New Score Simlator

[terren2000]

I love the score simulator.  but is it possible to redo it and a new feature?

I want a feature that will give us an idea of our score if a judgement, 30 day late, or 60 day late notation falls off.  This will give us an idea of how our score will look if we can get some of these late payments of judgements removed.

I know it wont be exact, but at least an idea based on our credit history.

Of a separate type of simulator that is a little more advanced, in which we can speak with some1 in a chat, and have them give us verbal ideas of what our score would look like if this and that were done.

The biggest improvement though would be letting us know how the late payments and judgements being drop would affect our score.

[fused]

FICO scoring has a lot of moving parts, so its always difficult to know what will happen when a baddie comes off of a report. The link below might partly answer your question. It gives score ranges when certain baddies hit reports. Check it out!

Credit Missteps

[smallfry]

That would be the FICO Estimator. Try it.

[haulingthescoreup]

+1

FICO Score Estimator

Be sure to enter the info from just one report. In other words, not all of your inquiries, but just the number on EQ, or on TU, or on EX. Same for any other questions. You want to reproduce the info on one report. If your reports contain a lot of different info, just run the Estimator 3 times, each time imitating one report.

[MidnightVoice]

It was spot on for me - TU at the bottom of the Range given, EQ at the top

[fused]

I forgot about that link. It was spot on for me too. My TU score was right in middle of the score range given, while EQ was a little bit higher.

[MattH]

---

@fused wrote:

I forgot about that link. It was spot on for me too. My TU score was right in middle of the score range given, while EQ was a little bit higher.

---

For me the Estimator predicts 765 to 815, which is pretty close to my actual scores.

Experimenting with scenarios I find that if I add one major ding in the past six months while keeping every other answer accurate then it says 695 to 745, or about 70 points lost; a similar ding 1 to 3 years ago comes out 720 to 770, or about 45 points lost.  So a single major ding would cost me a substantial number of points but its impact would diminish fairly quickly if I then went a year or two with no further dings.

[RobertEG]

You wont  ever see such refinements.

The FICO scoring algorithms are very closely guarded trade secrets upon which a multi-million dollar industry is based.  That is why the FICO simullators are broad.

To get too refined, the FICO simulator would just enable more refined reverse-engineering of their scoring algorithms.

Yeah, it would be nice, but no, it wont happen.

[MattH]

---

@RobertEG wrote:

You wont  ever see such refinements.

The FICO scoring algorithms are very closely guarded trade secrets upon which a multi-million dollar industry is based.  That is why the FICO simullators are broad.

To get too refined, the FICO simulator would just enable more refined reverse-engineering of their scoring algorithms.

Yeah, it would be nice, but no, it wont happen.

---

In my world, biomedical data mining, when we pay somebody else for access to their database usually the contract language will be very specific about how we may use their data: for the right to just their conclusions we pay X dollars but if we want to do further analyses on our own computers the cost will be a lot higher.  For similar reasons I'm sure FICO's contracts with lenders have pretty strong language to prevent them from reverse-engineering the algorithm.  Otherwise any big lender could come pretty close by pulling FICO scores on a few thousand of their customer files and feeding the results to their own statistical software.

PS: for a few especially high-value databases the vendors actually require each individual employee who will work with their data sign an individual license document; I've signed a few such non-disclosure agreements myself.

[MidnightVoice]

---

@MattH wrote:

In my world, biomedical data mining, when we pay somebody else for access to their database

---

I just use Google.