Toss all that initialization stuff. It's not just unnecessary, it's wrong. Here's a demo to show you a full load/use cycle.

#!/usr/bin/perl -w
use strict;

# Build the table
my $table_ref;
for my $row (1..5) {
   for my $col (1..4) {
      my @results = rand_vals(4);
      $table_ref->[$row][$col] = \@results;
   }
}

# Use the table
for my $row (1..5) {
   for my $col (1..4) {
      print "[";
      
      # Let Perl walk through the array for us
      #for (@{$table_ref->[$row][$col]}) {
      #   print "$_ ";
      #}
      
      # Or index our way through the array explicitly
      for (0..$#{$table_ref->[$row][$col]}) {
         print "$table_ref->[$row][$col][$_] ";
      }
      
      print "] ";
   }
   print "\n";
}

# Just a little tool for demo purposes
sub rand_vals {
   my $elements = shift;
   my @ary;
   push @ary, int rand(9) for 1..$elements;
   return @ary;
}
[download]

Prints out something like:

   [2 4 4 1 ] [8 6 2 5 ] [8 3 1 5 ] [6 5 3 7 ]
   [4 6 1 0 ] [6 6 6 2 ] [6 4 7 5 ] [8 2 6 8 ]
   [6 7 0 8 ] [7 8 6 2 ] [5 1 6 6 ] [1 7 8 3 ]
   [1 8 2 7 ] [7 8 2 1 ] [6 3 8 7 ] [2 2 6 4 ]
   [7 8 7 4 ] [5 8 4 7 ] [4 2 6 0 ] [3 2 3 5 ]
[download]

You can, of course, use the uncommented inner loop for accessing results values, or delete that loop and uncomment the other one. Whichever suits your needs.

------------------------------------------------------------
"Perl is a mess and that's good because the
problem space is also a mess." - Larry Wall