"It's astounding how few experts' confidence intervals included the correct estimate " is just wrong. some of the time it's perfectly reasonable for ALL expert confidence intervals to fail to include the correct estimate
-
Show this thread
-
Replying to @VesselOfSpirit
May be so, but then we’re severely over-provisioned in experts, beyond surge capacity in # of topics needing opinions.
1 reply 0 retweets 1 like -
-
Replying to @VesselOfSpirit
If-then. Why have experts concerned with prediction if they can’t try to be uncorrelated about it?
1 reply 0 retweets 0 likes -
-
Replying to @VesselOfSpirit @othercriteria
Vessel Of Spirit Retweeted Vessel Of Spirit
Vessel Of Spirit added,
1 reply 0 retweets 0 likes -
Replying to @VesselOfSpirit
If we’re asking experts for their opinion on what the two trailing digits of C*VID deaths are on, say, 2021-05-01, then sure.
1 reply 0 retweets 0 likes -
Replying to @othercriteria
my point is simply that some events are objectively surprising given finite information
1 reply 0 retweets 1 like -
Replying to @VesselOfSpirit
I think your argument has a reliance on monotonocity or ordering or implications holding across experts’ models.
2 replies 0 retweets 0 likes -
Replying to @othercriteria @VesselOfSpirit
But this can be changed by going from questions on, say, # deaths to ones on log(# deaths), or vice versa.
1 reply 0 retweets 0 likes
another analogy: maybe 99% of stock market experts on 2020-01-01 would have given 66% confidence intervals for the dow jones index that don't include the current value. this isn't because they were vastly overconfident, but because something surprising happened
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.