Honest questions in. A ranked diagnosis out.
Most stalled lifters are not missing information — they are missing the link between what they actually do and why it stops working. LiftDecode makes that link explicit in three stages.
Light to answer. Honest by design.
The first question picks your track — physique (you want to look different) or strength (you want the bar to move) — and the questions adapt. From there they cover the things that actually decide progress: your build and your training week (sessions and time per session for each muscle group), how you train and progress, how hard your sets really end, how your programme is built, how you sleep and recover, what you eat, and how much the rest of your life is asking of you.
Answering is meant to feel light: mostly one tap per question, either a point on a 1–10 scale or one of a few short choices. The handful of real numbers — height, weight, age, grams of protein and carbs — are there because the report reads everything else against them; if you don’t track your food, “not sure” is an honest answer and the report works with it. Answer for a normal recent week, not the version of yourself you would like to be; a diagnosis of that lifter is a diagnosis of nobody.
No question is a trap and none is filler. Each one exists because at least one finding rule depends on it.
Your training, effort, food, sleep, recovery and life load right now — not intentions or knowledge.
Your name, your lifts, or anything no rule uses. Every question feeds at least one finding.
Pick the option closest to a normal week. The engine weights, it does not punish.
The engine cross-references, it does not average.
Your answers are matched against a set of finding rules. A rule describes a bottleneck — say, sets ending too far from failure to drive growth — as a list of triggers: answer patterns that point at it, each with a weight. One answer can feed several findings; one finding usually needs several answers. When the summed weight clears the rule’s threshold, the finding fires, and it carries with it the exact answers that made it fire.
Then two things happen that a simple score sheet would not do. Findings are ranked by how strongly your answers support them, so the report leads with the thing most worth changing first. And weaker findings are suppressed by stronger causes: if your programme has no progression at all, we will not also lecture you about exercise order, because fixing the first makes the second irrelevant for now.
Alongside findings, the engine checks clearances — answer patterns that rule something out. These become the “not your problem” section, and they matter as much as the findings: most stalled lifters are quietly fixing the wrong thing.
- ✓Deterministic rules written and revised by LiftDecode. The same answers always give the same result, and every result can be traced to specific answers.
- ✓Versioned. A report is stored with the engine version that produced it and is never silently recomputed.
- ✓The rules decide the findings. After you pay, an AI model (Anthropic’s Claude) writes them up for your case from your answers, citing only a verified evidence base: it explains what the rules found, it does not replace them.
- ✕Not a quiz that maps one answer to one canned paragraph, and no model runs before you pay.
Everything the report contains.
Your result page names every problem the engine found, each with the answers behind it, and scores ten parts of your set-up, before you pay anything. If it does not describe you, walk away. If it does, the report explains it, written for your case:
- 01Verdict and first move
What is holding you back, in one sentence, and the single change to make this week.
- 02The chain
How your problems feed each other, from the root cause to the stall you feel, so you fix causes, not symptoms.
- 03Your scorecard
Ten parts of your set-up scored 0–100 from your answers, each with the lever that moves it.
- 04Each problem, in your case
Why it is happening to you, what it is costing you, and the fix in your numbers, with the answers that triggered it quoted back.
- 05Training-dose audit
Every muscle group's sessions and time turned into hard sets a week, against what grows muscle.
- 06Fuel audit
Protein and carbs per kilo, meal timing, sweat and minerals, and what your weight trend says about your goal.
- 07Recovery and conditioning
Sleep, stress and set-to-set recovery, and what cardio does and does not do for growth.
- 084-week plan
Week-by-week actions with check-offs, plus the numbers to track and what to leave alone.
- 09Sources
Every research claim links to the study behind it, from an evidence base we checked entry by entry.
A single report is $5 and is yours for as long as your account exists. It can be printed. Track-specific notes appear where the fix differs between physique and strength goals.
The report tells you what to change. Membership tells you whether it worked.
Training changes; so do bottlenecks. Re-run in four or eight weeks and see what cleared and what surfaced.
Log lifts and measurements weekly. The point is not the graph — it is knowing, with numbers, whether the plan is moving anything.
Two diagnoses side by side. Which findings disappeared, which persisted, which are new.
The full protocol for every bottleneck we can detect, not only the ones you were flagged for.
Training and nutrition education, not medical advice. If you have pain, an injury or a medical condition, see a qualified professional first.