[mickjohnson]

I’m still getting all on the Tapeta surface faster, except the 2m119y distance (203.55 to 201.52 seconds, from three runs) They are settling down though.
Going coefficients are still almost identical.
The old Wolverhampton poly track figuers are WOLVHMPTN(PT). The new Tapeta are WOLVHMPTN(AW)
http://www.racingsense.com/standardtimes.htm
http://www.racingsense.com/standardtimes.csv

[mickjohnson]

I just analyse all the variables, the course, distance, going, jockey, trainer, last run, best run, time since last run, max speeds, slowest speeds, pedigree, everything.
All I look for are horses that are capable of winning. They have been fast enough previously, recently, to beat the other horses in a race. No trainer ‘is on form’ crap or ‘some famous race is coming up, oooooh!’ . If they have been quick enough recently, they`ll win again.
I am sufficiently right more times than I am wrong. More than enough to make a good profit.
Two winners out of three today does for me. The third was a second! (mutter… mutter…)

[mickjohnson]

I’m not sure where you mean to get the results from. You can download the web pages to your pc but to convert them to CSV takes a lot of programming.
I’ve been getting the results and declarations daily from the company that supplies the newspapers for 20 years and keep the database updated. I have written some code that will allow the export of any of the data into CSV format, which should do what you want.
I`ve now allowed free access to the download again after upgrading both my servers` broadband to fibre. It can be downloaded from http://www.simplesoftware.co.uk/ssr1.htm or http://www.racingsense.com/download.htm
I’ve just had a look at your other posts. The database contains distance beaten, speeds in MPH and my original calcultaion’s were done in LBS.

[mickjohnson]

I do all the UK and Irish races and calculate all the going and speed ratings daily and have done for the last 20 years.
I get a results feed in the morning from the company that supplies the newspapers and a declarations feed early afternoon with the next day`s declarations.
It is pretty much automatic but there were three races yesterday that had no times and I had to go to the Racing Post web site and use (steal!) theirs.
I`ve a 32% win rate this month so far. Not my best, but 65% ROI does for me!

[mickjohnson]

I’ve just learned something! I use all the times of the winners to calculate my going coefficients, then I use all the times of the winners to calculate my speed ratings. I hadn`t thought about striping out the runs of the `for some reason, they didn`t run as fast as they could` races. As long as you have enough runs to do the sums on, it doesn’t matter if you ignore a few iffy ones. It has got to be better.

[mickjohnson]

Oh yes it can!!!!

For all horses having IMPERIAL in their name Total = -£781.76. 1958 Runs.

[mickjohnson]

Did some stats on SAM.
For all horses having SAM in their name. £1 level bet.
Total = £1875.36 profit. 5874 Runs in 20 years. 1/4 odds for 2nd and 3rd. Flat and Hunt.
This is quite a useful tool. Bit of an eye opener!
11/10/2014
Found some errors in my code last night. Missed out the bit where a place win on an e/w bet loses the win bet bit.
Recalculated and got :-
The last bit of the file…
Horse Finished T/Odds Runs Running Total
WHERE`S SAM(IRE) 7 50 1 -£2526.60
WISAM 1 0.67 1 -£2525.77
WISAM 3 2.25 1 -£2526.21
WISAM 4 22 2 -£2530.21
WISAM 6 8 1 -£2532.21
WISAM 7 25 1 -£2534.21
WISAM 9 35 2 -£2538.21
WISAM 13 12 1 -£2540.21
WISAM 17 12 1 -£2542.21
YVE SAM LA CHAMP 6 22 1 -£2544.21
YVE SAM LA CHAMP 10 166 2 -£2548.21
YVE SAM LA CHAMP 12 100 1 -£2550.21

For all horses having SAM in their name Total = -£2550.21. 5875 Runs.

Last 20 years, £1 level e/w bet.

[mickjohnson]

I’ve modified my program to do it for any entered horse name.

MOHEET had just one run, a winner at 8/1.
I’ve tested and retested my code and found I`d omitted the bit where an e/w place win loses the win part. I’m happy with all the numbers now.
For all horses having MIDNIGHT in their name

Total = -£627.12. 2532 Runs in last 20 years

for a £1 level e/w bet.
I tried it with a block on e/w bets on odds below 4/1 and it made little difference.
The button in my program exports a CSV file with the full data.

[mickjohnson]

My initial thoughts were that these sorts of systems are probably not much better than using a pin! But I thought, in the horse breeding system, it is usual for the foals of horses to have part of their parents name in theirs. There may be something in tracking parts of names!
I ran a query in my database on the last 20 years of looking at all the horses with MIDNIGHT in their names on the flat.
It looked pretty good in the early years but, for a £1 level bet to win on 1213 races, it came up with a loss of £74.59p, which is not too wild. With a look at the form at the same time, the obvious donkeys could be removed.
I’ll add it as a function to my program and include the Hunt races and sums for e/w bets as well.
I’ll then have a look at all the other ‘names’ in this section.

[mickjohnson]

Many years ago, there was a scam in Australian races where the delay in getting the results back to the UK bookies meant that UK punters could get a bet on a race when they new the result, if they had a phonecall from a friend in Oz before the bookies did.
With the internet, the timings are tighter, but the same method applies.
If you can watch a race at the course and get a bet on via the web during the run in, before the bookie closes bets, you have a better chance of winning. The odds will be shorter and shorter still the longer you leave it.
Obviously the exchanges know it too. I would guess that they freeze the bets before they officially say they do, to protect themselves. It is not beyond belief for them to falsely shorten odds rapidly on a horse that is not going to win, to catch running in bettors. You can see the amount of money that is bet very close to the finish.
All it needs is an employee at each meeting to push one of a couple of buttons on an app on his mobile.
The freezing the bets before they officially say they do happens on every race. The other, two or three times each meetings that I have spotted, but I only go to the meetings in my local area.

[mickjohnson]

I’m still getting Tapeta faster than the Wolverhampton old poly track, except the 2m119y, after 13 meetiungs:-
Yards Dist ST Seconds Runs Course PolytrackST Difference(-Tapeta Faster)
1120 5f20y 54.96 13 WOLVHMPTN(AW) 56.29 -1.32
1316 5f216y 67.09 18 WOLVHMPTN(AW) 67.12 -0.02
1572 7f32y 79.75 29 WOLVHMPTN(AW) 81.1 -1.34
1901 1m141y 98.03 19 WOLVHMPTN(AW) 100.37 -2.33
2083 1m1f103y 108.03 9 WOLVHMPTN(AW) 110.83 -2.79
2690 1m4f50y 142.11 10 WOLVHMPTN(AW) 145.7 -3.58
3054 1m5f194y 160.28 3 WOLVHMPTN(AW) 165.77 -5.48
3639 2m119y 203.76 2 WOLVHMPTN(AW) 201.39 2.37

[mickjohnson]

It does go back to Napoleon!
I only use 20 years data to analyse going corrections and course/distance coefficients so the way I’m doing it now is manageable. The ‘new’ horses, by definition, have little form so will probably not be assessed in a future race.
I use recent data on speed, adjusted for going, weight, etc., so pedigree will effect the posible performance of a horse but I measure the actual performance.

[mickjohnson]

I get my results and declaration data daily from a company that supplies the newspapers and the horse number is not included. I have loads of screen scraping stuff running on my Linux servers but none get that info either.
Mick Johnson

[mickjohnson]

After 11 meetings, I have my Standard times on the Wolverhampton Tapeta as:-
Distance Yards ST(secs)Runs
5f20y 1120 55.34 12
5f216y 1316 67.39 15
7f32y 1572 79.85 22
1m141y 1901 97.11 13
1m1f103y 2083 107.09 6
1m4f50y 2690 142.12 10
1m5f194y 3054 161.98 2
2m119y 3639 201.64 1
These as all faster than the Poly Track bar two which are onlt 0.5 slow, but we need more runs. I expect they will speed up too in time.
All standard times at:-
http://www.racingsense.com/frequentquestions.htm
Mick Johnson
Simple Software Racing

[mickjohnson]

I stumbled accross this post by accident but I have a useful reply, I hope. The problem with all the different Jockey names has been around for a while. It made it very hard to run queries looking at Jocket performance. I run a large program on the data in the Racingsense database every week or so to try and convert them to a standard form:- Surname Title Initials. It can’t be done locally on the database as it needs a bit of human intervention on the logic. Customers can download an up-to-date database including the changes, whenever they want.
Mick Johnson,
Simple Software.

[Author]

Posts