Re:Oppose the Earn IT Act
by godelski on Wednesday July 01, 2020 @04:38PM
(#23706332)
> I file it under "perfect being the enemy of the good"
This is 100% how you should think about it. But what I'm trying to say is that we already have substantially better methods, so why not use them? Perfect is the enemy of good, but if I have the choice of eating acid, a bowl of oatmeal, and chocolate pudding, I'm going with the pudding. Perfect is the enemy of good when we don't already have the choices sitting right in front of us.
> [IRV] still represents an order of magnitude improvement over FPTP at capturing political preferences
And here's where I'm disagreeing, but only slightly. It is definitely better, but not an order of magnitude.
> Approval voting
I love approval voting. It is great because of its simplicity. That's why they use it in systems like Netflix. Easy to collect and gather results from. Honestly I use it all the time. It is how I solve the "where do you want to eat" problem with friends. You just list things until you get a unanimous vote or if you exhaust your list you just take the most approvals. Simple and easy.
The reason I don't like it for elections is that it makes tying easy and I do not believe it has enough precision. The interview I linked Arrow talks about this so I'll deffer to him (he's clearly smarter than me). In elections you need a clear winner.
Why I like score voting (and Condorcet methods, like RP, which is again RCV) is because I actually hate party systems, but I recognize that they can't be destroyed. But these methods drastically reduce their power (I don't think IRV does).
And I highly encourage you to dig more into the subject and am happy to provide links. I also highly suggest looking deep into everything that the Fair Vote chart discusses, mainly because I believe you will be as frustrated as me (and others, like Condorcet proponents) when you learn what each of those topics are. Everyone I know that is well read in the subject is frustrated by Fair Vote's mischaracterizations.
This is 100% how you should think about it. But what I'm trying to say is that we already have substantially better methods, so why not use them? Perfect is the enemy of good, but if I have the choice of eating acid, a bowl of oatmeal, and chocolate pudding, I'm going with the pudding. Perfect is the enemy of good when we don't already have the choices sitting right in front of us.
> [IRV] still represents an order of magnitude improvement over FPTP at capturing political preferences
And here's where I'm disagreeing, but only slightly. It is definitely better, but not an order of magnitude.
> Approval voting
I love approval voting. It is great because of its simplicity. That's why they use it in systems like Netflix. Easy to collect and gather results from. Honestly I use it all the time. It is how I solve the "where do you want to eat" problem with friends. You just list things until you get a unanimous vote or if you exhaust your list you just take the most approvals. Simple and easy.
The reason I don't like it for elections is that it makes tying easy and I do not believe it has enough precision. The interview I linked Arrow talks about this so I'll deffer to him (he's clearly smarter than me). In elections you need a clear winner.
Why I like score voting (and Condorcet methods, like RP, which is again RCV) is because I actually hate party systems, but I recognize that they can't be destroyed. But these methods drastically reduce their power (I don't think IRV does).
And I highly encourage you to dig more into the subject and am happy to provide links. I also highly suggest looking deep into everything that the Fair Vote chart discusses, mainly because I believe you will be as frustrated as me (and others, like Condorcet proponents) when you learn what each of those topics are. Everyone I know that is well read in the subject is frustrated by Fair Vote's mischaracterizations.