A League of Football Teams · 2002–2026 · the 25th season
What a quarter-century of this league adds up to, before any argument about it. The league's own figures are exact for every season; the player figures cover only the seasons with captured lineups, and say so on the card.
Three populations, kept apart on purpose. Every game ever includes the playoffs; regular season is the 2,165 matchups that decided the standings. Mixing them is how you get a matchup count and a points total that do not belong to each other. Weekly lineups exist only from 2021 — CBS recorded standings, schedules and brackets but never a weekly lineup — so the passing, rushing and receiving totals cover six seasons rather than twenty-five, and say so on the card. The last four cards come back to the full twenty-five: the draft boards, the transaction log and the message board all survive, so those are counted in full. Where both eras can be counted, they are counted together — this is one league that changed platforms, not two leagues. An era is split out only where one side has no data at all: weekly lineups were never recorded on CBS, and Sleeper’s transaction log and message board were never captured. Nothing here is estimated; a number the archive cannot support is not shown.
Breakdown wins are what the field says you earned: each week you are credited with the share of the league you outscored. The gap to your actual record is schedule luck — and it is the argument the rest of this page is built on.
Every coach, every season. Jade where the schedule paid, brick where it charged. A blank cell is a season he wasn't in the league.
Thirty-six times in twenty-four seasons, across nineteen of them. Losses where exactly one team in the entire league outscored the coach that week — and he was playing it. Score the most points and you cannot lose, so this is the nearest thing the format has to robbery. Listed in chronological order, with the season, the week and the size of the field that week. Playoff weeks are excluded: with only a few games on the board, being beaten by the one team above you means nothing.
| Season | Wk | Coach | Scored | Lost to | Their score | Field |
|---|
Pick a coach above and this becomes his file: every season, his titles, the weeks the schedule robbed him, and the picks that paid.
Horizontal is field strength — the share of the league a coach outscored across his whole career, which luck cannot touch. Vertical is championships. The two disagree more than you'd hope.
The league worked this out in 2007, after its own commissioner went 98–71 against the field and missed the playoffs. Ever since, the last playoff berths have gone to a measure of points rather than a won-lost record. Run the schedule across twenty-four seasons and it separates the luckiest coach from the unluckiest by 23 wins.
In a fourteen-team league nobody plays the same schedule, so a won-lost record is partly a report on who you happened to draw. ALOFT has never seeded its whole field on it. The top four seeds go to the standings; the rest go to a measure of points. Which measure is the only thing that changed when the league left CBS.
Constitution, PLAYOFF RULES: “The top 4 teams will be decided by W-L record as always and will be seeded 1-2-3-4. The last 3 playoff spots will be determined by breakdown record, which is the record you have against every team counted every week.”
These are the seasons that argument is made of — where a coach's finish and his breakdown place disagree most violently. Negative means the schedule cost him places; positive means it handed them to him. Rows marked before the rule are from 2002–2006, when finishing order alone decided the field.
| Season | Coach | Finished | Bd place | H2H | Breakdown record | Bd % | Places +/− |
|---|
For the first five seasons the playoff field was simply the best finishing records. In 2006 that produced a result the league could not let stand: the Galloping Geckos made the playoffs on a losing breakdown record (46.7%, ninth of fourteen) while the commissioner's Blackbeards, with the third-best breakdown record in the league, finished seventh and missed.
“With the exception of the Gecko's all teams that made the playoffs had winning records in the all important Standings Breakdown column. My condolences to the commish who posted an impressive 98–71 record against the rest of the league but could not come up with the Gecko's formula of keeping their opponent to their worst offensive performance each week.”ALOFT message board, 8 December 2006
That 98–71 is a 58.0% breakdown record, third best in the league. The following season the rule existed, and the commissioner said plainly when it arrived and what it was for:
“Breakdown record is the basis for 5-6-7 seeds. The top 4 overall records, then 5-6-7 determined by breakdown record. This was announced before the season started… For the past three seasons, an unlucky team that was one of the best 4 teams in the league got left out of the playoffs because of the schedule. I had no idea how this would work out or who it would affect — it could have been me.”Terry Herndon, commissioner, 4 December 2007
It cut both ways, and the league said so at the time. The 2004 Slackers won ten games on the eighth-best breakdown record in a twelve-team league — at +3.6 wins it is still the luckiest season in ALOFT history, out of 302 on record — and nobody pretended otherwise:
“Slackers are the luckiest team in ALOFT. Vick gets −2 and you still win.”Dan Weissburg, 12 October 2004
“slackers opponents don't look much different: 35, 40, 59, 41, 27. That 41 was my lowest-scoring game of the year.”Kevin Herndon, 12 October 2004
“Everyone has their worst game against you. That kind of luck took him right to the Super Bowl. As far as I'm concerned I'd rather be lucky than good!!”Ray Barlow, the Slackers, 15 October 2004
The same coach had missed the playoffs the year before on the fourth-best breakdown record in the league. One system, both complaints, three years before it changed.
| Season | Coach | Record | Breakdown |
|---|---|---|---|
| 2003 | Chris Vogler | 7–7, finished 4th | 2nd of 12 |
| 2002 | Ben Byler | 6–8, finished 4th | 3rd of 8 |
| 2004 | Mike Pugh | 5–9, finished 9th | 3rd of 12 |
| 2006 | Terry Herndon | 6–7, finished 7th | 3rd of 14 |
| 2003 | Ray Barlow | 6–8, finished 7th | 4th of 12 |
Every one of them missed the playoffs. 2005 is left out: its postseason is only partly captured, so who actually played in it cannot be established.
Dated from the playoff fields themselves, not from the rulebook: seeding by record alone reproduces the field in 2004 and the breakdown rule does not, while from 2007 the breakdown rule reproduces it and record alone does not. 2002 and 2005 are too small or too incompletely captured to separate the two.
Sleeper keeps no breakdown record, so the instrument changed with the platform. Seeds 1–4 are the standings; seeds 5 and 6 go to the two highest season point totals among everyone left. The field also shrank from seven teams to six.
Every team below made the playoffs on points rather than on record. Seven of the ten would have missed a six-team field on their won-lost record alone — the column on the right is where the standings had them, and the column beside it is how much the standings were understating them.
| Season | Coach | Record | Points for | Points rank | Standings place |
|---|
This is the argument settled. A top breakdown or points-for record is not a consolation prize for a team that lost too much — it is the league telling you which teams were actually good, through a schedule that was telling you something else. Five of the nineteen champions crowned since the rule arrived would have missed the field entirely under a straight won-lost cut.
| Season | Champion | Standings | By the rule | Team |
|---|
Brian Pugh won the 2014 championship from thirteenth of fourteen in the standings, on the seventh-best breakdown record in the league. Under the field that existed in 2006 he would have been at home.
Twenty-five years of the same league, and almost nothing about how a point was earned survived intact. The yardage rates, the defence ladders, the lineup and the platform all moved — which is why a career total here is not a single currency, and why this page counts what the archive can support rather than what would be tidier.
Two kinds of line below. QUOTED is the league's own words, from the message
board or the captured rules page. DERIVED was reverse-engineered from results because no
record of it survives. The full workings, with the instruments and what is still open, are in
docs/aloft/cbs-scoring-history.md.
The single biggest change, and the league appears to have left no record of it. Both yardage rates got 2.5× more generous at once, preserving the 2:1 ratio between them. Mean team score jumps from 54.4 to 85.8 across that boundary — a +57.7% step that no change in football explains. That is block on block, 2002–2009 against 2010–2019; the single year either side of the line reads 54.0 to 84.0, +55.6%.
| term | 2002–2009 | 2010–2020 |
|---|---|---|
| passing yards | 1 per 50 | 1 per 20 |
| rushing + receiving | 1 per 25 | 1 per 10 |
| kick return yards | 1 per 20 | 1 per 20 |
| punt return yards | 1 per 10 | 1 per 10 |
| every touchdown | 6 | 6 |
Yardage is all or nothing: you score the point on crossing the threshold and get nothing for approaching it. Confirmed independently in 2026 by the lineup solver, a different instrument entirely — scoring 2009 at 50/25 instead of 20/10 cuts the team-weeks it cannot reconstruct from 63 to 23, and a sweep of every plausible divisor pair puts 50/25 at the floor.
Commissioner, Aug 2014, putting it to a vote: “Right now, it is too punitive if your defense gives up a single point or over 100 yards. Both of those things are pretty much guaranteed to happen, which means you're basically guaranteed to lose those 8 points every game.”
The worst case went from −12 to −6. The same vote added blocked kicks — 2 points for a field goal or punt, 1 for a PAT — which is why a defence's score before 2014 is not comparable to one after. An earlier change happened somewhere in 2007–2013: the 2006 table and the pre-2014 table differ, so the ladder moved, but the year it moved is still unknown.
From 2016 the defence is charged DSTPA — points allowed by the defence and special teams only — rather than every point the NFL team conceded. An offensive turnover returned for a touchdown stops counting against the fantasy defence that gave up nothing.
Every draft board from 2003 to 2019 carries whole-number season totals, 100% of them. Of the 2020 board's 163 picks that carry a season total, 116 are fractional — 71%. Nothing else about that season explains it: the league turned on decimal scoring, and a 2019 total cannot be compared with a 2020 one without saying so.
The move off CBS in 2021 reset the record-keeping but not the rules. The one change since is receptions going from 0 to 0.5 at 2024 — so a receiver's 2023 and 2024 totals are in different currencies, and half-PPR is the only scoring this league has ever published machine-readably.
Four configurations, so what counts as a startable player changed underneath all of the above: 2002–2011 three receiver slots with at least one true WR; 2012–2013 WR and TE lumped into one flex; 2014–2016 two true WR slots plus two flex; 2017–2020 two receiver-flex plus two full flex. The 2014 shape came from the same vote that fixed the defence ladder.
Every title in the league's history, oldest first. Thirteen men have won twenty-four of them; the ones marked in on the rule reached the playoffs on points rather than on record, which is the argument above paying out.
Not raw points, and not points against a positional average — both flatter the wrong
players. Value here is points above the replacement you could have started instead, using the real
lineup: QB 1, RB 1, WR-TE 2, RB-WR-TE 2, K 1, DST 1. Fourteen teams means the bar is
QB15, RB30, K15, DST15 — and because a tight end has never had a dedicated slot,
WR and TE share one pool at REC42. Shown as a share of an average team's whole season, so eras compare.
Late rounds only — rounds 11 to 14, where a hit is pure surplus rather than an expensive pick working out. Value is VORP: the player's season points minus the replacement level at his position that year, with WR and TE pooled because ALOFT starts them in the same slots. Quarterbacks, kickers and defenses are excluded — one QB starts, so a late quarterback's apparent value is an artefact of the single slot, not of scarcity. Draft boards exist for 2003–2020; 2002 and 2021 were never captured. Of 904 late picks, 445 resolved to a scored player, 5 were ambiguous and refused rather than guessed, and 52 could not be matched at all.
| Season | Round | Player | Pos | Points | Replacement | VORP | Drafted by |
|---|
CBS-era picks, each carrying the points that player actually scored that season. Value per pick is what a coach got beyond what that draft slot historically returns, normalised within each season so the four scoring eras cannot contaminate it. Coaches who joined after the CBS era have no draft record here.
| Coach | Picks | Value per pick |
|---|
Fifteen pairs have met fifteen times or more; the ten with the best win percentage are shown, each from the winning side and ranked by that percentage — so the Meetings column does not descend. CBS era only, 2002–2020: head-to-head here is counted from the CBS game log, so the Sleeper seasons are not in these records. The Draft and Steals above are CBS-only too, and say so.
| Coach | Opponent | Record | Win% | Meetings |
|---|
Per team-week: (teams outscored + half the ties) ÷ (teams faced), summed across weeks so every week counts once. Pooling opponents instead would over-weight 14-team seasons against 2002's 8-team field.
ALOFT has no ties: the constitution breaks them on reserve points, then passing, rushing and receiving
yards, and the CBS rules page says the same. games.json records the resolved side. Scoring
them as half-wins disagreed with coaches.md for 22 of 36 coaches; resolving them properly
matches all 36 of 36.
Replacement level comes from what actually starts. Measured across 1,260 Sleeper lineups: QB 1.00, RB 2.08, WR 2.49, TE 0.43, K 1.00, DST 1.00 per week — which reconciles to exactly the 70 RB/WR/TE slots the roster provides. The CBS rules capture shows the same eight-starter shape.
2002–2006 is INFERRED. The owner records that TE was once flex-only while the lineup required
WR specifically; that configuration is not in the archive, and the draft mix dates the change to around
2006–2007 (tight ends drafted per team jump from 0.17 to about 1.0). Pooling WR and TE across those
early seasons therefore overstates tight-end value. Player points before 2012 are
carried — no oracle exists to check them against.
Sleeper handles niceneasy205 and vman004 appear with
coach: null in every source in the archive. They are shown as handles rather than
guessed at — picking a plausible candidate would produce a wrong answer that looked right.
Keith33 was a third until the owner identified him as Keith Sipe on
2026-09-18; that name comes from him, not from the archive, which holds no record of it.
Resolving the handle merges a career: his CBS 2020 season with State Champs and his
Sleeper 2021–2023 seasons are one four-season record here rather than two strangers.
2026 is in season and excluded from every rate. Breakdown records for 2002–2020 are reconstructed from the 1,784-game log; 2021 onward ship natively. Head-to-head covers the CBS era only.
The CBS rule is quoted from the league constitution, but the constitution is the final version and does not say when the breakdown clause arrived. The start date is measured from the playoff fields. Each season's field was reconstructed by tracing the bracket backwards from the champion through the game log — CBS's own bracket form lists consolation teams too, so it cannot be read directly — then tested against both candidate rules.
The field is the best finishing records in 2004 and the breakdown rule does not reproduce it; from 2007 the breakdown rule reproduces it and record alone does not. 2006 is a clean negative: Damn Yankees made the field on the fifth-best breakdown while Blackbeards, third-best and on the same 6–7 record, missed. The reconstruction is corroborated by the commissioner's own all-time playoff list of 31 December 2006, which matches the traced fields for 2003, 2004 and 2006 exactly, and by two message-board posts either side of the change.
2002 and 2005 cannot be separated. 2002's field of three and 2005's of six are smaller than or equal to the four seeds both rules award on record, so the two rules predict the same field; 2005's postseason is also only partly captured — the commissioner lists six playoff teams and the game log holds winners-bracket games for four.
The Sleeper rule — seeds 1–4 by standings, 5–6 by season points — is likewise measured: applying it to each season's final rosters reproduces the exact six-team bracket the league played in all five completed Sleeper seasons. Sleeper standings include the league-average match, so a record there is out of 28, not 14.