Sending System Selections to Betfair

Bet Selector was updated mid 2023 with the ability in System Selector to create a special csv file for betting system selections with Betfair (note this works in both normal and expert modes). The ability to format selections for Betfair is an exciting development because as you may know Betfair’s Starting Price (BSP) is around 12% better than best tote. And in case you don’t know, Bet Selector’s top class and weight rated horses in selected races make a profit even at fixed price let alone best tote or BSP.

The new selection files for Betfair can be found in the “bets” subfolder in the “Price” folder. They are created after you run a system test and a key feature of them is that they identify selections by combining the horse tab number and horse name. This should make the files suitable for use with a number of third party Betfair bots but we have specifically formatted the file for use with BFBotmanager – click here for more info on BFBotmanager including a free trial.

When used with BFBotmanager, the files will work without the need for any modification but if you are already using another bot you may need to rename some of the columns or at the very least instruct your bot as to what the column names are and what they mean. With BFBotmanager you do not have to worry about that because the column names follow the BFBotmanager requirements.

Following is the listing of a sample file:

Provider,SelectionName,MarketType,BetType,Size,MinPrice
AVB1 SYS,3. Legal Esprit,WIN,BACK,1,4.33
AVB1 SYS,9. Rubiquitous,WIN,BACK,1,3.63
AVB1 SYS,2. Foxy Frida,WIN,BACK,1,3.93
AVB1 SYS,3. Military Expert,WIN,BACK,1,3.31
AVB1 SYS,2. Prince Of Boom,WIN,BACK,1,4.55
AVB1 SYS,2. Golden Boom,WIN,BACK,1,5.49

A file like the above will be created in the “bets” subfolder whenever you run a system or multiple systems within a folder/list. The name of the file is the same as the folder/list name but with “.csv” on the end to indicate that it is a comma separated vales (csv) file.

Note the file is only created if the system test start date and end date are the same. Also, files that were created before today are automatically removed from the folder.  Note you can still look at files for previous days provided you run the test today and only test one single day at a time (so you can test yesterday and still check out the file today).

The first column of the file (Provider) contains the system name. If you have only run one system then there will only be one system name listed (like in the example above) but if you have run multiple systems (either by using “Operate on All Systems in List Box” or by ticking “Test Next”) then multiple system names will appear in this column.

Having selections from multiple systems in the one file keeps things simple as you don’t have to worry about importing multiple files into BFBotmanager but the other big advantage is that in BFBotmanager you can set up different strategies for different systems even though they all come from the same import file.

The second column in the file (SelectionName) identifies the selection using its tab number followed by a period followed by a space followed by the horse name.

The third column in the file (MarketType) identifies the market type and defaults to WIN for win betting but if you put the letters PLC in your system name it will be changed to PLACE for place betting.

The fourth column in the file (BetType) identifies the bet type and defaults to BACK for backing a selection but if you put the letters LAY in your system name it will be changed to LAY for lay betting.

The fifth column in the file (Size) is the bet size and defaults to 1 unless you change the “Bet Size” on the system test dialogue from 1 to some other number like 5 or 10. Also you can bet to prices (BTP) by placing the letters BTP in your system name – bets will then be equal to 100 divided by the rated price from the fifth column (multiplied by the “Bet Size” on the system test dialogue if change from 1).

The final column in the file (MinPrice) is the minimum price at which bets can be placed on the selection. This defaults to the (original time rated) price displayed alongside selections but can be changed by using a MyCalc and setting the first letters of the calcname (or calccode) to AVB, PP or WGT (these select the main class, main time or main weight rated prices respectively). To distinguish between main, better or plus rated prices you can set the first letters of the calcname (or calccode) to AVB1, AVB2, AVB3, PP1, PP2, PP3,WGT1, WGT2, WGT3 (1 for main, 2 for better and 3 for plus).

NOTE: The use of codes in the calcname as just mentioned is designed so you can set the rated price to be used but still use the MyCalc for other purposes. If you use a MyCalc but do not use the above naming conventions then the actual calc value will be displayed in the system selection list and go into the MinPrice column. This is handy because it means you could set the minimum price (for example) to 1.5 times the rated price by using a MyCalc of 1.5*AVB1P (for 1.5 times the main class rated price).

IMPORTANT: If you intend to use them, please check the values in the bet size and/or minimum price columns are to your expectation. Note there is an option in BFBotmanager to ignore the bet size and/or minimum price columns so you do not have to use them. For example, you can easily set up a strategy with a different bet size or with a flat minimum price like $5.

New Features in Pro 3.0

Bet Selector Professional 3.0 was first released in July 2021 with new weight ratings in addition to its well established class and time ratings. The new weight ratings can be viewed in Price Predictor and can also be tested in the Expert Mode part of the new Professional 3.0 program.

Bet Selector Professional 3.0 underwent a major upgrade in January 2022 with the inclusion of trainer ratings to compliment the jockey ratings introduced in July 2020 in Pro 2.0 and 3.0. Just like jockey ratings, the trainer ratings are based on strike rates from the most recent 12 months adjusted for race class and range from 1 for the best trainers through to 9.9 for the worst. 

Another major addition was made in August 2022 when support for quadrella betting was added to both Race Day Live and System Selector.

Documentation on all these features can be found below.

Viewing the new weight ratings

The new weight ratings can be viewed on the main Price Predictor screen but for more detail are best viewed in the Extra Form HTML form guides as follows. Just go to Price Predictor off the pull down RUN menu in the program, select a meeting and click the VIEW EXTRA FORM button.

When looking at the field section for a race you will see three sets of ratings, first the Class Ratings, next the Time Ratings and finally the new Weight Ratings. Note that within each set of ratings there are Main, Better and Plus variants giving a total of 9 sets of ratings (3 for Class, 3 for Time and 3 for Weight).

Looking further down the page you will see the past form lines for each horse with ratings in the far left columns. You will see three columns of past performance ratings labelled as follows:

CRTG - past class ratings
TRTG - past time ratings
WTRG - past weight ratings (new)

Further to the right you will see the following headings:

AVB - the class value of the past race used to calculate past class ratings
DEF - the default class value of the past race used to calculate past weight ratings (new)

The AVB figures are the familiar AVBASE class figures used for class ratings (they can be thought of as a field strength based on the average of the time performance ratings of all horses in the race). The default DEF figures are based on a new table of classes that are hard coded into the program - they are the secret to the success of these new ratings and are based on the official race class and adjusted for things like day of the week, venue, prizemoney, age restriction and more.

System testing the new weight ratings

The new weight ratings can be tested in the Expert Mode part of System Selector. The variables available for this are as follows:

WGT1R – New WGT Main Ratings Rank
WGT2R – New WGT Better Ratings Rank
WGT3R – New WGT Plus Ratings Rank
WGT1V – New WGT Main Ratings Value
WGT2V – New WGT Better Ratings Value
WGT3V – New WGT Plus Ratings Value
WGT1P – New WGT Main Ratings Price
WGT2P – New WGT Better Ratings Price
WGT3P – New WGT Plus Ratings Price

So to test for the top Main weight rating horse you would use this rule:

WGT1R=1

If you try the above test you will confirm these new ratings are just as good as the existing ratings striking at just over 26% which is roughly the same as the class ratings and slightly better than the time ratings. Of course they do not always pick the same winners and this is the real benefit of using them as you will be able to pick more winners than users of older versions of the program.

Note also there is a new variable AVBDEF which gives you the default class value for each race as used for the weight ratings.

Trainer Ratings


Trainer ratings are displayed in the Price Predictor part of Bet Selector Pro 3.0 via "VIEW EXTRA FORM" where they appear to the right of the trainer names. They can also be viewed and used in system rules in System Selector Expert Mode. 

IMPORTANT: Trainer ratings are only available from 1 January 2022 onwards and only when form has been downloaded into the Pro 3.0 version dated 1.1.22 or later.

The new variables you can use for trainer ratings are as follows:


TRAT – trainer rating for each horse
TOPTRAT - top trainer rating for the race
TRANGE – range in kgs from top horse in race to current horse

Following are a few examples. First up, the rule for picking trainers rated between 1 and 2 (the very best trainers) is:

Within(TRAT,1,2)

If you want to pick horses where the trainer is rated within 1 kilogram of the top trainer in the race, the rule for this is:

TRANGE<=1

Note you can use TRANGE=0 to pick the horse with the top trainer in the race and this will include more than one horse if they are rated equal first.

Finally, the TOPRAT variable is less commonly used but can be used as a race selection tool. For example, to only consider races where the top trainer is no more than 3 use this rule:

TOPRAT<=3

System Refinery

New in January 2022 is a "System Refinery" feature that can dramatically speed up system tests. Some system tests can go from 2.5 minutes down to 30 seconds or even less. And this will work in both expert mode and normal mode!

This new feature allows you to “refine” a system by adding additional rules and so long as these rules are “within the bounds” of the previous test, the refined system will potentially run much faster.

To be clear, “within the bounds” means that if you first test for distances 1000m to 2000m you can subsequently test for 1200m to 1600m because these are “within the bounds” of the initial values (but you can’t test for 1000m to 2400m as these values are outside the bonds of the original values). This requirement to have rules “within the bounds” of the previous test only applies for this new System Refinery feature which can speed up your system tests. It does not apply to normal system testing.

This new feature is only implemented in special folders/list names that have the word “refinery” in them (note REFINERY, Refinery and refinery will all work). Once you have set up such a folder, you put your systems into this folder with the main one at the top and then subsequent “refined” ones get added one after the other to the list.

All systems in the list must be run over the same date range which is set by the date range you use for the first test in the list. Subsequent tests use the same dates as the first test and this can’t be changed as their date buttons and start/end date lists are disabled. Of course, you can change to a different date range but you would have to run all tests in the list again (or if you are happy with the final system on the list then you can also set it up in a different folder and run it on a different date range).

Each system you add to the list refines the rules of the previous test (adding new rules is also ok). Note that when you save a new system in a “refinery” list for the first time, its name will be prefixed with 01, 02, 03 etc to ensure they are listed in the order they are added (as this is the order they must be run int). The limit is 99 and if you think you need more than 99 system refinements then in all honesty you have too many. Note that system names could all be the same since they are preceded by unique numbers but I suggest you include a brief comment in each system name describing what rule change has been applied.

If after a number of systems, you wish to change or remove a system then that’s fine. However, if it is not the last one on the list, then you will have to retest all systems from that point to the end of the list.

In summary this system refinery approach speeds up system development because each time you change or add a rule you do it in a new system one down the list form the previous system. So, every system in the list represents a refinement to the one before it. The reason testing can be sped up so much is that each subsequent test only tests the horses and races from the previous test whereas traditional system testing always tests over all available meetings.

Each system idea needs its own folder but you can have multiple system ideas in multiple folders. Just remember to include the word “refinery” in the folder/list name (remember REFINERY, Refinery and refinery will also work) and then add and run systems in the order of entry.

If there is one single tip on how to best use this approach, it is to not restrict your rules too quickly. For example, if you are looking at short-priced horses in the betting market don’t start with a rule requiring them to be less than $3 because when you refine this rule in the next test you can reduce the figure to say $2.80 but you can’t increase it to $3.5. So, start with $4 or $5 and then reduce this figure in the subsequent systems which will test much faster.

Quadrella Betting (in Race Day Live)

New from August 2022 is the ability to create quadrella bets in Race Day Live via a new QD checkbox (between the FF radio button and Multi checkbox). When you click this checkbox (after selecting a race), you will be presented with a dialogue for picking standouts for the four legs of the quaddie or early quaddie.

Note: If on the main window you are viewing one of the last 4 races then those four races come up as the quaddie legs. If you are viewing an earlier race then the early quaddie legs will be displayed (and only if there are 6 or more races on the program). Note for race meetings with exactly 6 or7 races, you will have to be viewing races prior to the last 4 races to get to the window for the early quaddie.

The dialogue for selecting standouts for each leg is similar to that used for picking standouts for exotics in individual races. Both include checkboxes in each of the four columns so you can easily see which runners are selected.

IMPORTANT: You don’t have to click the checkboxes as clicking to the right of a checkbox will automatically select all runners down to the one clicked on – this provides a quick and easy way to select all runners above an individual probability or up to a certain cumulative probability. Note you can then use the actual checkboxes if you want to individually select or unselect any runner (eg you might want to remove the one at the top of the list if it is too short priced). Note also that the win standouts are initially set to whatever their settings are for each individual race and if you change them these changes are saved back to the individual races.

Once you have selected the desired runners for each leg you simply click the “Calculate Combinations” button in the top right of the window. This calculates and displays all the quadrella combinations based on the standout’s selected for each leg.

The program determines the bet sizes by dividing the quadrella takeout by the predicted quadrella price which in turn is based on the prices you have selected on the top (right hand) drop-down on the main Race Day Live window. The quadrella takeout is set to 100 times the WN/QN/EX takeout so the default is $10,000 (as the default WN/QN/EX takeout is $100). Note that prices, takeouts and bet multipliers can be set differently from race to race and it is the ones for the race displayed on the main Race Day Live window that are used (min-max ranges have not been implemented at this time but let me know if you think they are needed and what values you might use).

Below the list box on the right where bets are displayed are “View Bets” and “Copy Bets” buttons for viewing bets in Tabcorp bet-string format and for copying bets to the Windows clipboard so they can be more quickly bet via Tabcorp batch betting – for more on this see:

https://help.tab.com.au/s/article/Batch-betting-through-tab

Please note Bet Selector does not know the Venue Codes for non metrops so you must source these as per the above link and enter the correct code into the data entry box to the right of the Setup button on the main Race Day Live screen (it will show ?? for non-metrops). Note also you do not have to use Tabcorp’s batch betting feature as you can still read the bets off the list and bet them by other means.

Bet outlays and combination counts are displayed (below the above-mentioned buttons) as are the actual returns and profit/losses. Also, quadrella results and dividends are displayed down the bottom left of the dialogue if available.

Note the above relates to the more professional approach of betting individual combinations but for a simpler approach there is a "Calculate Boxed Bet" button for creating boxed bets plus an edit box where you can vary the bet percentage down from 100%. When using the boxed option this invokes traditional flexi betting by reducing the outlays and reformatting the bet strings. When using individual combinations, flexi betting does not apply but the reduction is used to reduce the size of each individual combination bet - in doing so combinations will start to disappear as bet sizes get lower than 1 and round down to zero (unless you have ticked "Round zero bets up" on the main window).

Quadrella System Testing (in System Selector)


Also new from August 2022 is the ability to test quadrellas in System Selector. They key to getting this to work is using Expert Mode and the reverse race number variable REVRNO which you may be aware has a value of 1 for the last race, 2 for the second last race and so on.

Since the quadrella is always the last 4 races the rule required to pick these races is:

2.REVRNO<5

That’s the only special rule you need as the other rules you use will be the ones you decide upon for selecting horses in each leg. You can also put in rules to eliminate races but keep in mind that if you do that you may eliminate some of the quadrella legs which will mean less quadrellas are selected. Note also that early quadrella will be addressed at a later date.

Results are collected by Race Day Live so should be picked up by your system tests as long as you have run Race Day Live for the days being tested.

IMPORTANT: For your convenience I have updated the master result files to end of last month with quadrella results going back to 1 July 2020 (you can download them via the pull-down FILE menu). The quadrella results were uploaded today so please download these files even if you already did so early in the month.

Outlays are determined by the number of selections in the first leg multiplied by the number in the second by the number in the third by the number in the fourth. If one or more legs has no selections, then the outlay is zero and the quadrella is ignored. The quadrella is also ignored if there are win results for the meeting but no quadrella results.

When system testing, the progressive quadrella testing result is shown at the bottom of the test dialogue (as the test is in progress) and the final quadrella result summary is shown on the main System Selector window (below the existing profit/loss summary in the lower middle list box).A summary of Quadrella outlays, returns and strike rates are shown via the “Log File” button while if you add a report to your system, the Report CSV will show you quadrella outlays and returns for each meeting.

A new feature here is that if a report does not exist, an empty copy with that name will be created to contain the quadrella outlays and returns. Note this means you no longer have to go into MyCalc/MyReport off the pull-down RUN menu as simply referencing the report will create it and clicking the “Edit Report/Calc” button in System Selector will let you edit it (eg if you want to add extra report items).

Speed Maps


Speed maps can be viewed in Price Predictor via the VIEW EXTRA FORM button. The speed map shows where each horse is expected to settle in the race. It also displays the adjusted barrier positions for each horse and a speed rating where 100 is neutral and horses above 100 settle towards the lead and those under 100 settle towards the back.

The graph goes from 90 to 110 as most horses end up in this range. Any horses lower than 90 (but not zero) get graphed the same as those on 90. Similarly, any above 110 get graphed the same as those on 110.

Note there is a rank column alongside the speed column so you see both the points score and the ranking for each runner. At the bottom of the ranking column is a number showing the percentage of (non-zero) runners above 100. Also new, the betting price is displayed after the horse’s name and barrier position (it is the Race Day Live fixed price if available, if not it defaults to the “Pre” early market price from the form downloads).

Note also you can now view the speed map in barrier position order. To do this you must use the new BP radio button on the main Price Predictor window – this switches all reports in Price Predictor to barrier position order except for the Place Odds or Top 6 reports (these can’t be sorted by barrier so will switch back to TAB order if you try to order them by barrier). Having selected BP order, Extra Form including the speed maps will now be in barrier order.

NOTE: Just like when you sort by rated order instead of TAB order, when you sort by barrier position, everything in Extra Form is in the new order – not just the speed maps but the ratings tables above and the detailed horse form below.

And here are the new speed map variables.

INRUN – the speed map points score for a horse
INRANK – the ranking of the horse’s points scores
INSPEED – the average points score across all horses with non-zero points


New Features in Pro 2.0

This post describes all the exciting new features in Bet Selector Professional 2.0.

Bonus: Free Best Bet systems are available via the “Download Free Systems” button once you go into System Selector and click the Expert Mode check box but only in Pro ver 2.0. Both versions 1 and 2 of these systems are available and they are saved in separately created folders NEALES BEST BETS V1 and NEALES BEST BETS V2 (if these folders already exist then system files are just added/overwritten).

Most of the systems are based on the top rated main avBase class ratings (AVB1) and/or the top rated main time based ratings (PP1) although there is one based on the top rated avBase Better Ratings (AVB2). As you might expect, some horses will be picked by more than one system but this is a good thing as these duplicated picks are the best of the best and my recommendation is to bet them as many times as they are selected.

Please make sure you enter your new password into the program as only then will it appear as “Bet Selector Professional 2.0” in the title bar and give you access to new features including:

You can now view new avBase class figures for races in Price Predictor and Extra Form
You can now use new avBase class figures in Expert Mode system tests
You can now view new class and time based ratings in Price Predictor and Extra Form
You can now test new class and time based ratings in Expert Mode system tests
You can now view new class and time based ratings in Race Day Live

IMPORTANT: The new class ratings are based on new avBase class figures for past races. These new class figures are saved when meeting are downloaded or updated but for old meetings you have to run a system test over at least the past 2 years or more to force them to update – it needs to be a system that picks at least one horse in every race at every meeting (try an Expert Mode system with just one rule eg PP1R=1). If you see lots of blank (zero) ratings in the past form CRTG column then you know this hasn’t worked.

New Ratings and avBase figures in Price Predictor


Bet Selector Professional 2.0 now has three sets of ratings and rated prices which means there are 3 lots of main, better and plus ratings giving a total of 9 ratings. If you go into the Price Predictor part of the program you will see a button towards the top right that says “Original Ratings”. If you click this button it will change to “Class Ratings” and the displays in the large box will change to reflect the new class based Main, Better and Plus ratings. Click the button again and it will change to “Time Ratings” and the displays in the large box will change to reflect the new time based Main, Better and Plus ratings.

Note that the above not only applies to the default “Prices” display where you can see the three sets of Main, Better and Plus ratings but also to the “Ratings” display that has been improved for the new ratings to also include the Main, Better and Plus ratings (the original version remains as before in for backwards compatibility). Also new you can now see first up stats via the Stats display and the Top 6 and Place% displays also adjust according to which ratings you are using.

You will notice the “Class Ratings” are labelled Main AVB, Better AVB and Plus AVB. The AVB reference indicates that they are actually based on the avBase class figures for a race (rather than the par time for a race). They are not however based on the old avBase figures you are used to but on a new set of avBase figures that much better reflect the range of classes. One problem with the old figures was that lower class races like maidens, early 2yos, country Class 1 and 2s etc  were rated too high (often avBase figures were in the high 40s or low 50s when they should have been in the high 30s or low to mid 40s). As a result, the new Main AVB, Better AVB and Plus AVB ratings are much better than before.

Similarly the “Time Ratings” are labelled Main PP, Better PP and Plus PP. The PP reference indicates a connection to the original Price Predictor time based ratings where a standard time formulae was used for all tracks (ie not different times for different tracks). While these might be the last set of ratings, they have been improved tremendously due to adjustments reflecting the difference between races of different class and races at different locations (eg metrop vs provincial vs country).

New Ratings and avBase figures in Extra Form


The new class ratings do not appear in the LOAD DATA TABLES part of Price Predictor but they do appear in Extra Form via the VIEW EXTRA FORM button. Now is the time to start using Extra Form if you haven’t used it before. Not only does it provide heaps of extra form info but it is the only place in the program where you can see all 9 ratings in the one report.

Please note that the “Original Ratings” which were previously based on par times are now based on the new times and are the same as the new time ratings. They have been relabelled “Saved Ratings” because they are saved in the meeting files and can be changed via edits to base and last start ratings in LOAD DATA TABLES.

Note also for past races on the far left in Extra Form you can see the past par times ratings, class rating and PP ratings under the headings TRTG, CRTG and PRTG ratings in the TRTG column will now be the same as in the PRTG column except for old meetings imported into the older version of the program).

A final note regarding Extra Form is that the Selection Poll now reflects which ratings you were viewing in the Price Predictor part of the program immediately before clicking the VIEW EXTRA FORM button.

New Ratings and avBase figures in Race Day Live


The new rated prices are also available in Race Day Live via a new “Toggle Ratings” button at the top of the Race Day Live window – it allows you to cycle from the old Main, Better and Plus ratings to the new  avBase Main, Better and Plus ratings to the new PP Main, Better and Plus ratings. Upon each click you will notice the column headings change in the main display to tell you which ratings are being displayed. In addition to the rated price columns changing, the highlighted top rated horse (grey highlight bar) now reflects which rating you are viewing (viz Main, AVB Main or PP Main). Also the ratings you end up displaying will be the ones available in the right hand drop downs for generating bets.

As you toggle between the different ratings in Race Day Live, you will find these now flow through to the reports via the buttons at the bottom of the window. And in something of a breakthrough for those interested in the Place Odds report, these predictions can now be based on TAB prices, TAB fixed prices, any of the 9 rated prices, early market (AAP) prices and even prices derived from the AAP 100 point ratings! To select which odds are used just use the drop down in the top right that is normally used for betting functionality.

Updated Original Ratings


Previously the “Original Ratings” changed every 1 to 2 years when the par times used by the program were updated so it should come as no surprise that they are being updated again (the last update was mid 2018). This time however they are not being updated via an external database file but rather are being updated in line with the internal times used for the PP time based ratings. You may recall the PP time based ratings are an improvement on the original time based ratings used in Price Predictor (hence the PP label) before we used the external par times tables.

Whether you understand the history or not does not matter. What does matter is that the “Original Ratings” are now updated and improved and the confusion of having two sets of similar time based ratings has been removed. You will still notice two sets of time ratings  displayed in Price Predictor and Extra Form and they will (initially) be identical. Leaving this duplication in the program may seem strange but keep in mind that you cannot change the PP predicted ratings for a horse but you can edit the “Original Ratings” via LOAD DATA TABLES (eg you can change the base rating, peak rating and any of the last 3 start ratings). For this reason the “Original Ratings” are now labelled “Saved Ratings” in Price Predictor.

NOTE: While the “Saved Ratings” are updated and improved, the updates to the past run ratings (for Saved Ratings) only go forward in time as they are implemented when form is downloaded. If you want to force all you past “Saved Ratings” to be updated then you need to run an Expert Mode system test over the meetings you want updated with the first line of the system being” 0.NEWPEAK,NEWBASE

As a result of the above changes a couple other things have also changed.

First the old AVBASE figures for each race and for past races are now somewhat redundant and are no longer displayed in Price Predictor or Extra Form (but you can still use them in Expert Mode system tests as they have not completely disappeared).

Second there were still some old “Peter Bent” form patterns present in the program and these have now been removed (most were removed a while ago as was the impact of improver and winning form settings on the ratings). They were of course based on the old ratings so now the plan is to have new form pattern comments based not on ratings which can change but on the raw form. These will be purely for display purposes and won’t affect the ratings. For example  you will now see comments like “failed last 3 starts”, “won or near win last 3 starts”, “unbeaten first up” etc. If you have any pet comments like this you think should be included please let me know (note they can only be based on the career to date stats and last 3 start form lines).

New Ratings and avBase figures in System Selector


The new avBase figures and new ratings are all available via Expert Mode variables as follows:

AVBPP – New avBase class figure for race
AVB1R – New AVB Main Ratings Rank
AVB2R – New AVB Better Ratings Rank
AVB3R – New AVB Plus Ratings Rank
AVB1V – New AVB Main Ratings Value
AVB2V – New AVB Better Ratings Value
AVB3V – New AVB Plus Ratings Value
AVB1P – New AVB Main Ratings Price
AVB2P – New AVB Better Ratings Price
AVB3P – New AVB Plus Ratings Price
PP1R – New PP Main Ratings Rank
PP2R – New PP Better Ratings Rank
PP3R – New PP Plus Ratings Rank
PP1V – New PP Main Ratings Value
PP2V – New PP Better Ratings Value
PP3V – New PP Plus Ratings Value
PP1P – New PP Main Ratings Price
PP2P – New PP Better Ratings Price
PP3P – New PP Plus Ratings Price

Note that all the above give you the improved values and overwrite the old values with the exception of the old race avBase which is still available via the same variable as before viz AVBASE (note AVBASE and AVBPP are now the same going forward since we are now using new times for the Original/Saved ratings).

Note also you can now display your choice of new rated price on the selection list and also export your choice of new rated prices to the bscsv.txt file that is created each time you run a system test (it can be used to export rated prices to other platforms like Dynamic race Odds or Racemate Pro). To change from the default rated odds to new odds just requires the use of a MyCalc – in fact you can put anything in a MyCalc and have it displayed and saved in the export file but as far as the new rated prices go just set your MyCalc to one of the “Price” variables from the above list (viz AVB1P, AVB2P, AVB3P, PP1P, PP2P OR PP3P).

Note you can use the CALCCODE as in the following example to quickly set a calc without having to refer to a saved MyCalc. The following example sets the calc and therefore display and exported prices to avBase main rated prices:

0.CALCCODE=AVB1P

Finally whilst on the topic of System Selector, if you use a MyCalc with one of the AVBxx or PPxx variables then not only will the relevant data be displayed in system selection lists but the relevant ratings are now available via the following variables for Expert Mode system use:

RPTOP - Price of Race Top Rated
RP2TOP - Price of Race 2nd Top Rated
RP3TOP - Price of Race 3rd Top Rated
BR2TOP - Better Ratings Pts of 2nd Top
BR3TOP - Better Ratings Pts of 3rd Top
BSRANK - Base Rating Rank
PKRANK - Peak Rating Rank
LSRANK - Last Start Rating Rank

For example if you use a MyCalc set to AVB1P then the rated price for the Main AVB ratings will appear on selection lists and the relevant Main AVB rated prices and ratings will be available via the variables listed above.

Note I have turned off the code to generate the new class (AVB) and time (PP) ratings to speed up system tests in normal mode. However in the event that you want to use AVB and/or PP variables in normal mode MyCalcs you can do so provided the first 2 or 3 letters of the calc are PP or AVB. This switches their calculation back on and in Expert Mode (as before) sets the default rating for variables like base rating rank, peak rating rank etc. An alternative option is to start your calc name with PP$ or AVB$ (so you don’t have to change the calc itself but may need to change the name in other system that use the calc).

Example Systems using New Ratings


As mentioned above the new avBase figures and new ratings are all available via Expert Mode variables. This means you must create an Expert Mode system to use them. This is not as hard as the name suggests as Expert Mode systems are simply text based rules that you can cut and paste from anywhere you like. Following is a system to pick the top AVB Main rated horse at metrops:

1.LOCATION$"SMBAW"
AVB1R=1

As you can see above there are just two rules, the first line for selection metrops and the second line for selecting the top AVB Main ranked horse.

TIP: You do not need to understand the above Expert Mode code. To get it into Bet Selector just copy it and paste it into the Expert Mode editor in System Selector (which you get to by clicking the Expert Mode check box and then the “Add Blank System” button). Make sure you give the system a name, eg AVB MAIN METROP, when saving it.

If you want to consider all Australian locations just change SMBAW in the above system to SMBAWX. You can also add N for New Zealand, H for Hong Kong and G for Singapore.

If you want to test for the top PP Main rated then use PP1R=1 in the second line.

Please note that since the new ratings refer to the new avBase figures, you will need to run a system test over all meetings going back at least a number of months (and ideally a few years) to force the creation of these new avBase figures. If you look in Extra Form at past form and see zero CRTG figures then you will know the new figures have not yet been created for these races.

Finally please note you can run the new ratings in System Selector’s normal mode by referring to a MyCalc set to the rating value you want to use, eg AVB1V for the new Main Class ratings or PP1V for the new Main Time ratings. Just remember to set the Calc min and max boxes both to 1 to get the top rated horse.

New feature in System Analyser


When you update to Bet Selector Pro 2.0 you will get the latest version and it has something new I wasn’t planning to release quite yet but I have included it anyway! The new feature is in System Analyser where you can now apply a filter when analysing. More details follow below but keep in mind it is still in the test stages – it might be possible to make it more user friendly down the track but it is very powerful and worth a look.

Say you analyse a system (like the Best Bets last Saturday) and see that Perth produced no winners. Now if you wanted to analyse the results without Perth then normally you would have to adjust the system rules, re-run the test and then re-analyse the results. But now you can filter the selections in System Analyser without having to re-run the system!

What you do is add “filter code” into the new text box on the System Analyser dialogue and then tick “Use Filter” to enact it. The code uses the CSV names not the Expert Mode names so for LOCATION you must use the name C and you must then put an @ symbol in front of the name like this:

@C<>”W”

The above will redo the analysis ignoring Perth (basically it reads ‘location not equal to Perth’). Note when you go into System Analyser you can see all the CSV column names (CSV ID) on the left in the long list box of criteria descriptions.

Because all columns in the CSV file are treated as characters in System Analyser you can’t use numbers. This is something I might be able to improve upon down the track but for the time being there is a way around this. For example to analyse PP1R=1 you can use the following:

@PP1R=”1”

And if you want to apply both the above filters at the same time you can do so like this:

@C<>”W”.AND. @PP1R=”1”

Note that your filter is remembered for the next analysis but you still have to tick ”Use Filter” to enact it.

More new variables in Expert Mode


Also new are the following variables for use in Expert Mode:

NEWBASE – used in the zero line to force rating updates in old meetings
TABRACE – used to pick TAB races and exclude non-TAB races
WAPR – weight adjusted peak rating rank
BTTIME - Last Trial Time
BTRTG - Last Trial Time Rating
STATE – The State abbreviation for the meeting (eg NSW, VIC etc)
UNIVAL - The AAP 100 point rating adjusted for track condition
NUMUNR - The number of unraced horses in a race 

As alluded to earlier the “Saved Ratings” are updated and improved but updates to the past run ratings for them only go forward in time as they are implemented when form is downloaded. If you want to force all you past “Saved Ratings” to be updated then you need to run an Expert Mode system test over the meetings you want updated with the first line of the system being” 0.NEWPEAK,NEWBASE

The TABRACE variable allows you to pick only those races identified as TAB races from the program’s TAB data. For it to work you must have already run Race Day Live over the day or days in question (or have TAB results from the monthly result files). This new variable is ignored for future days as the full TAB info isn’t usually available until race day morning. The main use of this new variable is to avoid non-TAB meetings appearing in your selection lists on race day (remember my recommendation is that you not delete these meetings as they are needed for calculation of AVB ratings for horse moving from non-TAB to TAB meetings). The way to use the TABRACE variable in a rule is as follows:

2. TABRACE=.T.

The other new variable WAPR is a ranking variable from 1 to 24 that ranks horses according to their peak distance rating but it does so by first adjusting each horses peak distance rating by the weight it is carrying. I have found that this ranking is best used as a refining rule rather than a main rule namely because not all top WAPR horses are in form but when looking at in form horses or horses in the top rated positions it often helps to confine selections to say the top 4 WAPR as the following example rule does:

Within(WAPR,1,4)

To test one state, eg QLD use this:

1.STATE="QLD"

To test multiple states, eg QLD and NSW use this:

1.STATE$"QLD,NSW"

To test country QLD use this:

1.LOCATION$”X”
1.STATE="QLD"

Note Canberra is considered NSW and codes will appear for overseas tracks like NZ, HK and SG (Singapore) although these are just to make it look nice as we can already use LOCATION to pick these overseas venues in system tests (viz N, H and G). Note also you can view a list of all meeting names with state codes via the “List Track Codes” button on the Edit Rules screens (if you see any marked as ??? please let me know what they are).

Note: UNIVAL is an old variable that used to be the same as the AAP 100 point dry rating (viz RTGD). Now it reflects RTGD for dry tracks and RTGW for wet tracks (soft and heavy).

Now that we have automated jockey ratings in form downlaods (as of July 2020) there are also new variables in Expert Mode to help with testing jockeys. They are:

JRANGE – kgs of horse jockey ratings as compared to jockey rating of top jockey in race
TOPJRT – top jockey rating in race (eg lowest number)

You will mainly use JRANGE. For example the following rule picks the top jockey:

JRANGE=0

The following rule picks jockeys with 1 kg of the top jockey:

JRANGE<=1

TOPJRT is used internally by the program to calculate JRANGE but you can also use it. For example the following rule picks all jockeys rated up to two times the rating of the top jockey:

H:JRT<=2*TOPJRT

Note you can still test for absolute jockey ratings. For example the following picks only horses with jockeys rated under 2

H:JRT<2

Note: I use the same code you have access to in the Jockey Analyser part of System Analyser to work out the jockey ratings and I do so based on the past 12 months and plan to update them around the first or second of every month. They are of course based on strike rates and importantly adjusted for race class and since all this is now done for you there is no need for you to create your own (although you may still wish to use the Analyser feature to look at stats and analyse trainers).

Finally some more variables added mid 2020:

AVB1P1 - Rated Price of Top AVB1 Horse       
AVB1P2 - Rated Price of 2nd AVB1 Horse       
AVB1P3 - Rated Price of 3rd AVB1 Horse       
PP1P1 - Rated Price of Top PP1 Horse         
PP1P2 - Rated Price of 2nd PP1 Horse         
PP1P3 - Rated Price of 3rd PP1 Horse         
MAXAVB1V - Top AVB1 Rated Value              
MAXAVB3V - Top AVB3 Rated Value              
MAXPP1V - Top PP1 Rated Value                
MAXPP3V - Top PP3 Rated Value     

The first six variables above will help you pick gaps between rated prices of the top 3 rated. The final four variables above will help you determine kilogram margins in ratings from the top rated horse (note you need to use them in conjunction with rated values for each horse like AVB1V, AVB3V, PP1V and PP3V).

Note also there are new variable names for the Ladbrokes fixed prices viz WN1, WN2 for the first and most recently collected fixed win prices and PL1 and PL2 for the first and most recently collected fixed place prices.  Ranking variables for these are also available, viz WN1R, WN2R, PL1R and PL2R. Having said that you do not need to write these into your old systems and calcs as the old Sportsbet and Tattsbet variables will still work going forward (they will pick up the new data if available otherwise they will pick up the old data). Note however you can no longer use both Sportsbet and Tattsbet variables together in the same system.

Breeding Tests and Analysis

If your program versison is “Ultimate Pro+” or "Pro 2.0" or above then you will have access to the following breeding features:

You can now analyse breeding with separate reports for sires and damsires
You can now system test, analyse and report by sire, dam and damsire   
You can now system test based on the position something appears in a list (eg top 5 sires)
You can now load sire and damsire analysis into spreadsheets so you can sort by bets, wins, strike rate, profits, ratings etc.
You can now load jockey and trainer analysis into spreadsheets so you can sort by bets, wins, strike rate, profits, ratings etc.

The remainder of this email includes a number of breeding based systems plus documentation on how to analyse sires and damsires (sires of dams).

BREED=Y Command


If you add BREED=Y in the first line of any Expert Mode system (the line starting with the number zero) this tells the program you want to analyse sires and damsires. The analysis is in the System Analyser part of System Selector and looks exactly the same as the jockey and trainer analysis – in fact the same radio button is used that normally selects jockey/trainer analysis so it is a case of doing sire/damsire or jockey/trainer depending on whether you set BREED=Y or not (note you can have other commands on the first line in addition to BREED=Y).

New Variables for Testing Dam, Sire and Damsire


The new variables available in Expert Mode for testing sire, dam and damsire (sires of dam) are as follows:

SIRE - Sire of horse being considered
DAM – Dam of horse being considered
DAMSIRE – Damsire of horse being considered

Following is an example system testing for certain sires at Saturday Randwick meetings:

0.BREED=Y
1.DAYCODE="7"
1.V:TRACK="RANDWICK"
Number(SIRE,"Snitzel,Not A Single Doubt,Fastnet Rock,Exceed And Excel,Encosta De Lago,Medaglia D'Oro")>0
BPA<7
H:CPRICE<30

TIP: You do not need to understand the above Expert Mode code. To get it into Bet Selector just copy it and paste it into the Expert Mode editor in System Selector (which you get to by clicking the Expert Mode check box and then the “Add Blank System” button). Make sure you give the system a name, eg RANDWICK SIRES, when saving it.

Note that the first line contains BREED=Y which tells the program you want access to the new breeding analysis. Note also the use of a new function Number() which gives the position a variable occurs in a comma delimited list of character strings. These two new features are described in the next two sections.

New Number() Function


You will find this new function very handy when testing to see if a character string occurs in a comma delimited list of character strings. In the example system given earlier it is used to test to see if the sire name appears in a list of 6 sires.

The function is called Number partly to remind you that it returns a number. The actual number returned is the numeric positon the first string appears in the list. If no match is found the function returns zero which is why a test for greater than zero (>0) as in the above example gives the desired result of a match being found. However you can do more than that since the function will return 1 if found in the first item on the list, 2 if the second item and so on – those of you into points systems should appreciate the benefit of this.

Following is another example system where the new function is also used to test for track names. This system considers the main trainers at the main metropolitan tracks.

0.BREED=Y
1.DAYCODE="7"
1.Number(V:TRACK,"FLEMINGTON,RANDWICK,DOOMBEN")>0
2.Within(R:DIST,1200,1700)
2.Within(GNUM,3,4)
2.NUMFUP<6
Number(SIRE,"Exceed And Excel,Artie Schiller,I Am Invincible,Commands,Al Maher,Fastnet Rock,Flying Spur,More Than Ready")>0
Within(BPA,1,10)
H:CPRICE<30

Note the Number function has two more features to make it easy to use. First it converts everything to upper case before doing the string comparisons so it does not matter what case you use. Second it trims trailing spaces so for example when using V:TRACK in the above example we do not have to worry about the fact that it is normally 16 characters long with trailing spaces.

Sire/Damsire Analysis


As previously mentioned adding BREED=Y in the first line of any Expert Mode system tells the System Analyser part of the program you want to analyse sires and damsires – which it does in place of jockeys and trainers. When you have a system set up this way (with BREED=Y in the first line of the Expert Mode rules) System Analyser displays a new sire/damsire radio button where the jockey/trainer radio button would normally be. Select this sire/damsire radio button and click “Start” and you will end up with a sire/damsire analysis with columns of numbers for bets, wins, outs, strike rates, profits and ratings – just like what you are used to with jockeys and trainers except that of course this report now refers to sires and damsires.

Note that the ratings for sires and damsires follow the same scale as those used for jockeys and trainers, namely from a minimum of 1.0 to a maximum of 9.8 with 1.0 being the best and 9.8 the worst. The reason 9.8 is used as a maximum is because when accessing these ratings in system tests we use a value of 9.9 to indicate a jockey was not found in the file. Note also that damsires (that is sires of dams) are analysed in preference to dams as sires give rise to higher winning strike rates than dams.

In addition to viewing the HTML report of sire and damsire performance you can also click on the CSV button in System Analyser to view a spreadsheet of the same information. The advantage of doing this is that you can sort the spreadsheet for example with highest strike rates at the top or highest wins at the top etc. Of course if you are not familiar with sorting in Excel then you can stick to the HTML report (it will just take you a little longer to get a feel for the better sires and damsires). Note these new spreadsheets are also now available for the jockey and trainer analysis.

Usually you would apply a few system rules before doing a breeding analysis. For example in the systems given earlier a system was run over metropolitan tracks and then sires where analyses to see which ones produced better results at those main tracks. The following system was the result for quickly identifying the best sires for wet tracks.

0.BREED=Y
1.LOCATION$"SMBAW"
Within(GNUM,5,8)
Within(H:RUNS,0,5)
F:MARG>-1
Number(SIRE,"Dubawi,Lope De Vega,Onemorenomore,Savabeel,Street Sense,Commands,Artie Schiller,Foxwedge,Authorized")>0

Bet Selector Task Scheduling

You can now schedule Bet Selector to download form at a certain time each day, get scratchings and results at a certain time etc. These tasks are done completely in the background with no pops ups to tell you what is going on (reason being that it is intended for unattended use when you may not even be logged into your computer so we don’t want pop-ups trying to get in the way of it working).

Three functions can now be scheduled to automatically occur each day, namely Bestform form downloads, Sportsbet fixed price updates and Tattsbet price, scratching and result updates. Special command line arguments are used when firing up Bet Selector to tell it what to do and the scheduling part is taken care of by the Windows Task Scheduler. This will work even if you are away and logged off your computer (but it must be left running and connected to the Internet).

Using the Bet Selector Command Line


The special command lines can also be called from other programs and specially configured shortcuts on your desktop but they are really designed to be called from the Windows Task Scheduler as described in detail below.

Let's start with the command lines. The one for downloading the current days form is as follows:

C:\Price\BETSEL.EXE AUTO FORM

Basically the above is the normal command for calling Bet Selector but with two additional arguments viz the words AUTO and FORM. Note that each of these words has a space before it. The word AUTO is used in all examples and tells bet Selector that the command line is designed for automation. The word FORM tells Bet Selector that the task to be automated is the downloading of form files.

The above only downloads form for the current day but you can download form for tomorrow as follows:

C:\Price\BETSEL.EXE AUTO FORM+1

You can also download form two days in advance using the following:

C:\Price\BETSEL.EXE AUTO FORM+2

Note: The above options were for the old Bestform doanloads so they no longer work given Bestform is no longer operational. They may be updated at a later date.

The command line for updating Tattsbet information is as follows:

C:\Price\BETSEL.EXE AUTO TAB

As above, the word AUTO is used in all examples and tells bet Selector that the command line is designed for automation. The word TAB tells Bet Selector that the task to be automated is the downloading of Tattsbet information. The above works for the current day but can be used to collect data for the previous day as follows:

C:\Price\BETSEL.EXE AUTO TAB -1

Task Scheduling Overview


Now to the scheduling of the above functions which is done by the Windows Task Scheduler. The Windows Task Scheduler is very powerful as it allows you to schedule things to be done say once a day at a certain time. The essentials on how to use it to automate the above functions will be outlined below. If you want to learn more I suggest you Google "Windows Task Scheduler".

To locate and run the Windows Task Scheduler just enter "Task Scheduler" in the search bar off the Start menu. Then click "Create Task" in the right-hand Actions pane. Note if you click "Create Basic Task" then you won't get full control over creating the task and will have to go back in and edit the tasks properties - hence it is best to go straight to "Create Task".

You will be presented with a popup displaying a number of TABs for creating your task. These are General, Triggers, Actions, Conditions and Settings. These are to be set as follows.

General Properties


First type in a name. I suggest you start all names with BS followed by the arguments. This way they will be easy to locate later in the task list and you can see at a glance what they do. For example use bsautoform for a task to automatically download form and bsautotab for a task to automatically download TAB data.

Then go down to the "Security options" and click the "Run whether user is logged on or not" radio button. Also make sure you tick the "Run with highest privileges" check-box.

Finally under "General" at the bottom where it says "Configure for:" make sure the correct version of Windows is selected. Then click the "Triggers" TAB (not OK).

Trigger Properties


Click the "New..." button to set a trigger for the first time. Leave the top drop down on "On a schedule" and click the "Daily" radio button (if not already selected). Then set the time at which you want the task to be run each day. Then go down to "Advanced settings" and tick "Stop task if it runs longer than:" and select "30 minutes" - this is important because if a task drags on because of internet issues or if the program freezes for some reason then you don’t want it left "open" as it will stop future tasks from running. The only other check-box in the advanced section that should be ticked is "Enabled".

Note that you can add more than one trigger which is handy for example if you want to automatically schedule TAB updates at a few different times a day. This avoids you having to set up multiple task for each time - you just have one task with multiple triggers set.

Scheduling tasks to occur exactly on the hour might best be avoided in case lots of other users all attempt to do the same thing. So rather than everyone scheduling end of day TAB updates at say 10pm or next day form updates at say 10pm it would be better to set a slightly different time like 10:11pm etc.

Important: Please do not unnecessarily schedule tasks to run more than you absolutely need to. Providers like Bestform, Tattsbet and Sportsbet may block your access if you access their data with too high a frequency.

Action Properties


Click the "New..." button to set an action for the first time. Leave the top drop down on "Start a program" and enter the following in the "Program/script" box:

C:\Price\BETSEL.EXE

Then enter the arguments in the next box for example AUTO FORM to automate form downloading from Bestform. Note you can also enter the full Bet Selector command line with AUTO and FORM on the end like in the earlier examples and you will just be asked to confirm the arguments.

Finally there is a box that says "Start in (optional)". This could be considered an optional setting that is used when you want to run the program in a separate location to where it resides (something it is best not to try with Bet Selector). However it is recommended that you enter the actual path the program resides in. This is just the beginning of the full path listed above for example C:\Price

Important: If you do not set "Start in" then the Task Scheduler may attempt to run the program in the Windows folder which you definitely do not want to occur.

Conditional Properties


Tick "Wake the computer to run this task". This should ensure your scheduled tasks run even if the computer is asleep. If you have problems with this not working then go into Control Panel and edit your Power Management settings by selecting the option to "Never put your computer to sleep".

Troubleshooting


Scheduling Bet Selector to do things is designed for when you are away from your computer or for times like during the night when you are not using it. Note that if you leave Bet Selector running (even if idle) then scheduled tasks most likely will not work. So make sure Bet Selector is closed during times when you want scheduled tasks to run.

There are other reasons why scheduling something might not work and following is a recap on some of the more important settings that if incorrect could cause tasks not to run:

  • Under "General Properties" make sure you select the "Run whether user is logged on or not" radio button and the "Run with highest privileges" check-box
  • Under "Action Properties" make sure you enter C:\Price (or whatever else your default location is) in "Start in"
  • Under "Conditional Properties" make sure you select "Wake the computer to run this task".

Note that once you locate the Task Scheduler (eg via Start>Search) you can right click it and select "Pin to Start Menu" so it is always easy to find. That way you will only be a couple moue clicks away from checking, changing or adding properties.

How to Test Ratings, Prices and Rankings

The most common test you are likely to want to do is to test for top rated horses. All ratings have ranking variables available in Expert Mode so the rule to test for the top rated is simply to require the ranking equals one.

To date there have been 3 sets of ratings, namely Main , Better and Plus. Better is a variation of the main predictive rating that emphasizes class and Plus is a variation of the main predictive rating that emphasizes consistency.

Now with the Ultimate Professional upgrade we have 3 more sets, namely avBase based variations of Main, Better and Plus.  These are based on the avBase class values compared to the original 3 sets of ratings which are based on  par-times. The benefits of using avBase figures is that they class a race based on the recent ratings of all horses in the race (mathematically weighted towards the higher ratings). The best way to think of the difference is that the original ratings are time based and the avBase ones class based.

In this post we will look at how you test rating rankings and also look at testing other rankings for things like pre-post markets, AAP ratings, TAB markets and more.

Testing Bet Selector Ratings by Ranking


The original "Main" rating rank is saved in the variable H:POS so the rule to test for the top rated is as follows:

H:POS=1

Note the "H:" prefix indicates this ranking is actually saved in the horse data file for each meeting. This is the only ranking saved in this way as it was the original ranking back in the days when the program only had one rating (of course it has been improved along the way but it is still saved in the horse data files).

All other ratings are calculated on the fly so to speak and have variable names as follows:

  • BRRANK - Better Ratings Rank
  • PRRANK - Plus Ratings Rank
  • AVB1R - AVB Main Ratings Rank
  • AVB2R - AVB Better Ratings Rank
  • AVB3R - AVB Plus Ratings Rank

So to test for the original Better Ratings top horse the rule is as follows:

BRRANK =1

Note that in Expert Mode it is preferable to use the above rather than the old way of putting BR$1 in your system name. Note also you can use the special "Within" functions to test for rankings within a range like in the following rule:

Within(BRRANK,1,3)

To test for the new avBase Better ratings top horse the rule is:

AVB2R=1

You may wish to remember that AVB stand for avBase based, the letter R on the end stands for ranking and the number refers to the rating type (1 for Main, 2 for Better and 3 for Plus).

In total there are six rating ranking variables you can use as outlined above. Please note that the avBase ones will only work correctly if you regularly download form each day from Bestform and regularly update your TAB results (downloading the master TAB result files after the end of each month and running a test over all the past months races with "force results checking" ticked is important in this regard).

Important: Make sure you never delete meetings.

Tip: The avBase ratings for each past race used to formulate the avBase ratings can be viewed in the AB column in Extra Form. The only past runs that should be missing AB figures are barrier trials and non-Bet Selector meetings (eg international races). If you find any recent Australian or New Zealand runs without AB figures then you may have to re-download these meetings or at least refresh their results.

Finally there are three more Bet Selector rating related rankings you might like to test for, namely if you want to pick the horse with the top base rating before bonuses and penalties, the horse with the top peak distance rating or the horse with the highest last start rating. Variables for these are as follows:

  • BSRANK - Base Rating Rank
  • PKRANK - Peak Rating Rank
  • LSRANK - Last Start Rating Rank

Testing Bet Selector Ratings by Value


Testing rankings is the easiest way to pick selected rated positions but there are also rated margin and/or price variables for most ratings. For example associated withe Main rating rank H:POS there are the following additional variables:

  • H:PRICE - Rated Price
  • H:LEN - Predicted Lengths

Note that H:PRICE is a text variable (as it used to display in bookie format like 5/2) so to convert it to a rated dividend you must use BOOK2DIV(H:PRICE).

Rated value/price variables for the other ratings are as follows:

  • BRVAL - Better Ratings Value
  • PRVAL - Plus Ratings Value
  • BRDIV - Better Ratings Rated Price
  • PRDIV - Plus Ratings Rated Price
  • AVB1V - AVB Main Ratings Value
  • AVB2V - AVB Better Ratings Value
  • AVB3V - AVB Plus Ratings Value

Remember when using a value variable you must compare it to a number to create a rule or use it inside the Within function. Here are some example rules that require rating values to be above a certain number:

BRVAL>95
PRVAL>55

The first example above tests for Better Rating values above 95 while the second example tests for Plus Ratings values above 55.

You normally would not test for an exact value as rating values are usually (real) decimal point numbers. The exception are Better Ratings which are (integer) whole number values up to 100. Following are some examples of exact value tests:

BRVAL=100
BRVAL=99

The first example above tests for Better Rating values of exactly 100 while the second example tests for Better Rating values of exactly 99. Note that testing for 100 is not quite the same as testing for the top rated horse as sometimes there is more than one horse rated 100 in a race.

Finally you can use the Within function to test for ranges as in teh following examples:

Within(BRVAL,95,99)
Within(AVB2V,97,99)

The first example above tests for Better Rating values between 95 and 99 (inclusive) while the second example tests for avBase based Better Rating values of between 97 and  99 (inclusive).

Testing Other Ratings and Rankings


Bestform downloads include two additional ratings namely the AAP 100 point ratings like you see in many newspaper form guides and the official handicapper ratings that can be found on Racing Australia's web site.

The ranking variables for these are as follows:

  • HCRANK - Official Hcp Rating Rank
  • RDRANK - AAP Dry Rating Rank
  • RWRANK - AAP Wet Rating Rank

Note there are additional variables for the actual ratings themselves as follows:

  • HCP - Official Handicapper Rating
  • RTGDRY - AAP 100 Point Dry Rating
  • RTGWET - AAP 100 Point Wet Rating

Note that where possible you should use the AAP 100 point rankings in rules as they are adjusted for scratchings whereas the ratings are not. Rules to pick the top AAP rated horse on dry and wet tracks respectively are as follows:

RDRANK=1
RWRANK=1

If you want a rule to pick the top AAP rated horse using the dry ratings for dry tracks and wet ratings for wet tracks then the rule is as follows:

IF(GCODE$"SH",RWRANK,RDRANK)=1

Note the above treats slow/soft and heavy tracks as wet and others as dry.

Ranking variables exist for lots of other things too like the early newspaper market in the Bestform downloads, early/final TAB prices, Sportsbet fixed prices, win and place strike rates and more.

  • MPRANK - Early Market Price Rank
  • APRANK - API Rank
  • WPRANK - Win % Rank
  • PPRANK - Plc % Rank
  • SPRANK - Starting Price Rank

Testing Fixed Prices and Rankings


Sportsbet fixed price market rankings and prices can also be tested but only if collected each day via the Sportsbet option in the Race DayLive part of the program. The available variables are as follows:

  • SB1W - Sportsbet 1st Market Win Price
  • SB1P - Sportsbet 1st Market Place Price
  • SB2W - Sportsbet 2nd Market Win Price
  • SB2P - Sportsbet 2nd Market Place Price
  • SB1WR - Sportsbet 1st Market Win Rank
  • SB1PR - Sportsbet 1st Market Place Rank
  • SB2WR - Sportsbet 2nd Market Win Rank
  • SB2PR - Sportsbet 2nd Market Place Rank

Similarly Tattsbet fixed price market rankings and prices can also be tested but only if collected each day via the Tattsbet option in the Race Day Live part of the program. The available variables are as follows:

  • TB1W - Tattsbet 1st Market Win Price
  • TB1P - Tattsbet 1st Market Place Price
  • TB2W - Tattsbet 2nd Market Win Price
  • TB2P - Tattsbet 2nd Market Place Price
  • TB1WR - Tattsbet 1st Market Win Rank
  • TB1PR - Tattsbet 1st Market Place Rank
  • TB2WR - Tattsbet 2nd Market Win Rank
  • TB2PR - Tattsbet 2nd Market Place Rank

In addition to the above we have TBWIN (Tattsbet Tote Win Price) and TBPLC (Tattsbet Tote Plc Price). All this data is saved when you run Race Day Live and also when you collect TAB data via the AUTO TAB scheduler option. TBPLC will be of particular interest to those of you who want to access the place div for every horse in a race but remember it will only be available if you have collected the data.

You will note the numbers 1 and 2 above which refer to the "first" and "second" Sportsbet market. These are collected and displayed in Race Day Live and the way it works is that the SB1 price variables are set the first time you collect data for the day and then on subsequent use the SB2 prices are set. Note however the SB1 market is automatically set regardless if not previously set plus after the first collection of prices you get prompted if you really want to collect a second set (normally you answer Yes but you could answer No if you wanted all "first" prices to be reset).

Typically you would collect your first set of prices after scratchings and track conditions are known say around 10am. Then during the day you could update the second set of prices or you could just leave things until after the races when the results are in so you can then test morning markets against final markets.

Note that during the day you may see a warning pop up indicating if any races don’t have fixed prices available. Note also that final prices can only be collected on the day or the very next day. After that they are no longer available so you need to remember to collect them each day or use the new scheduling features to collect them automatically each night.

Concluding Remarks


All the above variables are available in the drop down lists in the Expert Mode editor so you do not have to remember them all. The Main Rating ones saved in the horse data files are in the "Horse File Data" drop down list while the remainder are in the "More Horse Data" drop down list which is split into sections to make it easier to read (eg strike rates, ratings and rankings, Bestform data and Sportsbet data).

Note: There is a "List All Items" button on the far right of the Expert Mode editor that you can use to view and/or print out a full list of all available variables (not just horse ones but also the meeting, race and form ones).

As rating variables are either rankings (eg 1-24), ratings (eg points of kgs) or rated prices (eg dollar amounts), they all return numbers which are easy to use in rules. Just think of it as you would in plain English. For example if you want a ranking to equal one your rule is simply the ranking variable followed by an equals sign followed by the number 1 (eg H:POS=1 or AVB2R=1). If you want the top 3 rankings then this is the same as saying you want the ranking to be less than 4 so you would use the ranking variable followed by a less than sign followed by the number 4 (eg H:POS<4 or AVB2R<4).

Keep in mind there is a special "Within" function for testing ranges where you need to set both a minimum and maximum value for a ranking, rating or price. It has three arguments namely the variable you want to test followed by the minimum bound followed by the maximum bound. For example to test for Main Rating rankings of 2 or 3 you would use Within(H:POS,2,3) as your rule.

Remember you can have as many rules as you like one line after the other. Two you will most likely use are as follows for picking Saturdays and metrops receptively:

1.DAYCODE$"7"
1.LOCATION$"SMBAW"

Note the above rules are character based not number based. The DAYCODE variable is a character representation of the day of the week (eg 7 for Saturday) while the LOCATION variable is a character representation of the city code (eg S for Sydney). The dollar sign is an operator (a bit like an equals sign) that in English translation means "is contained in" - so the above rules test if the contents of the variable on the left is contained in the character string on the right.

Another special variable you might often use is GNUM that represents the track condition grading (eg 1-10). Note it is a number so it needs to be compared to numbers like in the following examples:

2.GNUM<5
2.Within(GNUM,3,4)

Note meeting rules can be prefixed with "1." and race rules with "2." to speed up testing. For more details on this and other rule writing examples please check out the earlier Expert Mode blog posts via the right hand menu.

Selecting Places

Predicted Place Prices


New in the more recent Pro versisons of Bet Selector is the ability to test predicted place prices in Expert Mode. There are two variables for this as follows:

PLCPREDSP - Predicted place price based on TAB prices
PLCPREDRP - Predicted place price based on rated prices

Note the above are not yet in the Expert Mode editor drop downs but can be typed in directly. For example if you want the predicted place price based on TAB prices to be less than $1.20 the rule would be:

PLCPREDSP<1.2

Note we do not have the final place divs to do value analysis but I am sure you will still find these variables useful. The above rule for example over metrops can even be used for win betting.

Some of you may recall my Place Odds Report from 20 years ago where we had a special calculation for determining true place odds (note they are not linearly related to win odds) and then used these for place and trifecta purposes. The above variables give you those predictions in terms of dividends. In theory a figure less than 1.20 should equate to an 80%+ place strike rate and a figure less than 1.1 should equate to a 90%+ place strike rate.  More background can be found in the original Place Odds Report which is available on request.