• 1

Statement: Implementation of the Linear Appraisal System

31 January 2017

The SABBS Board has noted a number of allegations regarding the implementation of the linear appraisal system and the appraisal held in Cape Town on 28 January 2017. We would like to set the record straight:

Allegation: Linear Appraisal is not appropriate for the appraisal of Boerboels

The previous system – developed in the late 1980’s and first applied in 1990 – was based on the generally applied method of appraising production animals (e.g. cattle, sheep and goats) in South Africa and many parts of the world at the time. The fact that the appraisal system was adapted for the use in Boerboel appraisals, and thus gave its appraisal a unique edge over other dog breed assessment methods, has always been a source of pride for the Boerboel community.

Linear classification is a simple process of assigning a numeric value on a sliding scale (1 – 9) to the variable expression of a characteristic. It is a simple scientific method of recording observations that can be used to a wide variety of applications.

Linear classification has been specifically adapted for the appraisal of animals, and in our case, the Boerboel. This method has been successfully applied in the appraisal of numerous animal breeding programmes, including among others specific breeds of horses, cattle, and sheep. It is recommended as a classification method by the International Committee for Animal Recording (ICAR). The ICAR has published extensive guidelines in this regard, which both SABBS and SA Stud Book adhere to.

The modernisation of our appraisal system to internationally accepted genetic improvement methods is well aligned to one of the core objectives of the Animal Improvement: the promotion of the breeding of genetically superior animals. This development should again be a source of pride for SABBS and its members.

While the linear appraisal system for the Boerboel has been in development for a number of years, and has been tested extensively, the large-scale roll-out thereof may show some minor technical issues, which will be addressed as they emerge.

Allegation: The “disproportionate attention to the head” will lead to the breeding of larger heads in the Boerboel

As already indicated, the linear appraisal chart does not introduce undue attention to the Boerboel head in appraisals.

The appraisal chart records the values representative of a dog’s exhibited traits as evaluated against the breed standard. It is the breed standard that determines the trait or characteristic, not the appraisal chart.

The recorded value reflects the dog’s measure of conformity to the breed standard.

Once sufficient linear appraisal scores have been collected, the aggregated values will reflect the state of the breed as measured against the breed standard.

Allegation: The Linear Appraisal system allocates a disproportionate emphasis on the head.

This is not true. In fact, the head, face, and neck have been combined in the linear appraisal system and the joint value has been reduced by 4%.

A summary of the old vs the new aggregated values:

20170131 linear appraisal statement table 1a5c8
  • The previous appraisal chart duplicated a number of measurable characteristics. These duplicates have been consolidated.
  • Non-conformation categories have been removed. In particular, the reference to the dog’s condition and condition vs. age (which are a reflection of general health care rather than conformation) have been removed.
  • Characteristics and/or conditions listed in the Disqualification List have been removed from the chart as measurable variables. Provision is made for these conditions to be listed as reason for a dog to be registered on the PET Register. Therefore, categories such as “genitals” have been removed from the appraisal chart.

These categories in the previous appraisal system afforded dogs guaranteed marks, which inflated mean appraisal scores.

Allegation: The Linear Appraisal system causes a dramatic reduction of a dog’s appraisal score

The removal and consolidation of categories outlined above, has brought about the removal of about 3% in “automatic points” from the mean appraisal score.

Having removed non-conformation categories, the score is a more realistic reflection of a dog’s conformity to the breed standard.

Allegation: The average score at the appraisal on 28 January 2017 in Cape Town was 77%

This is not true.

The mean appraisal score of 172 dogs appraised in Worcester and Kuilsriver in the Western Cape in 2015 and 2016 is 83.4%. The 30 dogs appraised on 28 January 2017 achieved a mean score of 80%. The mean deviation of the 28 January appraisal from the average therefore is 3.4%, which is within the expected parameters.

The highest score achieved on the day was 85% and the lowest score was 75.1%. Three dogs were disqualified. In keeping with the practice of the previous system, their appraisal scores were not considered in the calculation of the mean.

Five dogs were reappraised. The mean score on the previous appraisal system was 83.8%. The mean attained using linear appraisal was 80.4%. Thus, the mean difference between the two methods again resulted in a reduction of 3.4%, as expected.

Conclusion

Linear appraisal is a scientifically sound and internationally accepted method of appraising animals’ conformity to breed standards.

Changes made to the content of the appraisal chart comprise consolidation of duplications, and the removal of disqualification criteria and non-conformation criteria. This represents about 3% in previously “automatically attained” points.

Articles to further explain the linear appraisal and Best Linear Unbiased Prediction (BLUP) will be published in due course, as indicated in the Boerboel eNews of 26 January 2017.

Enquiries in this regard may be directed to the Secretary at This email address is being protected from spambots. You need JavaScript enabled to view it..


Kenny van der Merwe
Treasurer