Ratings Information & Help

Contents

1. Getting Started
2. HRB Example Set
3. Creating New Ratings
4. Backdating Ratings
5. Editing Existing Ratings Settings
6. Help With Settings

Creating horse racing ratings can be a simple task or one of extreme complexity - taking into account each individuals opinions on the factors of relevance in compiling a rating only increases the difficulty. No matter what formula is used, manually applying the settings to each horse in a race is time consuming and tedious work.

The purpose of the Ratings Machine is to provide a platform for all horse racing enthusiasts to create horse racing ratings to suit their needs and a presentation area to call upon and instantly view the compiled ratings. It has been designed in line with members feedback and assistance.

The first thing to be aware of is exactly what a rating is and this is covered in the ratings basics article.

With the basics mastered you will now wish to create some ratings! The Ratings Machine V2 provides a step by step platform for you to record your setting preferences which will be used to create the ratings for each race. At every step there are help links to talk you through the stage you are at, the purpose of the settings and calculations used. It is not completely necessary but reading through this help section in its entirety would be of huge benefit before you begin.

The main purpose of the example set is to illustrate a working demonstration of the machine and to give people who feel they could benefit from one, a starting point.

The example set is available to all members. These should be treated carefully as it is extremely difficult indeed to produce a set of ratings to cover every single race. Additionally the example set do not cater for any criteria which may be open to change, therefore offering no score on many factors which, if applied correctly, will certainly increase the potential of any ratings set you may wish to create (example jockey etc). Another point worth mentioning is because the example set is being applied to all races, settings such as Speed and Stall have been left blank as they would be useless in a National Hunt scenario. The same theory applies to some of the other settings left blank.

Despite these issues the example set do offer an insight into the race and can be very useful indeed for narrowing down the field. But clearly if you look into specific types of races and apply relevant and detailed settings it is highly likely they will be more effective.

To create a completely fresh set of ratings choose the option from the Main Menu by clicking your mouse in the circle to the left and then pressing the blue Go button. You will then be advised to give your new ratings set a name - this is very important as the name you choose will be the indentifier for applying your ratings. The name chosen is entirely up to you but only letters and numbers should be used and other characters on the keyboard may return an error. A good name will be something recognisable such as '3yo Handicaps' or 'Veterans'.

As creating a brand new set of ratings involves applying settings to all scenarios it is quite common to use an already existing ratings set as a template and then make amendments to these settings. To create ratings from a template the process is almost identical to that described above however you will instead choose the 'based on a template' option from the main menu.

At the point of naming your ratings you will also be presented with a dropdown menu listing ratings sets available to you as templates - these will include all of your own ratings sets plus communal sets available. I have created several ratings sets, which are communal to all members - these are identifiable by the prefix HRB.

Once you have chosen a name (and if appropriate a template) click the blue Go button and you will receive confirmation the settings have been created and will automatically be taken to the edit area so you can begin to apply your preferences.

The ability to backdate ratings settings is a massively powerful function, meaning you can tweak ratings to suit the circumstances and test as you go. It is vital to be aware that when backtesting, all criteria of the race is definitely known, whereas advance ratings are obviously based on what is known at the time.

There are two areas whereby you can backtest. The first is on any race, by viewing the race result and choosing the 'Backdate Rating' option. You may then apply any ratings set to the known race criteria and see where the top rated working downwards actually finished.

The second option is by mass backtesting the last 500 races matching any criteria of your choosing. This could mean including all races, handicaps, claimers, favourites or whatever you deem worth investigating. To test all races there is an option within the main Ratings Machine V2 layout. To add in criteria, use the System Builder and when you have loaded the criteria, choose the 'Rating Test' option from the main screen.

You may find yourself wishing to tweak and adjust your ratings settings in line with the ratings being produced and other factors such as their performance in races you are applying them to. To do so choose the 'edit existing ratings settings' option from the main menu and press Go, choose the relevant ratings setting from the dropdown and again press Go.

All rating calculation settings are applied by using dropdown menus. By reading the description of each setting you will then decide on and apply a value by selecting a figure from the dropdown. Each setting is bracketed into a group and these can be navigated by using the instructions at each stage. Within each of the settings areas there is a blue Change button at the bottom, which when clicked will store your settings in the database.

In order to calculate a rating the Ratings Machine must perform several functions and its operation has been designed to be instant. From a user perspective the complexity and coding in operation behind the scenes need not be considered other than to appreciate the fact settings do need to be separated out and follow some sort of general ruling and flow in order to produce a final rating for each horse.

Each area and setting calculation is explained within the relevant help content but at a very top level the ratings are broken into 3 groupings. Each of the these is detailed below.

Race Score

Factors pertaining to each individual horse race, which are shared by all those who competed in it such as the prize money on offer and the type of contest.

Horse Score

Factors concerning each horse who competes in a race, which are individual characteristics such as the position finished and weight carried.

Additional Factors

All remaining settings, which are not bracketed into the two above groupings.

To be able to assign each horse race a quality score, ten brackets have been created and every race ran in the UK/Ireland falls into one of these. As there is no class system in place in Ireland, a method is used based on prize money to slot each of the races into an appropriate bracket.

Starting with Group or Grade 1 races and going all the way down to Class 7 races it would appear to make sense to apply a higher score to the better quality brackets but this is of course at your own discretion.

This setting is used to translate the value of each horse race by dividing the prize money awarded to the winner by 1000 and rounding the value then multiplying this figure by your setting.

The maximum rounded value before applying your setting is 20 (therefore a race worth 500,000 would still have a score of 20). This is because as the race value gets higher it becomes extremely hard to control and races of a high value automatically always giving higher ratings regardless of placings. To overcome this issue you should use the Class Settings and award values dependent on Group/Graded/Listed status etc.

For some examples assume a setting of 5 -

• A race worth £8,345 would be assigned a value of 40 (rounded prize money of 8 * 5).
• A race worth £15,000 to the winner would have a value of 75 (rounded prize money of 15 * 5).
• A race worth £950 to the winner would have a Prize Money Translator value of 5 (rounded prize money of 1 * 5).

The Top OR score adds on a figure calculated by multiplying the highest official rating in the race by your setting. For example if your setting is 2 and the highest official rating in the horses race was 82 then 164 would be added to the race score. It is vital you are aware in races where there are no horses with an allocated official rating then the top OR is 0

The Win OR score adjusts the race score by adding on a figure calculated by multiplying the winners official rating in the race by your setting. For example if your setting is 3 and the winners official rating in the horses race was 90 then 270 would be added to the race score.

The purpose of the Prior Form settings is to boost the race score in line with the previous form of the horses competing in the race. This is done by multiplying the summed number of wins and places of all horses in their previous race by your settings. As an example if there were 4 horses in a race who had won last time out and 6 who had placed and your settings were Prior Wins - 2, Prior Places - 3 then 26 would be added to your race score.

Subsequent Form settings work in the same way as the Prior Form, the difference being any horses who have won or placed on their next outing are counted. An example would be if 3 horses from the race had won next time out and 8 had placed with your settings being Subsequent Win 0.6 and Subsequent Place 0.8 then a further 8.2 would be added to the race score.

In certain situations you may deem a handicap or non handicap race to be less deserving of a rating in the context of the race you are evaluating. This setting allows an adjustment to be made for these circumstances. By setting a percentage figure for both handicap and non handicap races the Ratings Machine can make an alteration on this basis.

As an example if two horses had a race rating of 200 going into this stage however one was based on a handicap and the other on a non handicap and your settings were 100% (handicap) and 90% (non-handicap) the new figures taken forward are 200 for the horse with the handicap race form and 180 for the horse with the non handicap form.

A further adjustment to the race score can be made dependent on the sphere of racing. This can be useful if you believe for example National Hunt form should not be treated with the same ruling as Flat form when judging a Flat race. By placing percentage values in each of the possibilities the Ratings Machine will be able to automatically adjust the race score in line with your settings. For example if you had set 80% as the adjustment for Hurdle races and the race score was 150 the new race score would be 120.

To only include races of exactly the same sphere (therefore meaning horses last ten runs will only include a database search for like for like race types) use the Like For Like dropdown - this will make the other sphere settings irrelevant as all races evaluated will be of the same type.

Obviously the finishing position of the horse is of vital significance and by applying values to different scenarios we can make an adjustment to the race score dependent on where the horse finished.

There are two ways of applying a placing percentage - either on the basis of an exact placing (first, second, third etc) or by using a placing ratio dependent on finishing place in comparison to number of runners (because 2nd of 3 is in theory not as good as 2nd of 12). The Ratings Machine shall use only one method, working on a first applied basis looking at Exact Placing settings and if there are no set values stored then Placing Ratio settings will be used.

For the distance beaten adjustment there are 3 ways of calculating your deduction, again the Ratings Machine will operate on a first applied basis checking through each setting and applying the first recorded. More details on each of the position settings is below.

Horse Score - Position (Placing)

You may decide you wish to award a rating on an exact finishing position basis. By doing so you would apply values to each of the settings as required with logically the better the finishing position the higher the setting. If you opt to use this method of calculating a placing score then the Ratings Machine will look up the finishing position of the horse and amend the race score accordingly. For example with a race score of 100 and a setting of 59% the new race score is 59.

Horse Score - Position (Placing Ratio)

The second option for amending the race score is by using a placing ratio. This method does not only look at the finishing position of the horse but also the number of horses in the race. Logically the higher the place ratio the higher the setting should be. For example with a race score of 200 and a setting of 9% the new race score is 18.

Horse Score - Position (Non Finishers)

Non finishers can be troublesome to evaluate as there is no distance beaten figure which can be used to amend the race score. Therefore one overall deduction setting is used to apply to all non finishers. As an example if you had set this figure to 10% then a race score of 150 would be amended to 15 for a non finisher.

Horse Score - Position (Max Distance Beaten Deduction)

When making a distance beaten adjustment it may be the case you do not wish to go too far. This setting allows you to manage the deduction made by setting a percentage or numerical cap on the maximum amount to be deducted. Using a scenario where a horse was beaten by 50 lengths (and you were using exact distance beaten settings) and their race score was 45, if you had a Max Distance Beaten setting of 80% in place then instead of their rating becoming a negative (45-50) it becomes 9 (45-(45*80%)).

As another example if your settings indicated a maximum numerical reduction of 12 then no more than this amount will be deducted irrespective of how far beaten.

Horse Score - Position (Dist Bt Option A - Lbs/Length Scale)

The lbs/length scale is a standard way of calculating a deduction to be made from each horse. This is done by multiplying the distance beaten by a figure dependent on the distance of the race (as detailed below) and then taking this value away from the race score.

• 5f -> exactdistbt*3.5
• 5.5f -> exactdistbt*3.25
• 6f -> exactdistbt*3
• 6.5f -> exactdistbt*2.75
• 7f ->exactdistbt*2.5
• 7.5f -> exactdistbt*2.25
• 8f -> exactdistbt*2
• 8.5f -> exactdistbt*1.9375
• 9f -> exactdistbt*1.875
• 9.5f ->exactdistbt*1.8125
• 10f -> exactdistbt*1.75
• 10.5f -> exactdistbt*1.6875
• 11f -> exactdistbt*1.625
• 11.5f -> exactdistbt*1.5625
• 12f ->exactdistbt*1.5
• 12.5f -> exactdistbt*1.4365
• 13f -> exactdistbt*1.375
• 13.5 -> exactdistbt*1.3125
• 14f -> exactdistbt*1.25
• 14.5f ->exactdistbt*1.1875
• 15f -> exactdistbt*1.125
• 15.5f -> exactdistbt*1.0625
• 2m+ -> exactdistbt*1

Horse Score - Position (Dist Bt Option B - Exact Dist Bt)

By using the exact distance beaten settings you will be deducting the number of lengths the horse was beaten from the race score (and if chosen multiplied). For example if you have chosen exact distance beaten and race score is 78 and the horse was beaten by 17 lengths then new score is 61.

Horse Score - Position (Dist Bt Option C - Personalised Percentage)

The personalised percentage method works on a rule basis with the Ratings Machine checking to see which rule applies to the evaluated horse. You set up the rules by using your logic and may for example wish to award 90% of the race score if the horse was beaten by less than a length but award only 60% if the horse was beaten by 10 lengths or less. The final rule is used for all horses whose distance beaten does not match any other rules.

The date of the horses runs can be an important factor and you may wish to reduce the race score if for example a run being judged took place over a year ago. By setting up some date rules you can award a higher percentage to races within a certain timeframe over those within another. For example if the race score was 120 and the relevant rule stated a percentage setting of 50%, the race score would be adjusted to 60.

As a special setting for Rule 1 you may specify to only apply this rule once, the reason being you may like to award a higher score to horses who have ran within a specified number of days, but not wish to amend the score dramatically if they have actually ran several times during this period. For example if your rule 1 was award 100% for runs within 30 days your intention may be to boost the scores of horses to have run within this timeframe however if a horse had actually ran 4 times in the last 30 days and you have set 'Apply All' then all 4 of those runs would be given this 100% setting. By specifying 'Apply Once' your ratings will be calculated applying the 100% to the most recent run but using rule 2 for all others within the specified timeframe.

As up to the last 10 pieces of a horses form can be rated you must decide whether the more recent runs should be treated differently to the more distant ones. These percentage settings allow us to control this. As an example you may decide the horses last run should be given 100% (meaning it will remain the same) whereas the second last is only to be awarded 80%.

A problem when creating ratings can be when compared horses have not competed the same number of times, for example if you are rating the ten last races of each horse but some of the field have not raced this many times then it is highly likely the accumulated ratings of more experienced horses will be significantly higher. The Inexperience Leveller setting tells the Rating Machine how to treat such a scenario. Using the example of a horse who has ran only 8 times and the presumption you have applied the setting to be Use Average Form Score, the Ratings Machine will then use the average score of its last 8 runs to perform calculations for the 9th and 10th last ratings. The same logic and approach will be used and your settings for 9th and 10th Runs Ago will be utilised.

An adjustment to the race score can be made by analysing the weight the horse carried. By looking at the horses weight compared to the winner and the top weighted horse you can amend the score accordingly.

This setting is multiplied by the weight difference between the evaluated horse and the top weighted horse in the appropriate race. For example if the horse received 10lbs from the top weight then your setting is muliplied by -10 and the race score amended accordingly.

Your setting here is multiplied by the difference between the weight of the winning horse and the horse being rated. In a scenario where the horse gave 4lbs to the winner your setting would be multiplied by 4. In this example if your setting was 6 then the race score would be increased by 24.

Creating speed ratings is notoriously difficult and when and how to create and apply them is a subject of great debate. The Ratings Machine will create a speed rating automatically for 18 different scenarios all of which are performed only on races the horse has taken part in within 2 furlongs either side of todays race, your settings will decide how important each of those scenarios is. Each actual speed rating (which your multiplier uses) is calculated using averages in the class, distance, going, track and all combinations of these factors. The process is complicated however it is worth noting the more data there is to go on (in other words how many similar types of race have been ran since 1997) the more accurate the assessment can be made and this is taken into account when HorseRaceBase applies a score. (Meaning in a scenario where there is little data to go on, less weighting will be given compared for example to a common distance/going/class/track).

By awarding an importance factor of between 1 and 10 for each of the settings, the Ratings Machine will then perform a further calculation using each multiplied setting by speed rating and producing a final speed rating based on the average. For example if the speed rating for Avg Distance Rating was set at 8, Best Distance Rating was set at 4 and Last Speed Rating was set at 3 with all other settings left at 0 then the final speed rating would be calculated as ((Avg Distance Rating * 8)+(Best Distance Rating * 4)+(Last Speed Rating * 3))/15.

To formulate a bonus rating for the jockey, trainer, stallion or horse the Ratings Machine uses the strike rates they have achieved in various scenarios and creates a managable figure from this between 0 and 10. By applying a multiplying factor to each of these figures a total rating for each of the factors is reached.

Because the number of actual rides/runs can dramatically affect how good the strike rate actually is (example is 1 win from 2 rides as good as 8 wins from 16 rides) each of the strike rate multipliers can be set up to 3 rules. These work on a first come first applied basis and you should apply rules from most to least.

As an example say you wish to award a multiplier of 3 when there have been 10 or more rides, multiplier of 2 when 5 or more and multiplier of 1 when there has been 1 or more rides. You would input the data in this order as if you entered it in reverse the first rule would mean the others were ignored (as stating 1 or more ride covers all scoring scenarios). The last rule (noted by 'else *') will be called upon to multiply all scenarios not covered by either rule 1 or rule 2.

The weight being carried by the horse in todays race can be of huge significance, these tests allow you to alter your overall rating dependent on your settings. Each of the tests store a value with the relevant information in and this figure is then multiplied by your setting. For example your settings may dictate that the weight received from the top weight is to be multiplied by 6 therefore a horse receiving 7 pounds would have a score of 42.

The VDW method is a widely discussed concept with some people even questioning whether its based on a myth. Nonetheless enough people have asked to include the option that if you wish to you can create a multiplier setting for each of the main VDW methods.

VDW1 works out an ability rating by adding all prize money won and dividing by total wins. This figure is then divided by 100, rounded and multiplied by your setting.

VDW2 simply adds up the horses last three finishing positions but with all runs of 10th or worse counting as ten. To make the rating in line with the machine where higher equals best, this total figure is then deducted from 30. So the best possible score is 27 multiplied by your setting.

Based on statistical analysis the draw at certain tracks does indicate a slight bias, the Ratings Machine looks up a stored procedure to test the stall and produce a score out of 10 in terms of benefit, your setting is then multiplied by this to provide the stall rating.

The main purpose of the ratings display is of course to provide you with the ratings for each horse in each race in line with the settings you have created. The default screen will always be the Totals page which will show you each of the ratings sub-totals and the overall rating for each horse. Each of the sub-totals can be broken down further by selecting the appropriate option from the 'Change View' dropdown. At any stage if you are unsure what a figure relates to, hover over the appropriate column and a description will appear.

If you have numerous ratings settings stored you may wish to change the ratings in line with another setting. To do so simply select the appropriate ratings settings from the 'Change Settings' dropdown.

The download to excel option will place all ratings in a spreadsheet for further analysing on your part.

The display is initially sorted by the total however you can sort the data in any way you wish by clicking the column heading of your choice.

This page loaded in 0 seconds.