Valid visualisation of partisanship in US Congress?

It’s real preddy - as visualisations often are - and a lot would agree with the conclusion, but does it stand up to scrutiny:

Is that a valid representation of the the past 24 years?

Since Lucioni uses publicly available voting data, and pretty clearly explains his methodology, why wouldn’t it be valid? What do you mean by “valid,” anyway?

OP’s linked-to graph has one big advantage over a graph that I linked to recently. The latter graph depicts dissimilarity between parties’ voting and might arise if, for example, there was left-vs-right polarization all along which was masked in party-based graphs because the “left” was split between the parties and so was the “right.”

Lucioni’s method demonstrates that that is not the case; that voting now shows stark bipolarization where once there was shifting issue-based voting.

More elections than Great Debates, really.

I give up.

There is reason for skepticism. network plots like that have a large number of tuning parameters and depending on how those are set can give very different visualizations. I usually view such plots with a cup or two of salt.

This says nothing about the correctness of his conclusions. I think its clear that congress has become more polarized in the last few years, but there are clearer ways to demonstrate that. They just don’t look as fancy, and so don’t generate the coolness factor that the network graphs do.

The video says “The placement of the points is determined by an algorithm”, but doesn’t say anything at all about what that algorithm is. That’s the essential point we need to know, here, to say how valid it is.

:confused: Did you click again? The only oddiity I noticed was

If earlier sessions had more “procedural votes” that might invalidate the comparisons. And, even without that fudge the portion of partisan votes will vary due to procedural or political changes. But even if study of the data reveals that procedures have become more partisan, that would confirm the over-all point.

For me, the chart just confirms what readers of objective blogs and magazines already knew.

No, I assumed that the link in the OP was to the thing we were supposed to be commenting on. I didn’t even see the nested link to the actual content.

I see this again and again here at SDMB. We don’t care about ideas, but need to rate each others’ exposition and debating skills.

Maybe I don’t fit in here.