@Rich Val

Just so you guys know, a few years ago I was struggling to win. Rich valentine came out of the woodwork and offered to coach me. We did 7 sessions over a couple of weeks that put me back on track.

Remember that, Rich?

You guys gotta know that winning players never share their secrets.....

I just gave you the most important piece of the game....

Head it and go forward — HorseSense

My experience is that winning players - and especially professional players - absolutely do share what they do.

After all, if WINNING PLAYERS really are that smart (and we are), how would anyone know if information wasn't shared?

Ping @Tony Kofalt @Rich Val

In my opinion, I would want at least 80% potential winners in any contender methodology and I am skeptical that Dave’s approach would accomplish this goal; however, using a “gap” approach has given me several ideas to pursue…. — Jack Price

I would contend (small pun coming) that the issue is not the win% on the top contenders.
The issue is figuring out which low odds horses to remove BEFORE selecting those contenders.

As you can imagine, the low odds horses will ALWAYS be at the top of the PSR list. If they aren't, then toss them as well.

Of course the low odds horses are the favorites I'm mentioning......
Does HSH have any proprietary factors to combat this? — HorseSense

Not so much proprietary , but of course.
We build OBJECTS - which are from 1 to 6 factors weighted together.

That was from Fraser Rawlinson.

Actually, you've described the basis of my handicapping - although for me it all begins with the low odds horses.

With this complicated software , your resorting to PSR with Early and Late? — HorseSense

The game is more difficult than ever and I have no interest in playing without long-term profitability.

I'm always ready to be taught something - simple or complicated - that allows me to continue to win.

Do you have anything to offer?

My point was that maybe the horse has a great f1, but later in the race stops. Would you consider the great f1 or would you toss the race entirely? — RanchWest

I don't consider "races." All of the ratings stand on their own.

I'll find some time to post my actual paceline SELECTION approach in the next day or two. It will necessitate showing how the system works - with some screenshots.

But the short version is that I want the TIGHTEST FINISH(es).

My favorite paceline would be a race that was won or lost by a head/neck, same surface, same (or similar) distance.

The rationale is that this race represents exactly how good this horse is. If he was any better, he'd have won by more; if he was worse, he wouldn't have been so close.

I don't know anything about Pickmaster, but, as Dave said, it takes a lot of factors to have consistent eliminations or inclusions. And, sometimes I use composite data. — RanchWest

Actually, ultimately, it boils down to a few meaningful objects with the right application.

Still the game isn't easy.

You have no idea. LOL

Have you tried searching for these scenarios through Pickmaster? If so, do you believe it would be a worthwhile endeavour? — Tony Kofalt

Yes, that is one of the tools.
We also have one called the OBJECT MASTER that allows you to turn on as many as 300 factors and optimize and object using a pretty cool approach.

Imagine 300^4 = 8.1 billion permutations - searching for the best object for (say) $net in the #1 rank. You can run something like that against 20,000 races in about 3 hours.

No, the closest thing I have done is similar to the RDSS CSR rating. CSR uses a Fibonacci allocation for the speed ratings of the last 4 races. It's a somewhat helpful number. So, I did the same thing with the BRIS pace figs. It gives me another way of looking at pace. As I have mentioned before, I am always looking for whether a number is real or phoney and this gives me another point of comparison. — RanchWest

We have a set of TIME DECAY factors for that.

Heck, we've got 4,000 factors per horse, so we sort of have everything. LOL

Let's say a horse stops and is vanned off, for example. Do I assume correctly that you would not use any computations from that race? It seems to me that if, for any reason, a number like total Energy cannot be computed, i would not use any computation. — RanchWest

Since the horse got horrible ratings, they'd not be used for any of the top-whatever. So, no. They'd not be used.

Perhaps I am missing your point, though.

Interesting to see what Dave says about composite pace. — Tom

Composite Pace is a slightly different animal.
Take all of those 8 pace ratings mentioned above...
F1,EP,SC,FT,FW,PWR,SP,LP

... and look at each one through the lens of the 13 ROWS.

Thus, F! has...

• Last Race
• Best of Last 2
• Best of Last 3
• Best of Last 4
• Best 2 of Last 3
• Best 2 of Last 4
• Best 3 of Last 4
• Average of Last 2
• Average of Last 3
• Average of Last 4
• Best Ever (last 10 races)
• Best 2 Ever (last 10 races)
• 2nd Best Ever (last 10 races)

Then we take those 13 and average them to make Composite F1.

When calculating factors for each paceline(s) such as E1, E2 etc has anyone ever experimented with a metho Dave calls synthetic pacelines? Dave maybe you could add a brief explanation of the synthetic paceline calculations? That way I won't lead anyone astray. I have to say that synthetic calculations have led me to many lone speed opportunities especially in maiden race. — Tony Kofalt

Sure!

What Pace Handicappers Do
Let's take the typical horse with 10 pacelines.
And let's imagine that we're using the basic Sartin approach of EP, SP, W.

So, if one considers the 10 pacelines, with each having 3 ratings, we have the potential for 30 ratings. Or, if you prefer, 10 rows with 3 ratings each.

What Synthetic Pace IS
So, we would have the same 3 columns - EP, SP, W.
But we have 13 rows and those rows are labeled as follows:
13 Rows

• Last Race
• Best of Last 2
• Best of Last 3
• Best of Last 4
• Best 2 of Last 3
• Best 2 of Last 4
• Best 3 of Last 4
• Average of Last 2
• Average of Last 3
• Average of Last 4
• Best Ever (last 10 races)
• Best 2 Ever (last 10 races)
• 2nd Best Ever (last 10 races)

Let's consider "Best 2 of last 3."
If we used the Sartin approach (i.e. EP, SP, W), we'd have 3 of those:

Best 2 of Last 3 EP
Best 2 of Last 3 SP
Best 2 of Last 3 W

So, 13 rows x 3 columns = 39 ratings.

Our modeling tool goes through the races that are LIKE THIS ONE and picks out the best of those 39 factors.

Thus, doing it this way, we might get:

• Best 2 of Last 3 EP
• Best 3 of Last 4 W
• Best Ever SP

In other words, we didn't pick a single ROW (or paceline). Instead, we created a SYNTHETIC PACELINE.

In Reality
In reality, we actually have 13 rows x 8 columns = 104 ratings.
F1,EP,SC,FT,FW,PWR,SP,LP

In addition, we can mix a bunch of other columns in with it, such as PSR, Jockey Ratings, etc. just to see if they have an real value.

Hope this helps.

@Tom & @RanchWest,

Great answers. Love how both of you have worked to "systematize" your approaches.
I'll have to grab some screenshots of what I do - will work on it tonight.

Well, let's get somebody else to start the thread and then we can tell them what we do.

Of course, I don't CURRENTLY use it because I don't "select pacelines," but my system is built into HSH where anyone can see it.

(HSH has a programmable paceline selection mechanism. Well, it has a couple of dozen, actually, and the users can create their own.)

And, I have my program select my pace lines. — RanchWest

"HOW TO SELECT PACE L;INES" would make for a great topic.
Would like to see someone else post it.

@Colty - How about you? Do you have any opinions?

I do have a spot play that has hit 9 of 45 (20%) and returning only 39 cents on the dollar. — RanchWest

Ironically, THAT is what I generally look for in a play-against horse: the overwhelmingly obvious.

Losing 17% is just not strong enough.

The bidding starts at -35%.
But it won't be a single factor that produces that. It would be a combination of factors.