You weren't too far off, i prefer to use a for loop instead and avoid explicitly using $1 whenever i can. Here's a little scheme i developed for myself a while back. Start with a simple for loop - grab the files you want and print them out:

perl -le'for(<*html>){print}'
[download]

Next grab a copy of $_, perform substitution on $_ and print

perl -le'for(<*html>){$o=$_;s/^new/U/;print}'
[download]

Finally, drop the print and do some damage

perl -e'for(<*html>){$o=$_;s/^new/U/;rename $o,$_}'
[download]

The first step is just to set up a working one-liner. The second step is important, as you will see what the results will look like, but i still like to live dangerously and just hammer out the third step unless i am using data of any importance. Be safe. :)