Rdss2 & NewPace

[Ted Craven]

Richard,

If I understand the VDC rank recording protocol correctly from the email you sent me, you would have recorded the winner above as VDC rank 2 even though there were 2 VDC rank 1's above it. And if the winner had been horse #1, you would have recorded the winner's VDC as rank #1, despite there having been 2 rank 1's.

What this means, if I understand correctly, is that out of the 200 races in your study, there were (likely) more than 200 horses with a VDC rank1, so the hit rate of VDC rank 1 was actually 83 out of 200+ rank 1's to choose from, depending on how many rank 1 ties there were. And likely the same for rank 2 and rank 3 at least . Did I understand you correctly?

It has been a long-challenged, oft-asked question: why are there VDC ties. The answer as always: to elevate deserving horses who otherwise rank poorer on the basis of weighted line score (i.e. BL/BL) alone.

The above discussions on treatment for betting purposes versus study purposes, I think, answer how to actually make use of VDC.

If I misunderstood your method, please do let me know.

cheers,

Ted

[rmath]

Ted, Yes you are correct that I do not distinguish the ties for the purpose of record keeping. I have never been able to find a way to separate them with any degree of accuracy. My figures are correct for the number of wins per ranking.
I have found that if I have a tie for the #1 ranking ,then those two are my main win contenders that I use to decide if I am going to bet the race or not.
If I have only one top and one second then I look at them as if they were tied.
I have found that my win % for my top two horses stays pretty much the same regardless.
If I have 3 or more horses tied in the top 2 rankings I use post time odds to find my 1 or 2 bets OR I just pass the race.
Like I said in my email , recording and betting are two totally and separate items.
I rarely bet more than two horses in a race and really prefer to bet only one horse to win and to place if the odds are high enough.
I have found that if I restrict my biggest bets to a top rated horse at 3/1 or higher, I will cash in 28% to 34% of my actual plays (one horse) to win and between 50and 55% place.
I hope this helps to clarify the difference between record keeping and my actual betting. About 90% of the races that I actually bet only have 1 top horse. This is why I rarely bet more than 2 or 3 races per day at any one track.
rmath

[Bill Lyster]

Originally Posted by Ted Craven

Thanks again Richard!

One thing to remember regarding V/DC: there are frequently tied ranks, e.g. 2 horses ranked 1, or 2 horses ranked 2 or 3, etc. Up to 50% of the time? This makes direct comparisons to factors which always have discreet ranks (such as Total Energy, or NewPace) a little more fuzzy.

So, it might make this V/DC related study a bit clearer to know how you dealt with those ties when counting the number of #1s and#2s etc.

For example, how would V/DC in this race be ranked in your tabulation process?

Attachment 23563

Bye the bye, this race is a perfect example illustrating the distinction Doc tried to make between ranks and tiers. The #1 and #6 are in the same exact tier, both for BL/BL and for V/DC (the display order is goverened by the Total Energy tie-breaker) and should be considered equal on our odds line. The idea is to break these ties (if you must) according to bet-time odds. The 3rd listed #5 horse is effectively equal to the top 2.

Eliminate a low odds horse (from Win betting) below a certain threshold if it does not justify its public odds, compared to your betting line. The #1 at 3/2 (1.4 - 1) is no better (though also no worse) than the other horses in our top 3. For Win, we are offered better odds on the net of the #6 and #5, and the #6 should also be bet to Place as well. The #1 belongs in an Exacta box.

I believe these instructions (more or less) will also be found among Doc's words written in Follow Ups #70 and forward.

Attachment 23564

One of the nice things I have been finding here is that there is a nice elimination corollary if the fifth VDC horse is also the 5th or 6th DCL horse.

Also in turf sprints or routes where you sometimes have to use a poly or dirt line for some horses. If the TE of the dirt or poly horse is greater than 3rd and the DCL is 5th or 6th you can eliminate such horses from win and place a lot of the time.

Bill

[So Cal Al]

Hi, Ted, Richard, and Bill:
I've been following this thread with interest. Great stuff, Richard, and I agree with Ted that, because of ties, there will be a total number of V/DC contenders that is larger than the the total of 4-horse contenders in NP. Nevertheless, the test information is very viable. Ever since my PIRCO days, I have always believed that the key was a proper pace line selection process. As we know, in some races it is easier said than done because of layoffs, surface changes, etc., but we can learn from experience (and by sharing that experience with others). Dave S. has indicated that the Sartin Methodology is interested in current form, and NP is not so constrained. In fact, I get a sense of the Pizzolla "way back" approach to turf races, when reviewing some NP race contenders. For myself, I'm still evaluating NP and V/DC from a standpoint of how each may possibly aid the other, and in the strengths and weaknesses of each. Good work, all!

[Bill Lyster]

2011 Delmar ONLY, 163 races through August 28th WITH filters:

Hid all route horses away 75 days; hid all sprint horses away 95 days
Hid all horses with ML odds of 20:1 or greater.
Hid all horses without a race or workout in 40 days - but there weren't any!!
No more than 2 FTS in maiden races; no more than 2 foreign horses making first start in States.

[Interesting info from Delmar website handicapping data. From 2007 to 2010 93% of route winners returned in 77 days or less; 92% of sprint winners returned in 88 days or less; 92.2% of turf route winners returned in 81 days or less; ONLY 83% of turf sprint winners returned in 145 days or less!]

Hid horses but only down to five contenders. Used top 5 TE unless horse was top three VDC rated; kept the top 3 VDC rated no matter their TE rank.

With remaining lowest ranked horses looked at both TE and DCL info.

In these 163 races there were 28 instances where the top 4 VDC had 5 horses, so about 1 out of every 6 races as many as 5 horses. I did not record rating ties so if three horses were #1 VDC I entered the data in my results table according to TE.


Results: Betting or boxing all top 4 VDC contenders

Wins: 136, avg odds 5.25:1, avg $2 mutual : $12.51; ROI 27.4%
Exactas: 96 , $1 Exacta average $33.53; ROI 60.6%
Trifectas: 59 , $1 Trifecta average $194.81; ROI 186.8%
Superfectas: 21 , $1 Superfecta average $511.90; ROI 173.1%


Good luck to all of you---

[rmath]

320 races run at 9 different tracks. No filters of any kind.
Pacelines were chosen using best of last 3 comparable for all entries adj sr.
Sorted down to final 5 using Primary Line scores.
#1 vdc 112
#2 vdc 74
#3vdc 55
#4 vdc 33
total wins 274 .856
Exactas in top 4 box 220 .688

New pace
E1 81
E2 48
L1 68
L2 45
total 242 wins .756
Exactas in all 4 161 .503

[Bill Lyster]

Nice work

And I note the use of the Adjusted Speed Ratings, NOT the raw speed ratings.



bill

[rmath]

Thanks Bill, and yes ever since the adj. sr ratings were made available I have followed DOCs guidelines. These guidelines are just as potent today as they were when Doc and Trackmaster first released them.

[rmath]

Ted , I thought you might be interested in how the vdc % has been doing since I started recording it also.
In the last 120 races of my testing the top 3 vdc won 94 or .783,
85 of the 94 were in the top 3 vdc%.
One thing that has caught my attention is that when I have 3 or 4 horses tied in the top 2 vdc , one of the top 2 vdc% horses will run 1st or 2nd over 75% of the time.
Not sure if this is any help in deciding on the usefulness of the vdc% or not.
Since I rarely play the exoctics anymore I am more interested in getting as many double digit winners in my top 2 or 3 as is possible.

[BJennet]

Originally Posted by rmath

320 races run at 9 different tracks. No filters of any kind.
Pacelines were chosen using best of last 3 comparable for all entries adj sr.
Sorted down to final 5 using Primary Line scores.
#1 vdc 112
#2 vdc 74
#3vdc 55
#4 vdc 33
total wins 274 .856
Exactas in top 4 box 220 .688

New pace
E1 81
E2 48
L1 68
L2 45
total 242 wins .756
Exactas in all 4 161 .503

Hi Rich,

Thanks again for continuing to provide these results on VD/C. I see that, in terms of hit rate these results are converging even more closely with my stats on TE, which convinces me more than ever that VD/C, Adj. SR, and TE, like many of the other Sartin factors are almost completely correlated, i.e. that VD/C has no statistical independence. I know that everyone's time is at a premium, but if you could check this correlation - how often is VD/C #1 also Adj. SR #1, and so on down the line, I would be grateful. If it turns out that they are not correlated, we have a new, tremendously powerful Sartin factor. However, I suspect that they are.

Thanks again.

Cheers,

B Jennet