An SAT score looks like a percentage but behaves nothing like one. Nobody scores “73% on the SAT” — you get a number between 400 and 1600, built from two section scores of 200 to 800 each (Reading & Writing, and Math). The path from the answers you got right to that final number runs through a conversion table unique to your test date, and understanding that table is the difference between reading a practice score correctly and misreading it by 50 points.
Start with the raw score, which is the simplest part: one point for every question answered correctly, zero for every question skipped or answered wrong. There is no guessing penalty — that was removed in 2016 — so a blank and a wrong answer are worth exactly the same: nothing. On the digital SAT, Reading & Writing has 54 questions and Math has 58. Your raw score on a section is just your count of correct answers, from 0 up to the section total.
The scaled score is where it gets interesting. Each test administration gets its own raw-to-scaled conversion table, produced by a statistical process called equating. Equating exists because no two test forms are exactly equal in difficulty: if your form's math section was slightly harder than last month's, the conversion table is more forgiving, so the same raw score earns a slightly higher scaled score. The College Board's goal is that a 700 means the same level of ability no matter which date you tested — the table is the mechanism that makes scores comparable across administrations.
Here is a worked example with realistic numbers. Say you take a practice test and get 44 of 54 correct in Reading & Writing and 50 of 58 correct in Math. On a typical conversion table, 44/54 might scale to roughly 640 and 50/58 to roughly 720, for a total near 1360. But on a harder form, that same 44/54 could scale to 660, and on an easier form to 620. That 40-point swing from the identical raw performance is entirely the equating at work — which is why the SAT Score Calculator treats conversions as estimates and why you should never compare raw scores across different practice tests directly.
To use practice scores properly, take at least two or three full practice tests and look at the scaled-score range, not a single number. If your three tests land at 1290, 1340, and 1310, your honest current level is roughly 1300–1340, not the 1340 peak. Then set your target 50 to 100 points above your average — ambitious enough to require real work, close enough to be reachable. Break the target down by section: most students gain faster in Math, where missed questions cluster around a handful of fixable topics, than in Reading & Writing, where gains come slower.
Three misconceptions to clear up. First, the guessing penalty is gone — on the digital SAT you should answer every question, because a guess has positive expected value and a blank has none. Second, you do not need a perfect raw score for an 800: on many forms, missing one or two questions still scales to a perfect 800, depending on the curve. Third, skipping hard questions strategically is not “giving up” — since every question is worth one raw point, spending four minutes on one brutal question instead of banking three easier ones is bad arithmetic. Mark it, move on, come back if time allows.
A final practical note: track your raw scores by question type, not just your totals. If your last three math sections show the same pattern — say, word problems and data analysis dragging while algebra is clean — that is a study plan writing itself. The scaled score tells you where you stand; the raw breakdown tells you what to do next. Run each practice test through the SAT Score Calculator, log both numbers, and watch the trend line, not the noise.