Although the prime minister wants to hold another "significant vote" on his Brexit plan next week, it is not at all certain that, when it comes to the crisis, he will choose to do so.
His closest colleagues told me that two conditions must be met to get them to proceed with the vote, probably Tuesday.
First, the DUP of Northern Ireland must say that in the end he changed his mind and decided to vote with her.
To be clear, there is no logical reason why they should do so, since there will be no alteration of the last minute to what they hate most about his agreement, namely the backstop that was designed to keep the border open on the island of Ireland and is enshrined in the revocation agreement.
Thus, the theoretical risk that the backstop will be forever, and will drive a regulatory wedge between Northern Ireland and Great Britain – which the DUP sees as toxic – is still implicit in the May plan.
But DUP deputies and politicians appear increasingly embarrassed by their central role in derailing Brexit, and ministers are therefore confident that they will only eventually cease their opposition if – for example – they play a central role in the negotiation of any agreement that eventually it will reduce backstop or short duration.
And when I ask the members of the government what else could appease the main Northern Ireland union party, they send me an emoji text message of dollar bills.
That said, the DUP tells me that they will not be correct, or at least not this time. And that they will vote for the agreement only if persuaded there is no risk that the Brexit agreement will break the union between Northern Ireland and Great Britain.
Honestly, I don't see how they will be able to save the PM, because – as I said from the beginning – nothing fundamental has changed or will change to support such a turnaround.
But since we live in such strange times, I am reluctant to bet my life savings on 10 parliamentarians of any persuasion by sticking to those who claim to be inviolable principles.
So, let's do the heroic – and probably wrong – hypothesis that the PM somehow talks about the DUP round. What then?
Well, she would still have only this third significant vote if the clear majority of the Brexiteer rebels in her own party, those belonging to the European research group, have a belated change of heart and even her agreement.
Again, there is absolutely no principle ground for any of them, since the agreement has not substantially changed. But some are conspicuously on the brink of capitulation, in the fear that they will not, Brexit would suffer significant delays (and a number will show solidarity with the DUP and will do everything the DUP decides).
So the second – and obvious – condition that should be met in May to get a third blow to the ratification of his agreement is that his whips, led by Julian Smith, would need to be sure that they can significantly reduce the margin of defeat or (less plausibly) actually wins.
The reason why reducing the loss to – say – 10 or 20 votes would be so valuable to her is that at that point she will feel encouraged to hold a fourth significant vote (because you guessed it!) Because she might be able to persuade maybe six or even 12 additional Labor MPs who want Brexit to support his agreement – meaning that in the end there would be a point for them to put up with all the hatred that would have been accumulated about them by their colleagues to challenge their leader's orders, Corbyn.
As one minister told me, "it's no reason for Labor MPs to get votes from members to vote with us unless they think we can actually win the vote."
At this point I imagine that you will be in agreement with me that it is very far from the fact that the Prime Minister secures his agreement next week, due to the widespread expectation that he is about to snatch victory from the jaws of late Brexit.
So let's briefly evaluate two other elements of Downing Street planning.
First, if he loses, he will tell the leaders of the EU at the Council on Thursday that the next two weeks will be devoted to probing parliamentarians' opinions – through so-called "indicative votes" – on what kind of Brexit they could support.
To be clear, the Prime Minister will try to reassure the rest of the EU that this does not imply the opening of the revocation agreement or an additional push to change the security lock. It would simply be a narrowing of the range options for the future relationship, or the purpose of the so-called Political Statement.
And both the leaders of May and those of the EU would assume that if the MPs could then merge on any reworked Brexit it would be a much softer Brexit – probably involving direct or approximate members of the customs union and the single market – and hence a Brexit that is even less appealing to Brexiter purists (those who would expect it to be hoarded).
This of course would be music for the ears of perhaps a quarter of the Cabinet, led by Rudd, Hammond, Clark, Gauke and Mundell.
But that would mean a long delay to Brexit, beyond 1 July – and therefore would require the participation of the United Kingdom in the May elections to the European Parliament.
Apparently, several conservative ministers and parliamentarians are less afraid of this than I had expected. "It's an inconvenience, not a disaster," says one.
So, if these scenarios are accurate, what does it tell us about what kind of Brexit – if any – will we have?
Well, this is still an impossible prediction.
Here are three reasons why.
First, if May gets its agreement approved next week, the Brexit ERG could still secure their beloved Brexit no-deal at the end of May, talking about the withdrawal agreement and the bill that would transform the vote significant in the reality of Brexit.
Second, if your agreement is stifled, the process of turning indicative votes into a viable Brexit agreement would be effective only if Labourism cooperated in an institutional sense, and this could not be guaranteed.
Thirdly, everything I have written is superfluous if the President decides that the Prime Minister's desire to keep the significant vote a third time is a clear and unacceptable violation of parliamentary conventions – and therefore prohibits it.
I said earlier and I repeat: the first Brexit law is "anything can happen".