The science behind
every feature.

Skills decay without practice. SatSpike tracks every skill's last-practiced date and schedules re-checks at scientifically optimal intervals: 3, 7, 14, and 30 days after first mastery. You never over-review strong skills or neglect weakening ones.

Not all wrong answers are equal. SatSpike classifies errors into four types: concept gap (didn't know the rule), careless (knew it but rushed), time pressure (ran out of time), and trap (fell for a designed distractor). Each type needs a different fix.

Most platforms track skills. SatSpike tracks patterns within skills. "Algebra" is too broad — "linear equations in two variables with fractional coefficients" is actionable. We decompose each skill into 2–6 specific patterns, each with its own mastery score.

The Digital SAT uses adaptive testing: Module 1 score determines whether you get a hard or easy Module 2. A score above ~60% in Module 1 routes you to the hard module, which has a higher scoring ceiling. SatSpike tracks your Module 1 control rate and optimizes for it.

Mastery isn't binary. We use an exponential moving average (EMA, α=0.3) to update skill mastery after each answer, weighted by recency. A recent wrong answer counts more than an old right answer. This gives you an honest, current picture of where you stand.

Our 4-week sprint system mirrors how professional coaches structure training: 3 weeks of learning + 1 repair week. Week 3 ends with a full mock. The sprint report analyses 3 data points (2 topic tests + 1 mock) to identify which patterns need the repair week.