Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
soundsop
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
76 ms
·
901.
▲
DES Encryption Part 1: Encrypting the Blocks
(scienceblogs.com)
2 points
by
soundsop
18y ago
|
0 comments
902.
▲
One-click upload: A jquery plugin
(michaelmitchell.co.nz)
4 points
by
soundsop
18y ago
|
0 comments
903.
▲
Amazing 70-year-old color photos
(mentalfloss.com)
1 points
by
soundsop
18y ago
|
0 comments
904.
▲
by
soundsop
18y ago
Request Haiku: I just want On Lisp, to improve my coding style with advanced techniques
905.
▲
The Closures are Coming
(notes-on-haskell.blogspot.com)
1 points
by
soundsop
18y ago
|
0 comments
906.
▲
Prions jump species barrier
(nature.com)
13 points
by
soundsop
18y ago
|
4 comments
907.
▲
Fuzzmail
(fuzzmail.org)
3 points
by
soundsop
18y ago
|
1 comments
908.
▲
Goodbye MapReduce, Hello Cascading
(blog.rapleaf.com)
30 points
by
soundsop
18y ago
|
4 comments
909.
▲
Roger Ebert: How to read a movie
(blogs.suntimes.com)
3 points
by
soundsop
18y ago
|
0 comments
910.
▲
Chrome replacing Firefox
(blog.statcounter.com)
12 points
by
soundsop
18y ago
|
6 comments
911.
▲
Weblog 1.0 - Simple blog software
(henry.precheur.org)
6 points
by
soundsop
18y ago
|
0 comments
912.
▲
One-Line Javascript memoization
(osteele.com)
8 points
by
soundsop
18y ago
|
2 comments
913.
▲
Delicious in Google Chrome
(fernandoacorreia.wordpress.com)
4 points
by
soundsop
18y ago
|
0 comments
914.
▲
Syntactic Sugar
(wordaligned.org)
3 points
by
soundsop
18y ago
|
1 comments
915.
▲
by
soundsop
18y ago
An opposing view: http://nion.modprobe.de/blog/archives/651-Google-chrome-code...
916.
▲
Google Chrome code reuse nightmare
(nion.modprobe.de)
1 points
by
soundsop
18y ago
|
0 comments
917.
▲
Libv8 + libfastcgi = server-side Javascript
(tokuhirom.vox.com)
24 points
by
soundsop
18y ago
|
6 comments
918.
▲
First look at Cappuccino and Objective-J
(theocacao.com)
10 points
by
soundsop
18y ago
|
0 comments
919.
▲
Three different implementations of MapReduce
(dbms2.com)
1 points
by
soundsop
18y ago
|
0 comments
920.
▲
Introduction to Block Ciphers
(scienceblogs.com)
3 points
by
soundsop
18y ago
|
0 comments
921.
▲
Computer resources are there to be used
(tech.slashdot.org)
2 points
by
soundsop
18y ago
|
0 comments
922.
▲
Why Spam Cannot be Stopped
(trevorblackwell.com)
3 points
by
soundsop
18y ago
|
3 comments
923.
▲
Young people have unique sense of Facebook privacy
(cbc.ca)
2 points
by
soundsop
18y ago
|
0 comments
924.
▲
Some assembly required
(dadhacker.com)
2 points
by
soundsop
18y ago
|
0 comments
925.
▲
Anatomy of a Heisenbug
(blogs.msdn.com)
3 points
by
soundsop
18y ago
|
0 comments
926.
▲
Hands-on with Yotify, your personal web secretary
(arstechnica.com)
1 points
by
soundsop
18y ago
|
0 comments
927.
▲
Google Chrome: Straight out of Compton
(arstechnica.com)
1 points
by
soundsop
18y ago
|
0 comments
928.
▲
by
soundsop
18y ago
The part about randomness is a red herring. They are choosing from a larger set of plays than is usual. The article implies that this is somehow akin to randomness, but I think that's a bit of a stretch. I do remember some research on using
929.
▲
by
soundsop
18y ago
I don't think having fewer than 5 linemen is a legal formation. The article is confusing in it's explanation of the formations. If you look at the slides of the plays (link at bottom), there seem to be at least 5 player on the line of scrim
930.
▲
by
soundsop
18y ago
The article's title refers to randomness, but most of article deals with the fact that the A-11 formation allows more possibilities of who receives the ball, which seems to be the critical factor that improved the teams win rate.
More ›