EPDs and EBVs explained

A photograph shows you what a bull looks like. A breeding value tells you what he is likely to pass on. For anyone buying semen, that second thing is what you are paying for.

Two names, one idea

• EPD — Expected Progeny Difference (used by most North American associations) predicts how the average calf of an animal will differ from the average calf of an average animal in the same breed evaluation.
• EBV — Estimated Breeding Value (used by BREEDPLAN in Australia, New Zealand, South Africa and elsewhere) estimates the genetic merit of the animal itself.

Because a parent passes on only half its genes to each calf, EPD = EBV ÷ 2. Both are useful as long as you only compare like with like.

Reading the numbers

Breeding values are expressed in the units of the trait and as a difference from the breed base, with a plus or minus sign. If Bull A has a yearling weight EPD of +65 lb and Bull B +45 lb, Bull A's calves are expected to average 20 lb heavier at a year, when both are mated to similar cows. The absolute number means little on its own; the difference between animals is what matters.

Common traits you will see:

• Calving ease and birth weight — critical when breeding heifers.
• Growth — weaning weight, yearling weight, sometimes mature cow weight.
• Maternal — milk, days to calving, scrotal size.
• Carcass — carcass weight, eye muscle (ribeye) area, fat depth, and marbling / intramuscular fat.
• Selection indexes — a single dollar value combining several traits for a stated breeding goal. Use the index that matches your goal (for example a terminal index for calves you will sell for beef, a self-replacing index if you keep heifers).

Accuracy: how much to trust it

Every breeding value comes with an accuracy figure. Accuracy rises as more information is added — the animal's own records, its relatives, genomic (DNA) data and, above all, the performance of its own calves. A young bull with a high EPD but low accuracy may move up or down considerably as his calves are recorded; a proven sire with high accuracy is a safer bet, though usually not the highest-ranked animal. Many breeders deliberately use proven bulls on most of the herd and a few high-ranking young bulls on the rest.

Percentile ranks

Associations also publish percentile tables. The 50th percentile is breed average; "top 10%" for marbling means only 10 percent of animals in that evaluation have a more desirable value. Percentiles are often the quickest way to see whether a number is actually good, because "+2.1" means nothing until you know the breed's range.

The mistakes buyers make

1. Comparing across breeds or associations. Each evaluation has its own base and its own population. An EPD from one breed association cannot be compared directly with an EPD from another breed, and an EBV from one evaluation cannot be compared with an EPD from a different one. (Researchers publish across-breed adjustment factors for some major breeds; they do not cover every breed.)
2. Using old figures. Evaluations are re-run regularly. Ask for the current values from the association's website, not last year's sale catalogue.
3. Ignoring accuracy. A spectacular number with low accuracy is a hope, not a promise.
4. Chasing one trait. Extreme growth can mean big calves and difficult births; extreme anything has trade-offs. Indexes exist to balance them.

Key takeaways

• EPDs and EBVs predict what an animal passes on. EPD = EBV ÷ 2.
• Compare animals only within the same breed evaluation.
• Always look at accuracy and the percentile rank, and use current figures.
• Pick the selection index that matches your own breeding goal.