Carlos Jasso / Reuters

We are now entering the third week of the staring contest between House Speaker Nancy Pelosi and Senate Majority Leader Mitch McConnell over when the House of Representatives will deliver its articles of impeachment to the Senate, and under what circumstances.

It’s time for some game theory.

The standoff stems from McConnell’s proposal for how to proceed with the impeachment trial of President Donald Trump: Pelosi has accused McConnell of violating his oath of office, while Senate Minority Leader Chuck Schumer has suggested a trial procedure that would guarantee that at least four witnesses are called to testify. McConnell has balked at agreeing in advance to call witnesses. Instead, he has argued, the Senate should begin the proceedings without such an agreement and address the question of witnesses only after the House has made its opening argument against the president and Trump’s team has responded.

For this reason, the articles of impeachment remain stalled between the House and the Senate, with Pelosi refusing to deliver them to the other side of the Capitol until McConnell reaches some kind of accommodation with Schumer. There’s no reason, the Democratic leadership argues, to provide the articles to the Senate only for them to be quickly quashed. So Pelosi has withheld the articles until, as she puts it, McConnell agrees to proceed “in a manner worthy of the Constitution.”

Who will blink first?

This sort of strategic tangle is quite different from the debates that led up to the president’s impeachment, which were charged with a kind of moral and legal urgency. The questions then presented a challenge of immediate importance: What sort of behavior by a president so steps outside the bounds of what Americans should consider acceptable that the House should act to remove him from office? To this question, there was a clear right answer, and the House acted to establish its view that Trump had breached his oath of office.

The trouble is that, in the absence of any real doubt over the president’s impending acquittal by the Senate, the questions that remain are not moral. They are strategic questions about what the goal of a Senate trial should be—and tactical questions about how best to achieve it. For Democrats, the goal appears to be to use the trial to erode the president’s prospects for reelection in November. This goal includes presenting as much new evidence unavailable during the House proceeding as possible. For the Senate Republican leadership, by contrast, the goal is to make the trial go away as quickly as possible and to prevent it from metastasizing into a larger search for evidence of presidential misconduct.

In the tactical pursuit of such goals, the answers are far less clear-cut than they are when dealing with constitutional first principles. A lot of commentators are speaking about the current standoff as though it were not a negotiation between savvy legislators but some great moral crusade. But when and how the House delivers articles of impeachment to the Senate depends, for both Democrats and Republicans, on complicated cost-benefit calculations with many different inputs. And crucially, some of these inputs are not publicly known—and probably not known to McConnell, Schumer, and Pelosi, either. So the best way to understand the current contest is as a game of poker. Some of the cards the public can see; some it can’t. Some the players can see; some they can’t.

The big unseen cards are the votes of swing Republicans on the questions of early dismissal and calling witnesses. A motion to dismiss requires 51 senators, as does a motion to block a witness. That means that if the 53 Republican senators stick together, they can sustain a motion at any time to dismiss the trial. And they can also, even if they don’t stick together for a motion to dismiss, block the impeachment managers from calling their witnesses. Conversely, if three or four Republicans (the exact number depends on how a specific motion is presented) vote with Democrats, the impeachment managers could keep the trial from being dismissed and could call their witnesses.

McConnell has not ruled out witnesses altogether. Rather, his pitch is that senators should vote on the question of whether to hear testimony following opening arguments by the House impeachment managers and by the president’s defense—the way the Senate handled matters in President Bill Clinton’s impeachment trial. He hinted in remarks on the Senate floor on Friday that the Senate might take up a motion to dismiss the case against the president following those opening arguments, referencing a similar unsuccessful motion by Democrats during the Clinton impeachment.

This posture on McConnell’s part could reflect his confidence that he has the votes to dismiss the case. If he knows the votes are there to make this go away, he would not need to give Pelosi and Schumer anything, because he can make short work of a trial if Pelosi ever delivers the articles without an agreement.

But McConnell’s stance could also be a bluff. That is, by pretending to be confident of the votes of senators like Susan Collins, Lisa Murkowski, Mitt Romney, and Cory Gardner, McConnell forces Pelosi into a situation in which she either doesn’t get a trial at all (if she does not turn over the articles), or has to move forward not knowing whether the articles will be quickly dismissed—and hopes of a full presentation and elaboration of the evidence against the president will be just as quickly dashed.

The reasons Pelosi and Schumer would object to McConnell’s proposal are obvious. Without ensuring that they will hear from the witnesses most important to them, there’s a real risk that the trial might adjourn without the testimony of key witnesses such as former National Security Adviser John Bolton and Acting Chief of Staff Mick Mulvaney. Certainly McConnell, trumpeting his “total coordination” with the White House, has not exactly presented himself as an honest broker. So by not delivering the articles, Pelosi prevents quick dismissal, and thereby generates pressure for McConnell to reach an accommodation.

Or does she?

Here’s the problem: If Democrats do not have the votes to move forward with bringing in witnesses, they also in effect lack the votes to beat back an effort by McConnell to dismiss the trial or to move directly to the final verdict after opening arguments. In other words, if they don’t have the votes to beat McConnell’s procedural motions, they ultimately have no leverage over the trial, whether the articles show up today or a month from now. In this sense, Pelosi’s threat to withhold the articles may be an empty one.

Except for two important things.

First, if McConnell is bluffing—that is, if he does not, in fact, know whether he has the votes—Pelosi can call his bluff by delivering the articles. The worst case for McConnell, after all, is if the Democrats turn out to have enough votes to win repeated procedural motions and he has not negotiated a deal to limit the number of witnesses. Then the trial could end up hearing from all kinds of people with evidence relevant to the case against the president—Rudy Giuliani, for starters. If McConnell lacks the votes to limit witnesses, a trial without some kind of framework deal to limit the possible damage would be a little like the U.K. careening into a no-deal Brexit scenario. It would be very dangerous. Such a trial could also be protracted, which is highly undesirable for the president in an election year.

Second, holding on to the articles gives Pelosi the ability to trigger a trial—if only a brief one—at any time of her convenience, and Trump’s inconvenience, while he is running for reelection. Imagine, for a minute, that it’s March or April and the Democrats have effectively settled on a nominee who is not a sitting senator. That candidate is now visiting swing states in a triumphal bus tour, and Pelosi all of a sudden decides to “capitulate” to McConnell’s proposal and sends the articles over to the Senate. The country then has to watch a spasmodic and bitter few days or weeks of argument over the president’s conduct even as Democrats are unifying behind someone far away from the skirmishes. Now imagine Pelosi does this, say, the week before the Republican Convention—or after Labor Day. Remember also that such a trial will require vulnerable Republican senators running for reelection to cast tough votes.

These two problems, particularly in combination with one another, will likely force McConnell to the table if he is bluffing.

It is quite clear that Schumer and Pelosi do not know how many—if any—Republican Senate votes they can count on. After all, if they knew they had four, and they knew they could hold on to all the Democrats, they would have no reason not to agree to McConnell’s proposed course of action. There is nothing intrinsically wrong with having a motion to dismiss before calling witnesses. The problem with that approach, for Democratic purposes, arises only if they don’t know that they have the votes to defeat such a motion. So we can infer that Pelosi and Schumer either don’t know where they stand, or know they are playing a weak hand once the articles arrive in the Senate.

The key question, therefore, is what McConnell knows about his colleagues and their intentions. We will know soon enough. If he is not confident he can count to 51, it is in his interests to negotiate a deal that limits the scope and duration of the trial, and to do so quickly. If he is confident he has the votes, however, expect him to hold firm and insist that the only deal he will accept is one that allows what amounts to a final vote without hearing from witnesses.

That will leave Pelosi with no control of the trial—except as to its timing, over which she would retain effectively absolute control. This would give her the ability to force the Senate at any time to have arguments over the president’s conduct and then hold a vote that would give opponents of the president material to campaign with in 2020, against those senators who vote to scotch the proceedings without hearing from witnesses. It might also give her the ability to stretch things out by way of generating pressure on swing senators to support a compromise—something Democrats are contending that her withholding the articles has already done, with Republicans like Collins and Murkowski expressing discontent with McConnell’s approach. And now that Bolton has expressed his willingness to testify before the Senate if subpoenaed, the pressure on moderate Republicans to break with McConnell may be even higher.

So even if McConnell has the votes, Pelosi doesn’t have nothing. The card she holds, the card of timing, might be especially potent given that a majority of Americans—including a significant percentage of Republicans—think the Senate should consider new evidence during the trial. So Pelosi is actually holding out for a position with majority support.

In other words, if the standoff continues, it will be a sign both that McConnell is confident and that Pelosi is choosing not to play her remaining card now.

Conversely, if Pelosi swiftly turns over the articles without any kind of deal, that would suggest one of two contradictory possibilities: either that she overplayed her hand in the first place and is now cutting her losses, or that she has developed a certain degree of confidence that she will have the votes she needs—enough confidence, at least, to take her chances.

A deal, on the other hand, is a pretty sure sign that none of the negotiators have a strong instinct about what the swing senators really have planned and that both sides, consequently, are minimizing risk.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.