Let’s call that a conjecture about an investigation based on conjecture. A new poll out from CNN this morning shows confidence in Robert Mueller eroding to a virtual tie at 41/39, its lowest level in the series. This follows other polling that suggests voters are souring on the special counsel, if not the probe itself. But why? CNN thinks that Donald Trump’s strategy of attack has paid off:

Most Americans continue to believe that the Russian effort to influence the 2016 presidential election is a serious matter that should be investigated, but the constant criticism by President Donald Trump of special counsel Robert Mueller is taking its toll, according to a CNN poll conducted by SSRS released Friday.

The number of Americans who approve of how Mueller is handling the investigation has dropped from 48% in March to 44% in May to just 41% now, the lowest it has been in CNN’s polling. Mueller has a lot of company; no one connected with this matter is coming out of it in a positive light. Mueller’s favorable rating is just 32%; former FBI director James Comey’s favorability is just 28%; Rudy Giuliani, one of Trump’s lawyers in the Russia investigation, is viewed favorably by only 31% of Americans.

Well, maybe, but that gives an enormous amount of credit for influence to a man who gets hammered in the same poll on this very issue. There has been almost no change in respondents’ perception of how Trump has handled the Russia probe over the past ten months. The first CNN poll taken after Mueller got appointed special counsel came last August, when respondents had a 31/59 rating for Trump’s handling of the investigation. That has changed to, er, 29/55 — almost the same gap as ten months ago. The only significant change in the series on that question in the intervening time has been an increase of “no opinion” from 9% to 16% of respondents in the latest iteration, a new high. Trump’s best rating on this question came in November, and that was only a 34/55.

It therefore seems pretty unlikely that Trump’s the cause of Mueller’s erosion. An even better measure of Trump’s influence comes in Question 20:

In the same time period as above, the gap on the question of the import of this matter has changed from 22 points to 20. That’s within the margin of error, and the stability of this question strongly indicates that Trump’s PR campaign against the probe as a “witch hunt” hasn’t had much impact at all. Undecideds again have risen outside the margin of error, but not by a much more significant amount than that.

The previous question in particular reflects on Mueller and his performance. If support for this probe being a serious matter has not appreciably changed in the past ten months, why has the perception of Mueller’s performance changed significantly, and especially why does it trail support for the probe by double digits?

It’s much more likely that Americans have begun to run out of patience for the probe. Mueller’s had it for a year, and the only indictments he’s made public are for unrelated crimes committed by Paul Manafort and Rick Gates (which could have been prosecuted in 2014 by the Department of Justice, and probably should have), a few alleged Russian entities that will never come to trial, as well as a handful of process crimes by fringe campaign figures. If there was collusion to be found, Mueller should have more indictments on that point by this time. After all, we’re not talking about the Corleones here, although sometimes it did seem as if there were a surfeit of Fredos involved.

This could all change, of course, if Mueller produces indictments on the core theory of the investigation rather than on ancillary actions. His approval would likely spike upward on the sheer politics involved. But with this much time passing without any such actions, it’s getting easier and easier to believe that there isn’t actually much to find — and that the investigation has been permanently compromised by the political bias of the investigators who opened it in the first place.