Explain This to Me like I'm A five- year-old

[Spot & Bubby]

 EPD:Expected Progeny Difference

I keep running into this term on the boards. Every time I think I've got my head wrapped around it, I can't make it make sense when the context changes. I think my confusion boils down to identifying the contributing factor.

In my head (and my own words), the definition sounds like:

• EPD is a score by which offspring within a generation (or other defined control group) can be predicted to perform above or below the measurable average of every other member of the generation (or other defined control group) for a specified trait based solely on
  • the genetic contribution of either or both progenitors, or
  • the verifiable performance of one or the other progenitor's collective progeny only.

If I had to explain it in my own words, I would say:

• The likelihood of a child to out perform or under perform his/her entire peer group (or generation), for a specific genetic trait, predicted only by either
  • the measurable performance of one or both parents for the specified trait, or
  • the performance of all of one parent's offspring as a whole.

There's a good chance that I'm over thinking it and making it harder than it needs to be because I don't understand where the the EPD comes from, who sets the standard, how it's maintained, or even how it's conveyed. I think traits can be (officially?) added or removed, but I'm not sure. Is it a document that "comes with" the progenitor?

On one thread, it was said that a good EPD doesn't mean it's a good bull and a good bull might not have a good EPD, if he even has one at all. If so, why's it a big deal?

Seriously, talk to me like I'm in kindergarten and help me understand this.

Kati

 

[SBMF 2015]

EPD:Expected Progeny Difference

I keep running into this term on the boards. Every time I think I've got my head wrapped around it, I can't make it make sense when the context changes. I think my confusion boils down to identifying the contributing factor.

In my head (and my own words), the definition sounds like:

• EPD is a score by which offspring within a generation (or other defined control group) can be predicted to perform above or below the measurable average of every other member of the generation (or other defined control group) for a specified trait based solely on
  • the genetic contribution of either or both progenitors, or
  • the verifiable performance of one or the other progenitor's collective progeny only.

If I had to explain it in my own words, I would say:

• The likelihood of a child to out perform or under perform his/her entire peer group (or generation), for a specific genetic trait, predicted only by either
  • the measurable performance of one or both parents for the specified trait, or
  • the performance of all of one parent's offspring as a whole.

There's a good chance that I'm over thinking it and making it harder than it needs to be because I don't understand where the the EPD comes from, who sets the standard, how it's maintained, or even how it's conveyed. I think traits can be (officially?) added or removed, but I'm not sure. Is it a document that "comes with" the progenitor?

On one thread, it was said that a good EPD doesn't mean it's a good bull and a good bull might not have a good EPD, if he even has one at all. If so, why's it a big deal?

Seriously, talk to me like I'm in kindergarten and help me understand this.

Kati

EPD’s are just a tool. To compare individuals within a breed. They are important to try to predict an individuals outcome for a specific trait. I.e. calving ease or growth.
Pure bred seed stock producers use them much more than commercial producers.

 

[Travlr]

EPD:Expected Progeny Difference

I keep running into this term on the boards. Every time I think I've got my head wrapped around it, I can't make it make sense when the context changes. I think my confusion boils down to identifying the contributing factor.

In my head (and my own words), the definition sounds like:

• EPD is a score by which offspring within a generation (or other defined control group) can be predicted to perform above or below the measurable average of every other member of the generation (or other defined control group) for a specified trait based solely on
  • the genetic contribution of either or both progenitors, or
  • the verifiable performance of one or the other progenitor's collective progeny only.

If I had to explain it in my own words, I would say:

• The likelihood of a child to out perform or under perform his/her entire peer group (or generation), for a specific genetic trait, predicted only by either
  • the measurable performance of one or both parents for the specified trait, or
  • the performance of all of one parent's offspring as a whole.

There's a good chance that I'm over thinking it and making it harder than it needs to be because I don't understand where the the EPD comes from, who sets the standard, how it's maintained, or even how it's conveyed. I think traits can be (officially?) added or removed, but I'm not sure. Is it a document that "comes with" the progenitor?

On one thread, it was said that a good EPD doesn't mean it's a good bull and a good bull might not have a good EPD, if he even has one at all. If so, why's it a big deal?

Seriously, talk to me like I'm in kindergarten and help me understand this.

Kati

I think you've got the gist of it. With more experience, using bulls for AI based on EPDs you'll see results you can correlate to how you chose them.

Most of my experience is in producing commercial cattle. My college education included EPDs but not at the expense of developing a good eye. So I'll stress this. EPDs are a tool to be used... but the product is where you make your money. The most important thing is to produce something the buyers want in a uniform package they can see with their eyes, and you don't get that with EPDs. You do get it by developing an eye for the type of cattle the buyers want. Consistent uniformity is phenotype.
Now that said, I've used EPDs to select bulls to breed the heifers I sold as replacements because some buyers were impressed with numbers. And I've selected bulls before I sold replacement heifers based on EPDs, but never until I selected for type/conformation. If the bull doesn't look like what I want my calves to look like, he's not selected no matter what his EPDs are.
I've seen some sorry bulls with great EPDs. I've also seen some great looking bulls with lousy EPDs. Why would anyone choose either when there are great looking bulls with EPDs to improve the flaws in your cows?

 

[Dsth]

I have been looking at EPDs for 40 years when selecting AI bulls for my herd and I still don't understand them. this how I kind of understand them. a zero EPD means that particular bull is likely to pass that trait to his offspring that is the average of all the animals of that breed that they have data on. some traits are better with a + and some are better with a - before the number posted. for example, BW (birth weight) depends on what you want. A smaller calf for a first calf heifer you will want a - number. if you want a larger calf, you look for a + in front of the number. the larger the number, the more that trait will change from the average of the offspring. Now I am confusing myself so I better quit. hope it helped a little bit.

 

[Mark Reynolds]

@Spot & Bubby. Kati, if I were to try to explain EPDs to you like a kindergartener, I'd have to explain it to you from the aspect of the English language and spelling with its horrific list of rules and then, worse yet, the list of the exceptions to the rules. EPDs are different for each breed of cattle. Additionally, they are guides to individual traits to which different producers have different preferences. In other words, what might be a great EPD for one producer may very well be a terrible value for another producer. That said, an EPD is a predictor/guide. It's not an absolute and you shouldn't necessarily expect exactly what the EPD is predicting. EPDs are muddy at best and potentially misleading and erroneous, although usually only muddy. A better option over EPDs is to talk to a breeder about characteristics you are specifically looking for when a novice or starting out.

 

[wbvs58]

The baseline 0 EPD is from when EPD's first started so an EPD figure of say 100 for WW does not mean an animal is 100 lbs/kg better than the one standing next to it, it might have an EPD of 95 for WW so the expected difference in performance is only 5lbs/kg. Will you actually see that in reality? Sometimes you may, mostly you won't, a lot of other factors play into it such as the cows milk EPD and environmental conditions , it is a figure that statistitions come up with however if you select for it over a period of time one would expect that your herd would move in the direction you are looking for but you need to be happy with the phenotype as well.

Ken

 

[Spot & Bubby]

EPD's are just a tool. To compare individuals within a breed. They are important to try to predict an individuals outcome for a specific trait. I.e. calving ease or growth.
Pure bred seed stock producers use them much more than commercial producers.

I don't expect that I will need to understand EPDs in the near future, but I would hate to be at a table where the topic of discussion turns to this bull selection tool and have nothing to contribute. Having enough understanding to ask meaningful questions is a good start.

 

[SBMF 2015]

I don't expect that I will need to understand EPDs in the near future, but I would hate to be at a table where the topic of discussion turns to this bull selection tool and have nothing to contribute. Having enough understanding to ask meaningful questions is a good start.

So for me the hardest part of EPDs is that they are not the same across breeds. A low number say in birth weight is a good thing for most people in the angus breed but in Charolais you want a higher number in birth weight. So if you trying to be fluent in several breeds EPDs it’s a lot to remember.
I can talk Angus EPD’s fairly well. A very little in the Charolais and I can’t even hold a conversation with the Simmental folks on here when we start talking numbers.

Probably the most important things for commercial herd owners to know are; calving ease/ birth weight, weaning weight, yearling weight, and docility. When choosing a purebred bull for a commercial herd.

 

[Spot & Bubby]

Consistent uniformity is phenotype.

I have a soft gaze for misfits. Understanding this tool, in theory, is one thing, but developing that eye will probably go a long way toward selecting bulls that will improve my flaws.

When we're talking about on "flaws" in my cows, though, we don't mean a permanent failure, do we? Age, physical development, and inexperience would be considered "flaws" in my heifers that could be improved upon by a bull with a negative EPD for birthweight. am I understanding that correctly?

 

[Spot & Bubby]

I have been looking at EPDs for 40 years when selecting AI bulls for my herd and I still don't understand them. this how I kind of understand them. a zero EPD means that particular bull is likely to pass that trait to his offspring that is the average of all the animals of that breed that they have data on. some traits are better with a + and some are better with a - before the number posted. for example, BW (birth weight) depends on what you want. A smaller calf for a first calf heifer you will want a - number. if you want a larger calf, you look for a + in front of the number. the larger the number, the more that trait will change from the average of the offspring. Now I am confusing myself so I better quit. hope it helped a little bit.

Where do those plus and minus marks come from? The breeder? Are they standardized? Is it a proven number or is it assigned the bull before he ever starts doing bull things? If I am going to have to trust the person who assigns it, right? Who is that person?

 

[Travlr]

I have a soft gaze for misfits. Understanding this tool, in theory, is one thing, but developing that eye will probably go a long way toward selecting bulls that will improve my flaws.

When we're talking about on "flaws" in my cows, though, we don't mean a permanent failure, do we? Age, physical development, and inexperience would be considered "flaws" in my heifers that could be improved upon by a bull with a negative EPD for birthweight. am I understanding that correctly?

Flaws are permanent, if I'm understanding what you are saying. Bad feet, udders, post legged, light muscle, dairy influence, colors that will be discounted and passed to calves, etc.

 

[TCRanch]

Probably the most important things for commercial herd owners to know are; calving ease/ birth weight, weaning weight, yearling weight, and docility. When choosing a purebred bull for a commercial herd.

Agree. But not to confuse calving ease with birth weight. I want calving ease bulls for my heifers, but the average birth weight is 70-75 lbs. And the same bull used on my mature cows, the average BW jumps to 85-90 lbs.

Docility is huge in my world, but I don't put a lot of trust in the DOC. Too subjective to gauge by looking at a piece of paper and I believe it's situational/the environment. I also look at CEM because I retain my own heifers.

 

[Spot & Bubby]

@Spot & Bubby. Kati, if I were to try to explain EPDs to you like a kindergartener, I'd have to explain it to you from the aspect of the English language and spelling with its horrific list of rules and then, worse yet, the list of the exceptions to the rules. EPDs are different for each breed of cattle. Additionally, they are guides to individual traits to which different producers have different preferences. In other words, what might be a great EPD for one producer may very well be a terrible value for another producer. That said, an EPD is a predictor/guide. It's not an absolute and you shouldn't necessarily expect exactly what the EPD is predicting. EPDs are muddy at best and potentially misleading and erroneous, although usually only muddy. A better option over EPDs is to talk to a breeder about characteristics you are specifically looking for when a novice or starting out.

I might not be hearing what you are saying. What I'm hearing is that EPDs are kind of junk science that limit the beginner's ability to develop a skill neccessary for superior production quality. at the same time it provides veteran producers peace of mind that what they "see" in a particular bull quantifiabley agrees with their expertice. Either way. if the end cresult misses the mark, it's not the producer who was off target, it was the EPD.

or does that go too far?

 

[Spot & Bubby]

The baseline 0 EPD is from when EPD's first started so an EPD figure of say 100 for WW does not mean an animal is 100 lbs/kg better than the one standing next to it, it might have an EPD of 95 for WW so the expected difference in performance is only 5lbs/kg. Will you actually see that in reality? Sometimes you may, mostly you won't, a lot of other factors play into it such as the cows milk EPD and environmental conditions , it is a figure that statistitions come up with however if you select for it over a period of time one would expect that your herd would move in the direction you are looking for but you need to be happy with the phenotype as well.

Ken

I think I have been mathing wrong. (watch me confuse myself in real time)

I thought if the baseline is zero and the EPD is 95 for WW, that bull's offspring would perform better for that trait than 95%, of the field. I was not seeing the weight trait values as actual WEIGHT on the calf. I was seeing it as a comparison to the average.

 

[kenny thomas]

First thing is with most breeds 0 isnt the base line.

 

[Travlr]

Agree. But not to confuse calving ease with birth weight. I want calving ease bulls for my heifers, but the average birth weight is 70-75 lbs. And the same bull used on my mature cows, the average BW jumps to 85-90 lbs.

Docility is huge in my world, but I don't put a lot of trust in the DOC. Too subjective to gauge by looking at a piece of paper and I believe it's situational/the environment. I also look at CEM because I retain my own heifers.

When I went to school docility was considered one of the most highly heritable traits. But that may have changed in the centuries since I was learning. There sure are a lot of docile animals that seem to have the odd renegade.

 

[GoWyo]

I pulled a screenshot of Angus sire Sitz Resilient 10208, which is a bull I use every year to some extent. The format may not have worked real well, but we can go through a couple EPDs. Starting with CED (calving ease direct) he is +4 with .92 accuracy (92%) based on 2657 data points (data turned in on heifers he was bred to), which is in the 70% percentile of the breed for calving ease. Next column is BW (birth weight) which is +.6 and is in the upper 35% of the breed (lighter BW in this case) with 97% accuracy based on 11,157 birth weights turned in. Based on this info I can predict that the calves will be born on the lighter end of breed average, but their shape is such that some heifers may have trouble calving to him despite the lower BW. WW (weaning weight) is +79, which is top 20% of the breed with 97% accuracy based on 9755 weaning weights of progeny turned in. That should get you started.


[H4]Sitz Resilient 10208[/H4]
[H4]Reg: AAA *19057457[/H4]
[H4]Bull[/H4]
Share
[ AMF-CAF-D2F-DDF-M1F-NHF-OHF-OSF-RDF ]
Birth Date: 02/15/2018 Tattoo: 10208
Parentage: SNP
Genomic: Angus GS
Genomic Prog: 8287
Parents Qualified
Breeder: 517050 - Sitz Angus Farm, Harrison MT
First Owner: 517050 - Sitz Angus Farm, Harrison MT
Owner(s):
25508 - Bar J V Angus Ranch, Fairview MT277427 - Dan E Ingalls, Casper WY346411 - Lunds B Bar Angus, Wibaux MT

Benfield Substance 8506	AAA #*16207085
Mohnen Substantial 272	AAA #*17292558
Mohnen Glyn Mawr Elba 1758	AAA 16396059
Sitz Stellar 726D	AAA #*18397542
Connealy Final Product	AAA #*15848422
SITZ Pride 200B	AAA *17776820
Sitz Pride 308Y	AAA #+*17050218
GDAR Game Day 449	AAA #+14691231[RDF]
Sitz Top Game 561X	AAA #*16559105
Sitz Pride 88T	AAA #15636920
Sitz Miss Burgess 1856	AAA *18395931
Sitz Rainmaker 10899	AAA 16266466
Sitz Miss Burgess 4381	AAA 16934054
Sitz Miss Burgess 1609	AAA 13263870
# Pathfinder + Embryo Transplant * Parents Qualified

EPD Percentiles

As of 12/17/2024​

Production
CED Acc % Prog	BW Acc % Prog	WW Acc % Prog	YW Acc % Prog
+4 .92 70% 2657	+.6 .97 35% 11157	+79 .97 20% 9755	+137 .95 20% 5340
RADG Acc % Prog	DMI Acc % Prog	YH Acc % Prog	SC Acc % Prog
+.27 .63 40% 31	+1.51 .63 80% 31	+.5 .93 50% 1002	+1.14 .94 35% 2096

Maternal
HP Acc % Daus	CEM Acc % Daus	Milk Acc %	MkH MkD
+18.5 .83 4% 713	+6 .86 70% 990	+22 .84 75%	175 990
MW Acc % Prog	MH Acc % Prog	$EN %
+56 .71 60% 74	+.4 .73 50% 44	-11 40%

Click to show/hide Management EPD Section

Management
Doc Acc % Prog	Claw Acc % Prog	Angle Acc % Prog	PAP Acc % Prog	HS Acc % Prog
+6 .91 95% 870	+.33 .88 4% 1034	+.36 .88 5% 1034	-1.76 .79 4% 564	+.18 .76 10% 138

Carcass
CW Acc %	Marb Acc %	RE Acc %
+41 .75 70%	+.94 .71 20%	+.53 .72 65%
Fat Acc %	Carc Grp Carg Pg	Usnd Grp Usnd Pg
+.031 .69 70%	8 54	694 2447

Click to show/hide Angus-on-Dairy $Values Section

Angus-on-Dairy $Values
$AxH %	$AxJ %
+112 45%	+80 50%

$Values
$M %	$F %	$B %
+93 3%	+76 80%	+140 60%
$W %	$G %	$C %
+87 3%	+64 25%	+275 30%
Phenotypic traits: CE, BW, WW, YW, SC, Claw, Angle, PAP Bold Phenotypes are included in the National Cattle Evaluation. More Info

25508 02062020; 277427 02062020; 346411 02062020
EPD & $Value Definitions | Why No EPDs?

 

[Ebenezer]

"WW (weaning weight) is +79, which is top 20% of the breed with 97% accuracy based on 9755 weaning weights of progeny turned in."
Weaning weight is a misnomer. It is a catchall EPD to include whatever is gained in some way but not totally in genetic potential to merely gain weight.

To me, EPDs are guardrails. You don't drive on or against guardrails - you stay on the road. If you hit a guardrail or merely rub it that is not good. My goal is never to breed outliers but to consistently create a core of cattle that are more alike (not exactly alike) but all with preferred function and performance. This requires either having goaled ranges of acceptable EPDs or trial and error use of animals with various EPDs to see if they work or not. The trial and error is expensive.

 

[simme]

Suggestion - Forget about the epd number for a trait and look at the percentile instead. For every trait in every breed, 50 percentile is breed average for the trait. Percentiles lower than 50 are better than breed average. Percentiles greater than 50 are worse than breed average. A 1 percentile epd is top 1% of the breed (only 1% of the animals have a better epd). A 99 percentile is bottom 1% of the breed (99% of the animals have a better epd). So, you never have to remember what the breed average epd number is for every trait. In the angus example above for Resilient, weaning weight epd is 79 pounds. I don't have angus cattle and don't know the breed averages. I don't know if 79 is good or bad (high growth or low growth). But easy to see from the epd info posted that 79 pounds is 20 percentile. So top 20% of the breed for weaning weight - so high growth. Who can remember breed average values for 5 different breeds when a breed might have 20 or more epd's (traits).

Even more important than the epd value or percentile is the accuracy. High accuracy epd's have a lot more value than low accuracy epd's.

 

[Caustic Burno]

@Spot & Bubby. Kati, if I were to try to explain EPDs to you like a kindergartener, I'd have to explain it to you from the aspect of the English language and spelling with its horrific list of rules and then, worse yet, the list of the exceptions to the rules. EPDs are different for each breed of cattle. Additionally, they are guides to individual traits to which different producers have different preferences. In other words, what might be a great EPD for one producer may very well be a terrible value for another producer. That said, an EPD is a predictor/guide. It's not an absolute and you shouldn't necessarily expect exactly what the EPD is predicting. EPDs are muddy at best and potentially misleading and erroneous, although usually only muddy. A better option over EPDs is to talk to a breeder about characteristics you are specifically looking for when a novice or starting out.

Well said .
Nothing takes the place of seeing the bull with his working clothes on and progeny standing in the pasture.