What we’re witnessing here is one of the most famously difficult maneuvers in combat, the attempt to carry out an organized withdrawal:

Armies are built to fight, not retreat, and success is built on and breeds confidence which is hard to sustain while in flight:

A withdrawal is a type of military operation, generally meaning retreating forces back while maintaining contact with the enemy. A withdrawal may be undertaken as part of a general retreat, to consolidate forces, to occupy ground that is more easily defended, or to lead the enemy into an ambush. It is considered a relatively risky operation, requiring discipline to keep from turning into a disorganized rout or at the very least doing severe damage to the military’s morale.

Sometimes, ordering a retreat can be a brilliant strategy. If you can find more advantageous ground to defend or buy time to bring up reinforcements then you may avoid disaster and later turn it into victory. In some cases, it’s possible to create overconfidence in the enemy and lure them into a trap. But it’s extremely hard to keep good order and morale when giving up ground to the enemy, and quite often retreats suffer from lack of discipline or even panic and desertion.

Trump has said “no collusion” dozens and dozens of times, but now he argues that there was collusion on both sides.  There may have been no point in trying to hold this ground which could not be defended in any case , but having ceded it to the enemy Trump has no prospect of ever regaining it.  He colluded (and lied about it), and now the question is whether or not this matters.

Trump, like any president, has many weapons. And he has a good infrastructure for sending out his talking points and orders. But his greatest asset is that he is presumptively immune from prosecution for so long as he remains in office, and that means that he can make extremely bad or even suicidal legal arguments without worry. If he has good reason to believe that these arguments will help him fend off impeachment, then they make strategic sense.

This retreat is highly unlikely to work as an ambush, but it does give him more defensible ground. He will now dig trenches on the “Hillary did it, too, and it was even worse” hill.  He’ll set up artillery posts on the adjacent “anyone else in my position would have done the same” hill.

These are last refuges of a scoundrel, no doubt, by they are still formidable obstacles.

The way to mount these obstacles is not to attack them frontally with factual or legal refutations, although a feint in that direction will have to be sustained. The way to take Trump out now is to outflank and encircle him with political and moral arguments about his fitness as a leader and his prospects for success. His defenders must become convinced that their position is hopeless and that their general has no prospect of extracting them.

This is one reason why it’s a particularly bad idea to demonstrate doubt, fatalism, or defeatism about the prospects for a successful rout.  It signals to his defenders that they are on safe and defensible high ground.  Battles like this sometimes turn on small things, including demonstrations of confidence.

Trump’s new lies might work if they are given time to metastasize, which is why he must be pursued vigorously as he retreats.  The “no collusion” line has been overrun.  Whether this turns into a decisive rout remains to be seen.