The past is a different country: Difference between revisions
Amwelladmin (talk | contribs) Created page with "{{a|stats|}}The past misleads us into believing the world is solvable because the past is fixed, complete, all knowable information is known, there can be nothing unexpected. ''We know what happened ''. When the data is in we can calculate, somehow the probability Lucy Letby there was a serial killer on the ward. Could we have predicted that in advance? All data is from the past. It can do a tidy job of helping us fill in the gaps about the past by triangulati..." Tags: Mobile edit Mobile web edit Advanced mobile edit |
Amwelladmin (talk | contribs) No edit summary Tags: Mobile edit Mobile web edit Advanced mobile edit |
||
| Line 1: | Line 1: | ||
{{a|stats|}}The past misleads us into believing the world is solvable because the past is fixed, complete, all knowable information is known, there can be nothing unexpected. '' | {{a|stats|}}The past misleads us into believing the world is solvable because the past is fixed, complete, all knowable information is known, there can be nothing unexpected. In the past ''we know what happened''. | ||
When the data is in we can calculate | When the data is in, we can calculate with ''some'' degree of confidence the probabilities of the various events that might have caused that cluster of collapses on the ward. [[Lucy Letby|there was a serial killer on the ward]]. But data won’t help us predict what will happen next, because all that closed, static finitude has gone away. | ||
All [[data is from the past]]. It can do a tidy job of helping us fill in the gaps about the past by triangulation: we can make sensible [[Bayesian inference]]s to work out what must have happened even if we did not see it. But here the sample space is closed, the boundaries are known — these are conditions for probabilistic reasoning. | All [[data is from the past]]. It can do a tidy job of helping us fill in the gaps about the past by triangulation: we can make sensible [[Bayesian inference]]s to work out what must have happened even if we did not see it. But here the sample space is closed, the boundaries are known — these are conditions for probabilistic reasoning. | ||
In the future, they are not. You cannot make expected value | In the future, they are not. You cannot make [[expected value]] calculations about an unknown | ||
sample space in the inchoate future of a [[complex system]]. We 'make'' the future when we are in a complex system. It doesn’t ''happen'' to us. | |||
This, in a sense, is the unfairness of | This, in a sense, is the unfairness of a referendum. We know about the status quo — it is from the past. We can sensibly judge it on its record. We know nothing about the alternative, which is in the future. We literally do not know what we are voting for. Full Brexit? EEA? Customs Union? Withdraw diplomatic missions? | ||
{{sa}} | {{sa}} | ||
{{Gb|[[The Bayesian]]<li>[[Brexit means Brexit]]<li>[[There is no data from the future]]<li>[[Complexity]]}} | {{Gb|[[The Bayesian]]<li>[[Brexit means Brexit]]<li>[[There is no data from the future]]<li>[[Complexity]]}} | ||
Latest revision as of 16:25, 26 September 2024
|
Lies, Damn Lies and Statistics
|
The past misleads us into believing the world is solvable because the past is fixed, complete, all knowable information is known, there can be nothing unexpected. In the past we know what happened.
When the data is in, we can calculate with some degree of confidence the probabilities of the various events that might have caused that cluster of collapses on the ward. there was a serial killer on the ward. But data won’t help us predict what will happen next, because all that closed, static finitude has gone away.
All data is from the past. It can do a tidy job of helping us fill in the gaps about the past by triangulation: we can make sensible Bayesian inferences to work out what must have happened even if we did not see it. But here the sample space is closed, the boundaries are known — these are conditions for probabilistic reasoning.
In the future, they are not. You cannot make expected value calculations about an unknown sample space in the inchoate future of a complex system. We 'make the future when we are in a complex system. It doesn’t happen to us.
This, in a sense, is the unfairness of a referendum. We know about the status quo — it is from the past. We can sensibly judge it on its record. We know nothing about the alternative, which is in the future. We literally do not know what we are voting for. Full Brexit? EEA? Customs Union? Withdraw diplomatic missions?