An Introduction to GTX Wintune
|
|
|
|
|
|||
|
|
|
|
|
An Introduction to GTX Wintune
|
|
|
|
|
|
|
|
Wintune is a tool that enables you to adjust historical Trakline, Formline or Trakform Ratings to produce more reliable predictive ratings for today’s races based on confirmed historical ratings. It must be stated that all changes made using Wintune are based on personal opinions. Wintune interfaces with historical form and gives you a graphical display of past ratings, similar to the Rating Graphs on each horse’s Form View, but for all runners in that race on a single screen. You can then fine-tune the numbers within past races using your knowledge of confirmed results, and optionally,the existing ratings may be replaced with your personal Wintuned ratings. If you elected to adjust the Trakform Ratings for example, you would end up with a set of personalised Trakform Ratings. These new numbers will be stored and will automatically be used in the calculation of future predictive ratings in the same way that the original numbers affect the calculation of the ratings in subsequent runs. However, you would no longer have Trakform Ratings to compare with unless you run a second copy of GTX on the drive without the Wintune option selected. Wintuning is recommended only for the form enthusiast wanting a result influenced by personal decisions. Used properly with commitment, this procedure can enhance rating performance dramatically. To reiterate, the Wintune module enables you to make adjustments to current ratings, store the resultant ratings, and use them forever. But remember it is only over time, with commitment, that this module will prove to be invaluable... once you master the art you will have a reliable set of numbers, or form indicators to quantify the value of each historical run. Rating horses is not a science, it is an art form. It can never be a science for many reasons, not the least of which is that horses are not mechanical things, and like other living creatures, they can become unwell, have physical or emotional problems, or just lose interest in the job at hand. Unfortunately, we observers can never know about all of these performance-affecting variables and how each of them affects each horse. Also, each and every horse reacts differently to trackwork, racing, rest periods, weather conditions, seasonal conditions or a host of other variables. Horses are trained and ridden by humans, each with a range of motives: some humans are influenced by other humans with their own personal motives. Let me reiterate: rating horses is an art form, no matter what rating method you choose to adopt. If you wish to join the small brigade of winners at this caper (estimated to be 2% of punters), you do not need luck or a big bank! You don’t need a good tipping service or ‘insider tips’… you do need to assess the real chances each horse has of winning today’s race, relative to the others in that race, and that’s before you consider wagering. One of the proven methods of winning regularly at the races is to use ratings. Predictive ratings are almost always derived from quantified historical data, so unless the historical data is completely accurate you start from a less-than-optimal base and can therefore expect inconsistencies. Wintune offers the ability to modify the effect of any known or suspected inaccuracies in historical data and build an accurate picture of each horse’s ability and class for future rating purposes. Inaccuracies can occur for the following reasons:
Additionally, a slow pace resulting in an overall relatively slow overall race time will always produce a low time rating which could be more useful if Wintuned even marginally.
Before you start WinTuning, or even just to understand ratings generally, it is important to know a little about the Qualifying Run. For Trakline, Formline and Trakform Ratings, programmed logic, also known as an algorithm, is employed to select a Qualifying Run (QR) from each horse’s past runs. The rating for that chosen run is generally a rating that might reasonably be expected to be repeated by that horse today, so the QR rating becomes the Base Rating for today’s predictive rating. Depending on the rating in question and the user’s personal rating preferences, bonuses and penalties can be applied against each base rating to produce each horse’s nett rating, which is the predictive rating for today. A Trakform Rating for a horse can be selected as the ‘Qualifying run’ for up to three runs or 42 days, so an abnormally high rating for each horse in just one race can often adversely influence all runners from that race for this entire period.... 10 horses... 3 runs... potentially up to 30 future races You can see how important it becomes to ensure that historical ratings are as accurate as possible, as they heavily influence the predictive rating for today … A good set of reliable historical ratings will improve the performance of any predictive rating which is derived from historical form.
WinTune is a GTX add-on ‘module’ which, as stated before, enables you to adjust the ratings that have been assigned to historical runs, and to store the adjusted ratings. During the importing of today’s races, the original ratings will be automatically overwritten with your own adjusted ratings. This results in personalised historical ratings that have been adjusted and confirmed by you. Wintune enables you to adjust the Trakline or Formline or Trakform Ratings assigned to all runners within a given historical race. Wintune should only be used with your choice of one of Trakline or Formline or Trakform Ratings. In short, the aim is to produce reliable predictive ratings for today’s races based on confirmed historical ratings. At the risk of being repititious, the primary reason you Wintune an historical race is to adjust the ratings assigned to each runner in that race so that the adjusted rating becomes a more representative quantification of that run where the original rating may have been formed by inaccurate data. Any ratings that are derived from a formula will have anomalies. The anomalies will be more noticeable if the formula relies upon the input data being factual and accurate. Although there is no better indicator of ability, ratings based on times have are sometimes inaccurate and this can be for a number of reasons. In addition to the three ‘rating killers’ mentioned above, if track conditions were constant, grass was of a constant length and condition, and there was no wind, the Trakform Rating would generally be a reliable way to measure the value of a run for rating purposes, assuming the par times were updated regularly. The secondary reason to Wintune an historical race is to confirm an existing historical rating when it has proven to be accurate.
Wintune can be used at any time, pre-race or post-race. However, generally the most productive use will be after the results of a race are known. You can then enjoy the benefit of hindsight and use the known result to prove ratings from a recent historical race.
The short answer is when it appears that a rating is too high or too low. This can be determined pre-race through the Rating Graph or post-race through the known result. If you consider that there are on average ten runners in each race and the rating assigned to each runner can be used (in Trakform Ratings) for any of the next three runs it only requires one rating to be incorrect to produce less than reliable ratings for up to 30 future horse runs. That’s quite a few races. The reciprocal is also true; get the rating correct and you have up to 30 reliable predictive ratings for those horses. This of course does not make all predictive ratings for today’s race correct because there could still be others requiring some adjustment Below are before and after examples (applying User Adjustments and recalculate QR) using a horse named Another Neptune from a race on Wednesday 3rd March, 2004. Please note that for the purposes of this example we chose to use Trakform Ratings as the subject rating. In the graphic on the right note the ‘green socks’ on each column graph. The presence of these socks indicates that each of these runs has been ‘Wintuned’
This can be put into more perspective by noting the effect on the predictive Trakform Rating market after the two Wintune options (User Adjustments and Recalculate QR) are selected using previously Wintuned numbers. The market, before options were selected, showed the top rated selection as Another Neptune at rated odds of 8/11 and the eventual winner Lodger ranked second at 11/4. Compare this to the market after Wintune options were selected. The eventual winner Lodger was top rated at 7/4 (won at $2.30 NSW) and Another Neptune rated at 9/2 (unplaced at $6.30 NSW) A post-race review indicates that most of the Wintuned numbers in this race were almost accurate for the purposes of ratings. However, if the most recent run of Declare, which ran 3rd, had been slightly higher at 43.3* instead of 42.3 you could not have created a better set of pre-race numbers for the way the race unfolded.
This Wintune Screen reveals the historical graphs for all horses in the race at Sandown won by Careyes where Declare finished 9th, beaten 4.3 lengths at its most recent start. Note the presence of the ‘green socks’ for all runners in most historical runs? This gives an insight into the degree to which this user has used Wintune to date.
As with all challenges, you become good only if you have passion, or at the least you become singularly focused. The Wintune facility serves you best when you know why you make changes and are able to confirm your decisions. Its simple; the more you use Wintune, the better exponent of the art you become. The better you become, the more accurate your historical ratings become. The more accurate your historical ratings become the more reliable your predictive ratings become. The more accurate your predictive ratings are, the easier it becomes to assess risk accurately, and in this caper, risk can be well rewarded when you enjoy an advantage over the people you bet against.
Firstly, Wintune may be used with either the Trakline, Formline, Trakform or X1-X4 ratings. The X1-X4 ratings will appear only on the computers of the users that have designed their own rating, and have paid to have them programmed as an add-on for their own use. Wintune requires the presence of its own database and the Results Database to enable your personal adjustments to be written back to, so it is important that you keep regular backups of these files and copy them to a different hard drive or back up media in case you suffer a hard disk failure, otherwise your personal adjustments will be lost. See Wintune Databases Backup. The Wintune module contains most needed options, however, prior to using Wintune there are a couple of settings, external to the module, that need to be considered. Select Setup > Ratings (Extended) then select the rating to be tuned. For the purposes of this set of instructions the Trakform Rating will be used. Select the following options:
This option will force GTX to choose the QR in accordance with the adjusted ratings rather than simply using the previously selected run. The purpose of opening the Wintune screen is to enable you to view a rating graph for each runner in a selected historical run and adjust the rating for all runners in that race, or to adjust the rating assigned to that particular runner only. To open the Wintune screen on a particular historical race firstly display the Form View and hover your mouse over the rating graph on the left side. As you move your mouse across the coloured columns that represent past runs, the related prior runs in the historical form above the graph will highlight in green or red. If the form run is highlighted in green as in the image below then it simply means you have that race file on your computer. If it is highlighted in red it means you do not have that race file therefore you will not be able to Wintune that race. The green ‘socks’ at the bottom of each graph column in the Form View indicate that you have previously adjusted the rating for that run.
If the form highlighted is green, then left-click your mouse on that graph column and the Wintune screen will open at that historical race displaying graphs representing the last 16 runs of all runners in that race. It is not essential to the use of Wintune to adjust all runs but that’s another story. To put it simply, you should only ever adjust a rating when you know why you want to. When using the Wintune screen for the first time we suggest that you choose the button for the rating you wish to change (top left of Wintune screen) and click that button down. It will remain selected for future use. Before actively beginning the Wintune process please take some time to familiarise yourself with the purpose and functionality of the active components in the Wintune screen:
As stated above, ABias is the adjusted bias which is the result of Wintuning a time. It is expressed as a percentage. For example, an OTime might 71.96 and the Par Time might be 70.43, therefore the OBias would be 2.17%.
Additionally, please note the valuable static information displayed:
And the following for each horse in the race:
When you access the Wintune screen, unless you have previously created an adjustment for this run, only columns for historical runs previous to the run you are viewing will be displayed. You need to create the rating for the run being ‘Wintuned’ so it’s best to consider what the value of the generic rating is before opening the Wintune screen Before Wintuning an individual horse you should create a rating for all runners in that race. Assuming you want to Wintune a race as opposed to an individual horse, there are a few ways to create an arbitrary starting point but the most logical is to left-click your mouse on Otime. This will create a starting point. Your task is to toggle the ATime up or down using the spinner in between the ATime and ABias until you reach the desired rating that you want to represent that run for future rating purposes. You should be guided by the previous rating graph columns for all runners in that race. When you use the abovementioned spinner the increments will be in 1/100 of a second. However, if you hold the [Ctrl] key down and then click the spinner the increments will be in 1/10 of a second. It is important to note that the adjusted time will only be saved when you click the Close button if you have the button Automatically post changes depressed. If you also have selected the options User adjustments and Recalculate QR in Setup>Ratings (Extended)>Trakform, GTX will do the calculation in between closing the Wintune screen and displaying the Form View and will make the necessary adjustments and calculations for the ratings for today’s race and in confirmation, display a green sock at the bottom of the graph column representing the run that has been ‘Wintuned’.
If, at a later time you wish to adust a previously adjusted time you repeat the above process but obviously it would not then be necessary to commence from an arbitrary adjustment. A database backup and restore facility has been provided for your convenience. If you use Wintune, please save your work so you don’t lose it. Once you have backed up the files you can also copy the files to external media for additional safety if your work is important.
Select Results Database and Wintune Database and click the Backup button. The databases are now backed up in compressed (zip) files in the <.\GTX\Backup> folder. The files are <results.zip> and <wintune.zip> Those users who like to copy their backups to remote media such as CDs, DVDs or external hard drives can simply copy these zip files to their chosen location.
For more on backing up databases, see Backup - Restore Databases. |
|||||||||||
|
|||||||||||
Some thoughts on rating horses...
