Speculator Frequently Asked Questions (FAQ) - Page 4 - Pace and Cap

mufasa · 05-14-2007, 07:56 AM

Richie:

In responding to Turbulator (post #30) and while reviewing your RDSS screen shot of the horse in question (Starlet Sky) I have come across a problem when using my base principles. When responding in post #30 I selected pace line #3 as the one that I would pick from among the last 3-4 lines in using Speculator 160 and insisting when I could that it must have been a good form race at similar distance and surface, and more importantly that the chosen line was against the best pace time (base principles-BP). I am OK with that if I am going to limit myself to the last 3-4 races. In reality though I consider all 10 given pace lines when selecting my preferred pace lines and on the basis of my BP’s I would have selected pace line #6 for the horse-Starlet Sky. This preferred line has a Perceptor One ranking of 4 seemingly a disadvantage in Speculator 160 computations when compared to pace lines #3 ranked 2 or even pace line #4 ranked 1. As far as I can see pace line #6 has a better raw pace call time then pace line #3 and when adjusted, normalized, and equalized it has the better pace call time then either pace lines 3 or 4. A good choice of a pace line to my way of thinking but maybe not so hot as to Spec160 computations.

I have to rethink my use of the Perceptor 1 screen since my most important qualifying rule is anchored in good form race finish against the best pace time. Perceptor 1 uses all of the 8 factors in its ranking and I am at this stage most interested in pace call times.

I have plenty of work to do in better understanding how best to use this great tool-Speculator 160 (and equally I presume RDSS). I had such a great day on Saturday (5/12/07) at Belmont but now I must revisit that card and see why my pace line selections worked so well given that pace call time is just one of 8 factors considered in the ranking.

Back to the drawing board-

RR

Ted Craven · 05-14-2007, 12:58 PM

Mufasa,

Others may disagree, but i consider it critical to establish a consistent paceline selection method, complemented by a consistent analytical approach. Many rules, applied in a dimishing order of importance, may apply, but apply them the same way over a large number of races.

In your Starlet Sky example, applying a consistent approach of best of the last 3 or 4 lines, similar distance, surface, track (in that order), would land you on the 3rd line because the 4th line was a projected 7f sprint, not on Turf, even worse, from KEE polytrack, a notoriously fast projecting surface. If you think the horse is demonstably fit or could conceivably be rounding into fitness (e.g. after a recent layoff), and could run back to its previously demonstrated ability, going back 4 lines, or 6 lines, or keeping to a maximum time window (e.g. 180 or 240 days), whatever, all make sense. Zigging and zagging in the collection of choices to make will in turn make it harder to review a set of 20 or 50 or 100 past races to see how your contender selection method is working out, or your winner and money horse identification process.

I like your approach of considering all pacelines, but demanding evidence of fitness. Spec (or Val or RDSS) can tell you how the lines compare to each other, but you need to decide whether a given line best represents the horse in today's contest. I will often put in 2 lines for a horse and if I find they both put it in the top 5, I'll sometimes even choose the worse of the 2, knowing it is capable of better.

If you must use a projected sprint line in today's route race, or a dirt line in today's Turf race, go ahead if there's no other choice, otherwise: similar distance, surface, track if you can.

The Equalized, Normalized, Adjusted times seek to treat all lines as if they were run as today's race. An internal distance equalization table is employed, a portion of the TrackMaster supplied Daily Track Variant is applied to normalize faster and slower days, and the TrackMaster Inter Track Variant is applied to attempt to adjust times run at one track, surface and distance to a norm. The foregoing are all an imprecise science, admittedly, but they are all applied equally consistently, warts and all. That's why you should have some confidence using a GP turf line for an AP turf race, or a 1 turn BEL route in a 2 turn HAW route race, etc. You may get to know when the adjustments don't satisfy you, then you can tinker with them in the Manual Edit screen (but perhaps save that for the 2nd or 3rd hundred set of races analysed).

Other guidance to apply consistently could be: don't use a line (even a recent one) showing the best ever Speed Rating (either Trackmaster's or internal SR), and consider the entire PP set to determine this. Or, limit yourself to a number of days back for a paceline, e.g. 180, 240, etc. If you come to include any such guidelines in your practise, do so because you validated yourself that they produced a better or more consistent, or more profitable result than not including them - not because soemone else suggested it from their own observations.

Consider using the PreAnalysis tool (which ranks ALL pacelines from ALL horses), setting the filters to a reasonable timeframe and track/distance/surface settings, then look to see which horse owns the best Second Call rating (or multiple ratings), or best Total Energy. After you do this, the PreAnalysis ratings will appear to the right of each horse's Past Performaces and you can see how it compares to its competitors on an array of factors. If one horse has multiple lines better than the others, look closely!

Just a few ideas...

Ted

mufasa · 05-15-2007, 05:11 PM

Ted:

Thanks very much for your response. I definitely will seriously consider your suggestions.

I have a clarification question on two particular factors that are provided with the BLBL printout and are found in the Synthesis Power Rankings and Fractals information table. Could you explain conceptually how the factors LS (Line Score) and TPP (Total Pace Potential) are computed or more importantly I suppose what is their foundation. I know that from previous post there has been some mention of there importance so I would like to consider them in my analysis but need to know what makes them tick and how best to either apply them or consider them.

I just want you to know that to date I am enjoying the experience of using and learning the preferred use of Speculator 160.

Thanks,

RR

Ted Craven · 05-15-2007, 05:50 PM

Mufasa,

LS is easy: it stands for 'Line Score' (either the Line Score of Primary Factors, or the Line Score of Supplementary Factors). It is the sum of the ranks of the Primary Factors (or Supplementary Factors). The BL/BL number is then derived from the Primary Line Score at so many points for a #1, so many for #2, etc. It is the counterpart to Perceptor 1, or, Perceptor 1 is a refinement to Primary Line Score, if you will.

TPP is 'Total Pace Potential'. TPP (found on the Energy screen and the Synthesis Superscreen) is an aggregate of E/ep (Early Energy Potential) and L/ep (Late Energy Potential). E/ep is a kind of blend of the Early fractions velocity with a smidge of F3, and L/ep is a similar blend of later fractions velocity with a smidge of F1. You would have to satisfy yourself from record keeping whether these 3 factors produced any more winners, or extra ROI, (by track, distance, surface) than simply using EPR (compared to E/ep) or LPR (compared to L/ep) or CPR (compared to TPP). Some think they do - others think it's factor overload.

Re 'preferred use of Speculator' - the discussions so far talk only of the preferred use by those who have posted (me included)! There are so many tools in that toolbox (and in Validator as well), and some tools suit particular handicapping styles, or particular happy experiences. Some people use Early/Late almost exclusively, some the Computed Beaten Lengths screen. Some hate the BL/BL and love Perceptor. One guy goes straight for the PreAnalysis. If they choose to comment, you'll also get 'their' preferred usage, but lest your head begin to spin, remember to focus on a few factors until they regularly give you positive results, then expand, experiment.

Good luck!

Ted

mufasa · 05-19-2007, 07:31 AM

Ted:

I have another question on one of the Spec160 screens. I have been looking at the Long Shot Detector, screen #5 at the Ratings and Graphs section. When looking at the screen on the computer monitor you get PORSC, PORFT, POH2F, POH3F, and Total. When you print it out the columns come out as F1, F2, F3, SC, and Final. I know what POR and POH mean and I am assuming that it is a projected look at particular segments of the estimated forthcoming pace. So that I can be sure and not necessarily guessing can you tell me exactly what the numbers represent per column? Is it all call/fraction times vs. percents or ratings for the segments.

Thanks,

RR

Ted Craven · 05-21-2007, 10:55 AM

Mufasa,

The Longshot Detector screen is a Pace of Race time weighted by Pace of Horse time - Line Score thingy, expressed in seconds and 10ths. You're right that the screen captions do not match the printed captions - the screen captions are correct. Times are the adjusted call times from the Manual Adjust Screen: equalized to today's distance, normalized according to Daily Track Variant, and adjusted by any Inter Track Variant, plus any Manual adjustment you may choose to make. In other words, the adjusted times upon which all the rest of the velocity, energy and compound factors are based.

PORSC is Pace of Race 2nd Call time (with ranks)
PORFT is Pace of Race Final time
POH2F is Pace of Horse Fraction 2 time (i.e. POR F2 time adjusted by the horse's beaten lengths)
POH3F is Pace of Horse Fraction 3 time

TOTAL is a sum of the foregoing 4 times for each paceline (with ranks). As with Perceptor Total numbers, the Sum itself is not valuable per se, rather the rank and perhaps relative difference between pacelines.

Thus, the lowest TOTAL (rank #1) will go to the horse who ran faster (i.e. who had faster F2 and F3 times) against a faster 2nd Call and Final POR.

In the example below, the paceline I chose for Curlin showed he ran against a seriously faster POR than that of Street Sense, and his own F2 and F3 times against that pace (especially F3) made him a strong contender (at least off this readout).

Hope that answers it!

Ted

mufasa · 05-21-2007, 02:08 PM

Thanks for the explanation it cleared it right up for me.

Let me take this opportunity to tell you that I am definitely sold on Speculator 160. The software is easily the best tool I have ever had to work a race with. I definitely want the non-expiring license for Spec160 and look forward to using and working with RDSS.

Thanks,

RR

mufasa · 05-28-2007, 09:23 AM

Quote:

Originally Posted by mufasa

Thanks for the explanation it cleared it right up for me.

Let me take this opportunity to tell you that I am definitely sold on Speculator 160. The software is easily the best tool I have ever had to work a race with. I definitely want the non-expiring license for Spec160 and look forward to using and working with RDSS.

Thanks,

RR

Ted:

I have noticed most recently when trying to print out the Late/Early Difference Graph that the centerline (the one with the arrow depicting earl/late) does not line up in the printed product. That is, at least with my attempts to print, it aligns left of center (in the area of the late part of the graph). Perhaps this subject has been already covered.

As to printing in general, I hope that the new RDSS will allow the printing of final computation screens of choice rather then giving out a set/predetermined print. For example, I would of liked to be able to print out lets say the Incremental Match-up Graph and the Macthup Showing Computer Beaten Lengths screens on a 8.5 by 11 page.

Thanks,

RR

Ted Craven · 05-28-2007, 10:48 AM

Mufasa,

It has been ages since I printed anything from Speculator, so I hadn't noticed that the E/L centre line is offset - but you're right. I note that the length of the dashed-lines agrees with the screen presentation, and the numbers are the same, though. So if you can hold your nose with that printed readout and still make use of the info - I don't think I'm going to correct that for Speculator.

What I would do for Spec and Val since many years, is create my own custom printout by cutting and pasting from the DOS text screen into a Word document, or a text file, with font set to Courier New 10 (i.e. a fixed-width font).

That way, I would get 2 pages per race (usually), with the screens I most used in the order I desired. I used BL/BL followed by Perceptor, then on a 2nd page E/L followed by MU CBL (an a little wagering grid I made in a Word template document). Others use many other combinations.

I certainly intend to make printouts from RDSS customizable and in colour.

Ted

mufasa · 06-01-2007, 09:16 AM

Ted:

Can you confirm or help me understand what happens as to ranking/rating when a contender is scratched. My understanding from reading the Sartin material is that as to BL/BL LS (odds) placement that if lets say contender 3rd ranked is scratched that since the ranking/rating is not linear that the 4th ranked contender does not necessarily move up to now 3rd best but could move up to 2nd best with the original 2nd best becoming now 3rd best. That is a contender can leap frog over another as to BL/BL rankings/ratings when there is a scratch among the top 3. Could you verify my understanding in this area? In addition, as to Perceptor 1 rankings my thinking is that this is more linear and that if there is a scratch above all contenders move up the slot.

I know you are busy with RDSS so please respond at you best opportunity-thanks,

05-14-2007, 12:58 PM	#32
Ted Craven Grade 1 Join Date: Jun 2005 Location: Nanaimo, British Columbia, Canada Posts: 8,853	Mufasa, Others may disagree, but i consider it critical to establish a consistent paceline selection method, complemented by a consistent analytical approach. Many rules, applied in a dimishing order of importance, may apply, but apply them the same way over a large number of races. In your Starlet Sky example, applying a consistent approach of best of the last 3 or 4 lines, similar distance, surface, track (in that order), would land you on the 3rd line because the 4th line was a projected 7f sprint, not on Turf, even worse, from KEE polytrack, a notoriously fast projecting surface. If you think the horse is demonstably fit or could conceivably be rounding into fitness (e.g. after a recent layoff), and could run back to its previously demonstrated ability, going back 4 lines, or 6 lines, or keeping to a maximum time window (e.g. 180 or 240 days), whatever, all make sense. Zigging and zagging in the collection of choices to make will in turn make it harder to review a set of 20 or 50 or 100 past races to see how your contender selection method is working out, or your winner and money horse identification process. I like your approach of considering all pacelines, but demanding evidence of fitness. Spec (or Val or RDSS) can tell you how the lines compare to each other, but you need to decide whether a given line best represents the horse in today's contest. I will often put in 2 lines for a horse and if I find they both put it in the top 5, I'll sometimes even choose the worse of the 2, knowing it is capable of better. If you must use a projected sprint line in today's route race, or a dirt line in today's Turf race, go ahead if there's no other choice, otherwise: similar distance, surface, track if you can. The Equalized, Normalized, Adjusted times seek to treat all lines as if they were run as today's race. An internal distance equalization table is employed, a portion of the TrackMaster supplied Daily Track Variant is applied to normalize faster and slower days, and the TrackMaster Inter Track Variant is applied to attempt to adjust times run at one track, surface and distance to a norm. The foregoing are all an imprecise science, admittedly, but they are all applied equally consistently, warts and all. That's why you should have some confidence using a GP turf line for an AP turf race, or a 1 turn BEL route in a 2 turn HAW route race, etc. You may get to know when the adjustments don't satisfy you, then you can tinker with them in the Manual Edit screen (but perhaps save that for the 2nd or 3rd hundred set of races analysed). Other guidance to apply consistently could be: don't use a line (even a recent one) showing the best ever Speed Rating (either Trackmaster's or internal SR), and consider the entire PP set to determine this. Or, limit yourself to a number of days back for a paceline, e.g. 180, 240, etc. If you come to include any such guidelines in your practise, do so because you validated yourself that they produced a better or more consistent, or more profitable result than not including them - not because soemone else suggested it from their own observations. Consider using the PreAnalysis tool (which ranks ALL pacelines from ALL horses), setting the filters to a reasonable timeframe and track/distance/surface settings, then look to see which horse owns the best Second Call rating (or multiple ratings), or best Total Energy. After you do this, the PreAnalysis ratings will appear to the right of each horse's Past Performaces and you can see how it compares to its competitors on an array of factors. If one horse has multiple lines better than the others, look closely! Just a few ideas... Ted __________________ RDSS - Racing Decision Support System™

05-15-2007, 05:50 PM	#34
Ted Craven Grade 1 Join Date: Jun 2005 Location: Nanaimo, British Columbia, Canada Posts: 8,853	Mufasa, LS is easy: it stands for 'Line Score' (either the Line Score of Primary Factors, or the Line Score of Supplementary Factors). It is the sum of the ranks of the Primary Factors (or Supplementary Factors). The BL/BL number is then derived from the Primary Line Score at so many points for a #1, so many for #2, etc. It is the counterpart to Perceptor 1, or, Perceptor 1 is a refinement to Primary Line Score, if you will. TPP is 'Total Pace Potential'. TPP (found on the Energy screen and the Synthesis Superscreen) is an aggregate of E/ep (Early Energy Potential) and L/ep (Late Energy Potential). E/ep is a kind of blend of the Early fractions velocity with a smidge of F3, and L/ep is a similar blend of later fractions velocity with a smidge of F1. You would have to satisfy yourself from record keeping whether these 3 factors produced any more winners, or extra ROI, (by track, distance, surface) than simply using EPR (compared to E/ep) or LPR (compared to L/ep) or CPR (compared to TPP). Some think they do - others think it's factor overload. Re 'preferred use of Speculator' - the discussions so far talk only of the preferred use by those who have posted (me included)! There are so many tools in that toolbox (and in Validator as well), and some tools suit particular handicapping styles, or particular happy experiences. Some people use Early/Late almost exclusively, some the Computed Beaten Lengths screen. Some hate the BL/BL and love Perceptor. One guy goes straight for the PreAnalysis. If they choose to comment, you'll also get 'their' preferred usage, but lest your head begin to spin, remember to focus on a few factors until they regularly give you positive results, then expand, experiment. Good luck! Ted __________________ RDSS - Racing Decision Support System™

05-21-2007, 10:55 AM	#36
Ted Craven Grade 1 Join Date: Jun 2005 Location: Nanaimo, British Columbia, Canada Posts: 8,853	Mufasa, The Longshot Detector screen is a Pace of Race time weighted by Pace of Horse time - Line Score thingy, expressed in seconds and 10ths. You're right that the screen captions do not match the printed captions - the screen captions are correct. Times are the adjusted call times from the Manual Adjust Screen: equalized to today's distance, normalized according to Daily Track Variant, and adjusted by any Inter Track Variant, plus any Manual adjustment you may choose to make. In other words, the adjusted times upon which all the rest of the velocity, energy and compound factors are based. PORSC is Pace of Race 2nd Call time (with ranks) PORFT is Pace of Race Final time POH2F is Pace of Horse Fraction 2 time (i.e. POR F2 time adjusted by the horse's beaten lengths) POH3F is Pace of Horse Fraction 3 time TOTAL is a sum of the foregoing 4 times for each paceline (with ranks). As with Perceptor Total numbers, the Sum itself is not valuable per se, rather the rank and perhaps relative difference between pacelines. Thus, the lowest TOTAL (rank #1) will go to the horse who ran faster (i.e. who had faster F2 and F3 times) against a faster 2nd Call and Final POR. In the example below, the paceline I chose for Curlin showed he ran against a seriously faster POR than that of Street Sense, and his own F2 and F3 times against that pace (especially F3) made him a strong contender (at least off this readout). Hope that answers it! Ted Attached Images __________________ RDSS - Racing Decision Support System™ Last edited by Ted Craven; 05-21-2007 at 10:58 AM.

05-28-2007, 10:48 AM	#39
Ted Craven Grade 1 Join Date: Jun 2005 Location: Nanaimo, British Columbia, Canada Posts: 8,853	Mufasa, It has been ages since I printed anything from Speculator, so I hadn't noticed that the E/L centre line is offset - but you're right. I note that the length of the dashed-lines agrees with the screen presentation, and the numbers are the same, though. So if you can hold your nose with that printed readout and still make use of the info - I don't think I'm going to correct that for Speculator. What I would do for Spec and Val since many years, is create my own custom printout by cutting and pasting from the DOS text screen into a Word document, or a text file, with font set to Courier New 10 (i.e. a fixed-width font). That way, I would get 2 pages per race (usually), with the screens I most used in the order I desired. I used BL/BL followed by Perceptor, then on a 2nd page E/L followed by MU CBL (an a little wagering grid I made in a Word template document). Others use many other combinations. I certainly intend to make printouts from RDSS customizable and in colour. Ted __________________ RDSS - Racing Decision Support System™

05-14-2007, 07:56 AM	#31
mufasa Grade 1 Join Date: Mar 2007 Posts: 310	Richie: In responding to Turbulator (post #30) and while reviewing your RDSS screen shot of the horse in question (Starlet Sky) I have come across a problem when using my base principles. When responding in post #30 I selected pace line #3 as the one that I would pick from among the last 3-4 lines in using Speculator 160 and insisting when I could that it must have been a good form race at similar distance and surface, and more importantly that the chosen line was against the best pace time (base principles-BP). I am OK with that if I am going to limit myself to the last 3-4 races. In reality though I consider all 10 given pace lines when selecting my preferred pace lines and on the basis of my BP’s I would have selected pace line #6 for the horse-Starlet Sky. This preferred line has a Perceptor One ranking of 4 seemingly a disadvantage in Speculator 160 computations when compared to pace lines #3 ranked 2 or even pace line #4 ranked 1. As far as I can see pace line #6 has a better raw pace call time then pace line #3 and when adjusted, normalized, and equalized it has the better pace call time then either pace lines 3 or 4. A good choice of a pace line to my way of thinking but maybe not so hot as to Spec160 computations. I have to rethink my use of the Perceptor 1 screen since my most important qualifying rule is anchored in good form race finish against the best pace time. Perceptor 1 uses all of the 8 factors in its ranking and I am at this stage most interested in pace call times. I have plenty of work to do in better understanding how best to use this great tool-Speculator 160 (and equally I presume RDSS). I had such a great day on Saturday (5/12/07) at Belmont but now I must revisit that card and see why my pace line selections worked so well given that pace call time is just one of 8 factors considered in the ranking. Back to the drawing board- RR

05-15-2007, 05:11 PM	#33
mufasa Grade 1 Join Date: Mar 2007 Posts: 310	Ted: Thanks very much for your response. I definitely will seriously consider your suggestions. I have a clarification question on two particular factors that are provided with the BLBL printout and are found in the Synthesis Power Rankings and Fractals information table. Could you explain conceptually how the factors LS (Line Score) and TPP (Total Pace Potential) are computed or more importantly I suppose what is their foundation. I know that from previous post there has been some mention of there importance so I would like to consider them in my analysis but need to know what makes them tick and how best to either apply them or consider them. I just want you to know that to date I am enjoying the experience of using and learning the preferred use of Speculator 160. Thanks, RR

05-19-2007, 07:31 AM	#35
mufasa Grade 1 Join Date: Mar 2007 Posts: 310	Ted: I have another question on one of the Spec160 screens. I have been looking at the Long Shot Detector, screen #5 at the Ratings and Graphs section. When looking at the screen on the computer monitor you get PORSC, PORFT, POH2F, POH3F, and Total. When you print it out the columns come out as F1, F2, F3, SC, and Final. I know what POR and POH mean and I am assuming that it is a projected look at particular segments of the estimated forthcoming pace. So that I can be sure and not necessarily guessing can you tell me exactly what the numbers represent per column? Is it all call/fraction times vs. percents or ratings for the segments. Thanks, RR

05-21-2007, 02:08 PM	#37
mufasa Grade 1 Join Date: Mar 2007 Posts: 310	Thanks for the explanation it cleared it right up for me. Let me take this opportunity to tell you that I am definitely sold on Speculator 160. The software is easily the best tool I have ever had to work a race with. I definitely want the non-expiring license for Spec160 and look forward to using and working with RDSS. Thanks, RR

06-01-2007, 09:16 AM	#40
mufasa Grade 1 Join Date: Mar 2007 Posts: 310	Ted: Can you confirm or help me understand what happens as to ranking/rating when a contender is scratched. My understanding from reading the Sartin material is that as to BL/BL LS (odds) placement that if lets say contender 3rd ranked is scratched that since the ranking/rating is not linear that the 4th ranked contender does not necessarily move up to now 3rd best but could move up to 2nd best with the original 2nd best becoming now 3rd best. That is a contender can leap frog over another as to BL/BL rankings/ratings when there is a scratch among the top 3. Could you verify my understanding in this area? In addition, as to Perceptor 1 rankings my thinking is that this is more linear and that if there is a scratch above all contenders move up the slot. I know you are busy with RDSS so please respond at you best opportunity-thanks,