Wizard Daily Report and Research - Friday, 29 March 2024.

• Improving Wizard Rating win strike rate (NEW)
• Repost - Combining Wizard Ratings and Wizard Plus Factorials
• Wizard Ratings Report for Thursday, March 28

Improving Wizard Rating win strike rate

For more than two decades the Wizard rating (Wrat) has provided a consistent and reliable assessment of runners in gallops races run throughout Australia. The ratings have performed equally well across metropolitan, provincial, and country racing.

Year after year the Wrat has delivered a steady 26% result (+/- 1%, depending on the year), despite having to rate some 17,000 races annually, on all types of tracks, under all conditions.

When the new Wizard Plus service was introduced in November last year, I took the opportunity to add three new ratings methodologies. The new ratings were intended as a Wrat supplement (Wmod) and complements (Whcp, and Wexpk).

Particularly useful has been the addition of Whcp and Wexpk that are classical, pure, relative weight assessments. These are designed to focus on the inherent ability of the horse. In themselves, these ratings methodologies do not have a high strike rate, but they are very informative for those who rely on ratings to identify the main chances in a race.

Now that these new ratings have a permanent place in Wizard Plus it seemed a worthwhile exercise to see if there was any synergistic benefit to be gained by combining them. (Last Monday in the Daily Report we saw the benefit that might be gained by combining the Wrat and the Wizard Factorials. This report is reposted below.)

To research this question, I did the following:

• Processed each rating option (Wrat, Wmod, Whcp, Wexpk) individually for the 21-22, 22-23, 23-24 seasons.
• Processed each combination of two ratings methodologies for each of the seasons.

The matrix below summarises the results of this research.

As you can see: 

• Wrat over the three seasons has recorded win strike rates of 26.8% (21-22), 26.0% (22-23), and 26.4% (so far in 23-24).
• Whcp over the three seasons has recorded win strike rates of 21.4% (21-22), 21.1% (22-23), and 20.6% (so far in 23-24).
• BUT, when combined, Wrat/Whcp record strike rates of 32.7% (21-22), 30.3% (22-23), and 30.1% (so far in 23-24).

There is an overall annual improvement in the win strike rate of some 17.5% across the three seasons.

A preliminary analysis indicates there will be a slight reduction in average odds when the two ratings are combined, but it will not go anywhere near offsetting the 17.5% improvement in the win strike rate. I will look at the odds distribution for these ratings various combinations in a future Daily Report.

Also, it is worth noting that combining Wrat and Wexpk (that is, categorising a runner as a qualifier when it is rated 100 on both Wrat and Wexpk methodologies) returns equally impressive improvements over Wrat alone.

This means that a 'combined' Wrat qualifier would be a Wrat 100 pointer that is rated 100 by either Whcp or Wexpk, or both of them.

Finally, it is important to note the analysis for each category was done on races where there was a single 100 pointer first selection in that category and ignored a Wrat/Whcp or Wrat/Wexpk race which had more than one 100 pointer in the race. I will investigate the multiple 100 pointer situation later.

I will also delve deeper into other possible synergistic opportunities that are provided by the various new elements in Wizard Plus and report on them in future Daily Report articles.

REPOST - Combining Wizard Ratings and Wizard Plus Factorials

When on any given day there are dozens of races on which to bet, and hundreds of horses contesting those races, what is needed is a means of separating the wheat from the chaff.

Last Saturday I discussed the Wizard Plus Factorials and how they might relate to the bet-selection process.

I said, "We have found that there is value in combining the ratings-based and 'factorial' assessments when making final selections and this aspect will be addressed in following articles."

This is the first of the follow-up articles on this topic.

First, let me make the point that what follows is not a recommended selection strategy. It is a description of one possible approach to integrating the quantitative (ratings) element of Wizard with the qualitative (factorial) element. It also reflects some of the points I have made, and views I have shared, in these Daily Reports.

So here is one possible approach to combining Wizard ratings and Wizard Plus factorials.

With respect to races:

• Eliminate from consideration all maiden class races, races restricted to 2yo or 3yo horses, and all Group races (Groups 1, 2, 3 and Listed). Whilst the Wizard ratings work well in many 2yo and 3yo and maiden class races, the data needed for factorial analysis is inadequate, and would therefore be unreliable. With respect to the Group races, the Wizard ratings also perform well, but a disproportionate number of runners are likely to qualify under the factorials and that means a lack of potential separation.
• Ignore races that will be run on heavy tracks.

When it comes to the ratings:

• The base rating is the Wizard Ratings (Wrat) 100 pointer. Accept the Wrat 100 pointer if it is also rated 100 by either the Whcp or Wexpk assessments. (Here we seeking to have the Wrat 100 pointer's "current form" claims confirmed by either the Whcp or Wexpk ratings that focus on the inherent ability of the horse, and over a longer time period.) Note: This horse must be the only 100 pointer in either the Whcp or Wexpk ratings; if there are equal 100 pointers, ignore the race.)

With respect to the Wizard factorials:

• The Wrat/Whcp/Wexpk ratings qualifier must also be listed as a qualifier in at least one of the Factorial columns: Wjktr (jockey/trainer), WFilt (filter), or Wneg (negatives).

Finally, the recommended odds range is our frequently discussed $2.00 to $5.90 range. There will be occasions when the market is at the margins of this range and decisions would have to be delayed until the last few minutes of betting. Such is life.

Yesterday, I analysed the meetings run over the past week and applied the above approach. The results for the week were:

From the table: On Friday, for example, there 11 qualifiers, of whom 5 won and 10 were placed.

Below are the qualifiers for last Saturday, with the final 6 being determined by the odds range.

Again, I stress that this is not a recommended selection/betting strategy. It reflects a conservative, narrow-focused approach, seeking quality rather than quantity. It ignores the opportunities to support Wizard's longer-priced rated runners. It is posted here merely to suggest the type of approach subscribers might consider when combining the ratings and non-ratings content of Wizard and Wizard Plus.

Wizard Ratings Report for the meetings on Thursday, March 28

The following table shows the where the winner of each race run yesterday was rated by Wrat, Wmod, Whcp, and Wexpk. The Wrat analysis applies to both Wizard and Wizard Plus. The Wmod, Whcp, and Wexpk analysis applies to the ratings included in the Wizard Plus Plus Panel

Header - w m h e =  Wrat, Wmod, Whcp, Wexpk

The table shows the where the winner of each race was rated by Wrat, Wmod, Whcp, and Wexpk in that order. The Wrat analysis applies to both Wizard and Wizard Plus. The Wmod, Whcp, and Wexpk analysis applies to the ratings included in the Wizard Plus 'Plus Panel'.

For example, a cell entry of 1 2. 3 would read as:

.. 1 .. rated 1st (100 pointer) on Wrat

.. - .. not rated in the top three on Wmod

.. 2. .. rated equal 2nd on Whcp (the full stop after the numeric indicates it was rated equal 2nd)

.. 3 .. rated 3rd on Wexpk

As well as the results for individual races there a summary for each meeting which shows the number of winners for each of the Wizard and Wizard Plus ratings. An entry showing '4 3 0 2.' would be read as follows:

.. 4 .. 4 of the winners were rated 1st on Wrat

.. 3 .. 3 of the winners were rated 1st on Wmod

.. 0 .. no winner was rated 1st on Whcp

.. 2. .. 2 winners were rated 1st on Wexpk but at least one of these was an equal top-rater

With respect to equal-rated runners: This does not happen with Wrat as we have an algorithm that separates runners by adjusting the base ratings by career-record and form-related factors. The same procedure is applied with the Wmod, but not as extensively, and occasionally equal-rated horses could appear, but infrequently.

Where equal-rated runners are more common is in the ratings analysis that is basically pure weight handicapping, the Whcp and Wexpk analyses.

Warren - Wiz-Ed

If you have any comment or suggestions about the Wizard Daily articles please feel free to drop me a line at:

Wiz-Ed@everyrace.com