I'd lay good odds that randis not actually messing up your strings. You are probably making a mistake in the declaration of your variables.

Let's walk through the process.

I added one line to your code so we could see the results:

$randomNumber = int(rand(100)) +1;
if ($randomNumber < 100 and $randomNumber > 50) {
    $answer = "kbNtA";    
} else {
    $answer = "WiaNq";
}
print "Answer = [$answer]\n";
[download]

It seems to be working fine:

S:\PerlMonks>perl scope0.pl
Answer = [kbNtA]

S:\PerlMonks>perl scope0.pl
Answer = [kbNtA]

S:\PerlMonks>perl scope0.pl
Answer = [kbNtA]

S:\PerlMonks>perl scope0.pl
Answer = [kbNtA]

S:\PerlMonks>perl scope0.pl
Answer = [WiaNq]
[download]

However, you're not running strictand warnings, which are critical to avoid mistakes and many bad programming techniques. So I cleaned up the code a bit:

#!/usr/bin/perl
use strict;
use warnings;

$randomNumber = int(rand(100)) +1;
if ($randomNumber < 100 and $randomNumber > 50) {
    $answer = "kbNtA";    
} else {
    $answer = "WiaNq";
}
print "Answer = [$answer]\n";
[download]

That doesn't work so well, and the issues raised could be the source of your problem. (It's hard to tell since you clearly did not show enough code and data to reproduce the problem.)

S:\PerlMonks>perl scope2.pl
Global symbol "$randomNumber" requires explicit package name at scope2
+.pl line 5.
Global symbol "$randomNumber" requires explicit package name at scope2
+.pl line 6.
Global symbol "$randomNumber" requires explicit package name at scope2
+.pl line 6.
Global symbol "$answer" requires explicit package name at scope2.pl li
+ne 7.
Global symbol "$answer" requires explicit package name at scope2.pl li
+ne 9.
Global symbol "$answer" requires explicit package name at scope2.pl li
+ne 11.
Execution of scope2.pl aborted due to compilation errors.

S:\PerlMonks>
[download]

So I cleaned up the errors:

#!/usr/bin/perl
use strict;
use warnings;

my $answer = '';

my $randomNumber = int(rand(100)) +1;
if ($randomNumber < 100 and $randomNumber > 50) {
    $answer = "kbNtA";    
} else {
    $answer = "WiaNq";
}
print "Answer = [$answer]\n";
[download]

And it seems to be working again:

S:\Steve\Dev\PerlMonks\P-2017-05-20@2349-Variable-Scope-Failure>perl s
+cope3.pl
Answer = [WiaNq]

S:\Steve\Dev\PerlMonks\P-2017-05-20@2349-Variable-Scope-Failure>perl s
+cope3.pl
Answer = [WiaNq]

S:\Steve\Dev\PerlMonks\P-2017-05-20@2349-Variable-Scope-Failure>perl s
+cope3.pl
Answer = [kbNtA]
[download]

I'd suggest using strictand warningsand declaring your variables with myand then re-run the script. I would not be surprised if Perl tells you what you're doing wrong.

#!/usr/bin/perl
use CGI qw/:html3 :standard/;
use Fcntl qw/:flock/;
#
#################################################
#                                               #
#            My Capcha                          #
#                                               #
#          Starting Date: April 20, 2017        #
#        Last Modified: May 21, 2017            #
#                                               #
#          Written by Troy Hammack              #
#          Email: troylh@pacbell.net            #
#                                               #
#################################################
#
#
$ENV{"PATH"}='';
$ENV{"BASH_ENV"}='';

print "Content-type:text/html\n\n";
if ($ENV{'REQUEST_METHOD'} eq 'GET') {
    @pairs = split(/&/, $ENV{'QUERY_STRING'});
} elsif ($ENV{'REQUEST_METHOD'} eq 'POST') {
    read (STDIN, $buffer, $ENV{'CONTENT_LENGTH'});
    @pairs = split(/&/, $buffer);
} else {
    print "Content-type: text/html\n\n";
    print "<P>Use Post or Get";
}

foreach $pair (@pairs) {
    ($key, $value) = split (/=/, $pair);
    $key =~ tr/+/ /;
    $key =~ s/%([a-fA-F0-9][a-fA-F0-9])/pack("C", hex($1))/eg;
    $value =~ tr/+/ /;
    $value =~ s/%([a-fA-F0-9][a-fA-F0-9])/pack("C", hex($1))/eg;
    $value =~s/<!--(.|\n)*-->//g;
    
    if ($formdata{$key}) {
        $formdata{$key} .= ", $value";
    } else {
        $formdata{$key} = $value;
    }
#
# Test to see if Reset button was selected and clear all values in the
+ keys
#
if ($formdata{'reset'} eq "Reset") {
    foreach $pair (@pairs) {
        ($key, $value) = split (/=/, $pair);
        $formdata{$key} = "";
    }
}

#
# The answer < -- If I use this it doesn't work!
#
my @chars = ("A".."Z", "a".."z");
my $answer;
$answer .= $chars[rand @chars] for 0..4;
#
# $answer = "ghRuE"; <-- If I use this it works fine.
#
print <<EndHTML;
<html><head>
<meta http-equiv="Content-Type" content="text/html; charset=windows-12
+52">
<title>My Capcha</title></head>
<body background="http://www.astrologybythesea.com/images/bg_coastR.jp
+g">
<!-- <p><u><font size="5" color="#0000FF">My Capcha</font></u></p> -->
<form action="/cgi/mycapcha.cgi" method="post">
<table border="0" width="100%">
<center>
<H5>AntiBot Verifiication</H><br>
<div style='height:20px;width:70px;border:1px solid blue'>
<font size="2">$answer</font></div>
<!-- Now get users input -->
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
<input type="text" size="8" maxlength="8" name="useranswer" value="$fo
+rmdata{'useranswer'}">
<input type="submit" name="verify" value="Verify"><br>&nbsp;&nbsp;&nbs
+p;Type the above 
characters</center>
EndHTML

#
# $UpperCaseUserAnswer  is UserAnswer converted to  uppercase
#
$UpperCaseUserAnswer = uc $formdata{'useranswer'};
#
# $UpperCaseAnswer is Upper case of $anwser
# =================================================
$UpperCaseAnswer = uc $answer;
if ($UpperCaseUserAnswer eq $UpperCaseAnswer) {
    goto END;
    exit(0);
}

print <<EndHTML;
</table></form></body></html>
EndHTML

exit(0);
#=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
END:
print  "<br>Finished Capcha;
exit(0);
[download]

You should have used <code> and </code> to delimit your code. it makes it much easier to read

Despite that i dont think you understand the nature of how cgi works. Each time that runs because the user pressed verify it generates a new $answer, that bears no relationship to whatever was typed into $useranswer the previous time it was displayed to the user. This is what you test against the users answer so it is bound not to match this run.

I dont do capcha stuff much, but i would include a hidden string to return in the post. This hidden string would have been used to create the previous "$answer" in a way that the end user would not be able to replicate easily. When this parameter is found in the reply you would then use it to regenerate the previous answer and compare it to what was returned in $useranwser. I might pick CGI::Session as the mechanism to store the previous answer and have the sessionid be used as the returned value to regenerate the $answer.

Edit: add example code

#!/usr/bin/perl
use strict; use warnings; 
select STDOUT; $| = 1;

  my $session_dir='/home/huck/monks-sessions';  # must exist and be wr
+iteable by www userid 
  my $expires='+1m';


  use CGI;

  use CGI::Session;

  my $session; 

  my $cgi = CGI->new;

  my $tssid  = $cgi->param('tssid');

  my $sessiona1=undef; 
  my $sessiona3={Directory=>$session_dir}; 

  unless ($tssid){ 
      new_capcha('');
      } # no tssid 
  else { 
      $session = CGI::Session->load($sessiona1, $tssid, $sessiona3);
      if ( $session->is_expired )    { 
          $session->delete(); 
          $session->flush();
          new_capcha('Try again: took too long');
          } 
      elsif ( $session->is_empty )   { 
          $session->delete(); 
          $session->flush();
          new_capcha('Try Again:Session not found');
          } 
      else { 
          my $oldanswer=$session->param('answer');
          my $useranswer= $cgi->param('useranswer');
          if (uc($oldanswer) ne uc($useranswer)) { 
              $session->delete(); 
              $session->flush();
              new_capcha('Try Again:didnt Match');
              } 
          } 
      } 
# the only way it can get here is if answer is right 
      print $cgi->header(); 
          print <<EndgoodHTML;
<html><head>
<meta http-equiv="Content-Type" content="text/html; charset=windows-12
+52">
<title>My Capcha</title></head>
<body>
You won
</body></html>
EndgoodHTML
  CGI::Session->find($sessiona1 ,sub {} ,$sessiona3); # clean expired 
+sessions 

exit; 

sub new_capcha{ 
  my $reason=shift; 
  $session = CGI::Session->new($sessiona1, undef,$sessiona3);
  my @chars = ("A".."Z", "a".."z");
  my $answer;
  $answer .= $chars[rand @chars] for 0..4;
  $session->expires($expires);
  $session->param('answer',$answer);
  $session->flush(); 
  $tssid= $session->id; 
  print "Content-type:text/html\n\n";
  print <<EndHTML;
<html><head>
<meta http-equiv="Content-Type" content="text/html; charset=windows-12
+52">
<title>My Capcha</title></head>
<body background="http://www.astrologybythesea.com/images/bg_coastR.jp
+g">
<!-- <p><u><font size="5" color="#0000FF">My Capcha</font></u></p> -->
<form action="/monks-bin/capcha.pl" method="post">
<table width="100%" border="0">
<center>
<h5>AntiBot Verifiication</H><br />
$reason 
<div>
<font size="2">$answer</font></div>
<!-- Now get users input -->
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbs
+p;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
<input type="text" size="8" maxlength="8" name="useranswer" value="">
<input type="hidden" name="tssid" value="$tssid">
<input type="submit" name="verify" value="Verify"><br />&nbsp;&nbsp;&n
+bsp;Type the above characters
</h5></center></table></form></body></html>
EndHTML
exit; 
} # newcapcha
[download]

the other is input from a form, it called $useranswe.

Okay.

then compare it to $useranswer

Well, there's your problem. $useranswe and $useranswer are two different variable names.

…

Now, I know what you're thinking. I spelt the variable name wrong when posting on PerlMonks. It's correct in my code. But how do we know that? You haven't shown us the code where you read the user input. You haven't shown us the code where you do the string comparison.

Hi TroyH and welcome to the monastery. Your question is a bit vague and does not really give enough info to give a proper answer (see How do I post a question effectively?). Is the form a web form or some type of console input. If console, I'd hazard a guess that the input has a trailing newline that needs to be chomped. Other than that, I think you'd need to reduce your comparison code to a simple test case and post it back here ... could be a problem with your comparison (ahh ... the many times I used = instead of == or eq).

This doesn't speak to your problem, but FWIW there's a neat way to equally select at random from an array, as you seem to be doing:

c:\@Work\Perl\monks>perl -wMstrict -MData::Dump -le
"my @ra = qw(x y z);
 printf '%s ', $ra[ rand @ra ] for 1 .. 20;
 print '';
 ;;
 my %h;
 $h{ $ra[ rand @ra ] }++ for 1 .. 1_000_000;
 dd \%h;
"
y y y x x z y y z z z y x z x y x y y x
{ "x" => 333660, "y" => 332908, z => 333432 }
[download]

Can anyone explain to me why does the random statement corrupt the $answer

Can you pare your code down to a minimal (don't just post the whole program!), but complete, stand-alone example that we can run for ourselves which demonstrates the problem? That would allow us to answer definitively (because we can run your code, see it work incorrectly, then fix it, run it again, and see it work correctly), plus there's a very strong chance that you'll find the answer yourself in the process of trying to create the minimal example program.