Build #2 · Week 2
BuildingBJJMath
Trusted calculators for grapplers — weight cuts, points, gi sizing, belt timelines — checked against the actual IBJJF and ADCC rulebooks.
The idea
What it is & why I built it
BJJMath is the MyNubs playbook pointed at something I actually do on the weekends: Brazilian jiu-jitsu. Same core bet — calculators are one of the few kinds of page AI search hasn't eaten, because people want to plug in their own numbers, not read a summary. The difference is the niche, and it's a good one.
Grapplers Google weirdly specific things, constantly. What IBJJF weight class am I in? How do I cut six pounds by Saturday without wrecking myself? What gi size do I order from a brand I've never tried? How long until my next belt? Right now the answers are forum guesses, a coach's gut feel, or a sketchy weight-cut thread from 2019. That gap is the whole opening.
The thing lives or dies on credibility, so that's the spine. Every calculator cites the actual source — the IBJJF and ADCC rulebooks, real sports-medicine references — shows its methodology, and (this is the part I still have to earn) gets checked by an actual black belt. The weight-cut and injury tools especially lean hard on "here's the range, here are the limits, go talk to a professional."
The build
How it was built
This week was the launch set: ten calculators across three buckets. Competition — an IBJJF weight class finder, a tournament weight-cut planner, a live points calculator, and an age-division lookup. Gear — a gi size finder across the big brands and a gi shrinkage calculator. Health and training — caloric needs for grapplers, a belt-promotion timeline estimator, a return-to-roll calculator, and a training-load tool.
It came together fast because MyNubs already did the hard part. Same engine underneath: hand-built pages that load instantly, a methodology page as the trust spine, a cited source on every tool, and the structured data that makes the pages easy for an AI to quote. The math was never the hard part — getting the rules exactly right, citing them, and being honest about what a calculator can't tell you is.
The stack
Tools used
Hand-built static pages Build
Same fast, no-framework setup as MyNubs — ten calculators, instant loads, almost nothing to break.
The 10 calculators Tools
Original logic pulled from the IBJJF and ADCC rulebooks and sports-medicine sources, not numbers guessed off a forum.
Methodology + citations Trust
A methodology page and a source on every tool, plus practitioner review — that's the entire credibility play.
Search & AI plumbing Discovery
Schema, sitemap, llms.txt, open AI-crawler access — built to be the source AI answers cite for BJJ numbers.
The numbers
Analytics so far
Brand new, so this is honestly blank for now — nothing's indexed yet. It fills in the same way MyNubs did: impressions first, then positions climbing off page two, then the first clicks. I'll post the real numbers here as they show up, zeros and all. No projections, no borrowed screenshots from the other build.
The recap
What worked & what I'd do better
What worked
- Reused everything from MyNubs, so a ten-calculator site came together in about a week instead of a month. The second build is so much faster than the first — which is the whole point of doing this on repeat.
- Leading with methodology and real rulebook citations means the credibility is baked in from page one, not bolted on later.
- Picked a niche where the current answers are genuinely bad — forum guesses and dodgy weight-cut advice — so the bar to be the most trustworthy option is low.
What I'd do better next time
- I need a real black belt locked in to review the methodology before I lean on 'reviewed by practitioners.' That claim has to be true or the trust angle falls apart.
- The gi sizing and shrinkage numbers have to be validated against actual brand charts. Sizing is the one thing people will screenshot and roast me for if it's off.
- The weight-cut and return-to-roll tools carry real risk if someone treats them as medical advice — the disclaimers and 'see a professional' framing need to be loud, not buried.
- Zero traffic on day one, exactly as expected. Same slow SEO burn as MyNubs — the difference is now I'm running the play twice and letting them stack.