A quick benchmark to confirm/dispel performance concerns over using `.map(&:to_i)` instead of a literal block (`.map { |x| x.to_i }`).

Explicit Block vs Symbol#to_proc.rb

require 'benchmark'

n = 100000
a = %w{ 1 2 3 4 5 6 7 8 9 0 }

Benchmark.bmbm do |b|
  b.report 'Explicit Block' do
    n.times do
      a.map { |x| x.to_i }
    end
  end
  
  b.report 'Symbol#to_proc' do
    n.times do
      a.map &:to_i
    end
  end
end

# Rehearsal --------------------------------------------------
# Explicit Block   0.350000   0.000000   0.350000 (  0.371145)
# Symbol#to_proc   0.730000   0.020000   0.750000 (  0.784472)
# ----------------------------------------- total: 1.100000sec
# 
#                      user     system      total        real
# Explicit Block   0.360000   0.010000   0.370000 (  0.365922)
# Symbol#to_proc   0.730000   0.020000   0.750000 (  0.786993)

Using MRI Ruby 2.0.0:

Rehearsal --------------------------------------------------
Explicit Block   0.160000   0.000000   0.160000 (  0.159286)
Symbol#to_proc   0.150000   0.000000   0.150000 (  0.148407)
----------------------------------------- total: 0.310000sec

                     user     system      total        real
Explicit Block   0.150000   0.000000   0.150000 (  0.157774)
Symbol#to_proc   0.150000   0.000000   0.150000 (  0.149271)