# Are Dems losing people to the "Walk Away" movement?

**URL:** <https://boards.straightdope.com/t/are-dems-losing-people-to-the-walk-away-movement/820354>\
**Category:** Politics & Elections\
**Created:** [August 27, 2018, 1:44pm UTC](https://boards.straightdope.com/t/are-dems-losing-people-to-the-walk-away-movement/820354 "2018-08-27T13:44:39Z")\
**Posts on this page:** 2\
**Page:** 3

<div class="post-metadata">

**Author:** ![sfriver](https://avatars.discourse-cdn.com/v4/letter/s/dbc845/32.png) [@sfriver](https://boards.straightdope.com/u/sfriver)\
**Post date:** [October 9, 2018, 2:43am UTC](https://boards.straightdope.com/t/are-dems-losing-people-to-the-walk-away-movement/820354/41 "2018-10-09T02:43:21Z")

</div>

[https://securingdemocracy.gmfus.org/the-methodology-of-the-hamilton-68-dashboard/](https://securingdemocracy.gmfus.org/the-methodology-of-the-hamilton-68-dashboard/)

> [@](#):
>
> We expect that we would have faced criticism for identifying the 600 accounts, and we expected (correctly) that we would face criticism for not identifying the accounts. We choose not to identify the accounts for the following reasons:
> 
> As noted above, our metrics are very accurate, but not 100 percent accurate. We believe based on the performance of the metrics described above and the subsequent manual review, that the monitoring list is at least 98 percent accurate, but that leaves as many as 12 accounts that may not be relevant. We are not willing to publicly attribute even one specific account incorrectly.  
> A 98 percent accuracy rate is, however, more than adequate to perform analysis on the aggregate set.  
> If we identified the users in the dataset, they would certainly change their behavior and render the dashboard essentially useless.

Instead of outing Russians and halting their propaganda, it’s best to let the Russians operate so they can statistically report the damage they do?

> [@](#):
>
> We cannot at this time offer a breakdown of the relative amount of content from each of these categories, although we may pursue this question in future analysis. Nevertheless, three important caveats contained in the description above are worth emphasizing:

This is in reference to how they focus on some accounts are more than others, essentially a weighted scale. This is why you have to dismiss the entire data set. This is not how statistics work. These are FAKE statistics.

---

<div class="post-metadata">

**Author:** ![adaher](https://avatars.discourse-cdn.com/v4/letter/a/dec6dc/32.png) [@adaher](https://boards.straightdope.com/u/adaher)\
**Post date:** [October 15, 2018, 6:44am UTC](https://boards.straightdope.com/t/are-dems-losing-people-to-the-walk-away-movement/820354/42 "2018-10-15T06:44:01Z")

</div>

It’s real and not real. You know how you can always point to some company doing layoffs no matter how good the economy is and cite it as evidence that people are hurting? Well sure, because 300 million jobs are created and 300 million destroyed every year. Political allegiances have similar churn. I don’t have exact numbers, but it wouldn’t surprise me at all if a million Democrats have turned Republican since 2017 and a million Republicans turned Democrat.

[Previous page](https://boards.straightdope.com/t/are-dems-losing-people-to-the-walk-away-movement/820354.md?page=2)