Big 12 Tiebreaker Tracker 202620252024

Sixteen teams, nine conference games each, no round robin — which means two teams can finish with identical records having played completely different schedules. The Big 12 sends its two best conference winning percentages to the championship game, and when teams tie, a strict seven-step procedure decides who goes. This site runs that procedure, verbatim, after every game.

The two golden rules

Before any step-counting, two structural rules do a lot of work:

• Win percentage, not raw wins. The goal is the two best conference winning percentages — mid-season, a 5–1 team outranks a 6–2 team. (The standings table has a toggle if you want to see it both ways.)
• A two-way tie for first between teams that met is no tie at all. Both are in the championship game; the head-to-head winner is the #1 seed. The procedure below is for everything messier.

The seven steps

Applied in order; the first step that separates anyone wins. With three or more tied teams, the moment one team is seeded, the survivors go back to the top and start over (dropping to the two-team rules once only two remain).

1. Head-to-head. For two teams: did you play, and who won? For three-plus: winning percentage in games among the tied group — with one twist: a team that beat every other team in the group is seeded even if the group didn't all play each other.
2. Common opponents. Win percentage against the conference opponents every tied team played. Different schedules, same yardstick.
3. The standings walk. Compare records against the best-placed team everyone played, then the next, on down the standings. Tied placement groups count as one collective opponent — you're compared against the group, not its members individually.
4. Strength of schedule. Combined conference record of each team's nine conference opponents. The tiebreaker equivalent of "who had the harder road."
5. Total wins across the full 12-game season — with at most one win over an FCS opponent counting, so a cupcake-heavy September can't decide a title.
6. SportSource Analytics rating. A proprietary metric the conference buys; it is not public. When a projection reaches this step, the site flags it rather than guessing.
7. Coin toss. Yes, really. It has never come to this.

A real one: the 2024 four-way tie

The procedure's baptism by fire came immediately: in its first season at sixteen teams, Arizona State, BYU, Colorado, and Iowa State all finished 7–2. Two championship-game spots, four teams, none of whom had all played each other. Here is the full resolution — generated by this site's rules engine from the 2024 results, matching the conference's official outcome (Arizona State #1, Iowa State #2):

Worth noticing: head-to-head never fired (the four didn't all meet), Arizona State escaped on common opponents, the standings walk twice found information but never a single leader, and Iowa State's ticket was punched by strength of schedule — step four — while BYU and Colorado never trailed anyone by record. That's the procedure working as designed: it keeps asking narrower questions until exactly one team has the better answer.

The what-if simulator

Every unplayed conference game on the main page has two buttons. Pick winners — one game or all seventy-two — and the site replays the entire procedure on the simulated season instantly, in your browser. The matchup card, standings, and tie narratives all update; clear your picks and reality returns. Nothing is uploaded or saved.

The ★ marks each game's favorite under the selected rating model, with the projected margin (home field included) in the hover text:

The Championship race card

Three layers, in order of certainty. Clinch and elimination statuses are proofs: early in the season, strict win-count arithmetic that no tiebreaker can undo; from mid-November, exhaustive enumeration of every remaining outcome through the full procedure — which is also where the "clinches with a win + a Utah loss" scenario lines come from. Percentages are Monte Carlo odds: ten thousand simulations of the rest of the season, win probabilities from an ensemble of the public rating models, every simulated season scored with the real tiebreakers. Proofs always override odds. And the Chaos Index compresses the whole race into one number from 0 (decided) to 100 (sixteen-way pileup): 60% the entropy of the championship odds, 25% how many living teams are tangled in ties, 15% how many teams are still mathematically alive.

How we know the engine is right

The conference publishes final standings every year, but they list tied teams as a block without applying the procedure — 2024's four 7-2 teams are printed BYU, Arizona State, Iowa State, Colorado, while the teams actually sent to the championship game were Arizona State and Iowa State. The pairing itself is the only published artifact that encodes the tiebreakers' result, so that is what we test against: this engine reproduces all nine championship-game pairings of the championship era, 2017 through 2025, from the game results alone. It also reproduces the 2024 four-way tie exactly, which is the case that pins down the rule discussed above.

Where the policy is silent, and what we chose

The published procedure is specific about the ladder and mostly specific about the mechanics, but a working implementation has to answer a few questions the text doesn't. In the interest of showing our work:

• Teams that trail but don't lose outright. The policy removes a team from a multi-team tie only by seeding it ("after one team has an advantage and is seeded, all remaining teams … repeat the tie-breaking procedure"). It never says a team that merely trails at some step drops out. We carry the whole group forward — and 2024 confirms it, since bouncing Colorado at step (b) would have sent BYU rather than Iowa State.
• Whether tied teams count as their own common opponents. Step (b) says "all common conference opponents played by all other teams involved in the tie." We read the tied teams themselves out of that set, because games among them are precisely what step (a) already weighed; counting them twice would let step (a) decide step (b).
• A standings group that separates without crowning anyone. Step (c) says to compare against the next highest placed common opponent "proceeding through the standings." When a placement group splits the tied teams but leaves two or more still level at the top, we keep walking down the standings, and fall through to step (d) only after the whole ladder of opponents is exhausted.
• The championship game itself. Step (e) counts wins "in a 12-game season," so the title game — a thirteenth — is excluded from that count, and from conference records everywhere on this site.

Every one of these choices is testable, and the nine championship pairings above are the test.

Where the data comes from

Scores arrive from collegefootballdata.com; the procedure text comes from the official Big 12 tiebreaker policy. During the season the site rebuilds itself hourly through weekend game windows (and daily otherwise), so the projection updates within about an hour of a final whistle. The rules engine is validated against the 2024 and 2025 seasons and every displayed conclusion comes with its step-by-step reasoning — if you think it's wrong, the receipts are right there. Found a bug anyway? dept@big12ology.com.