4 ms·

I'm not impressed. Did anyone notice that Cambridge Analytics did not work for Ted Cruz? I mean he did win Texas and the great plains states, but those were his

by mjfl 10y ago

I'm not impressed. Did anyone notice that Cambridge Analytics did not work for Ted Cruz? I mean he did win Texas and the great plains states, but those were his most-winnable locations from the beginning due to his consistent conservatism and other nominees dropping out. He even lost the south, which should have been his other target, presumably due to being off-wavelength with southern voters, something the analytics data should have helped him with. Cambridge Analytics and their Facebook questionnaires don't seem to be very effective.

patcon 10y ago
I imagine we'd be equally unimpressed with other early tech. imho, our concern should be of an extrapolative nature. we can guess where this will be in 5 or 10 years, nevermind Ted Cruz' future electoral successes. and we know how easy people are to psychologically "nudge". (sorry for assuming our agreement there)
- rounce 10y ago
  > and we know how easy people are to psychologically "nudge" How easy? I thought "Nudge theory" was beginning to unravel under the burden of empirical evidence?
SnakePlissken 10y ago
Ted Cruz is probably more to blame in this case than Cambridge; he came across to many as supremely unlikeable/comically phony and clearly lost the mud slinging fights with Trump that dominated the Republican primaries. Whatever benefit Cambridge bestowed probably wasn't enough to turn Ted Cruz into a serious Presidential candidate. In more colloquial terms, perhaps even the most effective data targeting can't polish a turd.
forgetsusername 10y ago
1. There are many degrees between "didn't work at all" and "worked so well he became predident". 2. Cruz was an establishment candidate. This electorate wanted an outsider.
- bryanrasmussen 10y ago
  well, some geographically significant proportion of the electorate wanted an outsider.
drinkjuice 10y ago
> Cambridge Analytics and their Facebook questionnaires don't seem to be very effective. So, nothing speaks against nipping that in the bud, right? It's like a tiny person with no tools sneaking around your house to murder your whole family; sure you could ignore them, at least at the moment, but why would you?
narrowrail 10y ago
I'm late to this party, but I'd point to the fact that many of the states Cruz lost in the south have open primaries: https://en.wikipedia.org/wiki/Open_primaries_in_the_United_States#States_with_an_open_presidential_primary https://en.wikipedia.org/wiki/Open_primaries_in_the_United_S... Actually, the following states were lost by Cruz and have open primaries: Alabama, Arkansas, Georgia, Illinois, Indiana, Mississippi, Missouri, North Carolina (All races' primaries open for unaffiliated voters only), South Carolina

by mjfl 10y ago

patcon 10y ago

I imagine we'd be equally unimpressed with other early tech. imho, our concern should be of an extrapolative nature. we can guess where this will be in 5 or 10 years, nevermind Ted Cruz' future electoral successes. and we know how easy people are to psychologically "nudge". (sorry for assuming our agreement there)

rounce 10y ago
> and we know how easy people are to psychologically "nudge" How easy? I thought "Nudge theory" was beginning to unravel under the burden of empirical evidence?

SnakePlissken 10y ago

Ted Cruz is probably more to blame in this case than Cambridge; he came across to many as supremely unlikeable/comically phony and clearly lost the mud slinging fights with Trump that dominated the Republican primaries. Whatever benefit Cambridge bestowed probably wasn't enough to turn Ted Cruz into a serious Presidential candidate. In more colloquial terms, perhaps even the most effective data targeting can't polish a turd.

forgetsusername 10y ago

1. There are many degrees between "didn't work at all" and "worked so well he became predident". 2. Cruz was an establishment candidate. This electorate wanted an outsider.

bryanrasmussen 10y ago
well, some geographically significant proportion of the electorate wanted an outsider.

drinkjuice 10y ago

> Cambridge Analytics and their Facebook questionnaires don't seem to be very effective. So, nothing speaks against nipping that in the bud, right? It's like a tiny person with no tools sneaking around your house to murder your whole family; sure you could ignore them, at least at the moment, but why would you?

narrowrail 10y ago

I'm late to this party, but I'd point to the fact that many of the states Cruz lost in the south have open primaries: https://en.wikipedia.org/wiki/Open_primaries_in_the_United_States#States_with_an_open_presidential_primary https://en.wikipedia.org/wiki/Open_primaries_in_the_United_S... Actually, the following states were lost by Cruz and have open primaries: Alabama, Arkansas, Georgia, Illinois, Indiana, Mississippi, Missouri, North Carolina (All races' primaries open for unaffiliated voters only), South Carolina