“Lies, Damned Lies, and Statistics”
Wikipedia states: “Lies, damned lies, and statistics” is part of a phrase attributed to Benjamin Disraeli and popularised in the United States by Mark Twain: “There are three kinds of lies: lies, damned lies, and statistics.” The statement refers to the persuasive power of numbers, the use of statistics to bolster weak arguments…
A quick flick through February or March’s horse magazines will show you how statistics are used to bolster arguments. Quite often the statistics are useful, and meaningful. However, looking at the stallion and stud adverts in Horse Deals, and Horse and Hound, I am struck by two things:
1) Not a single stallion advert (as far as I can see) shows reliable/meaningful fertility statistics.
2)Those adverts that do have a “statistic” (if we can call it that), have used a misleading and essentially meaningless number to “bolster” their advert.
For fertility statistics to be meaningful, we need to know how many mares the stallion was bred to, and how many attempts/cycles it took for those mares to become pregnant. Now, these figures are most usefully represented by showing the number of mares (as a perecentage) who became pregnant on the first cycle, or as the average number of cycles it took each mare to become pregnant. Here are some -rough- averages to see what you might expect:
Type of breeding / Average first cycle pregnancy rate / Average number of cycles for pregnancy
Natural cover or Fresh AI / 70% / 1.6 cycles
Chilled semen / 60% / 2 cycles
Frozen semen / 40% / 3 cycles
These are rough averages, and what you might expect of each type of breeding.
I would imagine that if you’ve seen fertility statistics you’ve seen figures like (and I’m quoting these from Horse Deals) “100% in foal success rate” or “99% conception rate” or “high conception rate”. These statistics are not worth the paper they are printed on. Why do I say this? Because these numbers give you no idea of how many mares have been bred to that particular stallion/at that AI centre, and you have no idea how many cycles it took for those mares to become pregnant.
For example… Say the pretend stallion “Mr Dudd” covers 1 mare in 2008 via natural cover. She gets pregnant on the 5th cycle. What figure does the stallion owner quote? “Oh, well 100% of Mr Dudd’s mares went in foal last year” or “Mr Dudd has a 100% conception rate”. If they quoted the statistics properly, it would be an average 1st cycle conception rate of 0 (ZERO) and an average of 5 cycles for the mare to become pregnant. Now obviously, the stallion owner won’t want to quote that (and the number of mares makes the statistic almost totally meaningless (although you would question why it took 5 cycles, right?)), but quoting the “100% conception rate” figure is VERY misleading. The owner could quote:
“100% of mares pregnant within 5 cycles. Of those, each pregnancy required an average of 5 cycles. The overall percentage of cycles resulting in pregnancies was 20%”.
Let’s go for another example… Mr Dudd’s hypothetical sire, “Fye Ring Blanks” is actually quite popular, and covered 20 mares in 2008. 5 mares became pregnant on the first cycle, 5 mares on the 2nd, 5 mares on the 3rd, 2 on the 4th and 3 mares didn’t get in foal at all. Now, old Fye’s owner likes to give the impression that he’s still pretty fertile… so she looks at the numbers. 17/20 in foal… that’s 85% overall. So again, she quotes “85% fertility” which still sounds pretty good. Right? Wrong. His first cycle pregnancy rate is (5/20) 25% which is pretty low. On average it takes 2.24 cycles to get a mare pregnant, which isn’t so bad. (Don’t forget, all these figures are totally hypothetical, so you’re unlikely to see that combination of numbers). It would be fair for the owner to quote the 2.24 cycles average, but not “85% fertility” because that leads mare owners to believe that they have an 85% chance of their mare being in foal after the first cycle, which is simply not the case. This owner could quote:
“85% of mares pregnant within 4 cycles. Of those, each pregnancy required an average of 2.24 cycles. The overall percentage of cycles resulting in pregnancies was 34%”.
So… the next time you see “100% conception rate”, “99% fertility”, or “all mares in foal in 2008″, have a think about what these statistics are actually saying. Ask the stallion owner or stud what a stallion’s fertility figures were “per cycle” last season. If they can’t give you an answer, they either don’t keep good enough records, don’t understand fertility statistics, or they know the rates are low and don’t want to tell you. Any of these reasons should be cause for concern!
Statistics can be very misleading!
