EPD's are BS

[Bright Raven]

I started to use Generation 2 yrs ago, didn't personally like his calves so changed my mind.
Midten my question would be with milk at 20 how much milk do you want in your cattle? Are you just trying to raise that number on particular cows?
I
Numbers changed alot with the 1 step, but they readjust every couple of yrs and change lots of numbers.
Starting in the fall breeding season Im going back to breeding cows, not gonna cater to bull buyers and mess with all the extremes they think they require.

Barry, does that mean you are going to go back to using Phenotype with less use of the other tools bull buyers are emphasizing?

 

[J&D Cattle]

I have this on zero authority but was told that AAA had recently updated their EPD's with all current data and this caused the recent swings of some animals. My client heard that they had been updating some of their software or programming and hadn't actually input some of the recent reporting data for quite some time. Anyone else heard these rumors?

 

[Allenw]

I started to use Generation 2 yrs ago, didn't personally like his calves so changed my mind.
Midten my question would be with milk at 20 how much milk do you want in your cattle? Are you just trying to raise that number on particular cows?
I
Numbers changed alot with the 1 step, but they readjust every couple of yrs and change lots of numbers.
Starting in the fall breeding season Im going back to breeding cows, not gonna cater to bull buyers and mess with all the extremes they think they require.

"not gonna cater to bull buyers and mess with all the extremes they think they require."

I question if that isn't part of the numbers changing, data being manipulated by the breeder to make the numbers better. Then as more information comes in over time the numbers begin to average out.

 

[Son of Butch]

Case in point.....VAR Generation 2100, considered a high accuracy sire....right now.... at .41 accuracy.
I looked back at the spring catalogue and he was 44 at .27 accuracy.

Is less than 50% accuracy considered high accuracy?
To use epds a minimal understanding of math is needed, beginning with knowing the difference between high and low.

Winning a hand of blackjack vs a casino is a 48% probability, but in the long run crushes the vast majority.
41% will get you killed deader than dead

As Amarillo Slim said [probably to encourage a Tennessee opponent to draw to an inside straight]
Your odds are always 50/50... either you make it or you don't.

 

[dun]

The Red Angus Association considers 60 and below to be low to medium accuracy

 

[bse]

Ron
Yes I'm going to phenotype, the university here has brainwashed these people, they believe if a bull isn't high CED low BW and some crazy YW you loose thousands on calves.
With the advancement in sexed semen, that's all I'm gonna use and not bulls with crazy numbers, bulls that I like something in the pedigree.
My belief if I keep catering to these bull buyers in a short period of time I will give up more Maternal than I already have. After A I all out of cleanup bull will be steered.

 

[Son of Butch]

Ron
Yes I'm going to phenotype, the university here has brainwashed these people....
Tennessee University isn't doing a very good job of brain washing, IF they think .27 accuracy means something.

My belief if I keep catering to these bull buyers in a short period of time I will give up more Maternal than I already have. True, if buyers want a terminal bull and you want to only produce maternal it's lose lose for both parties.

 

[Chocolate Cow]

The majority of Angus bulls sold today are terminal sires. They're high milk, high growth, high input animals. Their offspring excel in the feedlot but not so much in the pasture. I've found I only need a +10 on milk to have acceptable calves. The females rebreed and stay in good condition while raising their calf. Their winter feed requirements don't hurt me so much either. I don't have big cows. 1,000-1,100 lbs. At one time, I chased the high EPD's with a vengeance. I paid dearly for it with open cows, defective calves, high feed costs and the cow herd looked like death warmed over. I bought a couple of bulls from Scott Shively in North Dakota. He told me to remember "Nature only tolerates average". I've got to have a maternal herd before I can use a terminal sire. And, all of those calves must be sold. Don't keep even one to put back into the herd.

 

[Bright Raven]

Ron
Yes I'm going to phenotype, the university here has brainwashed these people, they believe if a bull isn't high CED low BW and some crazy YW you loose thousands on calves.
With the advancement in sexed semen, that's all I'm gonna use and not bulls with crazy numbers, bulls that I like something in the pedigree.
My belief if I keep catering to these bull buyers in a short period of time I will give up more Maternal than I already have. After A I all out of cleanup bull will be steered.

Thanks.

 

[Midtenn]

Case in point.....VAR Generation 2100, considered a high accuracy sire....right now.... at .41 accuracy.
I looked back at the spring catalogue and he was 44 at .27 accuracy.

Is less than 50% accuracy considered high accuracy?
To use epds a minimal understanding of math is needed, beginning with knowing the difference between high and low.

Winning a hand of blackjack vs a casino is a 48% probability, but in the long run crushes the vast majority.
41% will get you killed deader than dead

As Amarillo Slim said [probably to encourage a Tennessee opponent to draw to an inside straight]
Your odds are always 50/50... either you make it or you don't.

I said var generation is considered a high accuracy bull. No, his milk is not high accuracy but overall he is a high accuracy bull. So, the question still remains....if it is possible for a number to go from 44 to 20 overnight, do we need to call BS on any milk numbers below .95 accuracy?????

 

[Midtenn]

According to AAA's own accuracy table..... At 30% accuracy, milk number should not change over 7 points. Butch here's some simple math for you ......44-20 = ?

I'm not here to argue, I just would like to understand how AAA can say one thing and do another. Should I totally disregard numbers below .95 accuracy? I'm starting to think maybe so....

To answer BSE's question....I am not after high milk at all. I would have never considered the generation bull at 44 milk and -30 something EN. He always looked to me like he would throw good saleable bulls for sure, and now , if his new numbers are somewhat reliable, I would keep his heifers. But that's the part I don't know whether to trust.

 

[Midtenn]

According to the aaa database , there's only 2 bulls born since 2005 with.95 milk accuracy. So do we all need to use only those 2 bulls for maternal sires, or do we lower our accuracy standards to 90% ? 80% ? What's the magic number? Or do we go back to phenotype, pedigree, and forget the numbers? Hence the origin of this thread.....maybe epd's really are BS ?

 

[Son of Butch]

No... because it was a base change update combined with improved accuracy formulation.
All progressive breeds have major epd changes about every 15 years or so to fine tune info and improve accuracy.
It's A One Time dramatic change for some individuals... ESPECIALLY unproven (LOW Accuracy) animals.

Just because an individual is high accuracy for 1 trait (birth weight) does not make him high accuracy for all traits.
The associations go to great lengths to collect and present their information and accuracy for each trait.
The AAA has been transparent in explaining the collection and processing changes with hows and whys.
It has to be frustrating for them having uninformed individuals call epds B.S. because a .27 accuracy trait changes
dramatically to a more correct evaluation with a higher, but still unproven, rating of .41 accuracy.

IMO
.49 and below is unproven, semi proven is .50 to .69 proven is .70 -.90 and highly proven is >.90
Farmers willing to pay as much for an unproven bull as a proven sire baffles me.
Guess they are motivated by fear... maybe afraid they'll miss out on bragging rights to the next hot thing.

In addition beef milk production is tricky to accurately calculate because...
All Beef Farmers Are Lazy... you want to fix it?
Run the cows into a barn to have their milk weighed and tested, then Milk edp accuracy would improve greatly.

jk

 

[Son of Butch]

if his new numbers are somewhat reliable, I would keep his heifers. But that's the part I don't know whether to trust.

IF you are comfortable with 41% accuracy and a 20% margin of error... go for it... I prefer > 69%

 

[wbvs58]

Butch, it is my understanding in Australia that the milk EBV in beef breeds is calculated from the weaning weights of a cows progeny.

Ken

 

[Son of Butch]

Butch, it is my understanding in Australia that the milk EBV in beef breeds is calculated from the weaning weights of a cows progeny.

Correct and other growth factors including 'milk thief' calves can affect an individual's weaning weight.
I was admiring a calf's growth from one of my heifers that I thought must be a good milker, until I saw him stealing an
extra meal from another smaller calf's momma, plus some calves will hit the creep feeder harder than others.
Making it hard to give credit where credit is due.

 

[talltimber]

And don't forget the old, "Garbage in, garbage out" saying. I'm satisfied that happens enough to cause inaccurate numbers in some animals. Assoc wide? idk. I would hope not.

 

[Chocolate Cow]

Middtenn-did your milk search in the AAA datebase show this bull? Jad Phelps Z1708 Reg: AAA 17472368

 

[bse]

Butch
Im not sure what you are talking about with UT acc at .27 all they require to qualify is .15 and back to what I said they have brainwashed everyone.
I think you hit on all the points to the milk EPD no doubt one of the toughest to get right.
AAA seems to redo numbers every couple of yrs, it pushes a lot to the top drops some to the bottom, so if you don't like the 1 step wait about 2 yrs it will change!
For me im still not calling BS on EPDs

 

[Midtenn]

No... because it was a base change update combined with improved accuracy formulation.
All progressive breeds have major epd changes about every 15 years or so to fine tune info and improve accuracy.
It's A One Time dramatic change for some individuals... ESPECIALLY unproven (LOW Accuracy) animals.

Just because an individual is high accuracy for 1 trait (birth weight) does not make him high accuracy for all traits.
The associations go to great lengths to collect and present their information and accuracy for each trait.
The AAA has been transparent in explaining the collection and processing changes with hows and whys.
It has to be frustrating for them having uninformed individuals call epds B.S. because a .27 accuracy trait changes
dramatically to a more correct evaluation with a higher, but still unproven, rating of .41 accuracy.

IMO
.49 and below is unproven, semi proven is .50 to .69 proven is .70 -.90 and highly proven is >.90
Farmers willing to pay as much for an unproven bull as a proven sire baffles me.
Guess they are motivated by fear... maybe afraid they'll miss out on bragging rights to the next hot thing.

In addition beef milk production is tricky to accurately calculate because...
All Beef Farmers Are Lazy... you want to fix it?
Run the cows into a barn to have their milk weighed and tested, then Milk edp accuracy would improve greatly.

jk

This got all off on a conversation about milk epds. I was just using that as an extreme example of how drastically inaccurate epds really are. I do not think ANY of them are close to accurate until.95 or better and even then they will change more than the AAA accuracy table suggests (one third of the time). ABS advertises Generation as a "high accuracy multi trait sire". You missed my whole point....my point being EPD's are way overrated. I feel more comfortable trusting the history of my own herd, cows and bulls. I feel that is much more accurate than any epd's. I don't measure milk production, but have a pretty good idea of it based on fertility, body condition, calf growth, and inputs.
I have to use epd's to some extent when bringing in new bloodlines, but it's secondary to phenotype, progeny data (if available and trustworthy) and performance (if my gut tells me its true).
The most aggravating part to me is that in order to sell bulls, the epds numbers have to be in the range. It doesn't matter to most customers that the yearling bull they purchase is only .05 accurate, but the ced, bw, and ww numbers matter big time. You try to tell someone epd's at .05 or even .35 don't matter and see how many bulls you sell.
As I said in my first post, I will be natural breeding heifers to a 0 ced and 3.0 bw yearling bull come fall because I trust my own witness to history more than someone else's numbers. After that he will go on 25 commercial crossbred cows as a terminal at least for one season...his ww is only 50. But again I trust my own eyesight and ability to read scales more than numbers someone else derived. Truth is, he is probably neither calving ease or terminal, but balanced. But in my opinion he's the best bull been on my place in 2 years, though possibly unsalable to a epd informed customer....ain't that funny.