Are Atlas' election predictions biased?

[DC Al Fine]

Like posters consistently overestimating how social democratic parties will do. That sort of thing.

[Xing]

Pretty much any prediction is going to be biased, unless it’s based entirely on a calculation of some kind, like the polling average.

[DPKdebator]

Pretty much any prediction is going to be biased, unless it’s based entirely on a calculation of some kind, like the polling average.

[Torie]

My problem is more that so many predictions seem too certain to me, without a sufficient error factor, particularly at this still rather early stage. Part of it is the over-reliance on past voting/polling behavior, which in my view is statistically invalid at the margins, because there are too few past data points, bearing in mind that in each election the circumstances are somewhat different, so each data point is a somewhat different fruit than the other ones, which muddies things statistically. In this election, I see a quite large range of reasonably possible outcomes.

[SWE]

As opposed to all the unbiased ways to do predictions?

[Alabama_Indy10]

Obviously

[Free Bird]

The real question should be "are the predictions of a site dominated by Democratic partisans possibly biased?" Tough question.

[DC Al Fine]

Pretty much any prediction is going to be biased, unless it’s based entirely on a calculation of some kind, like the polling average.

As opposed to all the unbiased ways to do predictions?

Obviously

Ok guys I get it. I'm more interested in the how we are biased, not if we are biased.

For example I almost always underestimate establishment centre left parties in favour of the populist left and right.