To paraphrase a former secretary of defense: There are known knowns, there are known unknowns, and there are unknown unknowns.

In policy, as in military strategy, the first two epistemological categories are acceptable: People either know exactly how a policy will work, or they can make educated guesses based on data parameters they can’t quite know for certain. The unknown unknowns—what we don’t know we don’t know—are the problems, the things that could derail an entire policy and, in the process, ruin lives. Traditionally, the goal in lawmaking has been to eliminate the mystery from legislation so that there are as few unknown unknowns as possible. But, as it turns out, tradition can be easily broken.

Much has been made of the unorthodox and secretive processes by which previous Obamacare-repeal attempts in the House and Senate unfolded. The last effort in the Senate, the Better Care Reconciliation Act, was marked by procedural issues; Byrd Rule hang-ups; text flash-written in secret lunches that even the bill’s most vocal supporters hadn’t seen; and by Arizona Senator John McCain’s now-famous plea to “return to regular order.”

But while the failure of those bills may suggest to some a flaw in legislators’ approach, to Republican leadership the low-information shock tactics on display have proved valuable in the new health-policy landscape. The latest repeal attempt, led by Senators Lindsey Graham and Bill Cassidy, is a perfection of the form—their law could very well pass next week with even its sponsors in the dark about its effects.

As of this writing, just about the only concrete information Americans have about the Graham-Cassidy proposal is the text of the bill itself. And legislators appear not to have grasped much beyond the basic structure of the law—which would replace the Affordable Care Act’s Medicaid expansion and premium tax credits with a block grant to states, and would establish per capita caps for base Medicaid. On Tuesday, while being interviewed by Vox about the bill, nine GOP senators barely, if at all, talked specifics. The lawmakers could point out the broad strokes of block grants and Obamacare repeal, but didn’t go much further than that. Their sentiments could be largely summed up by Texas Senator Ted Cruz, who said that “the heart of the legislation takes the policymaking role of Washington and sends it to the states.”

But the mechanisms of how that would work—and the economic and coverage effects those mechanisms would have—are still very much unknown. While the few independent analyses that exist (from Avalere, Manatt, and the Kaiser Family Foundation) agree on a few matters—that overall federal funding for coverage would be slashed, that the bill would redistribute funds from Medicaid-expansion states to non-expansion states, and that it would constrict Medicaid funding over time—there’s still a divergence on specific estimates.

That’s because the formula Graham-Cassidy would use to allocate Medicaid expansion and exchange funding to states is complex, and not completely outlined in the text itself. It would involve different inputs from year to year about state choices, cost-sharing-reduction calculations, inflationary factors, total federal expenditures, population-risk adjustment, the actuarial value of existing programs, and a bevy of other demographic factors and year-to-year economic factors. Additionally, the biggest unknown is what happens after 2026, when the block grant ends. Will states have access to funding at all? If the funding does suddenly end, are there guardrails to keep state-health infrastructures from collapsing? What pieces of the current law would apply or not apply in that scenario?

With all these question marks, it’s impossible to come up with a firm number of people Graham-Cassidy would cover, which is the key output for any health policy. Given the funding decrease and the already-modeled effects of ending the individual mandate and capping base Medicaid, it seems a safe bet to assume the bill will increase the number of uninsured by north of 12 million people in the next decade above current projections. And given that, if the block grants are not renewed after 2026, the outcome would be identical to straight Obamacare repeal—worst-case models assume an upper bound of 32 million or more additional uninsured people.

It seems uncertainty is the desired legislative state for Senate Majority Leader Mitch McConnell and the GOP. When faced with news that, with just a week for turnaround, the Congressional Budget Office’s methodical analyses—which would involve modeling and gathering as much information as possible on unknowns and assumptions—would only be able to provide a rough assessment of Graham-Cassidy, Cruz echoed longtime Republican criticism of the CBO. The senator called the independent scoring agency “ridiculously slow, unreliable, and based on policy assumptions that are demonstrably false.” But unlike during the last round of repeal with the BCRA—when Republicans relied on a quickly spun-up and fuzzy score from the Department of Health and Human Services to justify not waiting for the CBO—now GOP leadership seems content with having no numbers at all.

Normally, the legislative process between a bill’s drafting and passage would take, at a minimum, weeks—with extensive research from the private and public sectors, multiple rounds of hearings, roundtables, formal and informal stakeholder meetings, floor debate, a CBO score, and multiple rounds of revision in response to feedback from each of those steps. There will be almost none of that this go-around, with 72 total hours of debate next week to meet a September 30 reconciliation deadline.

McConnell’s window is even more abridged than during his previous repeal tries. If brought to the floor at the latest possible moment, Graham-Cassidy would have involved a staggering 13-day period between its public unveiling and the final vote. For comparison: The BCRA was revealed in mid-June and died in late July, a span that, at over a month long, seems like a saga next to Graham-Cassidy.

What the GOP learned in that attempt was that, despite McCain’s pleas, eschewing regular order is a pretty good strategy for minimizing opposition. The kind of grassroots mobilization that was so effective in fighting Obamacare repeal over the summer takes time and money, and now has to fight against fatigue. Messaging against the bill requires analysis first, and then media digestion of that analysis. And citizens can’t express their displeasure about the legislation in polls if polls can’t collect their samples before the vote.

The September 30 deadline is an effective tool for amplifying the fog of war, maximizing confusion, and moving toward passage. It’s also somewhat arbitrary: Republicans could pass legislation to continue reconciliation into next year.

Will all those things lead to good policy that the majority of Americans want and would find beneficial? That seems to be tangential at this stage. As Iowa Senator Chuck Grassley said on Wednesday: “I could maybe give you 10 reasons why this bill shouldn’t be considered. … But Republicans campaigned on this so often that you have a responsibility to carry out what you said in the campaign. That’s pretty much as much of a reason as the substance of the bill.” Whether the tactics of confusion result in that promise’s fulfillment or not, it seems unlikely that they won’t be used again.