant play the highest in the race if 4-1 or higher may the horse be with you

[headhawg]

The main problem with paper & pencil systems is that the author has to make it easy enough to use. The result is that there usually aren't many steps to the system or method. Now add in the fact that the steps are based on obvious factors like speed ratings and morning line odds, and chances are you get led to the favorite as the selection. I recently had a consistency-type system tested by my guy (Doesn't everybody have a guy?), and here are the results: 34,907 plays; 6615 wins (that pencils out to about 19%), with an ROI of .792. You will do better playing the chalk than to use this particular system.

Is all lost? Not necessarily. To paraphrase something that Dave Schwartz posted recently in the book section, there still might be an idea to work with regarding these systems. Maybe it's adding a step, or looking at speed/pace ratings in a different way, or limiting plays to certain tracks or purse levels. The only way to know is by research. Time consuming and boring yes, but a necessary evil in order to find something that works.

[ReplayRandall]

Originally Posted by headhawg

34,907 plays; 6615 wins (that pencils out to about 19%), with an ROI of .792.

You've got a huge data set to work with...

Break down everything into subsets:

1. Track

2. Surface

3. Distance

4. Male or Female

5. Class
A. Straight maidens
B. Maiden Claimers
D. Cheap Claimers
E. Mid-level Claimers
F. High-level Claimers
G. Starter Alw.
H. All other Classes of Alw.
I. Handicap and Stakes
J. Graded Stakes
1. 3 yr old and younger
2. 4 yr old and older

The rest is up to you as far as odds groups are concerned...Example
1. 4-1 and lower
2. 9-2 to 9-1
3. 10-1 to 19-1
4. 20-1 and higher

This is just a rough draft of subsets....

[headhawg]

Thanks, RR. Trust me -- I know how to create subsets. And I can assure you that the system that I listed the stats for cannot be saved. My previous post was more of a follow-up commentary to NCG's and sharkey's posts as opposed to a cry for help.

For fun -- and maybe as a teaching point -- I had my guy run a quick test of sharkey's "system". The only difference was that the data provider's speed ratings were used instead of SR+TV. (The only reason to use SR+TV is possibly to get a higher ROI because hardly anyone uses those numbers. But when testing a large dataset any perceived advantage will be washed out anyway.) Also, wet dirt races were eliminated in this test. What were the results? In about 15,000 plays there were 16% winners with a .78 ROI. Ugh. Still want to use that system??

Surely we can do better, right? If we eliminate turf races then we have about 11,000 plays, with ~17% winners and an .80 ROI. Better, but not better than playing the chalk. Sprints only? Fewer plays (~7000), but no change in W% or ROI. Is there anything positive about sharkey's system? How about MSW (no baby races) -- 450+ plays with a W% of 20% and an 1.00 ROI. But do you trust those results? Pretty shaky with only 450 plays. Slicing and dicing gets you better results, but obviously the more filtering of the data, the smaller the sample size. This research stuff is tricky.

[ReplayRandall]

I would never question your experience Crusty, but things aren't always what they appear to be with deep data dives.....I'm a Contrarian handicapper, a convert long ago of Mark Cramer's, and also I'm dedicated to Replays.

It's surprising how much statistics LIE.....I'm not joking, it's always the data in question.

[headhawg]

I agree that statistics can lie. But so do the eyes.