|
RaceBase will no longer be regularly updating the prediction statistics There are 3 main areas of use for RaceBase : 1) Jockeys,
trainers and other industry professionals who use it as a
database and analysis tool. All 3 groups should find the predictions useful. The statistics presented here should help to give an indication of their success although they are mainly aimed at those who wish to use RaceBase as a punting tool. In RaceBase there are 19 different factors the program uses to award points when making its predictions. The user is able to alter the amount of emphasis placed on each factor and consequently the way the points are awarded. In effect a user can input his/her own prediction equations. RaceBase now comes with 2 already pre-programmed alternative settings. The Margins method is the setting that over the past 8 years has produced more winners. This method has the best strike rate. The Value method, over the same period, has shown to produce less winners than the Margins method, but because it gives higher ratings to horses lower down in the weights, it tends to show a better profit. The predictions were last improved on February 18th 2001 and subsequently introduced with RaceBase release Version 01.02 or higher. Statistics up to February 17th 2001 are therefore retrospective only, and represent what would have happened if you had put $1 to win on every top rated runner if the new predictions had been applied. They should be the same predictions you will get when doing analysis in the Archive Fields module of the RaceBase program. The table below represents the performances of the top rated runner in each of the methods described above between 27th October 1997 and 28th February 2007. The only races that are left out are races where more than 1/5 of the starters in a race are horses that are having their 1st start at a totalisator meeting (eg 3 first starters in a 10 horse field).
You can considerably improve on these performances by narrowing down your selections and betting only when certain criteria are met. This can include things like considering the horse ages, combining the predictions with good time ratings or looking at the dividends on offer. From time to time statistics are posted to the RaceBase e-mail list that may help you with this. If you subscribe to RaceBase, you can also do this type of analysis for yourself. You are able to do predictions back to October 1997, determine profitable selection criteria, work out systems, and even formulate and use your own ratings equations. Races where more than 1/5th of the total starters are horses having their first start at a totalisator meeting have been excluded from the following statistics. Margins Method Click here to view all Margins selections between Oct 27th 1997 and Nov 22nd 1998 Click here to view all Margins selections between Nov 25th 1998 and Dec 19th 1999 Click here to view all Margins selections between Dec 22nd 1999 and Dec 3rd 2000 Click here to view all Margins selections between Dec 6th 2000 and Nov 30th 2001 Click here to view all Margins selections between Dec 1st 2001 and Jan 11th 2003 Click here to view all Margins selections between Jan 15th 2003 and Jan 31st 2004 Click here to view all Margins selections between Feb 6th 2004 and Jan 31st 2005 Click here to view all Margins selections between Feb 3rd 2005 and Feb 26th 2006 Click here to view all Margins selections between Mar 1st 2006 and Feb 28th 2007 Value Method Click here to view all Value selections between Oct 27th 1997 and Nov 22nd 1998 Click here to view all Value selections between Nov 25th 1998 and Dec 19th 1999 Click here to view all Value selections between Dec 22nd 1999 and Dec 3rd 2000 Click here to view all Value selections between Dec 6th 2000 and Nov 30th 2001 Click here to view all Value selections between Dec 1st 2001 and Jan 11th 2003 Click here to view all Value selections between Jan 15th 2003 and Jan 31st 2004 Click here to view all Value selections between Feb 6th 2004 and Jan 31st 2005 Click here to view all Value selections between Feb 3rd 2005 and Feb 26th 2006 Click here to view all Value selections between Mar 1st 2006 and Feb 28th 2007
The statistics below are more comprehensive compared to those above in that : 1) They provide a month by month breakdown. 2) They include all races where there are at least 3 horses with any form (ie at least 3 horses that have had at least 1 previous start). This means that only 2 or 3 races (mainly 2 year old races) are excluded from each month. 3) The statistics are for only the race predictions that were actually out in the market place as subscribers had them and had they used the MARGINS or VALUE prediction methods.
In MARGINS method, RaceBase therefore correctly selected the winner 26.1% of the time over the past 5 years. 43.3% of all winners were in the top 2 rated runners, and 56.5% of all winners were in the top 3 picks. The average human punter loses 16% of all money invested. Past RaceBase statistics show that the RaceBase computer program has performed considerably better than the average human. Although the statistics presented here are accurate, RaceBase should not be viewed as a get rich quick scheme. It is not proven that the predictions will continue at the same level in the future. I prefer to think of the predictions as only part of the program and as an aid along with the other database capabilities, although there are some, including myself, who use them as the basis for a mechanical punting system. RaceBase offers a free and uncensored e-mail discussion group with open exchange between subscribers. By default, there cannot therefore be any manipulation or misrepresentation of the RaceBase prediction performance statistics. Terms Of Use : The ratings are for the use of the registered subscriber only. RaceBase retains copyright control of the predictions (ratings) at all times. Forwarding of upcoming RaceBase race ratings to non-subscribers is strictly forbidden. Disclaimer : The statistics on this page are accurate but they are only a reflection of the past performance of the ratings. We have no control over their future performance. Subscribers should treat it as a gamble. Use of the ratings are at the subscribers discretion and RaceBase forfeits any responsibility for any losses that may occur.
|