In the first example you can see two distinct pairs of parentheses. In the second example there is a third pair of parentheses around the whole regex which becomes $1 because it has the left-most opening parens.

examples/regex_capture.pl

use strict;
use warnings;
use 5.010;

my $text = 'Some text with a number: 23 and an answer: 42 and more';
if ($text =~ /(\w+):\s+(\d+)/) {
    say $1;  # number
    say $2;  # 23
}


if ($text =~ /((\w+):\s+(\d+))/) {
    say $1;  # number: 23
    say $2;  # number
    say $3;  # 23
}

Substitution

Capturing can also be used in the replacement part of a substitution:

examples/regex_substitute.pl

use strict;
use warnings;
use 5.010;

my $text = '1234567890';
$text =~ s/(\d)(\d)/$2$1/g;
say $text;  # 2143658709