The right analogy is someone trying to make money selling defective weapons. Like selling a bow and arrow (a broken one) and claiming it’s a rifle.

How would you “notice a decline in the accuracy of the numbers”? Assuming I was given the right information, do the Jake figures for those two horses look “accurate” to you?

Just to give you a sense of what this means, as a practical matter— Conglomorate had a good trip. Others in that race did not, and if you’re using ground loss (as both “rifles” do) those horses earned better figures than the winner did. Meaning also-rans in the Valedictory ran as well or better than the runaway winner of the Cigar Mile did, on “bows and arrows”.

Does that make any sense to you? Do you think it’s “accurate”?

By the way— as those out there who have made their own figures can tell you, it’s always possible to blow a variant for an individual race with a field of lightly raced horses (that’s why we sometimes leave boxes). But these two races had a) large fields, of b) stake horses— figure makers had tons of data to work with. Which makes errors of this magnitude almost impossible to comprehend. You would almost certainly have to have the two circuits completely out of whack with each other (by a lot) to get there. And if you can make mistakes like that, with that much data to work with, what happens when you don’t have it?

Again— on TG the difference between the two figures is 10 points. On Jake it’s one point. That 9 point error is worth 15 lengths at a mile.