You are reading a single comment by @Dramatic_Hammer and its replies.
Click here to read the full conversation.
-
Oh I don't trust the methodology at all (particularly as they had it as ~75% of respondents voted leave), it was more the utter illiteracy (numerical in this particular case) of people commenting.
They don't understand numbers or stats, so things like facts, studies and science suddenly have as much weight as anecdote and personal feeling.
It's the same thing with people believing shiny websites over actual medicine.
Dramatic_Hammer
Most people will assume this means: 51.89-4+1 = 48.89% leave and 48.11+4-1 = 51.11% remain.
But if it's a percentage of the voters who have switched it remains a slim Leave majority:-
Leave: ( 51.89 * 0.96 ) + (48.11 * 0.01) = 50.2955
Remain: ( 48.11 * 0.99 ) + (51.89 * 0.04) = 49.7045
Although I suspect the figures are rounded to the nearest percentage and therefore all bets are off; it doesn't take much (4.4% and 0.8%) and the calculation swings to remain (just).
tl;dr meaningless stats are meaningless