Banding policy — field reference
Every field on Settings → Banding policy: what it means, what it accepts, what it ships with, and what happens if it is left empty.
Where this page is, and what it governs
Settings → Banding policy, at /Members/Settings/BandingPolicy. Tabs across the top select the age group — only development age groups appear, because only they band. At Warrington Junior Football League that is U6 to U11.
The policy is held per age group, per season. U8 and U9 are configured separately and need not agree. Saving records who saved it and when; the footer of the page shows that provenance, because a value nobody chose sitting where a chosen one should be is the whole problem this page exists to prevent.
The page has three groups of fields, and they are not equal:
- What only this league can answer — no defaults. An age group missing any of these cannot generate a review at all.
- The cycle — when a review happens and how much evidence it reads.
- Gates, cooldown and caps — these ship with working values. Change them only if this league genuinely differs.
An unconfigured age group refuses, it does not guess
Five values have no safe default, and the portal will not invent one. An age group without them shows:
“U6 has no banding policy, so no review can run.”
The reasoning is worth understanding, because it explains the whole design of this page: a plausible default would be silent. An officer looking at a plausible-looking set of recommendations would have no way to tell that the thresholds behind them were invented rather than chosen by the league.
Ratings, metrics and the competitiveness picture are still computed and still visible for an unconfigured age group. Only movement depends on this configuration.
Group 1 — What only this league can answer
These five are the ones with no default. Each is a decision about how this league runs, or about what its fixture platform can actually do.
Does movement happen mid-season here?
Dropdown. Required. Options: Yes — teams may change division during the season / No — the structure is set for the season.
Whether a team may change division part-way through the season in this age group at all. Some leagues band once in September and leave it. Choosing No does not switch the engine off — it still assesses every team and still shows you the picture — but its recommendations become advisory and no movement is proposed.
What can the fixture platform actually do?
Dropdown. Required. Four options:
| Option | Meaning |
|---|---|
| A — Block generation | Fixtures are generated in blocks. The league may regrade as often as it likes between blocks. |
| B — Phase generation | Once per phase only. |
| C — Full season with one mid-season reset | The whole season is generated up front, with exactly one opportunity to restructure. |
| D — Fixed | Fixtures cannot be regenerated. Reviews are advisory only and no mid-season movement is possible. |
This records what the platform permits, not what the league would prefer. Getting it wrong produces a structure nobody can key in: a cadence the fixture platform cannot deliver is a set of movements that never happen. Ask the fixture platform; do not assume.
Review interval (days)
Number, 0–400. Placeholder 84. Paired with a tick box: “This league has no stated interval.”
The league’s own stated interval between banding reviews — the standard it holds itself to. It drives the Overdue state on the readiness panel: once this many days have passed since the last review, that age group is reported as overdue.
If the league has no such standard, tick the box rather than inventing a number. The portal will then never report the age group as overdue. It will not invent a standard and measure the league against it.
Set one or the other. A blank number with the box unticked is the unanswered state, and the age group stays unconfigured.
Handover lead time (days)
Number, 0–120. Placeholder 10.
How far ahead of the fixture platform’s block generation a new structure has to be published. This is an agreed figure, not an estimate: it must include the manual keying step at the county’s end, which takes a person’s time.
It is what produces the Handover by date on the readiness panel. A review generated after that date misses the next block of fixtures entirely — the recommendations are sound and simply arrive too late to be used.
Club consultation window (days)
Number, 0–90. Placeholder 7. Paired with a tick box: “No consultation.”
How long an affected club has to comment on a proposed move before it takes effect. If this league does not consult — the committee decides and clubs are told — tick the box. As with the review interval, the tick box is a real answer and an empty field is not.
Contradiction threshold
A sixth field sits in the same group but behaves differently — it is only consulted by one kind of cycle.
Contradiction threshold (goals)
Number, 0–10, in steps of 0.5. Placeholder 3.
After the evidence window closes, a checkpoint game week is played and its outcome compared with what the rating predicted. This is how far the outcome must diverge from the prediction before it counts as contradicting the recommendation — measured in goals.
A value of 3 means: if the checkpoint result differs from the predicted margin by three goals or more, the engine treats the recommendation as challenged and applies whatever When the checkpoint contradicts says. Set it lower and the engine is more easily talked out of a move; higher and it holds its nerve.
Only used by a matchday cycle. The other trigger modes have no checkpoint, so the value is ignored.
Group 2 — The cycle
This group decides when a review happens and how much evidence it reads. The two are linked more tightly than they look.
When does a review happen?
Dropdown. Ships set to Matchday cycle.
| Mode | Behaviour |
|---|---|
| On demand | Nothing is scheduled. A review runs only when an officer asks for one. |
| Scheduled | Every N calendar weeks, regardless of how much football has been played. Uses Scheduled interval. |
| Matchday cycle | Every N game weeks — counting football actually played, not time elapsed. Uses the evidence window and checkpoint. |
| Block boundary | Whenever the fixture platform generates a new block. |
| Threshold prompt | When enough teams have drifted outside their band to be worth acting on. |
A game week and a calendar week are not the same thing. A postponed round moves the game week but not the calendar. At development ages, where fixtures are lost to weather, the difference matters.
Evidence window (game weeks)
Number, 1–40. Placeholder 9.
How many game weeks of football the engine reads when assessing a team. A window of 6 means the last six game weeks are what the recommendation rests on.
This is the single most consequential number on the page, for a reason given below: it also sets the cycle stride.
Checkpoint (game weeks)
Number, 1–5. Ships at 1.
After the window closes, this many game weeks are held out and played before the review is confirmed. The checkpoint is the engine marking its own homework: it predicts those results from the ratings it just computed, then compares. It belongs to two cycles at once — it closes the one before it and opens the one after.
Scheduled interval (weeks)
Number, 1–52. Empty unless used.
Only used when the trigger mode is Scheduled. Ignored otherwise, and may be left blank.
Stride override (game weeks, optional)
Number, 1–40. Optional.
By default the cycle stride equals the evidence window — they are not independent settings. A six-game-week window produces a review every six game weeks. This field breaks that link deliberately, letting a league assess over a long window but move more often (or the reverse).
Leave it blank unless you have a specific reason. Overlapping windows mean consecutive reviews read much of the same football, and a team can be moved twice on largely the same evidence.
Group 3 — Gates, cooldown and caps
These ship with working values. They are the engine’s caution, and changing them changes how readily it will act. The shipped values are shown as they stand in production.
| Field | Accepts | Ships | What it means |
|---|---|---|---|
| Minimum fixtures in the window | 0–40 | 3 | How many fixtures a team must have played inside the evidence window before it is assessed at all. A team below this is reported as not assessed rather than guessed at. |
| Minimum fixtures overall | 0–60 | 5 | The same test against the whole season to date. Catches a team with a busy recent fortnight but almost no season behind it. |
| Minimum proportion of teams (0–1) | 0–1, steps of 0.05 | 0.6 | How much of the age group must clear those two gates before a review runs for the whole age group. At 0.6, if fewer than 60% of teams qualify the review is refused outright — a structure decided by a minority of the age group is not a structure. |
| Maximum divisions moved per review | 1–5 | 1 | The furthest a team may travel in one review. At 1, a team can move one division per review no matter how far its rating has drifted. Raise it only where the league is comfortable with a team jumping two bands at once. |
| Cooldown (fixtures) | 0–40 | 4 | After a team moves, how many fixtures it must play in its new division before it may be moved again. Settling time — a team that has just moved has no record in its new band yet. |
| Cooldown (weeks) | 0–52 | 6 | The same settling period in calendar weeks. Both apply; a team leaves cooldown when both are satisfied. |
The readiness panel reports how many teams are currently inside cooldown, so you can see when a review would be thinly populated before running it.
Group 3 continued — shape of the review
Four dropdowns complete the group. These change the shape of the review rather than its caution.
| Field | Options | What it means |
|---|---|---|
| Assessment window ships: MatchdayWindow | MatchdayWindow, SinceLastRegrade, RollingWeeks, LastNFixtures, SeasonToDate | Which football counts as evidence. MatchdayWindow uses the evidence window above. SinceLastRegrade reads everything since the team last moved. SeasonToDate reads the whole season, so early form never falls out of view. LastNFixtures counts fixtures rather than game weeks; RollingWeeks counts calendar time. |
| When the checkpoint contradicts ships: Flag | Flag, ReduceMovement, Hold | What happens when the checkpoint disagrees with the recommendation by more than the contradiction threshold. Flag passes it to the committee marked as challenged. ReduceMovement proposes a smaller move. Hold withdraws the recommendation entirely. |
| A new structure applies ships: NextMatchday | NextMatchday, BreakAligned, FixedDates | When a ratified structure takes effect. NextMatchday is the next game week. BreakAligned waits for a natural break — Christmas, a half-term — so teams change division between blocks of football rather than mid-run. FixedDates uses dates the league has set. |
| The season’s final cycle ships: RunWithoutCheckpoint | RunWithoutCheckpoint, Skip | The last cycle of a season has no room left for a checkpoint. RunWithoutCheckpoint runs it anyway, unconfirmed. Skip does not run it, leaving the structure as it stands until the season ends. |
Saving, and policies that contradict themselves
Save policy appears at both the top and the bottom of the page; they are the same button. The whole page saves together.
Some combinations are refused rather than stored, because they cannot both be true. The clearest is permitting mid-season movement while telling the portal the fixture platform is regime D — Fixed: no regeneration means no movement, whatever the league would prefer. Saving that pair returns the page with an explanation and nothing written.
A saved policy records its author and date. The footer reads, for example:
This policy was set by the league, by 676bf331… on 15 Sep 2026.
Provenance is recorded because the entire purpose of this page is that every threshold behind a recommendation was chosen by somebody.
Show me the rhythm — use it before you save
The button at the foot of The cycle projects your settings onto this season’s actual calendar before anything is saved. It is the quickest way to catch a setting that is arithmetically fine and practically absurd.
The same age group — Warrington U8 — with only the evidence window changed:
| Evidence window | Projection |
|---|---|
| 1 game week | 20 structure changes — one every 1.1 weeks |
| 9 game weeks | 2 structure changes — 14 Nov, 30 Jan |
| 5 game weeks (adopted) | 4 structure changes — 17 Oct, 21 Nov, 9 Jan, 13 Feb |
Read the projected count against the regime you chose. Under A a frequent cadence is the design intent. Under B it should be roughly the number of phases in the season. Under C it should be one.
The same settings project differently in different age groups, because the projection uses each age group’s own calendar. U7 and U8 both run a 5‑game‑week window, and U7 projects 2 changes to U8’s 4 — U7 simply has fewer game weeks in its season.
Worked example — a window that made the gates unsatisfiable
The rule is arithmetic: minimum fixtures in the window can never exceed the number of fixtures a team can actually play inside the evidence window. For a league playing one fixture a game week, a window of N game weeks holds at most N fixtures per team.
Warrington’s U8 policy showed what happens when that is violated. It was saved with an evidence window of 1 game week and a minimum of 3 fixtures in the window. No team could ever satisfy it, so no review could ever run — and the two halves of the portal disagreed about it:
| Readiness panel said | Preview said | |
|---|---|---|
| State | Ready | The review was not generated |
| Teams meeting the gates | 167 of 167 | 0 of 167 (0 %), against a threshold of 60 % |
Where the two disagree, the preview is the one that decides. The readiness panel summarises; the preview runs the engine. The disagreement was a symptom of the impossible window, not a separate fault — once the window was corrected the two agreed again.
U7 and U8 now run the league’s long-standing cadence: evidence window 5, checkpoint 1 — game weeks one to five are assessed and game week six confirms. The readiness panel reports it in those words:
Not yet — Cycle 1: matchday 2 of the 1–5 evidence window has been played. 3 more game week(s) to go, then the checkpoint at matchday 6.
Still no review, but for a reason that resolves itself as the season progresses rather than one that never can.
Always run a preview after changing a policy. The projection tells you how often reviews will happen; only the preview tells you whether one can happen at all.
The four regrade regimes
This is the most consequential answer on the page and the one most often guessed at. It records a fact about a different system — the league’s fixture platform — not a preference of the league’s.
League360 cannot generate fixtures. It can only publish a revised structure and rely on the fixture platform to act on it. So the question behind this field is narrow: how much of the season does that platform commit to at a time, and will it generate the next stretch against a structure we hand it?
The answer sets the ceiling on how often a team can change division. Everything else in The cycle should be configured underneath that ceiling, not above it.
| Option | The platform must be able to… | Regrades achievable |
|---|---|---|
| A — Block generation | Generate a short run at a time, after each published structure | As often as the league wants |
| B — Phase generation | Generate in two or three phases per season | Once per phase |
| C — Mid-season reset | Discard and regenerate the remainder of the season, once | One, typically at the halfway point |
| D — Fixed | Nothing — the season is generated once and stands | None. Reviews become end-of-season evidence |
They are listed in descending order of value to the league, and the difference between each is simply how much football the platform commits to before it will accept a new structure.
A, B and C in depth
A — Block generation
What has to be true: the platform generates a short run of fixtures at a time — a block — and generates each new block after a structure is published, against that structure.
What it costs: nothing in flexibility. Each block is a fresh opportunity to act on a review, so a team found in the wrong band in October is playing the right opposition within a block or two.
Cadence: as often as the league wants. This is the design intent of the whole module, and the only option where the banding engine can do what it was built to do.
Configuring underneath it: the evidence window is a free choice, limited only by wanting enough football behind each judgement. A short window gives responsive banding and frequent restructures; a long one gives steadier judgements and fewer. Check the projected count against how many handovers the league can actually staff — every restructure is keyed in by hand at the county’s end.
The trap: choose A only if the platform genuinely waits for your structure before generating each block. One that generates blocks on its own schedule regardless of what you publish is not option A — it is B or D wearing A’s clothes.
B — Phase generation
What has to be true: the platform generates in two or three larger phases across the season — September to December, then January to May, for instance.
What it costs: fixtures for the phase in progress already exist against the structure that was current when that phase was generated. A structure published mid-phase cannot be acted on until the next phase boundary, however sound the recommendations. A team identified as mis-banded in week three of a fifteen-week phase plays the rest of that phase where it is.
Cadence: once per phase — two or three regrades a season.
Configuring underneath it: size the evidence window so the cycle stride lands near the phase length, and consider A new structure applies → BreakAligned so a ratified structure waits for the boundary rather than trying to take effect mid-phase. The projected number of structure changes should be roughly the number of phases; if it says twenty, the cycle and the regime are contradicting each other.
C — Full season with one mid-season reset
What has to be true: the platform generates the whole season up front but can discard and regenerate the remainder, once.
What it costs: a single shot. Spend it early and a mis-banding discovered in February stands until the season ends; spend it late and teams have already played most of the season in the wrong band. The halfway point is the usual compromise, which is why the option exists at all.
Cadence: exactly one regrade.
Configuring underneath it: this is the one case where you genuinely want a long evidence window — the review is a single considered judgement on half a season, not a running correction. Aim the projection at one structure change. FixedDates is often the right choice for when it applies, since the league already knows when its reset falls.
D — Fixed, and why it is still worth having
What has to be true: the platform generates the full season once and nothing about it changes afterwards.
Cadence: none. No team moves division mid-season.
It is not a failure state, and should not be treated as one
Under D the ratings still run, the reviews still generate, and the evidence still accumulates. What the league loses is movement, not analysis. That evidence still feeds:
- the U11 → U12 seeding, where the first competitive divisions are formed;
- next season’s opening structure, which is the largest banding decision a league makes;
- the competitiveness picture across the age group — which divisions are lopsided, and by how much.
A league on D gains the whole analytical apparatus without the ability to act on it mid-season. That is a materially better position than having no view at all.
But it has to be a decision, not a discovery
The damaging version of D is the one nobody chose: a league publishes a banding policy to its clubs in September, promises movement, and finds out in November that the fixture platform was never going to regenerate anything. A published policy that cannot be acted on costs more trust than having no policy at all.
If D is the answer, tell the clubs it is the answer. “We assess every cycle and act at season’s end” is a coherent thing to say. “We will move your team when it is clearly mis-banded”, followed by nothing, is not.
The one option the portal treats differently
D is the only regime the software itself acts on. Choosing it changes two things:
- The readiness panel stops offering a review and reports structure is fixed for this season — reviews are advisory only. The ratings, reviews and evidence still run and feed next season’s structure, rather than a button that leads nowhere.
- Any review that is run carries a warning on its face saying the structure cannot change mid-season and that this review is end-of-season evidence — so a set of recommendations cannot be mistaken for a set of movements.
Setting Does movement happen mid-season here? to No produces the same two effects. The difference between them is meaning rather than behaviour: D says the platform cannot, No says the league chooses not to. Both are worth recording accurately, because next season somebody will ask which it was.
How to find out which one you are
The label describes a capability, so the answer comes from the fixture platform rather than from the committee. Three questions settle it:
- How far ahead do you generate fixtures? A few weeks at a time → A. Two or three large chunks → B. The whole season → C or D.
- If we send you a revised structure mid-season, will you generate the next fixtures against it? Yes, every time → A or B. Once → C. No → D.
- How much notice do you need? This is the handover lead time, and it is the binding constraint on cycle length — not the statistics. A regrade ratified on the Tuesday before generation is useless if they need ten days.
Evidence you already have
If the league has been banding for a while, its own history answers question 2 better than anyone’s recollection. A league that has successfully regraded every six game weeks has a platform that accepts a new structure every six game weeks — which is block generation, whatever anyone calls it.
That is how Warrington’s U7 and U8 were settled: a long-standing cadence of assessing game weeks one to five and confirming in week six projects as four structure changes a season, and four is a number only a block-generating platform can absorb.
It can differ by age group, and often does
The same platform may generate U8 in short blocks and U11 as a full season. That is exactly why this is set per age group rather than once for the league — ask about each separately rather than assuming one answer covers them all.
Whichever option applies, the sequence is the same
- A review is generated, or a configured trigger fires.
- The committee ratifies; movements are recorded with a reason for any deviation.
- A structure publication is created and handed to the fixture platform.
- The platform generates the next block, phase or season against it.
- Fixtures sync back and are reconciled against what was published.
Step 5 is not optional. Two systems holding a division structure will drift apart, and a team moved here but not there is invisible until somebody notices a fixture against the wrong opposition.
What the portal does with your answer — and what it does not
The portal records all four regimes but only branches on D. A, B and C currently behave identically to one another.
So choosing B or C does not cap the cycle, and nothing warns you when the cadence you configure asks for more regrades than the platform can absorb. Under B and C the ceiling has to be honoured by hand, in the cycle settings, and checked with Show me the rhythm before saving.
The mismatch is quiet: recommendations are produced, the committee ratifies them, a structure is published — and the fixture platform simply cannot consume it until its next boundary. Nothing in the portal objects.