RDSS 2.0 Pacelines

[pilot]

Hi, could someone please tell us, how many paceline selection methods we will have to choose from, and maybe a little about each of them.


Thanks Wayne

[Ted Craven]

Hi, could someone please tell us, how many paceline selection methods we will have to choose from, and maybe a little about each of them.

Thanks Wayne

Wayne,

I'll tell you what I've done so far, and what I see immediately forthcoming. The pre-beta version a few folks are looking at has a Last Line mode (picks the last line for every horse) - nothing to write home about, but it gives a quick look at Fastest Pace Last Race (FPLR). It also helps if you tend to pick your lines from within the past 3 lines, as likely half the lines will be line 1 anyway - just review the horses' PPs and choose alternates where appropriate.

I wrote a version of best of last 3 comparable surface/distance (BLT/C) to see what the moving parts in such a paceline selection tool were. It works pretty well for me, but I'll have to rewrite it so that some of the decisions I hard-coded can be modified by the user: use SR or Total Energy or Perceptor as the base, go back a variable number of lines (not just 3), deal with wet surfaces, decide how to compare Turf with Poly, or Dirt with Poly, etc, etc.

I will likely deliver a user modifiable template to create several saveable versions of line selections, which you can display on the spot, or have the Data Centre create for you during its Task run. One feature I like is being to show multiple different Analysis sets (collections of pacelines) simultaneously, so you can swiftly compare Last Line to your version of BLT/C to any manual studies to may want to do (e.g. multi-lines per horse).

I have lots of other ideas for paceline selection ideas, but those will likely have to wait until after the initial release of RDSS2, and users' feedback.

HTH!

Ted

[Lasix1]

Dear All,

This thread is about a year old, but I couldn't find a follow-up and am hoping I can get an answer without bothering Ted with a PM (not that he would mind!)

One of the automatic strategies in RDSS 2.0 Beta 2.5 is called "Population Sample 2nd Call". Does anyone know how this is defined, calculated and/or how effective it is? I've experimented with it a bit and it seems to throw very different pace lines from the more common variations such as "Best Preceptor", Best B/L-B/L Last Three, or Best VDC, Last Three. My very short and limited scan of it shows that it doesn’t pick winners as often, but selects live place and show horses very well. I might be dreaming about this, but before I run a longer test, I'd like to have some input from the Board.

Many thanks for your help! If it has been discussed elsewhere and I missed it I apologize in advance.

[SilentRun]

Dear All,

This thread is about a year old, but I couldn't find a follow-up and am hoping I can get an answer without bothering Ted with a PM (not that he would mind!)

One of the automatic strategies in RDSS 2.0 Beta 2.5 is called "Population Sample 2nd Call". Does anyone know how this is defined, calculated and/or how effective it is? I've experimented with it a bit and it seems to throw very different pace lines from the more common variations such as "Best Preceptor", Best B/L-B/L Last Three, or Best VDC, Last Three. My very short and limited scan of it shows that it doesn’t pick winners as often, but selects live place and show horses very well. I might be dreaming about this, but before I run a longer test, I'd like to have some input from the Board.

Many thanks for your help! If it has been discussed elsewhere and I missed it I apologize in advance.

Hello Lasix1,

The last time I looked the "Population Sample 2nd Call" is defined in the release notes and it is calculated based on the 2nd call POH ADJUSTED velocity. The program scans the entire PP's of a horse and extracts the line with the best 2nd call POH. You have the option of doing it by an individual horse or all the horses. I tried a few races and then adjusted the results based on my best paceline selections. The few races I did had one or more of the horses ITM as WPS. So I am seeing similiar results on limited data.

In order to determine the profitability potential is to conduct your own study over as many races as you can eat....in the hundreds at the minimum. So you have to do the grunt work set up your spreadsheets and dig in.

Ernie

[Lasix1]

Thanks so much, Ernie, for your quick and comprehensive reply.

I now see exactly how the program calculates the pace line to be selected with this setting and thanks to you can get going on running it through my results. I've got a 400+ race data base on RDSS selections and I'll rerun them with this setting and see how it comes out. Will let you know as soon as I get some results.

Thanks again for your kind reply. It was just what I needed. best wishes--chuck

[Ted Craven]

Ernie, Chuck,

Here is a separate thread which elaborates a little bit on how the 2nd Call Population Sample PSS works:

http://paceandcap.com/forums/showthread.php?t=8311

Ted

[Lasix1]

Thanks, Ted!

[Alice0607]

I have been using the best SR in the past performance, and that did not work out well. I am attempting to use the best of last three pace line (dirt, turf, sprint, or route). Any further suggestions would be welcomed. I am finding the best in TE, Preceptor, BL/Bl, or SR are all the same horses.

[delmarscott2004]

Alice0607.............using RDSS2 . I too use the best of the last 3 lines , excluding off track lines , dirt to dirt to poly is o/k with me , turf to turf . But if the last line is the best ever ,never showing another line like it in the total PP, I use the 2nd best line in the last 3 comparable. Bits of knowledge gained from Doc ,The Hat and other Pirco teachers ,Ted , Bill , RichieP and other fine handicappers on this board. late blooming 2yr olds turning 3 can bring some of this into question...................Scott

[Alice0607]

I have been testing paceline selections, with approximately 90 races analysed. What I found is shocking. the races handicap was from the New York circuit, Woodbine, South California circuit, north California circuit, and gulf stream park, churchill downs, Kneeland, and fair grounds
The Highest win percentage for dirt sprint using best preceptor selections of paceline was picking preceptor rank. For dirt routes, the best vdc selections of paceline with picking bl rank, for turf route races using best Bl for the last 3 races with picking bl ranking.

[Alice0607]

Adding more information about the previous post. I selected each paceline selection method, and indicated wps and three horse exacta box of each paceline selection. The only elimination I enter for each paceline method was dslr.
The ranking for each was placed on spread sheet for wps and exacta box, with % won by each method. When I finished complying the information I will post my results.

[Ted Craven]

Alice,

So, are you comparing the same set of races against all the existing Paceline Selection Strategies and tracking the WPS and Exacta mutuel payoffs for each - i.e. for comparing the different PSS against each other, perhaps by distance structure (or surface)?

And, only eliminating horses from contention by Days - no other criteria (e.g. finish position or 'moves' made in the last race or so)?

If I got that right, it would indeed be interesting, even if the sample is rather small. I hope your spreadsheet identifies each race (Date, Track, Race #) so other interested parties can study your work on their own.

Thanks again,

Ted

[Alice0607]

The only other elimination I have been using is the APV. If the apv is less than 50 % the horse is placed in the secondary category. I had redo all my previous calculation to used this as part of my calculation. I will post the result in a couple of weeks

[Alice0607]

Test results uing rdss2.0 with seven paceline selection strategies.
138 test races with results for win percentage and thre horse exacta box. Using the specaltor system with the seven paceline selection strtegies. Tracked used Aqu, Bel, Sar, Cd, Kne, Mth, Gp, Wo, Dmr, Hol, Sa, and Gg. I have 138 tst races of which 64 are of dirt sprint, 38 of dirt routes, 3 of turf sprint, and 33 of turf route races.

The perentage of dirt is 73.9 % and turf is 26.1 %. The percentage of dirt sprint to the total is 46.2 %, dirt route is 27.5 %, Turf sprint is 2.2 %, And turf route is 23.9 %

My method of eliminating horses are: (1) dslr - greater than 45 days - the exceptions ar if the horse had 5f workout over 92 SR and within 14 days, (2) AVP value less than 50 percent of the average field rate, and the (3) eliminating horses with no form or out of form.

I will post the results in next few days.

[Bill V.]

sounds like you have done lots of work.
I say good for you .

some questions
1. Does your test include maiden races ?
2. Do you have a min. field size ?
3. What do you consider in or no form ?
This can be tricky and your results might be different than the next person
based on this eliminator

Here is an example
would you rate this horse ?
Its going back to a 6.5 sprint today
33792

[Alice0607]

To answer your question #1 - there are no maiden races in the calculations. Question #2- no minimum field size. I will handicap the races with four year olds and up. The number five horse, I would declare it as no form horse because, it only race in route races and only one race of 7 furlong nad it came in four 6 lengths behind the winner. It has no early speed in any listed races. In my spreadsheet- it will list:date, trk, race #, race type, field size, secondary contenders elimination (apv, dslr, & no form), win, place, show, win, place, and show rankings in total energy, total preceptor, and total bl/bl rank, percentage of win with the top 2 horses, percentage of exacta wins with the top 3 horses. A breakdown of dirt sprint, dirt route, turf sprint, and turf route. This my third time going through all the races I have handicap since Aug 2012. It has taken me long time to complete the study, fiting in this study between work and attending to family matters. Also I don't play any races for two and three old horses only.

[joseph]

VDC--PACELINE SELECTION--IS IT IN ANY AUTOMATIC WAY CONNECTED TO THE HORSES CLASS AND TOUGHER RACE'S PACELINE IN WHICH HE HAS BEEN ENTERED ??????

DOES IT EQUATE WITH PERCENT MEDIAN ENERGY ????????

BY THAT I MEAN THE HORSE WITH THE MOST GAS LEFT IN HIS TANK AGAINST TOUGHER AND NOW SITS AND RUNS DOWN AN EASIER FIELD.

Thank you,
Joseph

[joseph]

VDC PACELINE SELECTION--IS IT IN ANY AUTOMATIC WAY CONNECTED TO THE HORSES BEST CLASS AND TOUGHEST RACE HE IS COMMING OUT OF ??????

DOES IT EQUATE WITH MEDIAN ENERGY INDICATING THIS HORSE HAS THE MOST GAS IN HIS TANK FOR THE LAST FURLONG TO RUN DOWN CHEAPER EVEN THOUGH HIS VDC RACE WAS NOT AN IN THE MONEY FINISH ?????

Joseph

[Ted Craven]

Joseph,

As for an explanation of (roughly) what VDC is measuring, you can find previous discussions over the years using the Google Site Search (website menubar, left side) and enter a search term like 'VDC explanation'.

You can also read what Doc Sartin wrote about its evolution from the Entropy days in a 2 part article beginning in Follow Up 76 Page 64 'Futureworld ...' and continuing in Follow Up 77 'Success and Profit for the 21st Century: a Whitepaper'.

VDC is Velocity relative to Deceleration (or vice versa) - simply: how much or little was the horse slowing down (decelerating) over each segment of the race relative to the velocity in that segment and the relative Pace of the Race overall. VDC only measures the paceline(s) you designate for the horse, and does not include 'class' measurements of race conditions or competition levels: it includes only velocity/energy derived info from a given single race.

Thus, you must still (always) take care to represent the horse from an appropriate line or lines (e.g. energy distribution showing on the E/L graph is not atypical); the horse is in a current form cycle or you are using consistent judgements about horses returning to a new form cycle - in other words: all analysis you would make anyway to determine if the horse is a contender to win or finish in the money today (see many discussions recently re Paceline selection).

If you have chosen a consistently applicable paceline, VDC will indeed show horses who have decelerated least in the 3rd fraction while also having run against a fair 2nd call and 1st call pace of race. You will see many corollaries with other readouts which include F3, for example TS+F3 (segments screen), especially where the VDC rank is 'out of order' with the BLBL rank. A simple Deceleration measurement like DEC (F3/2nd Call velocity) usually ranks inverse to the 2nd call rank - that is, the slower the horse was gong to the 2nd call, the less it decelerates in the 3rd fraction - which isn't exactly a newsflash, or even useful much of the time.

The real interesting info is when a horse (relative to its competitors today) decelerates less, late, relative to having had a bit of speed earlier (1st and especially 2nd calls). That mix of deceleration relative to velocity and tempered by overall Pace of Race is the 'secret sauce' of VDC. But you MUST perform other analysis about current form cycle and suitability to race conditions today to know whether your representative line for the horse (e.g. one of the last 3 lines) is fair today.

I am not providing the specific formula for VDC (and I know you are not asking for it ;)), but the above general concepts plus Doc's more theoretical description and examples in the above noted Follow Ups are all you really need to know. I don't know why light behaves like both a wave and a particle, but I do know it helps my cell phone work.

If you study some of the recent reports by Rmath, you will find that Top 2 and ties VDC ranks together with Composite Speed Ratings (CSR) from consistently chosen and representative pacelines, provide some steady sets of contenders from which to seek value.

Hope that helps!

Ted

[Alice0607]

The purpose of this study is to find which paceline selection will give me my highest win percentage for win, and exacta. Using the two highest rank for win wager, and the the three highest for exacta wager. It is not large sample, only 138 races and counting. It is from NY circuit, Wo, Mth, Gp, Cd, Kne, Fg, Dmr, Hol, Sa, and Gg.

:1: The spreadsheet column's are as follows: Date, Trk, Race #, Race type
(dirt sprint, dirt route, turf sprint, and turf route), field size, Secondary contenders( Apv, Dslr, and no form), win, place, and show rank.

:2: Ranking for Total Energy (win, place, and show), Total Preceptor (win,
place, and show), Total Bl/Bl ( win, place, and show).

:3: On the win column - the red color in bold indicates win with the top 2 picks. The win and place column that is highlighted in yellow indicates winner for the exacta ( top three ranking).

:4: Secondary contenders that are missed for win, place, and show are highlighted in bold black with shading in the background. There is also percentage that are missed for win, place, and show to the percentage of total number of races. That percentage is one percent for wins, six percent for place, and nineteen percent for show.

More description to follow.

Here is the Spreadsheet --> 33970

[Alice0607]

There is also win percentage and exacta percentage for TE, Preceptor, Bl/Bl, dirt sprint, dirt route, turf sprint, and turf route. For each paceline selection, Last line, Best preceptor, Last 4 lines, Population, Top 3 energy, Best Bl/bl, and Bet Vdc


:1: Dirt Sprint - For win only - Paceline selected Best VDC last 3 lines - Win rank is the Preceptor. It average 52 % for 138 races, of which there were 64 sprint races.

:1a: Dirt Sprint - For exacta only - Paceline selected was Best Peceptor - Exacta was best preceptor win percentage for 3 horse exacta box was 52 % for 138 races of which there were 64 sprint races.

:1b: Dirt Route - for win only - Paceline selected Last 4 lines - win ranking was from Bl/Bl rank which were 50 % for 138 races, of which there were 38 route races on dirt.

:1c: Dirt Route - for exacta only - Paceline selected Last 4 lines - Exacta was Bl/Bl rank win percentage was 37 % for 138 races, of which there were 38 route races.


:1d: Turf Sprint - not usable because of very low % to the total races study.



:1e: Turf Route - for win only - Paceline selected Top 3 Energy - win ranking was from the preceptor was 73 % for 138 races, of which there were 33 turf route races.


:1f: Turf Route - for exacta only - Paceline selected Last line - win was from Preceptor was 36 % of 138 races, of which there were 33 turf route races.

[joseph]

Red, Green, Yellow, the ratings for Class and Purse. Do you have that factor available in your breakdown of Data. What you have shown is very revealing. I thank you and eagerly await your response to see the skewing of your data with the Red, Green, Yellow Factors included if that is of interest to you.

Joseph