Top 25 Most Overrated Performances of '06

[philcski]

Originally Posted by cmorioles

This isn't opinion, I don't know any serious figure maker that relies on pars. What one do you know that does?

About the only time I could think of using pars would be lightly raced horses or first time starters, things like that. Even then, you try to avoid those races if at all possible as part of making the variant.

Myself... for example, the par time for an older 20K claimer on the AQU inner track is 1:11.16 over the last 5 years (with a boatload of datapoints). Fact. What is incorrect about using this information to your advantage?? I'd rather calculate a variant, if possible, on what DID happen rather than what SHOULD have happened. I have firm reservations against the projection method. It allows human nature to "prove yourself correct" by matching the figures from previous races...

[cmorioles]

What is incorrect is that you need to build in some kind of variant to get any kind of useful pars. Horses are more consistent than track conditions. If you look at the history of horses that run well in a race, the pace and final time, as well as the pace and final times of other races on the card, you are going to come up with much better numbers than you will using pars.

Quick question...what figure did you give Invasor's BC Classic, and what figure did you give Magna Graduate a few weeks ago on the inner dirt?

[philcski]

Originally Posted by cmorioles

What is incorrect is that you need to build in some kind of variant to get any kind of useful pars. Horses are more consistent than track conditions. If you look at the history of horses that run well in a race, the pace and final time, as well as the pace and final times of other races on the card, you are going to come up with much better numbers than you will using pars.

Quick question...what figure did you give Invasor's BC Classic, and what figure did you give Magna Graduate a few weeks ago on the inner dirt?

Invasor: a 124 on my figures, 1 point = 1 length (which works out to approximately a 113 Beyer.) He had run a 126 (Suburban) and a 125 (Whitney) in his previous two, Bernardini had run a 124 (Jim Dandy), 124 (Travers), and a 122 (JCGC). Premium Tap had run a 117 (mini-stake against SK), 118 (Whitney), 119 (Woodward). Note: I was very happy with this number, as Premium Tap ran a 120 in the Clark.

Magna Graduate: A 118 (about a 103 Beyer). Obviously the BC day CD race was a throwout for comparison purposes but his one other race I have in the database was a 119 (the Discovery.)

[cmorioles]

I only asked as by most time charts, Magna Graduate's races comes back gigantic. You must be doing something right in my opinion, though I came up with similar relationship using mostly projections.

To be clear, I rarely if ever use projections to modify single races. I use projections to come up with the overall variant for the card. I don't like splitting races out just to make them fit prior history, I need additional evidence.

[philcski]

Originally Posted by cmorioles

I only asked as by most time charts, Magna Graduate's races comes back gigantic. You must be doing something right in my opinion, though I came up with similar relationship using mostly projections.

To be clear, I rarely if ever use projections to modify single races. I use projections to come up with the overall variant for the card. I don't like splitting races out just to make them fit prior history, I need additional evidence.

The 9.5F distance is (obviously) seldom used, which the time charts don't account for- that's probably why the strictly-computer-generated figures don't get a good number. If it's a REALLY wacky distance like that 2 mile race they ran a couple weeks ago... I don't even bother with it. Just pull it out of the day's formula and have it as an "N/A" figure (not like a 2 mile race figure is worthwhile anyways!)

What's the MINIMUM number of races you prefer to have a variant "group" on a specific day? For me it's 3. Once is an abberation, two is a trend, three is a defined set of correlated results. (Of course, occasionally you'll have a day like the Jim Dandy where the first 2 dirt races were an obvious split thanks to the weather.)

[cmorioles]

I prefer three as well, if it is at all possible.

[Danzig]

perhaps if you put lemons forever in a match race with seek gold, one of them would see the winners circle again. oh, and just for fun, drag colonial colony into the mix. maybe sinister minister as well!