How the club league comparison is built — the metrics, the reasoning, and the limitations · 2025/26 Season
The Club League Strategy Comparison report draws conclusions about six (and counting) managers — their strategies, their relative performance, and their prospects for the run-in. This document explains exactly how those conclusions are reached: where the data comes from, what each metric measures, how managers are classified into archetypes, and where the analysis has blind spots.
The intended audience is any manager who wants to go beyond the headline numbers and understand the reasoning. Nothing here requires a statistical background — the goal is transparency, not complexity.
Every metric in the comparison report derives from one source: the FPL history API endpoint captured as JSON files per manager.
| File pattern | API endpoint | Contains |
|---|---|---|
| fpl_*_history.json | /api/entry/{id}/history/ | GW-by-GW points, OR, squad value, transfers, hits, bench points. One row per gameweek. |
| fpl_overall_top10.json | /api/leagues-classic/314/standings/ | Global overall league standings — the top 50 managers worldwide by total points. |
The FPL website uses Datadome, a bot-detection system that intercepts requests made by injected scripts — browser extensions, MCP tools, DevTools console commands. The danger is not that Datadome blocks these requests. It is that Datadome silently serves wrong data.
The correct method is to open the FPL website in a normal browser session, open DevTools → Network tab, navigate to trigger the relevant API calls, and save the raw JSON response. The browser makes the request natively with full session context and real cookies — Datadome cannot distinguish this from a human user.
The history endpoint is aggregate only. It tells you how many points a manager scored each GW, what their OR was, and how many transfers they made — but not which players they picked, who they captained, or what specific transfers occurred. Player-level detail requires the picks endpoint (`/api/entry/{id}/event/{gw}/picks/`), which is not used in this analysis.
Every column in the standings table and every bar in the charts maps to one of these fields from the history JSON. The field name shown is the exact key from the API response.
| Metric | Definition | Why it matters |
|---|---|---|
| overall_rank (OR) | The manager's position in the global standings after each gameweek, ranked by total_points descending. | Normalises performance against the full player pool each week. A 70-pt GW during a high-scoring week might drop your OR; the same score during a blank week might raise it. OR captures context that raw points miss. |
| total_points | Cumulative sum of points scored across all played GWs. Hit deductions are already applied by the API — i.e., a 4-pt hit GW shows the net score. | The primary competitive metric. All other metrics exist to explain why total_points ended up where it did. |
| points (GW score) | Points scored in a single gameweek, net of any transfer hit deductions applied that GW. | Used to identify high/low-water marks and to correlate chip plays with scoring spikes. |
| value | Sum of the current sell prices of all 15 squad players at the end of each GW, in tenths of £m (i.e., 1065 = £106.5m). | Proxy for asset quality. Players rise in price as ownership increases; holding winners and selling losers early compounds the value advantage. |
| Squad Value Growth | value at GW31 minus the starting budget of 1000 (£100m), expressed in £m. E.g., value 1086 → growth of +8.6m. | Isolates the accumulated asset management advantage. Managers who started with the same £100m ended GW31 with squad values ranging from £104.9m to £108.6m — a 3.7m structural gap. |
| points_on_bench | Total points scored by the four bench players in a given GW, regardless of whether any were used as automatic substitutes. | Measures the cost of incorrect lineup ordering. High bench waste does not always indicate poor play — it can mean your bench was unexpectedly strong. Context is required. |
| Bench Waste (season) | Sum of points_on_bench across all 31 GWs. | Cumulative measure of lineup mis-ordering. Lower is better, but must be read alongside OR — Ibsen has the highest bench waste (380pts) yet holds OR #1, because his starting XI dominated. |
| event_transfers_cost | Point deduction applied that GW for additional transfers beyond the free transfer allowance. Each extra transfer costs 4 points. | The direct cost of reactive transfers. Shown both per-GW and as a season total. |
| Hit Cost (season) | Sum of event_transfers_cost across all 31 GWs. | Headline measure of transfer discipline. Zero is achievable — four managers in this club achieved it. |
| event_transfers | Total number of transfers made in a given GW, including free transfers. | Combined with hit cost, reveals whether a manager's activity was disciplined (high transfers, zero hits = rolling free transfers) or reactive (high transfers, high hits = impulsive churning). |
| H1 / H2 Points | Total points scored in GW1–19 (H1) and GW20–31 (H2) respectively. See Section 8 for why GW19 is the dividing line. | Reveals whether performance is structurally consistent across both halves or concentrated in one phase, often due to chip deployment timing. |
Each manager in the comparison report is assigned an archetype — a short label that summarises their dominant strategic behaviour across the season. Archetypes are not prescriptive; they are descriptive. They are assigned after the fact by looking at four independent signals and finding the dominant pattern.
| Season hit cost | Classification | Interpretation |
|---|---|---|
| 0 pts | Elite discipline | Never took a hit. Every transfer was planned within free transfer budget. |
| 1–12 pts | Near-elite | At most 3 hits across 31 GWs — occasional tactical exception, not a pattern. |
| 13–24 pts | Average | 3–6 hits across the season — reactive transfers were a recurring tendency. |
| 25+ pts | Reactive | 7+ hits — transfer activity driven by short-term performance anxiety rather than long-term planning. |
| Avg transfers/GW | Classification | Interpretation |
|---|---|---|
| < 1.0 | Surgical | Regularly rolls free transfers — deploys them in batches when genuinely needed. |
| 1.0 – 1.1 | Normal | Uses the free transfer most weeks. Balanced activity. |
| 1.11 – 1.19 | Active | Regularly uses accumulated transfers plus occasional hits. |
| > 1.19 | Churn | Consistent overactivity. More transfers than free transfers available, not justified by OR outcomes. |
| Pattern | Classification |
|---|---|
| All H1 chips played by GW14 | H1 Blitz — commits early, builds from a clear position |
| Chips spread across H1 and H2 | Structured — staged deployment aligned with the season calendar |
| Most or all chips concentrated in H2 | Contrarian / Late — high-risk high-reward approach; depends on timing being right |
| Chips used reactively without a clear plan | Mixed / Reactive — chips deployed in response to events rather than in anticipation of them |
| Pattern | Archetype indicator |
|---|---|
| Top 100 from early GWs, sustained all season | Early Detonator or Frontrunner |
| Gradual, consistent climb through all 31 GWs | Patient Accumulator |
| Flat/mid-table H1, then steep H2 climb | Late Bloomer / Contrarian |
| Volatile — large OR swings week to week | Reactive / Hit-Taker |
All analysis in this project is anchored to a four-layer decision framework, where each layer must be resolved before the next begins:
The report identifies the GW6 TC as the single most impactful shared decision of the season. This conclusion is arrived at retrospectively, not predictively: 5 of 6 club managers played TC on GW6, and all five saw significant OR improvements that week. This convergence — managers making the same call independently — is strong evidence that GW6 was a genuinely high-ceiling captain week, likely featuring a premium asset (Salah, Haaland, or similar) in a highly favourable fixture.
The estimated impact of 40–60 pts above an average manager assumes the TC captain returned approximately 3× their normal expected score (a realistic ceiling for a premium asset in a great fixture). At 10M managers, even modest captain differentiation compounds into large OR movements.
The optimal BB window is when the squad is deepest and most in-form. Early in the season (GW1–10), managers typically have 15 genuinely viable players before injuries and form divergence thin the squad. Playing BB on GW1 (Trundlers, 115 pts) or GW9–10 (Mirinda, Ibsen) captures the full 15-player depth. Playing BB in GW18+ risks having 3–4 bench players who are injured, blanking, or in poor form — shrinking the effective return.
Chips are not just about the immediate GW they're played — they are structural advantages for the rest of the season. As of GW31, the chips-remaining picture is:
| Manager | Chips Remaining | Run-in ceiling |
|---|---|---|
| Ibsen, Trundlers, Mirinda | WC2 + TC2 | Highest — can restructure squad entirely AND play an elite TC on any GW |
| Crème de la Crème | FH | Strong — FH is most valuable during a blank GW (fields a full 15-player squad unaffected by blanks) |
| Bruno | WC2 | Moderate — can reset the squad but no TC multiplier remaining |
| Howitts | TC2 | Limited — one elite captain play but no squad restructure available |
The common framing of a transfer hit is: "I paid 4 points to get Player B instead of Player A." This understates the true cost. A transfer hit has three components:
Howitts paid 28 pts in hits over the season. The average GW score across the club is approximately 65 pts. That means Howitts surrendered the equivalent of nearly half an average gameweek to transfer penalties — before accounting for opportunity and regression costs. In a season where the gap between OR #1 (Ibsen, 2061 pts) and OR #33,068 (Howitts, 1880 pts) is 181 pts, the hit tax is not the only explanation, but it is a measurable and entirely avoidable contributor.
Every manager starts the season with £100m. The ending squad value is a function of which player prices rose (held correctly) and which fell (sold before the drop or avoided entirely). A manager who ended GW31 with £108.6m (Ibsen) has effectively "earned" 8.6m in player appreciation — meaning his squad generates more points-per-pound than a squad worth £104.9m (Howitts), because each of his 15 positions is filled with a more expensive (and generally more in-form) asset.
The 3.7m gap between Ibsen and Howitts at GW31 equates to roughly 2–3 extra points per GW in expected scoring, compounded across 7 remaining gameweeks. This is a structural advantage that cannot be erased by a single good captain pick.
GW19 is not an arbitrary midpoint. It is the last gameweek in which Wildcard 1 (WC1) can be played — the FPL rules restrict WC1 to the first half of the season (GW1–19) and WC2 to the second half (GW20+). This makes GW19 a genuine structural boundary in the season: chip availability, squad composition philosophy, and the fixture calendar all shift at this point.
The GW-by-GW heatmap in the comparison report colours each cell based on the manager's overall rank that week. The bands are not evenly distributed — they are weighted towards the elite end because that is where meaningful differentiation occurs between the managers in this club.
| OR range | Colour | What it means |
|---|---|---|
| 1 – 100 | Bright green | Elite — top 0.001% of all managers globally. Only Trundlers (GW6–7), Mirinda (GW9), and Ibsen (GW29+) have spent time here. |
| 101 – 1,000 | Light green | World-class — top 0.01%. Consistent performance at this level signals elite squad construction. |
| 1,001 – 10,000 | Teal | Excellent — top 0.1%. Most elite players cycle through this band during their best weeks. |
| 10,001 – 100,000 | Amber | Good — top 1%. A single poor GW or unlucky blank typically pushes a top-10k manager into this band. |
| 100,001 – 1,000,000 | Light red | Average — top 10%. Where most managers live week-to-week during ordinary GWs. |
| 1,000,001+ | Dark red | Below average week — bottom half of the player pool. Common in GW1–2 before the field stratifies. |
Raw GW scores are not comparable across gameweeks. A 70-point GW during a blank round (fewer players scoring) might rank higher globally than a 75-point GW during a double gameweek (everyone scoring more). Overall rank normalises the score against the actual field each week, making comparisons meaningful across the season.
"wildcard" in chip history, distinguished only by the event number. Any tool that counts "wildcard" entries without checking the GW will report zero wildcards remaining even when WC2 was never played. In this analysis, chip status is always cross-checked against the event number (events 1–19 = WC1, events 20+ = WC2).