The value of inquiries: EX vs. EQ data points

[Revelate]

---

@Asilomar wrote:

---

@Anonymous wrote:
Yes each CRA has certain tweaks they can make to the algorithm such as when you get dinged for number of accounts reporting and various other things there are small tweaks that can be made and have been made to the algorithm between bureaus. That’s the whole reason we do all the testing on each algorithm and on each bureau because there are differences.

---

So are you saying Sarge is incorrect?  That EX and EQ do indeed weight inquires differently?  Or you do think the 1 point / 5 point differential is due to the slight differences in the two reports?

---

I'm saying the entire rest of the algorithm may well be weighting things differently so you can't simply point at inquiries as the sole difference and say "Ah ha, here's the point spread!"

It just doesn't work that way under FICO 8... maybe in FICO 9 it would actually be closer, but one of the major lender complaints is that the exact same data on all three bureaus scores differently on FICO 8.

[Asilomar]

---

@Revelate wrote:

I'm saying the entire rest of the algorithm may well be weighting things differently so you can't simply point at inquiries as the sole difference and say "Ah ha, here's the point spread!"

Ah, got it. 

It just doesn't work that way under FICO 8... maybe in FICO 9 it would actually be closer, but one of the major lender complaints is that the exact same data on all three bureaus scores differently on FICO 8.

Which is why I always thought each CRA had control/some control over the maths or that they each had their own version of "The FICO8 Algorithm".

---

[Anonymous]

+1 to what Rev said. I don’t know if they weigh inquiries differently or not. That’s something that would have to be tested, but I do know that identical data will not result in identical scores due to small differences in the algorithm at the CRAs. As you said, they do have the ability to customize the algorithm to a small degree.

[Asilomar]

Thanks for the clarification, Birdman.