cancel
Showing results for 
Search instead for 
Did you mean: 

Question about Score Reverse Engineering

tag
Neblett
Regular Contributor

Question about Score Reverse Engineering

I just did some math, and I'm wondering if I have any kind of case here.

My FICO score as of 6-15-07 was 668.

On 6-21, my score dropped 13 points down to 665, due to a hard pull inquiry - I applied for a Best Buy store card.

On 6-27, my score dropped 33 points down to 622, due to the fact that my Target card account reported a history of past dues/late pays, a collection account reposted to the account, and a paid in full installment account changed some basic information (this last thing is a positive item and actually didn't alter my score).

On 6-29, my score dropped 37 points down to 585, due to two hard pull inquiries for the COAF and RoadLoans auto loans that I applied for on 6-16.

On 7-3, my score went up 10 points to 595, because a new positive account reported to my file.

On 7-6, my score went up 6 points to 601, because a couple of information changes and a positive account was added, but the inquiries from 6-21 and 6-29 were deleted from my account (believe it or not, CSC just deleted them - AND the rest of my inquiry history - while they were investigating some other inquiries which were outdated). So, at this point I have NO hard inquiries at all on my EQ report!

Now, here's my question - and it has to do with reverse engineering of scores. If they eliminated the inquiries and other negative aspects of my file that dropped my score 50 points (13 on 6-21 plus 37 on 6-29), why wouldn't my score be UP 50 points at the present time?

What I mean is, if those inquiries had never dropped my score, but the Target and collection account still posted as they did on 6-27, then shouldn't the following be the way it should look?

6-15: 668
6-21: 668 (No Best Buy inquiry - therefore, no 13 pt. drop)
6-27: 635 (33 pt. drop - Target and collection account)
6-29: 635 (No car loan inquiries - therefore, no 37 pt. drop)
7-3: 645 (10 pt. increase)
7-6: 651 (6 pt. increase)

Is this logical?

Anyone?
Message 1 of 12
11 REPLIES 11
MidnightVoice
Super Contributor

Re: Question about Score Reverse Engineering

I think you are laboring under two delusions:
 
1  that is it possible to understand FICO scoring
 
2  that FICO scoring is actually logical. 
 
Smiley Very Happy
 
 
The slide from grace is really more like gliding
And I've found the trick is not to stop the sliding
But to find a graceful way of staying slid
Message 2 of 12
Neblett
Regular Contributor

Re: Question about Score Reverse Engineering

That's my fear.
Message 3 of 12
MidnightVoice
Super Contributor

Re: Question about Score Reverse Engineering

I just think there are so many variables it is guess work.  Ageing of accounts, ageing of baddies, slow and fast updating, everything.  I think one can make ball park guesses, but that is about that.
 
And time is your friend if you keep a clean sheet and have some idea of how the game is played
The slide from grace is really more like gliding
And I've found the trick is not to stop the sliding
But to find a graceful way of staying slid
Message 4 of 12
Anonymous
Not applicable

Re: Question about Score Reverse Engineering

the problem is that it is an algorithm for which we have no way of knowing all of the factors that are taken into account.  sure, we know some generalities, like the percentages of our scores that some very broad categories make up; however, we start getting into problems in reverse engineering the algorithm when taking into account factors like scorecards, and especially the interrelationships of those broad categories.  for instance, it is quite possible (and likely, imo) that the payment history factor is affected by certain criteria in the age of accounts which is affected by other criteria in utilization, etc.  certainly these are factors that cannot be reverse engineered because first we would have to identify the relationships and then we would need access to the vast amount of data that is used in the algorithm, i.e. individual credit reports.  as we all know, that data is impossible to obtain.  the good thing is that there are a lot of folks who have collaborated and gained enough knowledge to provide general info on factors that help and hurt, but i believe it is impossible to ever reverse engineer the process to any reasonable degree of certainty.  let's not even talk about the fact that the algorithm is likely ever-changing.
Message 5 of 12
Anonymous
Not applicable

Re: Question about Score Reverse Engineering

Even though I've had some weird FICO score effects, I do think FICO scoring makes sense -- it's just extremely complicated, and you're scored relative to other people, too.
Message 6 of 12
RobertEG
Legendary Contributor

Re: Question about Score Reverse Engineering

I have been working on an extensive Excel spreadsheet that simulates current and future FICO scoring based on various factors that are relevevant to  me, and I would be glad to share it for comments with anyone who would like to take a look at it. 
Message 7 of 12
Anonymous
Not applicable

Re: Question about Score Reverse Engineering

i would like to see it.
Message 8 of 12
Anonymous
Not applicable

Re: Question about Score Reverse Engineering

count me int too
Message 9 of 12
RobertEG
Legendary Contributor

Re: Question about Score Reverse Engineering

If you would like a copy, send me an email at robertegarrett@netscape.net.
Message 10 of 12
Advertiser Disclosure: The offers that appear on this site are from third party advertisers from whom FICO receives compensation.