The League Office Before the Web: What Fantasy Football Taught Me About AI Operations

In 1989, buried among NFL previews, team rosters, and player writeups in Street & Smith’s Pro Football Preview — the one with Mike Singletary on the cover — there was a small ad for a piece of software called Fantasy Football League Office. It ran on IBM-compatible PCs. It shipped on floppy disks. It cost $44.95. It promised roster management, scoring options, a timed draft, NFL team reports, league reports, printer support, password security, and even financial reports.

I know because I wrote it.

The product was published by SportSoft Systems out of Dallas. It ran for two consecutive years in Street & Smith’s — the 1989 and 1990 editions. By the second year, it was version 2.0, advertised at $64.95, with manual or automatic data entry and compatibility with USA Today Sports Center and Computer Sports World. That detail matters. The product was not just a calculator or a roster tool. It was already moving toward the pre-web data pipeline that later fantasy platforms would make invisible.

Before fantasy football was an app

If you’ve only known fantasy football as a phone notification, it’s hard to understand what the game demanded before the web existed.

There was no ESPN app. No Yahoo league. No auto-scoring, no live stat feeds, no league dashboards. Fantasy football in the late 1980s was a commissioner with a stack of newspapers, a notebook full of rosters, a telephone, a calculator, and a lot of patience.

The commissioner was the system. They tracked rosters. They enforced rules. They calculated scores by hand from Monday’s box scores. They settled disputes. They managed the draft. They produced standings. They handled the money. Every week, someone — usually the most organized person in the group — sat down and did the administrative work that made the league function.

Fantasy Football League Office was built to reduce that burden. It didn’t automate the game. It automated the office that ran the game.

The name was deliberate. It wasn’t “Fantasy Football Calculator” or “Fantasy Football Tracker.” It was “League Office” — because that’s what the commissioner was running, whether they recognized it or not. A small, invisible, recurring administrative operation that existed only because no one had built a system for it yet.

The opportunity I did not compound

I did not trademark the product. I did not patent it. I did not continue developing it as the market moved online. Fantasy football grew into a multi-billion dollar industry over the next two decades, and I was not part of that growth.

For years, I saw that as the moment I lost the thread. Maybe even the moment I left millions on the table. That interpretation is not wrong. But the older I get, the more I think the more useful lesson is not that I failed to protect a specific product. It’s that I failed to recognize the durable pattern underneath it.

I had discovered something — not about fantasy football, but about how recurring administrative burden works. And I kept discovering it, in different contexts, for the next three decades. I just didn’t see the through-line until much later.

The pattern I kept finding

Fantasy Football League Office was not the only “hidden office” I built software for. It was one in a series.

Bracket Manager. Before Fantasy Football League Office, I built software for my parents, who ran bowling sidepot tournaments. A good tournament weekend could generate $10,000 in sidepot revenue — but only if someone managed the brackets, tracked the bets, and reconciled the payouts. My parents were that someone. The software became their tournament office.

Abend Database. While I was at Electronic Data Systems, production support teams tracked system failures — abends — in a physical notebook. When a failure occurred, someone flipped through the notebook looking for a prior occurrence and its resolution. I built an electronic lookup system. It replaced the notebook. It became the incident knowledge base.

Journal Manager. I later tried shareware with a personal journaling product. It received a few registrations. It did not become a major commercial product. But the impulse was the same: someone has a recurring capture-and-retrieval process, and software can run it.

Each product shared a structure I didn’t have the language for at the time:

Product	Human bottleneck	What the software became
Bracket Manager	Bowling sidepot administration	Tournament office
Fantasy Football League Office	Fantasy league commissioner work	League office
Abend Database	Production failure log lookup	Incident knowledge base
Journal Manager	Personal record capture	Personal record system

The pattern was not “Bill builds software.” The pattern was: someone is trapped inside a recurring administrative role, and the right software becomes the office that frees them.

Every business has a commissioner

This is not a pattern limited to fantasy football leagues and bowling tournaments. It operates inside every growing business I’ve analyzed.

Every business develops, over time, a hidden office — a set of recurring administrative functions that someone performs manually, often without recognition, often from memory. That person is the commissioner of the business. They hold the system together. They know the exceptions. They remember which customer needs follow-up. They reconcile the systems that don’t talk to each other. They are the duct tape between process and chaos.

In a dental practice, the hidden office is the front-desk coordinator who knows which patients need reactivation, which insurance claims need follow-up, and which providers are running behind schedule.

In an HVAC company, the hidden office is the owner who remembers which quotes went out last week, which service agreements are expiring, and which technician should handle which call.

In a professional services firm, the hidden office is the account manager who tracks scope creep in their head, follows up on proposals from memory, and manually reconciles time entries against project budgets.

These are all commissioners. They run operational offices that don’t appear on any org chart and don’t have any documented process. They function on memory, habit, and personal commitment — and when they leave, go on vacation, or simply get overwhelmed, the system breaks.

What Alchemy Inside looks for

Alchemy Inside exists because I eventually recognized the pattern I had been building around for decades. The question is always the same: where is the hidden office, and who is trapped inside it?

When I run a diagnostic on a small business, I am looking for commissioners — the people silently coordinating follow-up, exceptions, customer context, reporting, and handoffs that should be running on systems but are running on individuals.

AI agents are not valuable because they are magical. They are valuable when they become part of the operational office. The things an AI agent does well are the things a commissioner does manually:

Monitor — watch for signals that require action
Remind — surface tasks before they fall through cracks
Summarize — compress scattered data into usable briefings
Route — send the right information to the right person
Qualify — assess inputs before they consume human attention
Follow up — maintain contact sequences that humans abandon after the first touch
Reconcile — compare systems that should agree and flag when they don’t
Report — produce the operational snapshot that no one has time to compile
Preserve context — remember what happened last time so the next interaction starts informed

Every one of these functions exists in a hidden office somewhere in your business. Someone is doing them manually — or more likely, they’re being done inconsistently, from memory, and with gaps that cost real money.

The real lesson from a floppy disk

The opportunity I missed with Fantasy Football League Office was real. But the more durable lesson is not “protect your intellectual property” or “ride the market while it’s growing.” The more durable lesson is this:

Don’t chase the technology. Find the hidden office.

The technology changes every decade. DOS gave way to Windows, which gave way to the web, which gave way to mobile, which is giving way to AI agents. The underlying pattern does not change. Every growing human system creates a hidden office — a recurring administrative burden that someone absorbs personally because no system exists to handle it.

In 1989, that office belonged to the fantasy football commissioner. In 2026, it belongs to every business owner quietly running too much of the company from memory.

The opportunity is finding it before everyone else notices.

Key takeaways

Every business has a hidden office — a set of recurring administrative functions performed manually, often by one person, often from memory. That person is the “commissioner” of the business, holding the system together without documentation, tools, or recognition.
The pattern is older than AI, older than the web, and older than the PC. Software has been solving hidden-office problems since before we had language for workflow automation. The technology changes. The pattern is durable.
AI agents are valuable when they become part of the operational office. Monitoring, reminding, summarizing, routing, following up, reconciling, reporting, and preserving context — these are the functions of the hidden office, and they’re exactly what AI agents do well.
Start by finding your commissioner. Who in your business holds the system together from memory? What happens when they go on vacation? The gap between “the business works because this person remembers everything” and “the business works because a system handles it” is the gap Alchemy Inside was built to close.