How It Works

From live scores to draft rankings — the engine behind your team, and how it stays fresh on its own all season long.

Live ESPN scoringSnake draftsValue-based rankingsCustom league scoring
Where the data comes from
Two live sources, zero spreadsheets
1

CollegeFootballData (CFBD)

FBS rosters, the full schedule, and team‑strength ratings — SP+, SRS, strength of schedule and strength of record. This is the structured backbone.

2

ESPN — live

During games we pull real‑time scores and box scores from ESPN, so points update while the games are still being played.

3

Auto‑keyed to now

Everything targets the current season and week automatically — no hardcoded dates to babysit.

Always fresh
The app updates itself — here's the cadence
Live

During games (Thu–Sat). Every ~1–2 minutes, scores, player fantasy points, and matchup totals refresh — your dashboard ticks up on its own.

Weekly

Every Tuesday. The schedule and team ratings re‑sync and projections regenerate for the upcoming week.

Drafts

On the clock. If a pick timer runs out, a server‑side sweeper auto‑picks the best available player — even if everyone closed the tab.

Alerts

Thursday and Tuesday. A reminder if your lineup still has holes, and your result once the week is scored. At most one a day — and nothing at all when there's nothing worth saying.

Still needs a human twice a season. An admin syncs rosters before the season, and pulls final box scores after each weekend — the live feed covers games while they're being played, but official stats come from CFBD afterwards.
The projection engine
A multi-factor model, tuned for college football
1

Performance baseline

40% recent form (last ~5 games, weighted toward the most recent) + 40% season average (outliers trimmed) + 20% trend.

2

Team‑strength normalization

Stats are normalized by a composite team rating (SP+ 60% / SRS 40%), so production against weaker competition is discounted and tougher slates are rewarded.

3

Opponent defense

Adjusts for the matchup — elite defenses pull projections down, soft ones push them up.

4

Volume & efficiency

Usage (touches/targets) and points‑per‑opportunity. Since the box score often omits targets, receivers use receptions as a volume signal instead of being under‑counted.

5

Schedule, environment & home field

Strength of schedule, game environment (blowout vs. nail‑biter), and a home/away nudge.

6

Self‑correcting calibration

Per‑position factors are re‑learned from actual vs. projected accuracy, so the model tunes itself as the season goes.

7

Manual adjustments

Injuries, weather, or depth‑chart notes can override any projection as a final step.

8

Honest zeros

A bye week shows as Bye and a player with no data as No data — not a misleading 0.0.

Preseason & draft rankings
Ratings before a single snap is played

At Week 1 there are no current‑season stats, so we blend four signals into a season‑long draft board and Week 1 projections:

1

Prior production

Last year regressed toward the positional average, plus an age/experience curve.

2

Team offense

SP+ offense percentile + how much production the team returns.

3

Usage / role

Team volume split by each player's prior usage share — starters beat backups.

4

Newcomers

Recruiting rating for freshmen, and a competition adjustment for transfers changing levels.

Scoring
Half-PPR by default — commissioners can customize

Passing

  • Yards1 / 25
  • TD+4
  • INT−2

Rushing

  • Yards1 / 10
  • TD+6

Receiving

  • Reception+0.5
  • Yards1 / 10
  • TD+6

Kicking

  • Field Goal+3
  • Extra Point+1

“1 / 25” = one point per 25 yards. Each league's commissioner can override any of these in Settings, and matchups are scored with that league's rules.

Confidence score
A 0–100 reliability read on every projection
Games playedMore history → higher
ConsistencyLow week-to-week variance → higher
Team dataSP+/SRS available → higher
PreseasonCapped lower until games are played
Manual overrideApplies a penalty

Higher confidence = the projection rests on more, cleaner data.

Your roster
5 starters + 5 flexible bench = 10 players

Start exactly one of each:

QB×1
RB×1
WR×1
TE×1
K×1
5 bench spots, any position. Load up on RBs, handcuff your stars, or stack WRs — the bench is yours to strategize, as long as your starting five is filled.

Positions: QB, RB, WR, TE, K. (No defense/special teams.)

Draft day
Snake order, 10 rounds, 90 seconds a pick
1

Create and share a code

The commissioner opens a room and everyone joins with a six-digit code. Draft order follows the order people joined.

2

Snake order

Pick order reverses every round, so the last pick of round one gets the first of round two. Ten rounds fills all ten roster spots exactly.

3

90 seconds on the clock

The board is sorted by value over replacement, not raw points, so a quarterback doesn't occupy the whole first page.

4

Nobody can stall it

If the clock runs out, the server picks the best available player. This runs on the server, so it fires even if every browser in the draft is closed.

5

Rosters build themselves

When the last pick lands, each team is created and slotted automatically — first player at each position starts, the rest go to the bench.

On your phone
Install it, and get told when it matters
1

Add to home screen

Installs as a real app: full screen, its own icon, no address bar. On iPhone that's Share → Add to Home Screen; Android and desktop Chrome offer a one-tap install.

2

Works with a bad signal

The app shell is cached, so it opens in a stadium. Live data always comes from the network — a cached score would be worse than none.

3

Two reminders a week, at most

Thursday evening if your lineup has holes; Tuesday morning with your result. Capped at one a day and three a week, and skipped entirely when there's nothing to report.

4

Off by default

Nothing is sent until you turn it on in Settings, where each reminder type has its own switch and a button to send yourself a test.

On iPhone, notifications only work once the app is on your home screen — an Apple restriction. Reminders still appear under the bell icon either way.