back in april i published my parts split plan — a 7-day arms & delts focused container. it’s been ~12 weeks. time to be honest about what happened.
the scoreboard
adherence: A+. 55 logged training days, 62 sessions since late april (the first two weeks were pre-app, more on that below). that’s ~5.5 days a week for three months, and the only thing that broke the streak was a week in phuket. the 7-day container question is answered: it fits my life. i don’t skip.
progression: honestly… unclear. and that’s the real story of this post.
some things clearly moved
| exercise | start | now |
|---|---|---|
| Smith Incline Press | 40kg | 80kg |
| Cable Rope Tricep Pushdown | 20 | 38 |
| Machine Hip Adduction | 100 | 110 |
| Hammer Curl | 15kg | 17.5kg |
| Dumbbell Seated Incline Curl | 21kg | 23kg |
pressing and triceps went up meaningfully. curls crept up the way curls do. good.
and some things i genuinely cannot tell
here’s the embarrassing part. i pulled my own first-vs-last numbers and half of them look like i got dramatically weaker: machine flat press 87.5 → 42.5, RDL 95 → 50, seated leg curl 87.5 → 45.
i didn’t lose half my strength. i trained across six different locations this cycle — my main anytime fitness, another AF branch, NUS utown, activesg, and whatever gym was near wherever i was. every machine stack labels weight differently. cable ratios differ. “87.5” on one leg curl and “45” on another can be the same effort.
i log everything, and i still can’t answer the most basic question — did i get stronger? — for half my lifts. the data is honest; it’s just not comparable.
and the parts that were supposed to be the point of this split — side delts — stayed at 4-6kg on lateral raises the entire cycle. isolation weights barely move by design, so the number tells me nothing there either. did they grow? i think so. based on… looking at myself in the mirror. very scientific.
how i actually feel
three months of near-perfect consistency, and my honest self-review is: i executed like a machine and evaluated like nobody was watching. because nobody was.
the funny thing is, i coach PT clients now. i watch their form, i audit their week, i adjust their loads based on what i see, not just what the spreadsheet says. every one of them has something i don’t have: a coach.
a plan tells you what to do on day 1. a coach tells you what to change on week 6 — when the machine numbers lie, when the mirror matters more than the log, when “add 2.5kg” is the wrong call and “fix your elbow path” is the right one. that judgment layer is exactly what my last 12 weeks were missing.
so here’s what i’m doing about it
i’ve been studying how hany rambod coaches — the guy behind 25+ olympia titles, who coached direct rivals in the same lineup by building each man’s own best version instead of copy-pasting programs. what strikes me most isn’t FST-7 or any specific method. it’s that he makes look-based, week-to-week calls, and he explains the mechanism behind every correction, not just the cue.
i can’t hire hany. but i’ve been living in claude code for everything else in my life — so i’m building a hany clone: a coaching protocol where claude reviews my week like he would. direct verdict first, reasoning second, one fix, then move on. weekly photo audits instead of trusting incomparable machine numbers. tough love included.
next post: what i actually learned from studying rambod’s method, and how the clone works.
12 weeks of executing alone taught me the limit of executing alone. time to get coached — even if i have to build the coach myself.