Rebuilding the Pool

Splash Sports bought OfficeFootballPool and rebuilt it around their own contest formats. Ours isn’t one of them. Rather than bend our rules to fit somebody else’s product, I started rebuilding the thing we actually had: a confidence pool, ten picks a week, ranked one through ten, against the spread or the total.

This is where that stands. There’s a working pick sheet you can click around in today, running on a real board of games with real numbers. What’s missing is the plumbing — accounts, saved picks, standings — and none of it is hard. The rules were the hard part, and the rules are done.

The pool, exactly as we play it

Every one of these is already enforced in the working version.

The card

Ten picks a week, each ranked 1 to 10. A correct pick earns its rank. 55 points is a perfect week.

What you can pick

Any side against the spread, or any game total over/under.

Doubling up

You can take the spread and the total on the same game as two separate picks. You can’t take both sides of the same bet.

The board

Every FBS-vs-FBS college game and every NFL game. No FCS opponents.

The split

Week 1 is ten college picks. Once the NFL starts, it’s five college and five pro.

No pushes

Every spread and total carries a half-point, so nothing ever ties.

Submitting

All ten go in before the card saves. After that, any pick stays editable until its own game kicks off.

Locking

Each pick freezes at its own kickoff — pick and rank both.

Showing up late

If games are gone, you play a short card off the top ranks. Five picks means 10 through 6 — a 40-point ceiling instead of 55.

Not showing up

Slots you never filled score zero.

Why the rolling lock matters

This is the rule that makes the pool interesting, and it’s the one worth seeing in motion. Because each pick freezes at its own kickoff, a rank you spend on a Thursday night game is spent for the week. Put your 10 on Thursday and you’re assigning 9 through 1 across everything left.

How the numbers get in

The part that would otherwise be an hour of typing every week.

The board is big — a full college Saturday plus a full NFL Sunday runs to roughly seventy games, and each one carries a spread and a total. That’s about 140 numbers. Nobody is typing that in by hand in October.

So it’s automated. The lines come off a sportsbook odds board, and a script does the rest:

It reads the board, throws out anything against an FCS opponent, adds the half-point to any line that came in on a whole number, and hands back a clean slate. This week that meant dropping seven games — Albany at Buffalo, Idaho at Utah, Long Island at Kansas and four others — and hooking 24 of the 40 that were left.

Then it’s reviewed before it goes live. Nothing publishes automatically; the board gets eyeballed and locked, and once locked the numbers can’t change — not even by whoever posted them. Everybody plays the same line all week.

Scoring works the same way in reverse. Final scores come from a free college-football data feed, get matched against the board, and settle every card without anyone entering a result.

What’s left

Roughly in the order it needs doing.

1. The pick sheet Working

   Full board, ranking, the 5-and-5 split, doubling up on a game, rolling kickoff locks, short cards. All of it, running today.

2. The lines pipeline Working

   Automatic pull, FCS filtering, half-point hooks, review and lock.

3. Accounts To build

   Members sign in with an emailed link instead of a password — one login in August that lasts the season. Nothing to forget, nothing to reset, no passwords stored anywhere.

4. Saved picks To build

   Cards persist to a database instead of living in the browser. This is what turns the demo into the real thing.

5. Scoring and standings To build

   Weekly results, season standings, and the pick-comparison grid so everyone can see who took what once games lock.

6. Commissioner tools To build

   Invite and remove members, post and lock the weekly board, and enter a card for somebody who texts you their picks from a tailgate.

7. Reminders To build

   A nudge before the first kickoff to anyone who hasn’t got a full card in.

What it costs

For a group this size, essentially nothing.

Item	Per year
Hosting and database	Free tier
Logins	Free tier
Lines and scores	$0
Domain name	~$12
Total	~$12

The free tiers are sized for tens of thousands of users. A pool of fifty doesn’t come close to the ceiling, and there’s no per-member cost as the group grows. If we ever outgrew them it would be a good problem, and it would still be a few dollars a month.

What I need decided

→ Open the working pick sheet

The rules are the hard part, and they’re done. Everything left is standard work on a well-understood problem. Tell me the hook convention and I’ll have the board settling itself.