Reverse engineering the FICO scores

count me in! I love this kind of a challenge and I would love to beat the idiots at their own game!

The first thing we would need to do is to gather raw data. You would need to know Score when CN started, info on good and derrog accounts to include dates/no of lates/when late. Then as their scores improve, enter the variables (the things that changed).

There would have to be foundation data as well. Time in residence, own vs rent, time on job, salary, number of children, marital status, demographics _ area of the country, ethnic group (they say that this is not included, but the FICO report to congress implicates otherwise).

We also need to have acceptors of the data.... somone with the time to create a database (access maybe) and then once we have that data, we can begin the analysis to reverse engineer.

Probably best to take it offline if we get enough folks involved.

 

Impossible or more correctly EXTREMELY improbable...even for Mr. Fair or Mr. Isaac.

Even if you could, the amount of daunting work compiling the data would be enough to take you far into the timeframe that Fair Isaac would have already updated the weights attached to the ever changing algorithm...so all you would solve would be last years risk model.

Nice thoughts, but it is unlikely we will ever be able to know with any certainty, what factors exactly into our score.

Now if someone could get their hands on the algorithm itself ) (anybody work at Fair Isaac?)

-Peace, Dave

 

RNG may be a JOKE...but HOW can you explain the $30,000 payment (NOT A BT)...and the POINTS I LOST???

 

I've noticed that FICO scores are far more concerned with conformity of behavior than financial capacity. A $30,000 payment is probably larger than the norm, therefore you are penalized, with no measurement of whether it is actually financially saavy.

For anyone who has no derogatory information on their report, yet receives a lower score due to credit usage, inquiries or payment patterns, I think the FICO models are basically saying they don't understand you because you deviate from the "norm," therefore you're receiving a lower score to compensate for the "risk of the unknown."

 

I HAVE ZERO NEGATIVES...BUT I CAN NEVER GET A F.I.C.O. SCORE OF 850...EVEN IF I PAY MY MORTGAGE AND ALL MY CARDS DOWN TO $0.00!!!

SOMETHING IS WRONG WITH THAT!!!

 

I agree GEORGE!!!!

I have the SAME exact score I had a year ago, I have gone from 50% available to 85%, accounts have aged and inquiries have gone down, result NO change.

I am going to just focus on paying down debt and screw the scoring system, in the process of cancelling all my monitoring services.

 

IT IS A LOSE~~~LOSE PROGRAM...

CAN'T WIN!!!

 

I agree with you totally, George. Your example fits my theory; if you suddenly paid off your mortgage and all your cards, that would be behavior which is outside the norm. Most people make continual monthly payments. Therefore your score would go down.

My theory is that FICO models are affected by the *style* of your behavior and have almost nothing to do with actual dollars and cents in a financial analysis way. Anything such as large, irregular payments, although not traditionally negative, are perceived as negative by the FICO models because they're "not normal."

The models get creepy for those with no traditional derogatory information because they're trying to predict "surprise" bankruptcies using very incomplete information.

You're absolutely correct, somthing is wrong with that.

 

I think the reason perfect history isn't always rewarded with the highest scores is that you could be a great a reliable person, but if they can't tell it by your credit report, its worthless to creditors. So I think the reverse becomes true--distiguishable reliable people will be rewarded. The most reliable person in the US doesn't have the best score, in fact he probably doesn't even have a credit file!

I've read that Fair Isaac use scorecards -kind of like like what marketing people use to help them dice up their Target Market. So for example I wouldn't be surprised if there were an engineer profile, or maybe a constant debtor, or repenting credit card user(I remember reading one online marketing site and they had profiles called "shotguns and pickups" and"minks and stationwagons". They could then define criteria for who will be included in each group. The groups could be statistically tracked in the real world, and their performance used to adjust the fico scoring. If this is how they do it, it wouldn't surprise me that if you do something way out of ordinary behavior and land in some ill defined catchall group which is punished, because it can't be tracked(The 'we don't know you from Adam' group?).

Oh, and I've been experimenting a lot lately...I've got a (legal) way to put an outlandish amount of credit accounts on a profile(I'm sorry I can't post how, if i did the method would disappear just as quickly, like the back door double or doc's trick). I guarantee there will be no report like it. lol luckily I have a guinea pig friend! I learned studying physics that you learn the most when you test at the extremes.

After all we already learned that having too many soft inquiries on equifax kills the scoring algorithm(which to me implies it is included in the scoring). I wonder if Experian got around that problem on their system by entering one entry and listing multiple dates(so looking at credit expert a bunch of times in considered one 'soft' inquiry). Lol I bet Equifax just didn't catch that bug in product developement and Experian did.

 

This may be controversial but I don't have any aspirations of obtaining the "perfect" score. Just want to get a "good enough" score. Mainly I feel this way b/c it seems like keeping a car loan open helps your score and keeping current debt alive also helps. And if I could pay off a home mortgage and own a home free and clear, I would. Getting out of debt may lower your score but it is also counterintuitive and doesn't make sense to me financially. Also, having a lot of obsolete accts on my rpt makes my credit file unwieldy even if they are positive tradelines. It's all nonsense. I can understand that negative info being reported can hurt your score but to have fiscal responsibility also lowering your score just doesn't make sense. After a point, I have to do what works for me and try to believe that a score in the good range is enough.

 

Another way Scoring hurts people is because it causes folks to make the wrong financial decisions often with disastrous results. just to appease the scoring model ( make that monster)

 

I have a theory. I know this sounds strange, but my score ALWAYS drops 7 points whenever the stock market goes down. So I'm thinking that we might be penalized whenever there's a sudden shift in the economy. I also think that demographics plays a role as well. I think the scoring system can probably identify certain demographic information about you based on where you live and that information is likely factored into your score as well. FICO claims to use hundreds of "factors" in determining your current credit score. I think the economy and the area in which you live are two of those factors that they don't advertise.


Calmest_LA

 

you are probably at least partly right, someone here reported that a friend who worked for fico siad there were some external factors that affected your fico score(we're not talking credit expert here). I think one of the factors mentioned was money supply or something like that... Anyways so maybe its not all your fault if your scores aren't the highest...its the economy stupid!

 

Ok: so how do you explain the same scores in a thriving economy?

 

FWIW, I was talking credit repair at work and one of my co-workers told me that when they had their report pulled for a car purchase it came up 871!

This person is in their late 60's, still working and collecting SS with only a mortgage and car payment.

Go figure.

 

This can't be done.

They constantly change the rules and how they play the game.
Even if some how such a data base could be assembled it would be out dated almost instantly because of how they play the game you would be back to square one very shortly.
This is why they operate the way they do=to make just such an attempt by us a futile effort.