Breeding value index
Weight profiles, reliability per trait, rank within the whole book and the expected trait shift per generation.
An index is a profile of weights over traits, not a single formula. A farm may have its own profile and the association another, and both are computed on the same data. The result is stored together with the version of the comparison base: without it, six months later there is no way to explain where the figure in an issued document came from.
Alongside the index we show the expected shift of traits over a generation: traits are genetically correlated, and selecting on protein moves fertility — something usually discovered three years later. This is a case where warning is cheaper than repairing.
A national index, once it exists, will enter as one more profile over the same traceable data — the book will not have to be rebuilt for it.
The profile is named and reliability stands next to the number. An index without a profile is a number without a unit: there is nothing to compare it with.
What is shown is not the number but what it is made of — including a contribution with a minus sign. An index without its breakdown cannot be checked and cannot be argued with.
Limits
- The comparison base is currently borrowed (Net Merit 2025, converted to metric units). Recalculation on our own population is written and waiting for data.
- What is computed is the spread of estimates, not the true genetic standard deviation: strictly, that comes from variance components under an animal model, which is the work of an evaluation centre. Our figures cannot be compared with official NM$ or TPI.
Other sections
See how it works
The Holstein book is open: every section can be opened and read on live data.