gistfile1.rb

# 9 out of 10 microbenchmarks agree: implicit return smokes explicit return
require 'benchmark'

def explicit
  return 1
end

def implicit
  1
end

n = 10_000_000

Benchmark.bmbm do |x|
  x.report("explicit") { n.times { explicit } }
  x.report("implicit") { n.times { implicit } }
end

gistfile2.txt

$ ruby test.rb 
Rehearsal --------------------------------------------
explicit   4.090000   0.020000   4.110000 (  4.143585)
implicit   2.610000   0.010000   2.620000 (  2.631262)
----------------------------------- total: 6.730000sec

               user     system      total        real
explicit   4.100000   0.010000   4.110000 (  4.805887)
implicit   2.630000   0.010000   2.640000 (  2.660820)