A second thought on what at first seemed like the perfect historical reference ...

Last month, we launched a new feature on TheAtlantic.com called the "Iran War Clock." We based it explicitly on the "Doomsday Clock," an iconic image that the Bulletin of the Atomic Scientists has variously deployed since 1947 to represent its view of how close humanity is at a given moment to complete ruin. It wouldn't be especially modest of us to say that our project is comparatively modest in scope, though we took it to be similar in spirit: "... the purpose is to gauge the chances of conflict with Iran in the hope of producing a more informed debate in the U.S. and abroad."

The original Doomsday Clock has typically been set somewhere between 11:45 p.m. and 12:00 a.m. Its "latest" setting was at 11:58 p.m., after the United States and the Soviet Union tested thermonuclear devices within nine months of each another in 1953; it's "earliest" was 11:43 a.m., after the U.S. and the Soviets signed the Strategic Arms Reduction Treaty in 1991.

Midnight signals, in the Bulletin's formulation -- with what might be the most gratuitous use of an adjective ever -- our "catastrophic destruction." That used to mean our annihilation by global nuclear war. But in recent years the Bulletin has expanded its definition of "doom" to encompass "climate-changing technologies and new developments in the life sciences that could inflict irrevocable harm." Or rather, since humanity is presumably risking these different kinds of self-destruction on different timelines, the Bulletin hasn't expanded its definition so much as multiplied its definitions of "doom" -- to categories beyond the traditional competencies of atomic scientists -- while somehow still representing the world's proximity to an overall state of destruction with a single indicator. The Bulletin has specified little over the decades about its process for deciding to move the minute hand -- and less about its methodology.

Our Iran War Clock was intended as an evocative homage to what we saw as a cool symbol of Cold War-era humanistic vigilance, reworked in ways that would disperse the bigger question marks hovering around the original Doomsday Clock. Our version ranged from 11:40 p.m. to 12:00 a.m., with each minute representing a 5 percent difference in the odds of either the U.S. or Israel attacking Iranian nuclear facilities sometime in the next year, as estimated in aggregate by our panel of 22 academics, journalists, and foreign-policy analysts with specialist knowledge of the region and issue. "Midnight" would theoretically equal certain war.

As we wanted to emphasize, the Iran War Clock was not intended to -- nor could it possibly -- represent a scientific or objective analysis. It was rather intended to serve as a comparable measure of a highly informed, ideologically diverse panel of subject-matter experts' "collective gut check," taken at more-or-less monthly intervals. As our Dominic Tierney explains:

Betting markets like Intrade.com rely on the wisdom of crowds. When many people wager real money the market can prove effective at predicting political outcomes. Another way to gauge the odds of war is to rely on an expert's opinion. [Our approach] combines elements of both models through the wisdom of a crowd of experts. On the one hand, the panelists are highly knowledgeable. On the other hand, there are sufficient members of the panel that any individual instance of bias or error should not have an overly negative effect on the aggregate prediction.

It was certainly not intended to promote military strikes on Iran -- any more than the Doomsday Clock was intended to promote doom. Neither was it intended to suggest the inevitability of war. After all, the Doomsday Clock itself has moved closer to and farther from midnight repeatedly over the past 65 years -- according to the Bulletin of the Atomic Scientists, in fact, we happen to be more than twice as far from "doom" in 2012 as we were in 1953 -- and the whole premise of our own project is precisely that neither war nor peace is inevitable with Iran. That's why we're doing this.

But it's also why, on reflection, the Doomsday Clock is a bad metaphor -- even if you straightforwardly define what the "time" means and are transparent about how you set it. Clocks don't tick both ways; and if they do, you should probably ignore them.

So we've reframed our project: This week, Tierney has posted the second installment in the series, which we're calling the "Iran War Dial." We lose the direct historical allusion, alas (historical allusion being the sort of thing we have an occasional weakness for here at The Atlantic). But we gain the clarity of an indicator that renders our panel's collective assessment of the odds of conflict with Iran in the same terms that the panel determines it: percentage. Of course we lose the Doomsday Clock's implicit ticking-hand drama, too, but we're only happy with that. This really isn't a subject that needs artificial drama.

The panel currently has the odds of conflict at 42 percent, down six percentage points from last month. Read the full story here.