Ed Coan's career is readable because he wrote it down.

Ask why Ed Coan lasted, and the answer is a notebook

Ed Coan's numbers are why people know the name. He is credited with 71 world records in powerlifting, per Wikipedia, which describes him as widely regarded as the greatest powerlifter of all time. ([source](https://en.wikipedia.org/wiki/Ed_Coan)) He deadlifted 901 lbs (408.6 kg) at 220 lb bodyweight at the 1991 USPF Senior Nationals in Texas. ([source](https://www.facebook.com/apemanstrong/posts/show-em-eddycoan-901-pound-deadlift-at-220-pounds-1991-seniors-deadlift-deadlift/1413677267460581/)) That is a pull most lifters never get near, at any bodyweight.

But ask lifters and coaches why he stayed at the top for as long as he did, and the explanation that keeps coming back is paperwork. Cycles written out ahead of time. Each one planned backward from the meet date. Each one built on what the previous cycle's numbers said.

That's the detail worth stealing. The deadlift gets the highlight reel. The logbook is what made a career that long repeatable.

The long ledger: a career is only as readable as its record

Call it the long ledger. A peak total is one data point. A career is the line drawn through hundreds of them, and you can only draw that line if the points still exist.

Coan's arc is legible because the record was continuous. When a cycle went well, there was a written account of the percentages, the jumps and the weeks that dragged. When one went badly, same thing. Over enough years that record stops being a diary and becomes the most specific programming reference a lifter can own: built on one body, under real load, at every age that body has been.

So the thesis is simple. The longer a career runs, the more the record matters relative to any single program. And a 30-year record has requirements most tracking tools were never designed to meet.

A peak total is one data point. A career is the line drawn through all of them.

Why copying a Coan cycle misses the point

The usual lessons people pull from Coan are genetics, work ethic, or a specific program. Genetics is unfalsifiable and useless to you. Work ethic is real but generic. The program is the interesting one, because Coan-style deadlift cycles have been reprinted and passed around gym to gym for decades.

Running one cold has a catch. A cycle like that is the output of a record. The percentages assume you know your numbers from the last time you peaked, how recovery behaved in the back half of the block, which assistance work actually carried over. Coan had that. Most lifters copying the spreadsheet don't, so they're borrowing the answer without any of the working.

Run published programs, by all means. Just run them on top of your own history, so the next block starts from evidence and the one after that starts from better evidence.

What a decades-long record demands from the tool holding it

Paper had one property most lifters never thought about: nobody could revoke it. A notebook on a shelf doesn't migrate, doesn't expire, and doesn't decide which years you're allowed to read.

Digital logs fixed the problems paper had (search, math, backups) and brought new ones. Across the category, four patterns break long records. None of them matter in week one. All of them matter by year ten.

Walking it through: one lifter, ten years, four tools

Take a hypothetical lifter who started training seriously in 2016. Two years in a notebook. Then a spreadsheet, because the notebook couldn't do percentages. Then an app. Then a second app when the first one changed its pricing.

It's 2026, and they're planning a squat peak. They want to ask Coan's question: what happened last time? The 2017 peak lives in a notebook somewhere in a closet. The 2019 block sits in a spreadsheet with no RPE column. The 2021 data came into the second app as one flattened note with the dates gone. The only clean, queryable history covers the last two years.

Ten years of training. Two years of usable evidence.

No single session went wrong. The record just didn't survive the tools, so the next cycle starts from a guess, the same place a beginner starts.

Pick a tracker the way you'd pick an archive

If the long ledger holds, the most important property of a training app in year one is how it behaves in year ten. Most of the category is built around a 12-week program: onboarding, a first PR, a nudge to come back tomorrow. Very little is built for the lifter whose log is older than the app.

Three checks before you commit years of data anywhere:

1. Can it import your past sessions with original dates, sets and loads intact? 2. If you stopped paying tomorrow, could you still open your oldest session? 3. Can you export everything, today, in a format another tool can read?

A no on any of them means the tool is renting you your own history. Coan's notebooks never asked that of him.

Where Platepusher sits

Platepusher is built for the lifter in that walkthrough. Import a CSV from another tracker and every lift, set and date lands as native data, so a block from five years ago reads the same as last week's. Full history, CSV export and the Apple Watch app all ship on the free tier, which means the record doesn't lapse when a payment does. Monthly, yearly or lifetime if you upgrade. The ledger is yours at every tier.

Then the history gets read back. The plateau flag marks the week your top set stops moving: math run on your own numbers, no coach voice attached. That's the Coan question, answered from the record you already kept. The longer argument for treating your log this way is in [Your training history is the experiment](/articles/your-history-is-the-experiment).

Thirty years of training deserves a record that's still there in year thirty.