How we source, score, and verify the NHL trade-deadline rumors you read in the Wire — and exactly where the numbers come from.
A long-form explainer on the desk's three-tier sourcing system, the 1,184-rumor verification log, the Nakano Index credibility model, and how our fantasy-relevant projection hit rate is measured. No marketing claims, no anonymous sourcing — just the methodology behind every rumor you've seen quoted back to us on Twitter.
Reading time: 9 minutes. Last revised 14 February 2025. Cite as: Flick or Nakano'd Media Inc. (2025). Methodology, v3.2.
The verification log: 1,184 rumors, 847 confirmed within 48 hours.
Every rumor we've published since the March 2019 launch lives in a single auditable log. Each row carries a timestamp, the source tier, the prediction, the confirmation timestamp (if any), and a verdict. The full log is downloadable by GM Kit subscribers. The numbers below are computed directly from that file — no editorial rounding.
-
Rumors logged since launch 1,184 Across six trade-deadline cycles, March 2019 – April 2024
-
Confirmed accurate within 48 hours 847 A 71.5% raw hit rate on the lifetime log
-
Newsletter subscribers, March 2020 → 2024 312 → 48,700 From the 2020 cycle to the 2024 cycle, no paid ads
-
Industry recognition #1 niche "Best Niche Fantasy Hockey Resource" — Fantasy Hockey Geek reader poll, 2023
Methodology note: "Confirmed accurate" means the rumored player move, team, or contract term landed publicly within 48 hours of publication. We do not score partial confirmations — a half-right rumor counts as a miss in the log. Rows are append-only.
Three sourcing tiers, each with a measured lead time.
"Ahead of the headline" only means something if you can put a number on it. Every rumor we publish is tagged with a tier before it goes out, and the tier is what determines how the desk frames it — not the headline. Lead times below are averaged across the 2024 deadline cycle and tracked against major-outlet publication timestamps by an independent volunteer reviewer.
-
01Tier 1
Beat reporters on the record.
Direct messages, phone calls, or in-person conversations with the 22 NHL beat reporters we track. We never publish a tier-1 item on attribution alone — we cross-check the claim against at least one independent beat reporter covering the same beat before it leaves the desk. Tier-1 items are published with the reporter's name and outlet where we have written consent; otherwise we tag them as "tier-1 sourced" without naming the reporter.
- Average lead time in 2024
- 6.1 hours
- Sample size
- 138 published tier-1 items
- Confirmation rate (48h)
- 82.6%
-
02Tier 2
Secondary sources and aggregators.
Agent chatter, equipment-manager text threads, and independent podcast hosts who surface a rumor before a major outlet does. Tier-2 items are treated as leads, not facts. They are published with a confidence range and almost always re-tagged within 24 hours once a tier-1 source confirms or denies. The 2024 cycle saw 2.3 hours of average lead time on tier-2 trades, but with a wider confidence interval and a higher rate of false positives.
- Average lead time in 2024
- 2.3 hours
- Sample size
- 241 published tier-2 items
- Confirmation rate (48h)
- 63.5%
-
03Tier 3
Unconfirmed whispers — labeled, never buried.
Anything from a fan DM, a Slack ping, or a Reddit thread we can't independently verify. Tier-3 items are published under an explicit "unconfirmed" tag in the Wire, with the original source noted where we have permission. They are excluded from our headline lead-time numbers and from our fantasy-relevant projection track record. The desk has a standing policy: we'd rather miss a rumor than ship an unverified one without the label.
- Average lead time in 2024
- Not published
- Sample size
- 189 published tier-3 items
- Confirmation rate (48h)
- 41.3%
Headline lead time against ESPN and NHL Network on confirmed deadline-day trades in 2024: 9.4 hours average. That figure is computed only over trades that were eventually confirmed — it is not a measure of how often we were first.
Scoring 22 beat reporters on a 0–100 credibility scale — as a model, not a leaderboard.
The Nakano Index is a rolling 0–100 score assigned to each of the 22 NHL beat reporters we monitor, recalculated after every rumor resolves. It is not a ranking. It is a credibility model with three inputs: hit rate over the trailing 18 months, mean lead time on confirmed trades (versus a reference panel of major outlets), and a decay function that punishes reporters who publish a high-profile miss and rewards those who publicly correct themselves within 24 hours.
The index updates the morning after a rumor resolves. A reporter who is first on a tier-1 trade but wrong gets a larger penalty than one who is third and right — we weight calibration higher than speed. A reporter who floats an unverified rumor and then walks it back within a day sees a small positive adjustment, because publicly correcting a wrong take is the behavior we want to incentivize in the beat ecosystem.
We deliberately do not publish the full 22-reporter table in the Wire. Subscribers can request their own reporter subset via the desk email. The reason is simple: the Nakano Index is meant to be a working tool, not a trophy list. Reporters who show up at the top of a public table tend to get more anonymous sourcing — which degrades their hit rate the next season. The index works better when reporters don't know exactly where they sit.
Three Substacks with combined circulation north of 25,000 readers have cited the Nakano Index tip sheet in their own methodology pages. We are flattered, and we ask that anyone citing the index notes its trailing-window nature — a 0–100 score from January is not a 0–100 score from March.
Every rumor we publish carries the tier it was filed under. Every projection we make is framed as a probability, not a promise. We will not tell you a player will score 28 goals next season — we will tell you there is a 41% chance, given his deployment and linemates, that he lands between 24 and 31. We will not call something "confirmed" until it is on the record. If we get a rumor wrong, the correction goes out the same day. If we are slow, we say so. This is a signal fire, not a press release.
71.3% hit rate on fantasy-relevant projections — measured, framed, and qualified.
Across the 2022, 2023, and 2024 trade-deadline cycles, the desk published fantasy-relevant projections on 842 roster moves (buy/sell/hold calls keyed to specific players in specific roster slots). The headline hit rate is 71.3%, computed against a 30-day post-deadline outcome window. The breakdown below is the full picture.
- Hit rate on fantasy-relevant projections 71.3% n = 842 plays over three deadline cycles
- Sample window 2022–2024 Three trade deadlines, 30-day post-deadline outcome window
- Tier-1 average lead time 6.1h 2024 cycle, beat reporters on the record
- Tier-2 average lead time 2.3h 2024 cycle, secondary sources and aggregators
Framing note: hit rate is a probabilistic summary. A 71.3% hit rate on 842 plays means a randomly chosen play from the desk would land within the projected range about 71 times out of 100 — it does not mean any individual projection is 71.3% likely to hit. Individual projections carry their own probability ranges, published alongside the call. We do not promise win rates; we publish ranges.
The desk was first to break four of the top eleven trades of the 2024 NHL trade deadline (independent verification by Daily Faceoff's published timeline). Trusted by the commissioners of 14 of the top 25 public keeper leagues on Fantrax and MFL. Cited as a sourcing feed by writers at The Athletic, Daily Faceoff, and EP Rinkside.