What's wrong with chasing numbers?

[3waycross]

This just goes to show that very, very, very few people understand statistics. And without understanding statistics you can not fully understand EPDS.

3Way is one of the better (smarter?) posters here and he has shown that he does not understand the basics of the bell curve.

Average Joe Rancher is being led astray with EPDS and that is my story and I'm sticking to it!

Well now I didn't say I don't understand it, but, you are right I don't understand it. I am a whiz with language, Logic(both inductive and deductive) and a few other things but MATH and all things related to it are beyond my scope. I am pretty good unless they start using letters instead of numbers or freakin graphs then I am cooked.

However I DID think I understood EPD"S. Apparently what I do not understand are EPD accuracies. I guess I'll just have to back to pickin the pretty ones and hopin they work out.


BTW WLM I owe ya buddy. Most people when refering to me as smart usually include the word a-- along with the smart part. Thanks buddy, if you are ever in Colo and need anything just call.

 

[3waycross]

I went back and REALLY looked at what they are saying there. At least it makes some sense to me but I have to admit I now have a raging headache. Thanks a lot Dun!

 

[dun]

I went back and REALLY looked at what they are saying there. At least it makes some sense to me but I have to admit I now have a raging headache. Thanks a lot Dun!

Take 2 aspirins and call me in the morning

 

[Chuckie]

I was reading along, and then SRBeef hit the nail on the head about using the numbers! So, no need to hammer that any further.

Then on the percentages, the higher the better. But, there are always going to be some variations. It can't always be exactly what is printed. A variation of good numbers has a lot of give to it. A low BW (Angus) from a +1.0 acc. 70% acc. Then for the numbers to range from +.05 to 2.0, I can live with. Hopefully it would not go to a higher number. Then the low accuracy, you can be thrown a curve ball.

We bought a heifer one year and her BW was in the sale paper as 4.3 and got her home, and received the reg. papers and the BW was a 6.1.
The accuracy was .05% and when I got the papers, it was 35%. Now that was not what I wanted to see. It still has 65% to change. So, low accuracy does make a difference. A cow with very high numbers makes a difference too.
Chuckie

 

[LFF]

Chasing numbers? I finally had to put in my 2 cents worth.

Epd's have really been bashed for too long. The problems with numbers is lack of proper use and understanding. Lets hypothetically say W bought cattle that had great numbers yet were not the correct numbers for his herd, and he did not understand epd variability.

W brings home some heifers that have high epd's for growth and milk believing that he will increase his weaning and yearling weights. He also purchases a bull from H that has even higher epd numbers for growth and milk and is proven. He then turns them out on poor pastures and expects them to out perform his older cows that have lower epds.

When calving season arrives many of W's new additions do not calf and the ones that do are poor and weigh less than his older cows calves that have low epd's. :cry2: W now realizes that his heifers epds may be incorrect because H's bull is proven. W then buys mature cows that have decent epd accuracies that again are high in growth and milk from a well know breeder S. Now he thinks :idea: that his next crop is surely going to be what he originally hoped for. Next calving season lo an behold he experiences similar results as the previous year. :shock:


Now W is mad :x at H because he had to be lying :???: when he turned in data to the association. W now bad mouths H and says that no one should buy his bulls. W next tries AIing his high growth cattle with different bulls that have high growth epd's, again he has similar results.

W now is convinced that epd's are a crock of bull. :nod:

There is nothing wrong with chasing numbers if you truly understand what numbers are ideal for your orperation. You must be prepared to sell cattle that do not meet your expections regardless if they have the epd target number's that you desire. Many will not accurately produce according to their stated epd's.

As others have said higher or bigger is not alway better. Finding your optimum epd's takes years of experience and trials.

One set of epd's Herefords have which can be utilized is the indexes. Don't use the CHB index for producing momma cows. Use the BMI index for herd building.

 

[alexfarms]

Start your search here:

I once sold a bull to a man who was very attentive to epd's . The bull I sold him had a negative bw epd and a bw ratio of over 100. Not surprisingly, he was dissappointed in the bull. He said the bull never threw any really big or really small calves, but his bull calves were in the upper 80's and just a little too thick for him. He went on to explain that he believed the bull would have been alright if he just would have thrown some small calves to bring the average down like most of his low bw sires did. I responded 'so you're criticizing the bull for being consistent?'. He sold the bull, but not before enough data had been sent into the TPR system to get him listed as a bw trait leader. I beleive one of the drawbacks of epds is they don't express a bull's consistency, only averages and that is kind of what your bell curves expresses also. Some pretty smart people have told me I am wrong and epds do express consistency. They may be right, but I don't see it. I think epds are displayed with the assumption of consistency, which often isn't there.

 

[Red Bull Breeder]

Only thing wrong with chasing numbers is carching them before you have enough cow to back them up.

 

[Julian]

I beleive one of the drawbacks of epds is they don't express a bull's consistency, only averages and that is kind of what your bell curves expresses also. Some pretty smart people have told me I am wrong and epds do express consistency. They may be right, but I don't see it. I think epds are displayed with the assumption of consistency, which often isn't there.

John,
This has to be one of the best passages I have read in a long, long time. Thank you for sharing it with us.
Respectfully,
Dwight

 

[KNERSIE]

One thing that it seems most people don't understand is.... bulls don't become more consistent or more prepotent with higher accuracy EPDs, they just become more predictable. There will still be outliers too, though.

 

[Santas and Duhram Reds]

Thought I would add that there are a lot of desirable traits that epds do not reflect or predict.

 

[pdfangus]

Thought I would add that there are a lot of desirable traits that epds do not reflect or predict.

don't make a list or someone will start creating epd's for em.

 

[Third Row]

Has any association actually released the methodology used in calculating EPDs? I've tried on several occasions to find out (to model for an educational tool) but always came up blank.

 

[KNERSIE]

Has any association actually released the methodology used in calculating EPDs? I've tried on several occasions to find out (to model for an educational tool) but always came up blank.

Its probably considered intellectual property.

 

[3waycross]

Has any association actually released the methodology used in calculating EPDs? I've tried on several occasions to find out (to model for an educational tool) but always came up blank.

Its probably considered intellectual property.

Well I gues on any given day THAT would have to be considered an Oxymoron.

 

[alexfarms]

Has any association actually released the methodology used in calculating EPDs? I've tried on several occasions to find out (to model for an educational tool) but always came up blank.

I think the mathematics used to be written at the beginning of the old sire summaries, when they used to be in print. If you have access to any of the early APHA sire summaries I believe it would be in some of them. It used to be done by the University of Georgia, they may still be involved in it.

 

[bigag03]

The math behind the calculation is called BLUP (Best Linear Unbiased Prediction). It is a series of matrices that combine individual performance, parent performance, full and half sib performance, and most importantly, progeny performance. It is a weighted calculation where obviously, progeny data is the most heavily weighted. The larger the database is behind the data, the stronger the math is and the more reliable the predictions are (this is why Angus EPDs are the most reliable...there is more data).

The common misconception with EPD's is that they represent some given value (1.0 BW = 75 lbs)...NOT TRUE. The truth is that they only express the average of calves compared to the average of some other bull's calves. A set of EPD's by themselves mean nothing...they have to be compared to something (e.g. another bull or breed average). As the accuracy of an EPD increases, the variability in production never changes, but we become more confident in how that bull's average compares to others.

 

[WichitaLineMan]

>>As the accuracy of an EPD increases, the variability in production never changes, but we become more confident in how that bull's average compares to others.<<

Good post.