Canvassing is a numbers game. Whilst standing at unanswered doors yesterday evening, I was multiplying fractions.
About one in ten people are in, or answer the door when knocked.
Only half of the people of that inner city ward register to vote.
Only one in five actually votes in local elections.
Of those that vote in that area, half voted Lib Dem at the last election.
That’s 1/10 * 1/2 * 1/5 * 1/2 = 1/200 chance that any given door I knock on will have someone behind it who will tell me they will vote Lib Dem.
I have to knock on 200 doors to find someone who will vote for us. And if I do, that means we are winning.
So why do we it? Because one in ten of those one in two hundred people might join the Lib Dems if asked. And one in ten of those might help out and deliver leaflets. And who knows, one in ten of those might want to become a Lib Dem councillor.
That gets us to 1/200,000.
There are only 280,000 people in Nottingham. We’re clearly screwed. Or I can’t do maths.
Niles's Blog 
Well, if you insist on only counting the two-leggers…
Your problem lies in the factors used below:
“About one in ten people are in, or answer the door when knocked.
Only half of the people of that inner city ward register to vote.
Only one in five actually votes in local elections.”
Whilst these factors are relevant for the purpose of finding Lib Dems when out canvassing, they’re not relevant for the purpose of finding members, activists or candidates.
1. People who are out when you canvass may still vote, join etc the Liberal Democrats
2. It might be less likely, but people who haven’t registered to vote might still join and become active in a party, especially in a General Election
3. Ditto, only 1-in-5 may vote in local elections, but this doesn’t prevent them from joining parties etc. Indeed, they may only become active voters after joining a political party around the time of a General Election.