The following weights each match by making better field matches much more important than lesser field matches:

use warnings;
use strict;

my $line = 0;
my @data = map {chomp; [split, ++$line, -1]} (<DATA>);

@data = sort {&doCmp} @data;
print join "\n", map {join " ", @$_} @data;


sub doCmp
{
calcWeight ($a) if -1 == $a->[-1];
calcWeight ($b) if -1 == $b->[-1];
return $a->[-1] <=> $b->[-1];
}

sub calcWeight
{
my $vector = \@{$_[0]};
my $sum = 0;
$sum += $vector->[$_] * $vector->[$_] for (0..@$vector-3);
$vector->[-1] = $sum;
}


__DATA__
0 1 3 3 1
0 2 2 2 2
1 1 2 2 2
1 1 1 2 3
0 0 3 3 3
[download]

Prints:

1 1 2 2 2 3 14
0 2 2 2 2 2 16
1 1 1 2 3 4 16
0 1 3 3 1 1 20
0 0 3 3 3 5 27
[download]

Note that the last two columns are the original line number and the calculated weighting respectively.