Exploration has targeting down to a science, and choosing down to a coin flip.
AI now finds more drill targets than any junior explorer can afford to test.
The bottleneck has moved from finding to choosing, and the industry has no discipline for choosing.
Compare the return per dollar spent
Exploration return: US$135B of estimated discovered value against US$196B of expenditure over 2012–2021, Schodde (2023), MinEx Consulting - IMARC, Sydney, 31 Oct 2023, slides 22 & 30 .
*Poker machine minimum return is set by state law, between 85 and 92 cents in the dollar depending on the state.
The tools exist.
Mining never imported them.
The oil and gas sector prices a prospect before the rig arrives. Venture capital sizes every cheque against the odds of failure. Private equity holds no asset without an exit thesis. Exploration never really imported that discipline.
The shared toolkit, expected value analysis, real options, Bayesian updating, is decades old and unglamorous.
Junior mineral exploration runs on the same uncertainty with little of the same discipline. Programs are approved on geological merit, market appetite and gut feel; rarely on the question of what a dollar in is expected to bring back.
Machine learning and AI is making this worse, not better: every improvement in targeting produces more drill-ready prospects per company than the budget can test, which makes the choice between them the decision where shareholder value is won or lost.
Can the odds actually be brought up?
Not the way a gambler would hope,
and that's the point.
No framework makes an individual drill hole luckier. The orebody is there or it isn't. What capital-allocation discipline changes is the portfolio:
- It stops money flowing into programs whose expected value is negative before the rig arrives
- It sizes each program to the value of the information it will produce
- It kills fading projects quickly instead of funding them out of habit
The same geology, the same drill holes, fewer wasted dollars around them. That is how $0.69 moves toward and past $1.00, not by changing the rocks, but by changing which cheques get written.
Three questions every board should answer before approving a program.
If we spend $500,000 on this program, how much shareholder value do we expect to create?
With four projects and $1‑million to spend, what allocation produces the best return on shareholders' capital?
If we ask new investors for $1‑million, what return should they expect once the program is run?
If these can't be answered, the program isn't a plan,
it's a position.
These three questions are the short form of the six-check board test, Before the cheque.
Requiring the six checks is the board's job. Producing them does not have to be. That is where Rocworth comes in. Rocworth runs the numbers, builds the models, and delivers the board paper that carries them.
Read more about my analysis of Exploration Economics
Benjamin Bell
Exploration geologist with thirty years and a discovery record that includes the 2.44 Moz Katanning Gold Deposit; former managing director of two ASX-listed explorers.
Rocworth is the integration of that experience with the capital-allocation toolkit, built for the boards, brokers and investors of junior explorers.
Presenting on this topic at the AIG ML/AI in Greenfields Exploration workshop, Perth, 29 July 2026.
Requiring the numbers is the board's job.
Producing them is ours.
Rocworth runs the analysis and delivers the board paper that carries it. Drill-program valuation, portfolio allocation, and the expected return your next raise can stand behind. Tell us about one program and we will show you what the six checks look like on your own numbers.