General Election 2019

I am not knowledgeable about such things, but I worry about:

I run one of the sites with a group of others, it's this one https://tacticalvote.co.uk/ which is the most popular and largest of the sites.

The people who run it come from a wide range of political biases (except Tory) and all feel unrepresented by the current voting system and govt.

So I hope you won't mind my going through your concerns:

It only works based on predictive information.

Yes, but in our case not data driven by machines... humans who look at the history of a seat, factor in local elements, Remain/Leave votes, prior elections, whether it's changed, etc.

But everyone is saying the predictions and polls are not reliable.

Predictions and polls are unreliable, but knowing where to tactically vote and which way to vote isn't unreliable because the recommendations are pretty reliable and unified. We know where tactical voting can make a difference and if it's done it will make a difference.

Different websites give different advice

Not so much: https://tacticalvote.co.uk/key-seats/

Virtually every website is aligned, as no tactical voting site wishes to split the vote and risk a bad outcome.

It kind of freezes things in the past since that's what a lot of the prediction is based on.

This is true, if people are always tactical voting then recommendations in the future echo off of that. But why to tactical vote is interesting and I personally do not wish this to become an election echo effect... I want tactical voting to show the need for progressive agreements (where candidates work together and drop out to better represent local votes) and then that be a step towards proportional representation.

Voting doesn't reflect actual preferences. Doesn't that fuck up the whole idea of democractic elections?

No, it makes it more reflective of voter preference:

Some areas go Tory when the majority in the area vote left/liberal/green.

But can tactical voting based on poor information actually screw things up? It's apparent popularity makes me nervous.

In the 2017 GE tactical voting did well: https://tacticalvote.co.uk/tactical2017/

At a glance:

• Seats where we could have affected the outcome: 100% accuracy
  
• Seats where we could not have affected the outcome: 89% accuracy
  
• All seats:
  
  • 565 correct
    
  • 69 incorrect (mostly safe Tory seats where tactical voting cannot change the outcome)
    
  • 16 no recommendation