GSV takes on English Maidens

[George William Smith]

The data involved in all 913 maiden races ran in England in 2004 is now posted on my web site at
http://www.members.shaw.ca/angoss1/maidenenglishraces.xls

The following is a summary of the results, but feel free to examine the data for accuracy and objectivity as part of a good learning experience. Needless to say, the results surprised even me, but I am one happy camper.

I encourage all, whether skeptic or follower, to visit my web pages and learn more about the GSV and its limitations. Enter my site at:
http://www.members.shaw.ca/thematchmaker

Thanks to AP for doing a study of all English Maiden Races in 2004. During his study AP entered into a database for all English Maiden races the date, track, name of runner, finishing position, race value, race distance, race time, weight, odds, GSV, GSV of sire, GSV of dam, year of birth, sire. I added the GSV ranking of each horse in the race and how many runners in each race. Here is a summary of the study:
There were 913 maiden races in England in 2004.
The races took place at 35 racetracks
The average field size was 10.8 runners; most maiden races had from 6 to 14 runners, but fields ranged from 3 to 24 runners.
The only means of selection was the maiden's GSV score. No consideration whatsoever to the jockey, trainer, owner, morning line odds, past performances, etc.

151 races were won by the maiden with the highest GSV [16.5%].
119 races were won by the maiden with the 2nd highest GSV [13.0%]
110 races were won by the maiden with the 3rd highest GSV [12%]
88 of 911 races with at least 4 runners were won by the maiden with the 4th highest GSV [9.7%]
81 of 901 races with at least 5 runners were won by the 5th highest GSV [9.0%]
92 of 881 races with at least 6 runners were won by the 6th highest GSV [10.4%]
62 of 819 races with at least 7 runners were won by the 7th highest GSV [7.6%]
49 of 750 races with at least 8 runners were won by the 8th highest GSV [6.4%]
45 of 659 races with at least 9 runners were won by the 9th highest GSV [6.8%]
32 of 580 races with at least 10 runners were won by the 10th highest GSV [5.5%]
27 of 489 races with at least 11 runners were won by the 11th highest GSV [5.5%]
20 of 410 races with at least 12 runners were won by the 12th highest GSV [4.1%]

Using a $1 betting unit, the profit or loss due to betting solely on the GSV ranking appears below [note: no info on takeout; profit/loss adjusted for the number of races as given above]
Betting the highest GSV score in all 913 races showed a profit of $26.68. It held its profit in each block of 100 races with the exception of the first block which showed a loss of $39.21.
Betting 2nd highest GSV showed a loss of $345.76
Betting 3rd highest GSV showed a loss of $162.32
Betting 4th highest GSV showed a loss of $322.33
Betting 5th highest GSV showed a loss of $332.15
Betting 6th highest GSV showed a loss of $254.23
Betting 7th highest GSV showed a loss of $350.50
Betting 8th highest GSV showed a loss of $180.87 [with a 100/1 winner!]
Betting 9th highest GSV showed a loss of $276.16
Betting 10th highest GSV showed a loss of $208.96 [with a 100/1 and 50/1 winner]
Betting 11th highest GSV showed a loss of $295.62
Betting 12th highest GSV showed a loss of $230.75

The highest GSV won 151 races
The highest GSV was 2nd 107 times
was 3rd 99 times
was 4th 109 times
was 5th 89 times
was 6th 82 times
was 7th 63 times
was 8th 56 times
was 9th 41 times
was 10th 32 times
was 11th 24 times
was 12th 16 times

Profit/Loss by Track betting the maiden with the highest GSV
Ascot, only 6 races, all lost
Ayr, 17 races, all lost
Bath, 29 races, excellent profit
Beverly, 21 races, loss
Brighton, 21 races, slight loss
Carlisle, 10 races, small profit
Catterick, 18 races, loss
Chepstown, 17 races, good profit
Doncaster, 18 races, loss
Epson, only 9 races, slight loss
Folkstone, 18 races, slight loss
Goodwood, 29 races, huge profit
Hamilton, 21 races, loss
Haydock, 27 races, huge loss [only 1 winner]
Kempton, 29 races, loss
Leichester, 24 races, slight loss
Lingfield [all weather], 83 races, huge loss
Lingfield [turf], 11 races, good profit
Mussboro, 19 races, huge loss
Newbury, 33 races, loss
Newcastle, 20 races, good profit
Newmarket [RM], 60 races, huge profit
Nottingham, 32 races, loss
Pontefract, 24 races, good profit
Redcar, 19 races, huge loss
Ripon, 19 races, slight loss
Salisbury, 26 races, slight profit
Sandown, 17 races, huge profit
Southwell [all weather], 29 races, huge loss
Southwell [turf], 2 races, good profit
Thirsk, 20 races, huge profit
Warwick, 24 races, good profit
Windsor, 40 races, slight loss
Wolverhampton, 57 races, huge profit
Yarmouth, 26 races, even
York, 21 races, huge loss

[Seabird]

George (I hope you don't mind the familiarity)

What fascinating data!

Very interesting to see that:

40% of tracks produce worse than slight losses
20% of tracks produce slight losses
6% of tracks produce slight profits
and
34% of tracks produce better than slight profits

Can you pinpoint any factors which may contribute to this pronounced split between the profitablility of tracks? I know our English (BTW Ayr, Hamilton Park and Musselburgh are in Scotland and Chepstow is in Wales) are very variable and I shall be trying to do some analysis which looks into this, but I would like to hear your comments on this clear dichotomy as I know nothing yet about your GSV rating (I will soon be remedying that!!)

Just a few comments to clarify some information for any readers interested in some of the UK tracks mentioned:

Kempton Park is now Polytrack although it was turf at the time of the analysis. It will be interesting to see if future results alter from loss to profit.

Newmarket (RM) stands for Newmarket Rowley Mile course, Newmarket also has a July Course which I take it has no maidens run at it.

You mention 2 races at Southwell Turf, looking at the actual data I count up 23, a typo I thnk.

Wolverhampton, Lingfield Park and Kempton Park are all Polytrack now and Southwell (AW) is fibresand. The Polytrack at Wolverhampton was opened in October 2004 so I guess the data spans both its use when previously Equitrack. Lingfield Park swapped from Equitrack to Polytrack in 2001.

I look forward to your comments and congratulate you and AP on this interesting study.

[Colonel]

Thanks George, excellent data. Being an American, I assume most, if not all of the results were on turf? Thanks again for your excellent work. I would like to ask if you can, at some point, post another daily cards results.

Regards

[Pierre LP]

Kempton now has all its flat racing on a polytrack surface but it is completely different in nature to the other 2 polytrack venues being right handed and the false rail in the straight means that it is much fairer.

The quality of horses in maidens varies a great deal, and there are different types of maiden races:

Ordinary maidens where the fillies get a 5lb allowance
Fillies only
Median Auction maiden races where the weight is decided upon the median value of a sires progeny at the sales
Auction maidens where the cost of each horse at the sales decides the weight
Maiden claiming races where the trainer decides the weight
Maiden handicaps

There are possibly other types.....food for thought......

[George William Smith]

Thanks Pierre:
Good information from you. If you've looked at the excel file you'll see everything that AP noted about the race including purse, wt. etc.

George

[Pierre LP]

George

It will be interesting to see wether the demographic spread of GSV values for the 2yo runners stays the same over a few years.

Unfortunately I couldn't monitor it last year, but it would be easy to extrapolate from the excel file each year.

It appears to me that breeders/owners in the UK are looking for sharper earlier types rather than horses which are longer term prospects.

Regards

Peter