Latest (and last) Derby Prep figures thru 4-20

[cj]

Count Fleet Aqu 1-4 8.3f mdy 84 86 Grey Comet
GG Derby GG 1-11 8.5f mdy 100 86 Standard Setter
Holy Bull GP 1-18 8.5f fst 96 99 Offlee Wild
Santa Ctlina SA 1-18 8.5f fst 102 102 Domestic Dispute
The Lecomte FG 1-25 8f fst 89 93 Saintly Look
The WEBN TP 2-1 8f gd 83 79 Champali
Sham S SA 2-7 9f fst 99 98 Man Among Men
Whirlaway Aqu 2-8 8.5f fst 98 93 Boston Park
Miracle Wood Lrl 2-8 8.5f fst 101 98 Gimmeawink
GG Mile GG 2-9 8f fst 89 92 Ministers Wild Cat
Fntn of Yth GP 2-15 8.5f fst 104 105 Trust N Luck
Risen Star FG 2-16 8.5f fst 101 109 Badge of Silver
Dptd Test Lrl 3-1 8.5f fst 81 89 Cherokee's Boy
Southwest OP 3-1 8f gd 93 104 Great Notion
San Rafael SA 3-1 8f fst 105 94 Rojo Toro
Battaglia TP 3-1 8.5f mdy 108 88 Champali
ElCmnoReal GG 3-8 8.5f fst 107 95 Ocean Terrace
SA Oaks SA 3-8 8.5f fst 87 101 Composure
LouisianaD FG 3-9 8.5f fst 109 105 Peace Rules
Fla Derby GP 3-15 9f fst 108 108 Empire Maker
The Gotham Aqu 3-16 8.3f fst 97 108 Alysweep
San Felipe SA 3-16 8.5f gd 107 101 Buddy Gil
Rebel S OP 3-22 8.5f fst 99 90 Crowned King
Rushaway TP 3-22 8.5f fst 88 82 Private Gold
Lane's End TP 3-22 9f fst 106 95 New York Hero
Prvte Trms Lrl 3-29 8.5f mdy 104 94 Sky Soldier
Ill Derby Haw 4-5 9f fst 118 110 Ten Most Wanted
SnAnta Dby SA 4-5 9f fst 118 104 Buddy Gil
Aventura GP 4-5 8.5f fst 96 102 Dynever
Wood Mem Aqu 4-12 9f mdy 111 111 Empire Maker
Calif Dy BM 4-12 9f sly 119 94 Mr. Technique
Blue Grss Kee 4-12 9f fst 118 104 Peace Rules
North Spur OP 4-12 8f fst 96 94 Mauk Four
Ark Derby OP 4-12 9f fst 123 110 Sir Cherokee
Lexington Kee 4-19 8.5f fst 107 90 Scrimshaw
Fdrc Tesio Pim 4-19 9f fst 86 92 Cherokee's Boy



CJ

[cj]

When looking at the route races from Keeneland on 4-19, it was very interesting. There were 5 routes run that day, but the variant was split. What was odd was the split was not a normal one. The variant did not split mid-card, but was different for 8.5 and 9f races. I have encountered this a few times before, but it is rare.

Here were the variants assigned by the Beyer team:

Races 1 and 8 at 8.5f
slow 15 points

Races 2, 4, and 5
slow 5 points

I decided to investigate a little closer as this would be very hard to explain in reality. How could a sixteenth of a mile differ track speed so much, about 4 lengths? Examining the pace revealed the answer. The pace of both 8.5f races was exceptionally quick for the final times, while the 9f races were slow to average. I decided the "slow 5" variant was probably much more accurate for all the races, thus the 90 figure you see above for the Lexington.

Any thoughts???

CJ

[keilan]

CJ – from what you have described I would agree. The variant for those races might be a slow 6 or 7 depending on the class level and strength of field being analyzed.

When I made variants I always considered these two variables -- how fast or slow the race went to the 2nd call and the reliability of the class level. One might refer to these observations as the art of making good variants. Simply taking all races and averaging is too mechanical and occasionally leads to faulty numbers IMO.

[turfspec]

CJ - The first thing I think of when I see a variant split by distance is the gate placement vis a vis the turn and/or the run-up to the timer. These two factors can have a great influence on the early fraction and thus the final time for obvious reasons. An analysis of early pace vs final time without knowledge of these factors at the track in question is a guess at best. Another explaination for Kee on that day could have simply been the stretch was deeper than the turn or turns. Kee being an 8.5f track configuration more of the race (early fraction) at that distance was run on the turn vs the 9f races. Then, of course, is the eternal question of sample size when making a variant. Even if all the races fall in line in a four race sample you can still be wrong due to the sample size. The best we can hope for is to be right enough of the time over the long haul while realizing any one day can be wrong even though we're forced to make a decision.

Best Racing Luck - Rob

[Tom]

Originally posted by cjmilkowski


Any thoughts???

CJ

Shhhhhh! :rolleyes:

[cj]

Originally posted by turfspec
CJ - The first thing I think of when I see a variant split by distance is the gate placement vis a vis the turn and/or the run-up to the timer. These two factors can have a great influence on the early fraction and thus the final time for obvious reasons. An analysis of early pace vs final time without knowledge of these factors at the track in question is a guess at best. Another explaination for Kee on that day could have simply been the stretch was deeper than the turn or turns. Kee being an 8.5f track configuration more of the race (early fraction) at that distance was run on the turn vs the 9f races. Then, of course, is the eternal question of sample size when making a variant. Even if all the races fall in line in a four race sample you can still be wrong due to the sample size. The best we can hope for is to be right enough of the time over the long haul while realizing any one day can be wrong even though we're forced to make a decision.

Best Racing Luck - Rob

The first thing I did think of was the gate, but I watched the replays several times, including the replays from the day before, and the gate positions were stable in the same place they always are. As for the more on the turn theory, it doesn't hold water here. I already have that calculated in my pars.

I am not saying I'm right in this particular case, but I know, as you say, I am right way more often than not.

CJ

[turfspec]

CJ - I'm sure you are right most of the time. To clarify, the turn adjustment in your par/variant is based on an average? I don't have any idea what the turns were like at Kee that day - just throwing it out there as a possible explaination - but due to the banking and track maintenance the cushion on the turns is subject to more daily variation than the stretch. Do you happen to know the run-up distance from gate to timer at 8.5f vs 9f at Kee.

Thanks - Rob

[cj]

No, I don't know the exact distance. I just know the position via landmarks on the track.

I agree, we can never be certain. I just take what I think is the most reasonable explanation and apply it. I look for any way to exploit what I think are bad Beyers, since they are such a heavy influence on the odds.

As for Keeneland, I've been doing the variants there for 5 years, 10 full meets, and this has never arisen at the track before. I have to assume it is more of a case of the Beyer guys trying to figure out why the race came up so slow, not taking into account the pace.

CJ

[Tom]

CJ,
Actually, I run into this a lot at Finger Lakes. It is not uncommon to see 6 furlongs go fast and 5.5 go slow on the same card.
Somedays, at 6 furlongs races go wire-to-wire, while 5.5 are won by off pace horses. No idea why, but when it does happen, take off the rubber bands and limber up that wrist - going to get a workout at the windows.

[cj]

I've seen it a few other places, namely Haw and TP at 6 and 6.5f.

CJ