Teaching tool
A simulator that models one lever and admits it.
Almost every credit score simulator on the internet is a lead magnet dressed as arithmetic. This one models revolving utilisation — the one factor with a publicly describable relationship — and prints the rest of the model as a list of things it cannot do.
- Factor moved
- Amounts owed
- Share of the model
- 30%
- Utilisation change
- −25 pts
- Range width
- 78 pts
What cannot be done
Four things we will never tell you
- That your score will go up. No one can tell you that. Not a number, not a range, not a “typical” figure. Anyone who does is telling you something they cannot know.
- That it will take a specific number of days. The statutory investigation window belongs to the bureau, not to us, and outcomes are not on a schedule.
- That accurate negative information can be removed. It cannot — not by us, not by you, not by anyone. Accurate, current, verifiable entries report until their clock runs out.
- That you need us. You do not. Every right we exercise is yours already, and exercising it costs nothing. See the notice above.
The refusal
Four things this tool will not pretend to model.
Every one of these is something you will find a slider for on a competitor’s simulator. There is no public formula behind any of those sliders, and a number without a formula behind it is a sales tactic.
Payment history — 35% of the model
The largest factor, and the one this tool refuses to touch. Its weight depends on how late, how recently, how many accounts and whether anything is currently delinquent. No public formula reproduces that, and any simulator that hands you a number for it is guessing and charging you for the guess.
Collections, charge-offs and public records
A single recent collection can dominate a file. Its effect depends on the rest of the file in a way a slider cannot represent.
Length of history, new credit and mix — 35% between them
Age moves on its own schedule. Inquiries are small and decay. Mix improves as a by-product. None of the three is a lever you pull.
Which score anyone is actually looking at
FICO 8, FICO 9, FICO 10T, the mortgage-industry classics, VantageScore 3.0, 4.0 — plus per-bureau data differences. There is no single number to simulate.
The arithmetic, printed
The utilisation contribution is modelled as 85 × (1 − min(1, u ÷ 70)^0.8) where u is reported revolving utilisation as a percentage. The difference between the contribution at your current utilisation and at your target is added to your starting score, and the result is widened by ±25 points plus a quarter of the movement — because a model this crude should show its own uncertainty rather than hide it.
That curve is a shape chosen to match the widely-reported behaviour that lower is better all the way down, with the steepest gains below about 30%. It is not FICO’s curve. Nobody outside FICO has FICO’s curve.