Weighing Averages and Medians

While the average is used to compare sale results for stallions, many number-crunchers believe the median is a more accurate figure to use.

One of the more widely used methods of comparing sales results for stallions whose progeny have been offered at a yearling sale is to look at average prices.

Comparing a stallion’s average from one year to another, between sales, or with the averages of other stallions, indicates the direction a sire's commercial value may be heading.  Many number-crunchers, however, believe the median price provides a more accurate picture. The median is preferred because it is not influenced by the prices at the top end of the market. It is the exact midpoint in the range of prices paid for a sire's progeny because it is the number for which half the horses sold below and half sold above. Extraordinarily high-priced horses skew averages that could, in some cases, lead to an inaccurate conclusion.

Regardless of which standard of market measurement you prefer, BloodHorseNOW.com's exclusive Data Digest has done the work for you. Included in the 26-page set of tables and data from the recently concluded Keeneland September Yearling Sale are lists of leading sires by average and median prices, as well as profitability tables for sires based on their yearlings’ averages and medians. The Data Digest is a downloadable PDF free to subscribers of The Blood-Horse magazine who have registered for BloodHorseNOW.com.

For example, A.P. Indywas leading sire (with 3 or more sold) at Keeneland by both average and median, with figures of $858,043 and $675,000, respectively. He stood for a $300,000 stud fee in 2005. Distorted Humor , however, led both the average and median profitability lists. According to Data Digest, after deducting two times his 2005 stud fee of $60,000 plus the $15,000 estimated expenses incurred by the breeder from the time the foal was born until sold, Distorted Humor’s Keeneland sale average of $414,000 produced an average profit of $279,000 (67.4%) for the 38 yearlings sold. His $300,000 median resulted in a $165,000 (55%) profit.

A.P. Indy was second by average profit, with $243,043 (28.3%) and was sixth in the median profit category, with $60,000 (8.9%). Second on the leading sire list for median profit was Spanish Steps, whose 2005 stud fee was $5,000 and whose four Keeneland yearlings produced a median of $97,500 (79.6%).

In comparing profitability by average with profitability by median for the Keeneland sale, the data reveals that 22 sires that showed profits based on their average prices also showed losses when profitability was based on their medians. On the other hand, there is only one stallion that generated a profit based on his median but showed a loss when profitability is calculated on his average.

And if you’re a real numbers devotee who doesn’t consider average and median prices and profitability sufficient bases on which to draw conclusions about the Keeneland sale, take a look at the “mode” computed by Data Digest.

Statistically speaking, the mode is the data point that occurs most frequently in a set of numbers. For this year’s Keeneland sale, the mode of $50,000 was unchanged from 2006. The mode last year, however, rose to $50,000 after being $20,000 for three consecutive years.

Regardless of the criteria one prefers to evaluate stallions and how their offspring sold at Keeneland, the Data Digest has done the work.

Download Data Digest: Keeneland September