Possible het's

[Emilio]

I guess i was trying to reinforce my decision on going the possible route LoL!!!!Great info guy's!!! 75% probability is awesome.

[frankykeno]

You know everytime I think about probabilities....I just can't help myself but to go to the Hitchhiker's Guide to the Galaxy and their improbability driven spaceship (bet that says something about me )

[aaajohnson]

You know it!!!

[hoo-t]

Neil,
(or any of you other statistics gurus)

To turn this around and look at it from a slightly different angle...

Lets say you breed a homozygous (visible) albino to a 50% probable het albino and get 6 eggs. If the probable het IS het, you would expect to get 3 albino babies (statistically 50%).

If you get any albino babies at all, your 50% probable het has just been proven, and is now 100% probable.

Lets say you don't get any albinos; all babies appear normal. Obviously, all babies are hets due to the sire being homozygous, but thats not the point of my post. Your confidence level in that 50% probable het just went down. It doesn't mean she's not a het, just less likely (see Ed's post re: cin x cin). How would you QUANTIFY your decreased confidence?

Steve

[dr del]

In sobs per second




dr del

[aaajohnson]

Lets say you don't get any albinos; all babies appear normal. Obviously, all babies are hets due to the sire being homozygous, but thats not the point of my post. Your confidence level in that 50% probable het just went down. It doesn't mean she's not a het, just less likely (see Ed's post re: cin x cin). How would you QUANTIFY your decreased confidence?

Steve

Well first of all I am faaaaarrrrrr from a stats guru but, there are formulas for figuring confidence intervals (CI) based on related events and sequences, so I believe there is way to figure the CI that possible het is not a het based on the example you have given. I would have to dig the books out, but the answer would be formatted something like this:

"Based on 6 non-homozygus babies being produced, there X% confidence (lets say 90%) that the possible het is indeed not a het."

Again, the chance still exists, so you would have to breed again, and each successive breeding with no homozygus animals would increase that confidence interval.

Any breeders have a rule of thumb when proving possible hets as to when you finally say "she is not a het"?

This is starting to make my hair hurt ......

In sobs per second

dr del

I agree!!!!

Neil

[hoo-t]

Any breeders have a rule of thumb when proving possible hets as to when you finally say "she is not a het"?

Neil

Thanks, Neil. Hmm. I'm starting to wonder if I have a stats book laying around somewhere! Its been 15 - 20 years since I took stats, but I kept most of those types of books. Anyone else know the formula?

Personally, if no homozygous animals were produced after 2 clutches, I think I'd give up and consider the "possible het" to be normal. My interest in this stems from the fact that we have 1.12 unproven "100%" het albinos (in addition to our homozygous pair), 1.0 unproven 66% het albino (daughter's pet), 1.1 unproven "100%" het pieds, and 0.6 unproven 66% het pieds. The males are all ready to go, but none of the females are big enough to breed yet. Just thinking about the future!

Steve

[jhall1468]

I guess i was trying to reinforce my decision on going the possible route LoL!!!!Great info guy's!!! 75% probability is awesome.

75% probability is good, however, it's wrong in your case.

What you are asking for is a probability of two independant events, meaning whether one snake is het or not has no regard on whether the other is het or not. So here's your probabilities:

Both Normal = 25%
Both Het = 25%
One het, one normal = 50%

Here's the formula for determining the statistical probability of the same outcome from two independant events (two normals or two hets). We use .5 because there is a 50% chance of het and 50% chance of a normal:

P(A and B) = P(A) * P(B)
P(.5 and .5) = P(.5) * P(.5)
.5 * .5 = .25

.25 or 25% probility of both being the normal, .25 or 25% probability of both being het and of course, 50% probability that one is het one isn't .

[aaajohnson]

True, but that means 75% probability that AT LEAST 1 is het (50% + 25%)


Neil

[Emilio]

This is a great thread!!!.... great info from everyone. Neil sorry about the hair ache. LOL