Any survey concerning a group of 16,692 where you only sample 60 of that group will yield results that will vary from the next sample of 60 by +/-12.63%.
You wouldn't trust a survey that inaccurate to accurately say which President candidate would win - normally a 5% margin of error is considered way too much.
And you want to use this to justify killing people?
This discussion involves different paradigms.
My paradigm is 'if killing people is wrong, then killing people is wrong'.
It means there is nothing you can say to convince me 'judicial' killing is just.
At the same time I would cheerfully kill anyone who harmed or tried to harm me or mine. Self defence is fine, and I wouldn't try and pretend vengeance was justice.
Now just as I come from a paradigm resistant to persuasion, I know many pro-'judicial' killing people come from equally resistant paradigms.
Only in discussions where the facts do not matter, but the principles do, will you see such poor 'evidence' entered to justify one opinion.
Some people might genuinely make cold clinical assessment of the situation. More think they do.
But at the end of the day many think not killing a terrible murderer is just wrong, and many others will think killing a murderer is terribly wrong.
Oh... when West Germany stopped executions after WWII, the politicians went against the opinion of the public. Now Germany has one of the lowest percentages of people supporting the deah penalty.
Maybe you feel not killing murderers is just wrong, because you grew up in a society where the State killed people. It was normal.
Just like those Germans did.