The record behind the ranges

Every forward range in this issue can be held against a plain historical record. The Environmental Registry of Ontario has decided mineral-exploration permit files, and each one left two facts behind: how it ended, and how long it took. Before trusting any range on the ladder, look at how those files actually went. The working files in today's queue of are walking the same road.

How decided files ended

Nearly every file that reached an outcome was issued, with terms attached: of , or . Another were withdrawn by the proponent and were refused outright. The registry's historical vocabulary differs from this issue's two-way split: the forward ranges fold refusals and withdrawals together into a single not-approved outcome, and count the risk that a file never reaches a decision at all separately, as the no-decision chance quoted on each permit page.

The lesson of the mix is that outcome is rarely the question for an Ontario exploration permit. Timing is.

How long they took

Two clocks matter: days from the application to the decision, and days from the close of public comment to the decision. The second clock isolates the stretch where the file sits entirely with the Ministry.

From application to decision

80% of decided files landed between and days after the application, and half were decided within days. The tail is long: the slowest file took days.

From comment close to decision

Once public comment closed, 80% of decided files landed between and days, and half were decided within days. of the decided files carry a recorded comment-close date; the rest appear only in the first chart.

Find your stage

Pick a clock and a number of days already elapsed, and read off how the decided files looked at that point.

How to read a credible interval

Each working permit in this issue carries a range: a decision is likely between one date and another. That range is an 80% credible interval: given the data and the method, there is an 8 in 10 chance the decision date falls inside it, a 1 in 10 chance it lands earlier, and a 1 in 10 chance it lands later. The most likely date sits inside the range, marked separately.

Those ranges come from a fitted probability model, and a range from a model is only worth what its performance record says it is worth. How often ranges like these contain the dates that actually arrive is tracked openly on the track record page as decisions land; the record is young, and it will speak for itself as it accumulates.

The percentiles on this page are a different kind of number: simple counts of what past files did, no model involved. They are the sanity check to hold any forward range against. A range that strayed far from this record would need a reason, and the permit pages name the factors when there is one.

The front page lines today's working files up from fastest to slowest against exactly this record, and the table carries every pending file, working or stalled. Historical outcomes from the Environmental Registry of Ontario, as of .