Are we too dependent on EPD's?

Help Support CattleToday:

Joined
Jan 12, 2022
Messages
11
Reaction score
16
Location
Michigan
As the industry continues to grow, and as we look for various ways to advance our own causes, EPD's have played a big part of this. The creation of EPD's decades ago now seem to be as prevalent as they've ever been with marketing tools paying homage to certain EPD categories (caving ease bulls as an example) and catalogs being littered with EPD's and projected mating EPD's are pretty standard now. However, am I (or we) too dependent on these numbers?

As an example, I will provide two sets of EPD's on two separate heifers (both PB Simmentals). Red = Below breed average (percentile rank)

CEBrthWeanYearADGMCEMilkMWWStayDocCWYGMarbBFREAShrAPITI
13.00.563.583.00.125.421.152.813.310.46.6-0.490.16-0.0910.79-0.41124.271.5
25%30%95%99%99%60%70%95%85%65%99%20%40%35%75%15%65%80%

CEBrthWeanYearADGMCEMilkMWWStayDocCWYGMarbBFREAShrAPITI
15.4-2.080.5124.20.277.526.666.822.018.439.0-0.430.34-0.0741.15-0.33168.991.1
10%4%40%30%30%20%30%30%3%1%20%50%15%75%10%60%3%10%

Looking at these two animals, what do you feel like you could deduce? For me, it's the difference in carcass traits and maternal traits.

I for one, as a purchaser of cattle across the country, rely on any and all data possible to help make decisions when certain animals are bough sight unseen. I am also a bit of a visual learner so have created graphs that pull from a spreadsheet of EPD's I input. One graph is the heifers/cows as standalone animals. The other two graphs are progeny EPD's from certain matings to bulls we have. This is all compared to breed average and used to see, again visually, where animals place compared to each other and to the average.

Understanding what EPD's are, how they're used, and that there is a level of change that will certainly happen, I can't help but look at these two sets of EPD's and be swayed toward one over the other. Even though from a visual aspect (in person, pictures, videos, etc..) they both look like sound heifers.

I am not one to think that these numbers are arbitrary or not based off science like some. I do believe these came from a place of scientific testing as a foundation and then sprawls out using mathematical equations to provide the best possible insight into these animals as possible (with variation). In short, I feel they should be looked at, but to what degree? Am I (we) too dependent on these numbers as cattle owners/breeders?

* I ask because I am looking at some spring heifers and utilize my sheets and charts to help narrow a decision. Also because I was reading last night on here and it appears certain people feel these are taken too serious and some even feel these numbers aren't based off science and are completely arbitrary. So let's start that conversation.
 
Last edited:
@Ranchin' Randy I've recently had some similar threads about how traits are passed on to the next generation (sometimes NOT passed on).
There's one very important thing to remember, EPD is Expected Progeny Difference, sometimes you don't get what you expect!

Lots of variables come into play and I was even recently introduced to a "new to me" concept of gestational programming. I definitely don't understand it all, but I feel like I'm learning along the way with the help of the forum members that are more knowledgeable about these things.

For very large operations I think it's a numbers game with playing the averages. For a smaller guy like me I would prefer a higher success ratio than just playing the EPD averages. That's why I try to see what has worked for others on this forum.

Eventually I think EPDs will become "Old School" and the new method will involve gene editing or gene splicing. In other words, for the right price, you can pick all the traits you want, and they will create it for you in the lab and implant it into your cow.
There I said it.......in the future you can alter the genetic makeup of your cows to create what you want. I may not agree with it, but some people will play God with their cows. Some with deep pockets may already be doing it.
 
EPDs are an average number of the assumed offspring. EPDs have more inflation than the national economy. The most unfair part of EPDs are the pivot year. It discounts older cattle and that is just not true. But it is good for sales so that tells me that sales are more important than truth.

Why are there discussions on this site and others about guarding bulls with bad feet, some about disposition and mothering instinct and the constant, constant question about "tell me about the daughters" ... if EPDs work? The point is that AI sires in catalogs and ranches are picked by EPDs and salability of the EPDs. Some fudge and force feed to get 1000 pound weaning weights that no purchaser can produce. I cannot imagine the pondering someone does for EPDs from a force fed bull with high EPDs and the lesser data on the next generation of normally fed animals. Others sources for AI studs are Wagyu-like selection of bulls so that the cow side is minimal in concerns. The dairy industry is not in decline if the EPDs for milk are anywhere near correct for the latest wonders of the semen catalogs and 4 color sales catalogs. I had a friend who told me he was rambling through some papers and found a 10 year old semen catalog. Not a single sire is still listed in the current catalog. The turnover rate is not due to success but due to failure.

You can use EPDs as bumpers or guard rails and know that you do not want to run off of the road. And you can note herds that have trends: high % low HP EPD, excessive CED in a lot of sires, negative CEM in a lot of sires... The skew of the herd is more important than the single #s of any one star bull they have.
 
I personally think epd are a waste and hurt more then they help. 10 year old epd # is equivalent to the same epd # today. Not realistic numbers based on real world facts.
There are enough advancements in genetics that their is no reason why genetic info should be what breeding decisions should be based on . Mating outcomes calculators should be based on genetics not random made up numbers. You can already do genetic testing for defects as will as hybrid vigor , milk protein so there is no reason that genetics can't be used. Then why isn't it already standard in all breed associations that 100 of all registrations must be genotype matched to both parents in order to be registered?
For the same reasons genetics are used and promoted in breed associations . Because they are afraid of what will actually turn up if the genetics are actually studied
 
I used to study and fret over EPD# but I have got over that, went to eyeballing bull and parents/siblings it all worked out.
Now I have been buying from one breeder for some time and I just call him and he sends pics of the type of bulls that I am interested in and I pick the ones I want. Don't believe there are EPD's for the breed of bulls that I have been using.
 
You can use EPDs as bumpers or guard rails and know that you do not want to run off of the road. And you can note herds that have trends: high % low HP EPD, excessive CED in a lot of sires, negative CEM in a lot of sires... The skew of the herd is more important than the single #s of any one star bull they have.
Agreed 100% here. The herd as a whole is an important aspect I feel get's lost at times to buyers. Another being the herd's environment and how it also impacts phenotype.


There are enough advancements in genetics that their is no reason why genetic info should be what breeding decisions should be based on . Mating outcomes calculators should be based on genetics not random made up numbers.
I guess I'm curious why you think these numbers are made up? Genuinely curious as I am not agreeing or disagreeing. However, I will add that a quick search shows that the formulas used for EPD's take into account not only performance numbers submitted by producers, but also any and all genomic details from their given pedigree.


Correct me if I am wrong, but you feel the numbers are arbitrary and are not based on genomic testing? Is that the long and short of it?

I used to study and fret over EPD# but I have got over that, went to eyeballing bull and parents/siblings it all worked out.
Now I have been buying from one breeder for some time and I just call him and he sends pics of the type of bulls that I am interested in and I pick the ones I want. Don't believe there are EPD's for the breed of bulls that I have been using.
I don't find myself there, but I find it very easy for someone to be there if they don't reel themselves in. Seedstock producers and breeders across the world have created this idea that the EPD alone speaks to the quality of the animal. And I just don't agree.

As I stated above, I do look at every animals EPD's, but it's just a matter and question of how much weight we should put on it. I am a believer of P=G+E (Phenotype = Genomics + Environment) so do what I can to understand the genetics behind the animal, but nothing beats getting eyes on (that specific animal, it's dam/sire if possible, the herd it comes from, etc...)
 
True breeders are few and far between at this point in time. AI has as much to do with were we are at today as epd's do. Everyone wants to use the latest and greatest bull Most of which are gone before they are proven to be replaced by the new generation's greatest.
 
True breeders are few and far between at this point in time. AI has as much to do with were we are at today as epd's do. Everyone wants to use the latest and greatest bull Most of which are gone before they are proven to be replaced by the new generation's greatest.
I definitely agree there. Like any industry, name recognition can persuade the public away from what's good to what's "hot" or popular. From an EPD perspective, I've seen plenty of bulls that could more than compete with the popular high selling ones. They were just owned by the wrong ranch.

It's a seller's world. Everyone wants to be ahead of the curve. No one wants to be that operation that has a herd based on "old" genetics. While I agree, to a degree, it has gotten a little ridiculous.

Certain Cattle Co's could sell a pig as a herd sire and still get $25000 for it. Does that point to a problem with the sellers, the buyers, or the industry? I don't know. Just a small guy trying to navigate the waters.
 
I guess I'm curious why you think these numbers are made up? Genuinely curious as I am not agreeing or disagreeing. However, I will add that a quick search shows that the formulas used for EPD's take into account not only performance numbers submitted by producers, but also any and all genomic details from their given pedigree.


Correct me if I am wrong, but you feel the numbers are arbitrary and are not based on genomic testing? Is that the long and short of
Why do I think the epd # are made up ?
Simple to answer. Because they are.
They are the result of a mathematical formula that is not based on genetics. They do not accurately predict the genetic potential across generations or across different environments. And are very easily manipulated.
If you took four clones of the same animal and sent them to four different regions of the us do you think the epds would be the same ? Do you think an animal raised in the Midwest and was feed huge amounts of creep would have the same epds as one that was raised in the west on a BLM permit with no creep feed? They should if they are identical genetically. But we all know they wouldn't even have close to the same epds.
Epds are too easily manipulated and are nothing more than a tool to put more money in certain breeders pockets and are not based on genetics.
 
Last edited:
I look at the sale picture first, if I like what I see I move on to pedigree, look at the dams production record, if I like what I see there I wait for the video to come out. Then I might call a sales rep or the producer if I'm interested and get another opinion. The sale reps know what I like and what I don't. Been dealing with the same few for years. At that point I will look at the EPDs. I also only buy January born females now bred to calve in January. That's when I calve the bulk of mine and I am finding it helps with fertility and fitting into my program. Same applies to bulls January born only. EPDs have very little influence on my program unless something sticks out real bad. That can make it hard to sell there progeny. But it has to be way out of whack. I'm also raising Simmentals
 
Randy, you really jumped right in on your choice for a first topic. My wife says it is time to eat now. But I will be back to get in on this. Just one point first.
Way more important than the EPD number is the accuracy of that number. If someone harvests 300 bushels of corn per acre with an accuracy of 20%, is that good or bad? If WW EPD is 90 with an accuracy of 0.20, good or not? I can't tell. But if WW EPD is 90 based on data from 1000 calves, accuracy will be high and that EPD should be used as part of your selection process. Unless you believe in conspiracy.
 
I wish more objective info like the the Circle A Ranch/ABS angus sire alliance was available:
Current breed associations epd data is reliant on breeders providing accurate data. I'm sure the majority of breeders report accurately, but it only takes a few clowns to ruin the whole system.. and then it gets worse when the clowns start colluding..
 
Randy, you really jumped right in on your choice for a first topic.
I tend to do that in life. In all reality, I don't have many questions at the moment. Cattle aren't new to us as we've done it all our lives. What peaked my interest most was the amount of people I read on here that disregard EPD's in their animal selections. This puzzled me as it is used by us as a selection tool. It is, by far, not the only tool but I do put some weight behind the ASA Herdbook.

I was more curious on where other's stood with EPD's being used as selection tools and how much weight others put on these numbers. Clearly some believe they're crap, others put weight into it, and some likely live by these numbers.
 
So where do EPD's come from and how are they calculated? Random number generator? I don't think so. It is an estimate, an expectation. It is not a promise or a guarantee. It is like a forecast based on data.
Take a trait that is easily measured like birth weight. You just need a scale. Simmental has a planned mating EPD estimating tool. Just plug in the sire and the dam that you plan to mate and hit the button. It is mostly just an average of the EPD's of the sire and dam. (I say mostly because there may be some "expected" adjustment based on heterosis depending on the mating.) We know that the actual mating will combine genetics from the parents with some variation. There is no equation that calculates the birth weight of the calf from the EPD's. But if both parent families have heavier than average birth weights, isn't it logical that there is a high probability that the calf they produce will have a higher birth weight than average for the breed? But also possible that it could be low birth weight?
EPD's will predict a higher expected bw. If the actual bw from the scale is lower than breed average, is the EPD wrong? Maybe or maybe not depending on environment. Let's ignore that environment thing for now.
So bw epd is high and the calf is one of the lightest in the pasture. Well maybe that EPD is wrong. Let's look at the accuracy. Say this is the first calf from this heifer. EPD's assumed she would genetically be a mathematical average of her parents. Suppose that no birth weights were recorded for any of her ancestors for several generations. That BW EPD can't be accurate, can it? But it will still be calculated with the best data available based on all the ancestors.
EPD's are most accurate when data is measured and submitted. So you start recording birth weights on her calves. First one has a lower than expected bw per the scale. The EPD calculating machine will take note of that and make an adjustment of the EPD based on what we will call a statistical model. How much adjustment? Well, she has only had one calf. Maybe it is a statistical outlier based on how those chromosomes combined. Her BW EPD accuracy will still be pretty low. Now after she has 10 calves with recorded birth weights, we and the EPD model machine will have a clearer expectation of her BW EPD and the accuracy of that EPD will be much higher. And can be trusted more.
Say there were no EPD's. Your evaluation of BW could only be based on what you see from the birth weights of the cow family and their calves. That's sort of the same way that EPD's work. The more data - the greater the accuracy and confidence in the number.
Don't expect that a yearling bull can always be counted on to match up with his EPD's. Accuracy is too low. But even then, would you rather not have the prediction? If you are going to cut hay, do you use the weather forecast even though the accuracy is not 100% or just look at the sky 4 days before you hope to bale?
My message is that EPD's with low accuracy are better than nothing, but understand that low accuracy means lower confidence in the number. High accuracy EPD's - if you don't use those, you are missing out. High accuracy comes from lots of progeny - older bulls that have been used a lot. As has been said many times, EPD's are just one tool for evaluation and selection. Don't make them the only tool but understand them to allow them to be a tool.
 
Why use a probability or possibly or maybe predictor when the genetics are available. Genetics don't change with creep feed or manipulating wean date to your advantage. Over fed cattle will not change the genetics.

 
If we know the genetics of meat quality why aren't we using genetics instead of using some mathematical equation that can be and is very often easily manipulated?
We are.

I can login to the ASA Herdbook right now and tell you which EPD's are based on Genomic testing (and their parentage genetics), and which are based solely off producer provided statistics (and their pedigree [which may also be producer provided]). I understand your skepticism to producer provided numbers. It's an industry with a lot of cash flow as a whole, so there's bound to be corruption.

But the American Simmental Association does genomic testing for their Enhanced EPD's. I won't speak on any other breed association, but feel it would be similar. the ASA uses 100,000 DNA markers to more clearly define EPD numbers and their accuracy.
 
Show me where on the ASA website (or any other beef registry) where it shows the results of the genetic testing for any other trait then lethals . Also show me what weight these genetic tests have compared to what weight producer provided info. Is given for that same trait in the"enhanced" epds
 
To answer the question are we too dependent on EPD's, my answer is YES. In the early stages of EPD's they were said to be used as just a tool for comparison. Now it seems like they have taken on a life of their own, as a stand alone marketing angle. I'm not convinced that genomic enhanced stuff is much better either. I think the oldtimers that spent their lives actually selectively breeding cattle using their own eyes and experience of what worked and didn't work laid a good foundation, and unfortunately all the fads, numbers, and AI bull of the month groupies have tried to make the fine art of cattle breeding into something more complicated than rocket science and in the process are taking focus out of the breeding pasture and onto graphs, boxes and shiny pictures, that make a few people more money, and overall hurt a lot of good traits.
 

Latest posts

Top