@SoonerSoldier33 wrote:It's pretty well known that each of the 3 CRAs have their own unique FICO algorithm for all the different scoring models. So, your scores may vary between the CRAs both bc of different info reported to each CRA, and the differences between the unique algorithms. As of today, my TU and EQ credit reports are 100% identical...not 1 single difference between either report. TU FICO 8: 678 EQ FICO 8: 669.
Thank you. Seems absurd to me, but then I'm not running Fair Isaac or any of the CRAs. Fair Isaac certainly seems to have a good scam going, though! The current snapshot model is ridiculous and could have been programmed by a high school student. I understand they may be moving to a trend model which would make a lot more sense.
FICO 10T (trended) is 'live' so to speak, and supposedly somehow factors in the past 24 months of utilization history, but I know of no lenders that are using it for lending decisions yet. If it (or a similar trended model) ever become the standard, we'll have to reinvent the wheel on optimizing credit scores.
@Anonymous wrote:
If you want to nerd out on the differences, you can look at some of the research fone on FICO5/4/2 scores. Several members have determined some of the factors that have more of an impact for one bureau, very cool data, but very hard to test. As far as FICO 8 goes, the differences in each bureau's algo is still pretty obscure, we know they exist we just don't know what they are.
Thanks. Like I said, I'm a tech nerd so these questions are more to satisfy my curioosity than to gain any real benefit! Plus the more I know, the more fun it is to play around and try to game the system. Currently I don't have a real need (except a psychological one) to raise my score so I can afford to experiment a little bit, one variable at a time. 😉
@Horseshoez wrote:
@NYC_Fella wrote:Thanks, that's the only possible explanation. I have no doubt that you're correct, but it bothers me that I've never seen that officially explained anywhere (the fact that Fair Isaac has three separate algorithms, not just a single one). No disrespect to all the highly experienced and helpful contributors here!
If I were a CRA, I would find it much more effective to build my own model and save a ton of money! The model can't be particularly sophisticated. Yes, I'm a tech nerd. 😊
LOL, only 3? How about 28 different models?
LOL!!!!! I was fixed on 8 to make it easier to understand. 
Currently, each of the three major credit bureaus uses their own version of the FICO scoring method -- Equifax has the BEACON score, Experian has the Experian/Fair Isaac Risk Model and TransUnion has the EMPIRICA score. The three versions can come up with varying scores because they use different algorithms.
Now if we dove into buckets @NYC_Fella . That would really throw a wrench into it. Your actually placed with a group of people with very close info. Now you can move up so far in that bucket and then to advance to another group of people in their bucket. Now you start at the bottom. Score may rise slowly or faster. Depends on the new group your in. It goes on forever in so many different ways we cant even explain all that goes into it to come up with a score. So theres more than just algorithms.
@FireMedic1 wrote:Currently, each of the three major credit bureaus uses their own version of the FICO scoring method -- Equifax has the BEACON score, Experian has the Experian/Fair Isaac Risk Model and TransUnion has the EMPIRICA score. The three versions can come up with varying scores because they use different algorithms.
Now if we dove into buckets @NYC_Fella . That would really throw a wrench into it. Your actually placed with a group of people with very close info. Now you can move up so far in that bucket and then to advance to another group of people in their bucket. Now you start at the bottom. Score may rise slowly or faster. Depends on the new group your in. It goes on forever in so many different ways we cant even explain all that goes into it to come up with a score. So theres more than just algorithms.
Thank you, now that's an explanation I can understand.
I'm too early in the learning process to start thinking about buckets!