Paceline Modes

[tbrown]

I'ver seen PL1,4, and 5 talked about, but I don't recall anyone mentioning PL2 or PL3.
Does anyone use these regularly, or for certain types of races?
Sitting here thinking up reasons not to go outside and face the 8 foot snow drift in my driveway:(

[Victor]

from Nov 01 Newsletter as posted by Ricks


quote:
--------------------------------------------------------------------------------
Below is a paceline mode summary of the methods used in HTR2001 (versions dated Nov 5, 2001 and beyond). Click the ‘paceline mode’ option box at the top right of the program screen to change modes at any time. The past-performance screens display the selected lines with a “#” symbol.

Method (0)
None selected – allows you to choose your own. If you don’t want to begin with a blank slate, then start with one of the other five methods and change the lines. The software will automatically switch to mode (0) for you when you change a line selection. See page 10 for more details on picking pacelines.

Method (1)
Selects the most recent line only. Will take the second line if the top line cannot be computed. Same as DOS version and the output and rankings should closely match except for decimal differences in the f/p/s.

Method (2)
Selects the best effort of the horse’s last three running lines. Same as DOS version.

Method (3)
Selects the best 2 of the last 3 lines and averages them. Same as DOS version.

Method (4)
Selects a single line that is the horse’s best effort at today’s distance or surface in the last 6 months. If the horse does not have a line that matches those criteria, then no line is selected for that runner. If no line is selected, the horse is considered a non-contender on the velocity screens. This is a change from the DOS version that forced a line to be chosen for every horse, even if it didn’t match the criteria above. For example, a horse with a two-year layoff will not have a paceline selected in HTR2001 mode (4), yet the DOS version will pick something regardless of the time away. The goal with method (4) in HTR2001 is to strengthen user confidence with the numbers. Imposing a strict set of parameters before a line can be chosen is certain to make the comparisons more accurate.

Method (5)
The ‘artificial intelligence’ method selects one or two running lines based on analysis of the horse’s form cycle and congruence to today’s conditions. Two lines are selected and averaged if the software perceives the first line unrepresentative of the horse’s normal effort. Similar to DOS version but some tweaking of the code has been done in HTR2001 to tighten up the accuracy.

Note on wet tracks: Modes (4) and (5) have been adjusted to allow use of a ‘wet’ running line if the user has changed the surface mode to “wet” in the Scratch screen. This is change from the DOS versions.

[Victor]

tbrown,

Eight feet? Must be that Global Warming again!

Have not used PL2 or 3 enough to say much-- but my question would be what is meant by 'best' of last 3? How best?

I suppose that PL2 by using a strict mechanical approach would uncover some overlooked longshots at times (used to use best of last three Beyers to choose my lines), but at other times, this type of line selection will not lead to good results. "Live by the sword, die by the sword." Seems unlikely to me that PL2 or 3 is superior to any of the others.

[km]

T'

PL-2 selects the line with the best Cramer figure from last 3. This is an ideal method for 2yr and lightly raced 3's where improvement is almost certain. Might as well look at their maximum best effort and then project how much improvement might occur. PL-5's algorithm is aware of this and usually picks the same line. PL-2 may also be a good choice with cheap claimers since it won't go back any further than 3 starts.

PL-3 selects the two best of the last 3 lines according to Cramer figs. This is a good method for horses that are consistent, such as Graded Stakes runners and high class grass horses. Not a good choice for horses with choppy or inconsistent form, cheapies and young horses. Don't want to average a good line with a bad one.

[ERNIE LOGSDON]

To prove what paceline mode is best "for you"
or is best suited "for your" betting habits.....run the number of races or tracks that you decide,but run all five paceline modes print each and compare....
For another way of looking at paceline modes:
First and foremost pl-5 is the only mode where the computer has free rein to select one or two pacelines and can go as far back as it takes to use a paceline that the horse normally runs expressed in feet per second.........

set it for paceline one and you are saying that the last line is the best....vanned off...stumbled over another horse.....finished 10 of 10 beat by 99 lengths but use the last line

pl-2 is the same you are instructing the computer that no matter what one of the last two lines have the best figure choose one

pl-3 the reins are a little looser ....you say I pay all this money every month....I read all the newsletters and am depending on this programmer to do his very best to get me a good program.....but don't you(the computer) go back more than 3 pacelines.....

PL-4 is explained in Ricks post above,it has good points if you never want to bet on a horse going long today,going short today on the dirt or turf for the first time today....the trainer is up to something can't possibly win so save me the ink....I don't want to see to many choices.

While I belive that one can see/handicap most races by using the last three lines I have my computer defaulting to PL-5 and am amazed at how many times it uses the last line only....
pl-5 has out performed my picks every time I try something new.......

I quit using PL-3 when PL-5 kept beating my selections and PL-3 in win % and r.o.i about 5 years ago.....
and the standard disclaimer "but I could be wrong" ;) Ernie

[Victor]

Paceline Zero is capable of being the best. In order to beat PL5, 4, 3, 2, or 1, attack first their greatest weakness, that is, they all will award lines or a line to horses that do not deserve them. That's the first step.

For spot plays I use PL5 because it works well in most situations. It could well be that PL2 exported for claiming or PL4 for stakes horses would do better under these circumstances. Right now I don't have the time to find out. Does anyone have databases for each paceline mode and could they report the results as they pertain to this discussion?

[tomcat]

I read the explanation of the pacelines from the last 3 posts. They seem a little different. Should we assume KM is the correct one?