Has anyone else ever looked at the sectional times in detail?

The reason I ask is because of the recorded information differences between sectionals and the 600m race time given in the official results. I am not just talking about a small variance but differences of sometimes well over 1sec (This is over 6Lens).

For example take ANAPINE who won yesterday at Rosehill. Look at the information provide on its last start at Randwick R4. The last 600m for the race was listed at 34.33sec in the results, and ANAPINE lead all the way. Therefore you would expect its last 600 in the sectionals to be around 34.3secs. But when you open the sectionals you see the 600 sectional of 35.7secs. When you listen to the race it seems the sectional of 35.7secs is closer to the money.

This is just one example of many and its not just Randwick. Just have a look for yourself on the ajc site or any other.

Can anyone shed some light on what is going on and which information can we trust if any? This is very frustrating when trying to do sectional analysis.

__________________
Regards,
Pete

Hi!
Yes, there are many mistakes in the data, after all these figures are entered by human data entry operators.
I personally depend a great deal on the final time of the race re. rating, but disregard the sectionals even when fully available, ie. for the turn 800, etc.
simply because it will only show you one thing, and one thing only; HOW the race was run!!!
This is only my opinion of course, so devotees of sectionals, if you want to buy an argument look elsewhere.
The time for Rosehill on a good track for 1400 M should be around 85 + sec.
I have a bench mark for all tracks all conditions all distances, and my programme will compare the racetime with the benchmark before rating, and if it is out of whack I get an error message.
Saved me a lot of dud bets I can tell you!
Hope it helps a little, good luck!

Beats me. Having just checked something clearly doesn't add up.

The AJC uses a (presumably fixed) beam to calculate the 600m time. Did they neglect to adjust for the 8.0m rail?

Thanks Marcus25 - When are they going to replace those silly human operators with computers.

I also have track models for all tracks under good conditions for open or = class. For example Rosehill is:
m ParTime
800 45.2
850 48.3
900 51.4
950 54.5
1000 57.6
1050 60.75
1100 63.9
1150 67
1200 70.1
1250 73.2
1300 76.6
1350 79.9
1400 83.2
1450 86.4
1500 89.6
1550 92.8
1600 96.2
1650 99.6
1700 102.95
1750 106.3
1800 109.65
1850 112.95
1900 116.25
1950 119.65
2000 123
2050 126.35
2100 129.65
2150 132.95
2200 136.25
2250 139.55
2300 142.85
2350 146.15
2400 149.45
2450 152.75
2500 156.05
2550 159.35
2600 162.65
2650 165.95
2700 169.25
2750 172.55
2800 175.85
2850 179.15
2900 182.45
2950 185.75
3000 189.05
3050 192.35
3100 195.65
3150 198.95
3200 202.25

I notice my Par for 1400 is a bit quicker than your's but it all relative.

I do similar benchmarking and thats why I posted but to date have been relucant to change data. I find myself adjusting daily variance factors but I will keep your thoughts in mind next time I come across something funny.

__________________
Regards,
Pete

A brief investigation suggests the discrepancy is due to the rail position.


Diff - Rail
0.00 true
0.19 +4
0.43 +8

The sportscolour 600m time seems to exceed the AJC time by ~:

5% of rail position.

The AJC time appears to be correct.

Sportscolour uses underground sensors and obviously it wouldn't dig them up and reposition every time the rail moves. It claims to normalise the times to 200m but apparently doesn't bother to here.

It's not a big deal to remember than when the rail is out 8m the 600m trench is really ~607m, then scale the observed time by 600/607.

Still this sloppy status quo should give false rail runners tremendous early speed figures for experts to enthuse about.

As accuracy seems a problem, is their any point in sectionals except for them often leading us up a garden path?

They seem like a good idea on the mostly small US dirt tracks and play a big part in handicapping there, but in OZ with the rails all over the show and the track times often different depending on very small differences in 'good' track condt. [even the length of the grass would make a difference], track records are often set in very middle of the road races.

Nice detective work jfc.
This is just another example of the crap data you have to contend with to handicap races. I compare data sources all the time and get different winning margins, weights carried, even jockeys from one place to another.

KV

Marcus - As someone who has been logging racetimes for ten years or so, just a query re the Rosehill 1400m (Good track conditions) as I have the standard time listed as 83.5s somewhat different to your own. Do you adjust your standard times in some way?

JFC - I think you will also find that the racetime for the winner will also be different to that of the winner by 0.05 (5/100ths) of a second because the timer devices are in the saddle of the horse while the racetime is judged as soon as the winner's nose crosses the line.

As an example in last Saturday's Orr Stakes, Sportscolour have the race time as 82.36s and the winner's time Perfect Promise as 82.41s

Thanks jfc but I don't think it is the rail position. Sportcolour state they adjust their sectionals for such things and going back to my example the difference in times was 1.37secs therefore the rail would have to be 27.4m out for the day.

La Mer it seems our track Par Time for 1400m at Rosehill is similar (mine 83.2s) - Depends on the class you choose to take average for.

__________________
Regards,
Pete