4 ms·
Peter Norvig: Lieberman, Egg, Sausage and Lieberman
- jibiki 18y agoThe assumption that people's political views are distributed bimodally (which is what he's really saying with his "two superimposed normal distributions") strikes me as being a little bit arbitrary. Is this based on some sort of evidence? (Of course, the original study makes an equally arbitrary assumption that political affiliation is distributed uniformly...)
- imp 18y agoI think he just started with that assumption because that's what 538 did. His last analysis assumed one distribution, so my guess is he thought that might be more accurate. Although he didn't give data either way.
- Tangurena 18y agoI'd say it is an accurate observation of contemporary US politics. Groups describe themselves with "who we are" just as much as "who we are NOT." Lately the "who we are NOT" has been drowning out "who we are" to the point where the process inside political parties has been to reject potential candidates that are "too close to the enemy." That's why terms like RINO or DINO ($PARTY In Name Only) as epithets/attacks to denounce candidates (and elected officials) as too similar to the $ENEMY. To give one example, many Republicans (at the party level) would rather lose an election than "dilute their brand" by aiming for bipartisanship. disclaimer/weasel words: I ran for election last November. It was for a "non partisan" office, so I was able to dodge the political catfighting and backstabbery.