Pedigree Analysis is limited to the info we have on hand at the time.
When we give someone an analysis of their pedigree we look at the pedigree and see if there are any nn (negative horses in the pedigree (we can then rule that line out). Then we look to see if there are any positives in the pedigree and then we look if there are any horses with positive foals. (A horse with a positive foal is not necessarily positive as the foal may have gotten the positive gene from the other side of his pedigree) and finally we look to see there are any suspects.
Obviously if a horse has 2 nn parents we can declare the foal as nn as well.
But if a horse has as an eg. two n/P1 parents - we can not declare that foal to be positive. Even if both parents are n/P1 the foal can still have a chance to be nn - so we can never guess, based on pedigree alone if a horse is going to be positive - all we can do is highly recommend testing.
There are two cases where we might guess a horse to be positive - one with 98% accuracy and one with about 70% accuracy.
Case 1 - a parent is homozygous for a genetic defect - eg. P1/P1 - the foal will automatically inherit a P1 gene - the only margin for error here is if the parent's result was recorded wrong.
Case 2 - a parent is n/P1 and a the foal has list of obvious PSSM1 symptoms. Margin for error here is that PSSM2 symptoms are very similar to PSSM1 - so the foal has a 50/50 chance of being P1 but it could be nn and positive for PSSM2 - further pedigree analysis looking for PSSM2 suspects can help increase the odds of this prediction to being more accurate.