50 Race Comparison: VDC vs NewPace

[rmath]

I just finished my first 50 races using RDSS2 with New Pace.

23144

There were 3 exactas in NP that were not in my top 4 vdc& 12 exactas in top 4 vdc that NP did not get.

So far , in my opinion RDSS2 is doing better by 10% on wins and 18% better in exactas.

For what it is worth the top 3 vdc won 26 exactas.

Ted: I ran all 50 races thru Rdss 1 as you asked and only that one I sent you had any differences. I really like the new layout in RDSS2.
Thanks for letting me be one of the testers.
rmath

[rmath]

When I wrote the posting I had separated the vdc results from the new pace results but when it was sent there were no spaces put in. Sorry

[Ted Craven]

Richard,

I cleaned up the layout. Thank you very much for reporting this. This confirms some of what my own experience is, and that of several folks who seem reluctant to post it online.

One of the most important attributes of NewPace, IMO, is that it is automated - you push a button and 4 contenders appear. With the rest of RDSS and the Methodology, you have to push a few more buttons and use your brain to get the contenders. Doing so, at minimum, does not give you worse results. Then, with either approach, you begin making betting decisions.

Are you able to elaborate on the relative profitability of the 2 approaches to getting contenders, using a common betting strategy?

Thanks again for this work!

Ted

P.S. I moved this thread out of the Testing Forum

[SilentRun]

These results , however small the sample, is showing that boxing the top 4 VDC in RDSS is better than New Pace.

And....does E1 and L1 outperform E2 and L2 in terms of winners ?

This is worth exploring further.

Ernie

[shoeless]

Rmath,


What were you using as a criteria for picking pacelines?It is really impressive that you had 36 winners out of 50 in your top two VDC.

[Ted Craven]

Just a reminder that the VDC ranking sometimes has ties, so Top 2 should probably read 'Top 2 and ties', which in less than 40% (??) of races may mean 3 horses (e.g. 2 horses tied rank 1, or 2 horses ties rank 2)

[shoeless]

Good point Ted

[rmath]

1) My criteria for picking pace lines is to use best of last three comparable.
2) I eliminate down to 5 using line scores, if I have a tie for #5 line score I the use total energy to get rid of the 6th horse.

New pace hit 50% exactas, a return of 460.95 less 600.00 bet for a net loss of 139.05
Top 4 vdc hit 34 or 68% for a return of 891.60 less 600.00 bet for a net profit of 291.60
Boxing the top 3 vdc hit 52% for 621.60 return less 300.00 bet for a net profit of 321.60.
These results look impressive at first glance but I rarely play any exotics and prefer to bet win if two or more of my top three are 4/1 or higher.
The reason for my saying that the results look great is that VDC had a 358.50 exacta and new pace did not. The highest new pace payoff was 120.75.
In my opinion both methods should be researched a lot more and if used The low payoffs below 20.00 should be avoided. There were too many 4.00 to 18.00 payoffs in both 4 horse groups.
I had 14 winners in the 50 races run that paid 10.00 and up. I passed a lot of these 50
because the odds on two or more were too low, alot of 4/5 and 6/5 top rated horses.
My play is concentrated on making a specific dollar amount each day and then quitting for the day. This has been my approach for the last 20 years. I came to this way of playing after losing to much money chasing tri and supers. After seeing how much I was actually spending on the gimmicks, when I could have bet to win and made more in the long run I totally changed my way of playing. This is what works for me, and as my grandmother used to say To each his own.
rmath

[BJennet]

Richard,

I cleaned up the layout. Thank you very much for reporting this. This confirms some of what my own experience is, and that of several folks who seem reluctant to post it online.

One of the most important attributes of NewPace, IMO, is that it is automated - you push a button and 4 contenders appear. With the rest of RDSS and the Methodology, you have to push a few more buttons and use your brain to get the contenders. Doing so, at minimum, does not give you worse results. Then, with either approach, you begin making betting decisions.

Are you able to elaborate on the relative profitability of the 2 approaches to getting contenders, using a common betting strategy?

Thanks again for this work!

Ted

P.S. I moved this thread out of the Testing Forum

Hi Ted,

Can't go into as much detail as I'd like to due to circumstances, but this is a very important point you make that no one else seems to grasp. The great value of NP, if it works, is that it's automated. Therefore you can make many more bets than is possible with RDSS, therefore a greater net is possible with a lower ROI than RDSS. It is for this reason that RDSS can't really be tested. As Ted points out, someone is still making decisions - all the 'tests' I see here include filtering for odds and for other reasons depending on personal judgement - makes perfect sense if you're just handicapping, but it invalidates the results as a test. Also, and I don't mean to seem like I'm haranguing anyone - small samples are no better than chance. If you want to base your betting decisions on them that's your business, but don't expect them to hold up going forward. For those who aren't believers, I suggest a reading of Nassim Taleb's 'Fooled by Randomness'. One of the reasons I get annoyed with Dave is that he understands stuff like this perfectly well, but ignores it when selling handicapping products. To put sample size in perspective many blackjack games require a sample of 50k rounds to achieve a 95% confidence interval. Since you theoretically can have a much larger edge in handicapping the necessary sample is smaller, but still much larger than most people seem to believe.

Cheers,

B Jennet

[rmath]

BJ, the 50 races I ran were all the races run at 5 different tracks. I ran sprints as well as routes in my testing of New Pace and RDSS2. I did not filter any horses out for any reason except to get down to the final 5 as I explained in my previous post.
I was trying to be completely fair in my test.
I did not bet all of these races, only was trying to complete a fair study.
I did bet about 20+ races that do meet my guidelines but these had no effect on the testing. I know that 50 races do not make a big enough base to make final decisions on, but that is why there are many others doing thetesting as Ted wanted.
I am quit sure that each one of us would not pick the exact same pacelines that I chose. We each see things differently. What I do gets the results that I have come to expect after many years of practice. Like Bill V. stated there are certain races he does not even consider. He does have the option to run them tho and not bet them.
I personally confine most of my plays to 5.5f upto 7.0f for betting purposes but not for testing.
I hope this answers your ?s.
I will continue to run more races and report my findings.
rmath

[Bill Lyster]

Rmath:

I took a cue from your recent comparison studies and added VDC columns to the recent DMR workup. there are 100 races in this sample.

I expanded the number of contenders to top four and ties in VDC after hiding the layoffs and long odds horses. See my notes at bottom of page. I chose contenders and let total energy be the primary guide to deletion.

Every once in a while a #5 or #6 TE horse would show up as VDC 1 or VDC 2, so in those cases I kept those horses and hid horses with higher TE, but lower VDC.

I did not bet races with more than two FTS, or foreign first US starts, or races with 4 horses entered.

By doing this I started seeing some amazing results - 100 race sample.
Wins: 83
exactas: 55
trifectas: 36
superfectas: 14

Great Idea, glad you decided to post your data.

The spreadsheet: 23214


Bill

[rmath]

Glad I could help.
That is what we are all trying to do. Help one another to be better.
I see you follow DOCs guidlines for playable races. I do too, But I raised his low odds of 5/2 to 4/1. Get less plays but make more money.
Thanks again for the compliment.
rmath

[Bill V.]

Thank you Bill L and Rmath

Im so happy Doc's VDC is doing so well for people
thank you for doing these test. VDC readouts and a solid betting method
seem very profitable in 2011 and into the future

one of the best vocals I have ever heard
Smith
http://youtu.be/8_uGlUxZ1J4

[Ted Craven]

I re-posted Bill's L's spreadsheet (in his post above) after there was a problem with the initial one. It's available now to view.

Ted

[Bill Lyster]

Race 50 , line 53 is actually a win; so total wins = 83. It does not change anything other than the win percentage and ROI. All other columns were correctly shown the first time.

Bill

[SilentRun]

Thanks for these studies they are valued contributions.....Ernie

[JeffH]

Just a reminder that the VDC ranking sometimes has ties, so Top 2 should probably read 'Top 2 and ties', which in less than 40% (??) of races may mean 3 horses (e.g. 2 horses tied rank 1, or 2 horses ties rank 2)

I was just going to comment on the VDC scores. Is there a way to break these ties using the program so top3 or top 4 are exactly that? I was running races on Friday, and found some races with 5 horses in top 3 VDC.

[rmath]

JeffH; Jeff I use odds to separate any ties beyond the top as my second and third horses win% is about the same. You could use bl/bl or if you keep track of the rankings that make up the primary and secondary line scores,you coulduse which ever rankings are coming up the most. Eric Penicka wrote about separating ties in several articles in the Follow Up. I have found that when I have a tie for top that usually one or both will be right there.
I only get ties in about 30% to 35% of my playable races and if two or more of the top 2 or 3 ,depending on ties, are bet down below 8/5 or less I pass that race. Lately I have been passing a lot of races that have less than 7 starters.

[Bill Lyster]

Here is the Delmar meet updated through yesterday. I also corrected some minor spreadsheet errors.

I allocated 4 slots for VDC horses and occasionally you will see five horses crammed into those 4 spaces. the horses entered are from the BL screen with highest BL that is VDC 1 listed first. sometimes two or three are tied as VDC 1, so they are entered, left to right in the 1-2-3 slots.

The column subtotals at the bottom of the sheet do not represent the number of VDC1, VDC 2, etc win totals as a result.

Bill

[Bill Lyster]

Here is the Delmar meet updated through yesterday. I also corrected some minor spreadsheet errors.

I allocated 4 slots for VDC horses and occasionally you will see five horses crammed into those 4 spaces. the horses entered are from the BL screen with highest BL that is VDC 1 listed first. sometimes two or three are tied as VDC 1, so they are entered, left to right in the 1-2-3 slots.

The column subtotals at the bottom of the sheet do not represent the number of VDC1, VDC 2, etc win totals as a result.

Bill

[Bill Lyster]

Has anyone found tracks where VDC is not as predictive as it seems to be at Delmar?

Does VDC work universally, or is it better suited to the major vs minor circuits?

I am curious to know if this factor is localized or not.


thanks,


Bill

[lsosa54]

Bill: My opinion is that the relationship between velocity & deceleration is valid everywhere. I think the issue with the smaller tracks and cheaper races is finding the predictive pace line, as form does not tend to hold for several reasons.

[Bill V.]

Has anyone found tracks where VDC is not as predictive as it seems to be at Delmar?

Does VDC work universally, or is it better suited to the major vs minor circuits?

I am curious to know if this factor is localized or not.


thanks,


Bill

Hello Bill L

I am having good success at both Del Mar and Monmouth using VDC only
I want to stay dialed in to what is working for me both financially and mentally. I'm doing well at Mth in the survival contest they are running
Plus my daily Del Mar selections are for all to peruse.

As soon as Del Mar and Mth end I will test other tracks
I truly believe things will always be good with VDC

Bill V.
Orangeburg NY

[rmath]

Bill, I have used vdc at Beu,RD, Tdn, delaware, Mth, Saratoga , Mnr,Suffolk, Lad,CD, AP, and several other high and low class tracks over the past 20+ years and it has always held up quite well. I do believe that the win % may be a little lower at the lowest class tracks, but not enough not to play them.
Hope this helps.
rmath

[JeffH]

JeffH; Jeff I use odds to separate any ties beyond the top as my second and third horses win% is about the same. You could use bl/bl or if you keep track of the rankings that make up the primary and secondary line scores,you coulduse which ever rankings are coming up the most. Eric Penicka wrote about separating ties in several articles in the Follow Up. I have found that when I have a tie for top that usually one or both will be right there.
I only get ties in about 30% to 35% of my playable races and if two or more of the top 2 or 3 ,depending on ties, are bet down below 8/5 or less I pass that race. Lately I have been passing a lot of races that have less than 7 starters.

Thanks for the info. I did use BL/BL to break ties on Friday. I did 50 races and the winner was in top 2 60% (30 winners). 70.84% were in top 3(34-48).

[rmath]

Nice going JeffH, keep up the good work
rmath

[Ted Craven]

Jeff,

There was a method to Doc's madness in having tied ranks in VDC and I think you can read his words somewhere in the Follow Up issues around the mid-70s or later. The whole point was that some horses may not have enough Total Energy to rank highly in BLBL (which is highly co-related with TE), but they are making a stong move and decelerating modestly compared to their incremental velocity. That's what VDC measures (deceleration relative to incremental velocity - segment by segment). So those horses are elevated higher than their BLBL rank would otherwise show. In other words, horses that on the surface might otherwise appear somewhat slow, but have hidden potential are brought to your attention by elevating their rank (when this has merit, do check other corollaries).

His advice was to separate the tied VDC ranks based on Odds, as Rmath suggests at the outset. Consider the top 2 and ties, or the Top 3 without ties as equal, then select 2 horses from those 3 when you can get your minimum odds. Your own hit rate of 60 - 71% for Top 2 and 3 VDC horses respectively indicates that if you pick theTtop VDC above your own odds minimum, say 5/2 or 3-1 plus one more horse from the remaining 3 (separate by Odds, pick the highest odds horse of the remaining 2) - you must eventually make a profit. I suggest you go back and check your 50 races to see what net return you would end up with if you followed this recipe. In effect, if you can consistently achieve a hit rate like this, you can dial your own ROI by restricting yourself to the minimum odds required to achieve it.

VDC is one thing Speculator users are missing, and a good reason to make the move to RDSS, IMO.

Ted

[JeffH]

Thanks for the input, Ted. I will go back and work more races with using same paceline method and using the odds criteria. I will post the results if they prove profitable.

[rmath]

I now have completed 100 races with rdss2. Here are the results :
#1 vdc 51 wins
#2 vdc 22 wins
#3 vdc 11 wins
#4 vdc 9 wins
#5 vdc 2 wins
eliminated 5 wins from top 5 final contenders.

there were 82 perfectas in final 5 horses.
71 perfectas were in top 4 vdc horses.

There were only 4 winners in the New pace horses that did not rank in the final 5 vdc contenders.
New pace only had 52 perfectas in its contenders.

My conclusions thus far are if any of the 4 new pace contenders are not in my final top 4 vdc horses they should be considered as non contenders.

One further note this test consisted of only playable races as per Docs guidelines, unlike the first 50 races I posted. There were quite a few races in the 50 that I would not normally run.
I hope that this info is a help to any and all who are interested.
rmath

[Mr.Pagine]

Ted,

Tested it yesterday with great results......

5 races to bet (I don't bet Maidens) - we had 5 winners.

3,5,9 - good prices
7,8 - way below what I would bet on

So far so good my friend.....

Happy Punting
JDL

[rmath]

First I missed counted the top vdc horse, the top only won 50 not 51 as I posted earlier. Sorry for the mistake.
I stated in that post that New Pace was not doing that well,but as I went over my records I noticed that if I compared the horses from both methods and only used the ones that were listed in the top 3 vdc and were also listed in new pace that the winner was in this group most of the time.
43/50 top horses or 86%
15/22 second
9/11 third horses.
In most of the races I got it down to 1 or 2 final contenders for win. Once in awhile I ended up with 3 win horses.
Instead of looking at 5 or more horses I am able to separate the ties using the combined method.
Dave Schwartz talked about using two different approaches to get it down to only a few horses in each race. He said you should use two very different methods, and IT is my thinking that RDSS & New Pace meet that goal.
I tried it today and caught three nice payoffs at Finger Lakes: 11.60, 15.40 and in the last a 34.20 horse. I played 5 races where I only had a total of 12 contenders. In the two races where I had 3 horses I bet the 2 best prices. One ran 2nd beaten by a 4.10 horse that even tho it was one of the final horses I eliminated it for odds.
Dave suggested that you should be able to get 2/3 of your winners from 1/3 of your contenders and I think this is coming pretty close.

[So Cal Al]

First off, nice work guys! With enough analysis , we might get down to some workable considerations. I have formed some "impressions" so far that I can share. Perhaps, others will be able to add or subtract from these "impressions". There is no question that what Ted has done in RDSS2 makes NP a faster handicapping approach. It also seems that those with accurate paceline selections can do well (or better) using VDC. If I understand some of Ted's recent input, I think he is looking for collaries, as well as differences in performance between VDC and NP. One early "impression" that I have so far, is that NP works better at lower class tracks, e.g. Mountaineer better than Del Mar. Has anyone come to a similar opinion? Another impression is that when NP and VDC top contenders are alike, there is a very strong betting opportunity. My last " impression" for this reply is that NP, although "automated" must, per se, ignore foreigh horses and first time starters (which also cannot be in VDC readouts), but can be shown to be viable contenders by checking other information found within RDSS. Thoughts?

[Bill Lyster]

The current testing module for NP offers Trackmaster SRs and adjusted SRs. If you choose the TM SRs you can get foreign first time US starters because their races have been given a speed ratings.

I think there have been one or more examples at DMR so far this meet.



Bill

[Bill Lyster]

In the 507 race sample only MNR and TAM would probably be considered as smaller or minor. the best results came from the first version of NP 34/48 wins at MNR and 42/50 at TAM

KEE hit 48/56 and AQU 50/61 as some of the best of the larger tracks, but be aware that hit rates where fields are smaller should be better just from a numbers standpoint (pointedly AQU). This would make KEE a standout.

OP hit 39/50.

The rest of the tracks studied hit about 2/3rds. Also no apparent bias between dirt or synth tracks.


Bill