Speculator V1.60

[Ted Craven]

An update to Speculator is now available, Version 1.60. Here are the changes compared with Spec150:

1. Upgraded Spec160 to make use of the TrackMaster InterTrack Variant (ITV) figures supplied by TrackMaster. This variant makes different tracks, distances and surfaces comparable. It was always used in Validator but the formula became degraded after the Feb 2006 Speed Rating and Daily Track Variant (DTV) scale changes by Equibase/TrackMaster. It was never used by Speculator. The ITV formula was corrected and is now used in Spec160.
2. Changed the default use of the amount of the DTV from FULL back to HALF, in keeping with historical use in Validator and pre Feb 2006 Speculator, on the theory that the Match Up, day of the week, quality of runners (and not simply inherent track surface speed) has a significant effect on final time, and thus on DTV. Large (high or low) DTVs can sometimes significantly distort final adjusted times. If you wish to continue using the FULL DTV (the Spec150 default), rename the file TVHALF.TXT to something else (e.g. _TVHALF.TXT), but be mindful of large variants.
3. Changed the value in seconds of a DTV point to be in keeping with how TrackMaster formulates them. Instead of valuing a DTV point at .1 seconds (1/10th second) at all distances, Spec160 now values 1 DTV point at .0125 seconds per furlong of paceline distance. Thus at 8 furlongs, a DTV point is worth .1 seconds (.0125 x 8); at 6 furlongs .075 seconds (.0125 x 6); at 10 furlongs .125 seconds (.0125 x 10); etc. Previously, sprints were being over-adjusted, routes over 8f under-adjusted.

As a result of using the TrackMaster ITVs, adjusted times are more similar to what Validator produced prior to Feb 2006, although Spec still projects and extracts differently and uses a constant value for beaten lengths. In Spec160, one turn routes seem to behave more normally now, though undoubtedly some tracks, distances and surfaces will still ship too fast or too slow and may require some further manual adjustment.

I am sending out the update today by email to registered Spec150 users and to Spec150 evaluators (July 31 expiry extended to August 31). If I've missed anyone, please just write me (ted@ArtofPace.com).

After a lot of discussions with Ellis Starr and with TrackMaster engineers, I really do believe this update solves all the problems the software ran into as a result of the Feb 2006 changes at Equibase/TrackMaster. There will always be debate over whether TrackMaster's Speed Ratings and Variants are good or not, whether they are better or worse than other vendors', and probably sometimes they will be worse and other times they'll be better. Or good enough. However, I firmly believe Spec160 is now utilizing the TM data and ratings the way Doc originally intended it, and perhaps a smidge better.

This Spec160 update is my response (for now) to discussions elsewhere about fixing the formulas in Validator. I have no rights to provide any version of Validator to anyone. If that changes, I'll let people know.

As always, I'd appreciate any comments on whether folks find this update valuable.

Just a note, if you are evaluating Spec150/160, I'm taking off for 5 weeks starting August 26, so if you want to make any payment/licensing arrangements, best do it before then!

Good luck, good skill!

Ted

[Ted Craven]

Someone asked me this question, a bit confused over all the 'variants' flying about:

'I thought Speculator had TV in it from the very beginning, but now you say it did not and you have just 'added' a new variant to Spec160. Huh?'

My answer:

There are 2 types of ‘variants’: Daily Track Variants (DTV) and Inter Track Variants (ITV) and we need to distinguish between the two. The TrackMaster Speed Rating consists of 3 parts: SR based on winner final time + DTV + ITV.

DTV is supplied overtly by TrackMaster and is displayed in Validator and Speculator. It is the variance on each day, by track and surface, of the average difference between actual final time and projected final time based on TM’s Class Rating (which is supposed to be a pre-race projection of the winner’s Speed Rating). The DTV is supposed to help us know how relatively fast (negative DTV) or slow (positive DTV) was on a given day so we can govern ourselves accordingly as to how to treat the final time for a given race on that day. But we know final time is affected by the Match Up in any given race, as well as calibre of comptetition on any day, so the average DTV for a day can sometimes be unrepresentative of inherent track speed that day (which is why we try to use only HALF of the DTV in our calculations).

The ITV is NOT supplied overtly by TrackMaster but they helped Doc and Guy figure out how to extract it from the Speed Ratings. This variant helps compare pacelines from different tracks, surfaces and distances to a hypothetical ‘norm’ - i.e. as if the race was being run today on ‘TrackMaster Downs’. We then take this inter-track variant, which ranges from a fraction of a second to several seconds (positive or negative) and add it to raw times to create adjusted times in an attempt to compare ‘apples to apples’ when using pacelines from dissimilar tracks, surfaces, distances. You will notice that even on, say, a BEL 6f dirt race today, using a BEL 6f dirt paceline with a 0 DTV for that day, there will still be a difference between raw and adjusted final time, even though there is no track/surface/distance change and no DTV to apply. That’s because the BEL paceline is being equalized NOT to BEL (in this example) but to ‘TM Downs’.

This ITV was always used in Validator but was omitted from Speculator (no idea why - but I think it definitely belongs). So, Speculator always used SR and DTV and displayed these ratings, but in the inner formulas, never used ITV. If you take the same pacelines and compare adjusted final times, same DTV portion (i.e. HALF DTV) for both Spec150 (or SpecPA) and Spec160, the difference in adjusted times will be mostly the ITV, and partly the more accurate variable unit of measure for applying DTV (per point #3 in my original post of this thread).

This use of ITV data hidden within TM’s Speed Ratings is unique to Sartin programs - they never gave this to anyone else. They say – just use the final time Speed Ratings to do your comparative analysis. But we use an incremental pace methodology, so we want lots more detail about how a race was run than just final time.

Ted

[Ted Craven]

As a follow up thought to the last post, it occurs to me that all this talk about variants and formulas and ratings is way too much information for probably most of us. Do the programs accurately rank contenders for us to wager on is the real burning question?

In case anyone was wondering, you don't have to understand any of this variant stuff to use the software. I discuss any of this to the extent that I'm able, to give some underpinnings to why the programs work and their historical progression, and for those who may be interested. If any of it helps give more confidence to your use of the software - that's great! If it makes your head spin, just forget about it and cash some tickets instead!

Ted