Accuracy scoreboard
Every past catalyst is graded after the fact against the announcements that actually followed. Nothing airbrushed — and when we can't yet confirm an outcome, it stays "pending" rather than being guessed.
Outcomes graded
3,460 decided · 2,251 pending
Hit-rate
landed ÷ decided
High-confidence hit-rate
2,506 decided
Avg timing error
landed events · |predicted − actual|
By catalyst category
| Category | Decided | Hit-rate | Avg error | Distribution |
|---|---|---|---|---|
| Drilling Results | 586 | 34% | 25.1d | |
| Regulatory Filing | 438 | 2% | 24.8d | |
| Other | 423 | 0% | — | |
| Exploration Update | 361 | 68% | 24.5d | |
| Capital Raise | 235 | 5% | 17.9d | |
| Production Update | 183 | 8% | 13.6d | |
| Earnings | 132 | 2% | 2.0d | |
| Clinical Trial | 117 | 21% | 33.7d | |
| Resource Estimate | 114 | 15% | 27.3d | |
| Guidance | 99 | 1% | 30.0d | |
| Project Development | 95 | 39% | 28.7d | |
| Contract Award | 76 | 3% | 22.0d | |
| Corporate Action | 74 | 12% | 15.9d | |
| Acquisition | 73 | 8% | 14.3d | |
| Financial / Cashflow | 57 | 9% | 9.4d | |
| Feasibility Study | 53 | 51% | 22.9d | |
| Management Change | 49 | 12% | 20.7d | |
| Permitting | 48 | 54% | 28.1d | |
| Investor Presentation | 43 | 9% | 14.5d | |
| Annual General Meeting | 41 | 15% | 20.2d | |
| Exploration | 37 | 0% | — | |
| Trading Halt | 29 | 0% | — | |
| Metallurgical Results | 24 | 46% | 17.0d | |
| Partnership / JV | 19 | 11% | 8.5d | |
| Debt Financing | 11 | 27% | 16.7d | |
| Commercialisation | 9 | 56% | 20.6d | |
| Legal / Litigation | 6 | 0% | — | |
| Grant / Non-dilutive | 5 | 20% | 38.0d | |
| Investment | 5 | 0% | — | |
| Product Update | 5 | 0% | — | |
| Corporate | 4 | 0% | — | |
| Drilling | 2 | 0% | — | |
| Drilling Program | 2 | 0% | — | |
| Strategic Review | 2 | 0% | — | |
| Dividend | 1 | 0% | — | |
| Exploration Results | 1 | 0% | — | |
| Shareholder Meeting | 1 | 0% | — |
"Decided" = landed + missed · withdrawn & pending excluded · Regraded twice daily · Last graded 1 Nov 2026
Method notes
Every catalyst we publish is stamped with its predicted date and confidence grade. After the window passes, we look for a later announcement of the same category from the same company near the predicted date.
Landed = a corroborating announcement appeared within ±60 days of our predicted date. The signed gap between the two is the timing error we report.
Missed = no corroboration appeared long after the window closed. Pending = still too early to call — we never guess a landing.
High-confidence = the subset of catalysts our extraction graded "High" at publish time, scored the same way.
Events flagged cancelled or withdrawn by the issuer are excluded from the hit-rate.
This scoreboard tracks timing accuracy, not outcome prediction. Catalyst Ahead doesn't predict share-price direction. See .
Not financial advice. Catalyst Ahead is an AI-driven research tool. All projections are inferences from public company announcements. AI-derived inferences from company announcements. Always do your own research and consult a licensed adviser before trading.