Did Nate Silver cave to the liberal attacks on his model?

Previously he had been way less bullish on Harris than other prognosticators and he caught heat from the left

On the 26th of October, Nate silver would have had Harris leading by 2.7 based on his change from last week. Meanwhile, that is the same number (2.7) that 538 had on that same day. So while 538 showed a 0.1 gain for Harris over the past week, Silver showed a 0.8% gain. That is a difference between 538's model showing almost no statistical movement, versus Nate Silver showing almost a 30% increase in Harris's lead over the same time period, entering the exact same data into the two models. If you were to go back further in the modelling, it was a pretty consistent theme that 538 had a higher Harris lead than Silver's model was showing. Occasionally they might come pretty close to matching, but Silver's model never favored Harris over the 538 model.

There had been much discussion (argument) over which side had changed models. Silver demanding that 538 had a new updated model that was showing things better for Harris, while he was sticking to the original model which was obviously better for Trump. For this Silver was catching heat and getting quite defensive about it on X and other places. He was suddenly an outlier and apparently some sort of traitor to the cause for showing Trump with a better chance to win than Harris.

But now something has changed. Not only has the Silver model moved in the direction of the 538 model, but somehow it has overtaken it and pushed Harris to an unexplainable 3.5 lead. What makes this so odd is that obviously both models are entering the same data, and quite frankly none of the polling is really moving at all. In the past week or so we have seen Activote, YouGov, Emerson, Leger, Redfield & Wilton, Morning Consult, Outward Intelligence and others. None of them have shown any movement towards Harris. Most have been flat or even showed Trump gaining a point. Meanwhile, the battleground state polling (according to RCP) has moved in Trump's direction by also a small amount. So the question comes... where does Silver see the nearly one point increase in support for Harris?

Apparently he must have crashed the Hill/DDHQ site where they can also add a poll that has moved in Trump's direction, but shows their overall average as increasing for Harris. The only thing that makes sense (again) is some form of cumulative effect being given to pollsters who poll more frequently, rather than looking at replacing the previous poll. As suggested, these busy pollsters (YouGov, Morning Consult) have been traditionally favorable and to the left and not necessarily high rated pollsters. So giving them more weight for doing more polling might explain this, but certainly does not defend it logically.

At the end of the day it certainly looks like Silver has adjusted his model to make things better for Harris. I know he subjectively changed his influence on a couple of pollsters that have been favorable to Trump, but that still doesn't explain the big movement in his model.