Discut, welcome!
A few things:
Actually, some points in Mitch44's post above are incorrect, re Daily Track Variant DTV. His comments certainly apply to DTV supplied in the
Daily Racing Form, not by TrackMaster which RDSS uses. Sartin discussed many studies he did back in the day on DRF DTV figs and indeed found that 17 was an average or mean.
When he/we started using TrackMaster data in the late 1990s, that changed. Whatever the average DTV was for a day, that was called
0. Faster than average is shown as negative DTV and slower than average is shown as positive DTV values with Average in the middle. That's the way the Configure Tab is organized to display the range and amount of a given DTV used in incremental and final time adjustments.
The time value of a DTV point supplied by TrackMaster is 0.1 second divided by 8 = .0125 seconds per furlongs per race (using 8 furlongs as a norm). In a 8f race, -5 DTV means the final time was -5 x .0125 x 8 = -.50 seconds
faster than average or 'par' for that track, surface and distance according to TrackMaster/Equibase's annual updated tables. +7 DTV in a 6f race means the final time was 7 x .0125 = .525 seconds
slower than par. Etc.
What RDSS does with those DTV adjustments is to slow down or increase final (and incremental) times by fast DTV time values, and to speed up or decrease times by slow DTV values.
Times posted may be affected by environmental issues such as rain, wind and track maintenance and thus the DTV adjustments equalize times for the track/surface/distance over the season. However, much of time variance (possibly the majority) is due to the match up of horses running styles and preferred running styles (and quantity and mix of early, pressing and late running styles). Hence, Sartin (and by default) RDSS advocates a default setting of using HALF (50%) of the time value of those DTV points in adjustments - the other HALF of daily variance being a function of the aforementioned in-race matchup, not the environment. (Some prefer to use 100%, or even 0% - to each their own).
Further, to smooth out wild fast or slow DTVs (such as -30 or +40) often due to single races on a surface, or sudden storm, or race full of plodders or speedballs, RDSS provides outer margins of DTV ranges, outside of which DTV points are ignored. So, at the recommended DTV cutoff range of -15 to +15, using 50% of DTV, a faster than par -5 DTV turns into the time value of -2.5 and a slower than par +20 gets cut off at +15 then half of +7.5 gets used in the adjustment process.
The final time variant is also applied incrementally to the first and second call points.
As Mitch said, you don't see those detailed adjustments, but that's what is used.
The raw times from the Original screen are projected to TODAY'S distance by a formula, then modified by a race distance par-gap concept of Sartin's (another discussion) which brings into the adjustment the final and incremental speed of the given paceline (which also adjusts for surface), then DTV is applied per above, then ITV Inter-Track Variant is applied.
Inter-Track Variant was your original question, I think! The ITV is supplied by TrackMaster with each race card from their periodically updated inter-track normalization studies (annually, or as soon after a new track or surface debuts as valid data collection allows). This ITV is not shown and is proprietary. It uses the same scale as DTV = .125 per point at 8f. Times from slower tracks get sped up, times from faster tracks get slowed down. According to the Config settings, you can either use those ITV adjustments or not use them (i.e. enough rope to lasso the horse or to hang yourself).
In fact, after the Original screen (moving from left to right in the set of Tabs for a horse and on the Analysis screen), ALL the data is adjusted and conceptually no longer is being run at the host track - it is being run at (affectionately dubbed) '
Sartin Downs' - a hypothetical location of the mind where (as best as possible yet never perfect) the pacelines of each horse can be compared to each other. Of course, the more adjustments you make to running lines, the more fiction you indulge in. Hence longer projections of sprints to routes or vice versa, or 8f to 10f or longer distances should carry mental caveats. Some people like to apply restrictions as to how many furlongs difference to accept in using a running line for a horse in today's race (e.g. no more than 1 furlong difference). Caveat emptor. Use a slower line to represent what a horse can do today from a line closer in distance to today's race, and the horse looks worse. Or - let the software do the work most of the time, and develop a sense as to when things are off.
Other tidbits - using the DTV to decide when the track is 'fast' or 'slow' can be a better decision (i.e. based on time) than whether the track condition is rated sloppy, slow, yielding, etc. (Of course, sloppy tracks do throw mud in the faces of pressers or unfortunately positioned horses, thus affecting their times.)
You don't have to make ANY adjustments to the Adjusted numbers you see, or the Velocities or compounded factors which follow from those adjusted times. If you are trying to gain insights which you can use in your own program, hopefully you can benefit from the above described CONCEPTS, though you'll have to create and maintain your own DATA to accomplish the same adjustments.
Meanwhile, you can just use RDSS as is!
Ted