use strict;
# strict is good ++
# Where's warnings? --

sub sum {my $x; $x += $_ for @_; $x }
# poorly formatted --

sub avg {sum(@_)/@_}
# poorly formatted --
# probably should use scalar @_ instead of just @_ here

undef $/;
# no need to slurp to do an average for a line --

open(my $F, 'skaters.txt');
# lexical filehandle ++
# please use three-argument open --
# no error check on a system call --

my @x = map{[split/,/]}grep{/\S/}split/\n+/,<$F>;
# very poorly formatted --
# uses split to achieve what while ( <$F> ) would have done without the above slurp--
# makes me think well of those who call Perl line noise
# combines two loops and two searches on one line += 0
#     (good display of power, but could have built up
#      the data structure over time more clearly)

for(@x){
    @$_ = ($_->[0],avg(@{[sort{$a<=>$b}@{$_}[1..$#$_]]}[1..$#$_-2]))
    # nearly incomprehensible --
    # partly due to unnecessary terseness and formatting --
    # partly due to painful twists of syntax --
    # (The hash solution on perlbuzz was simpler and clearer.)
}

@x = sort{$b->[1] <=> $a->[1]}@x;
# poor formatting again --

printf "%-5s %-20s %-20s\n",$_+1,@{$x[$_]} for(0..2);
# poor formatting again --