I scanned my bracket into PDF and I use the Cloud tool in Acrobat Pro to circle my correct picks and the Line tool to cross out my incorrect picks

3/19/2009 9:37:55 PM

k

3/19/2009 11:12:32 PM

i printed mine and used a pen.

3/19/2009 11:39:10 PM

i did both

3/19/2009 11:48:23 PM

i used a probabilistic model to make all of my brackets on espn and the two money pools i'm in. i've done this for the past few years with little tweaks to the model each year. so far this year it seems that my tweaks were not very good. my brackets are almost all doing terribly through the first day.

3/19/2009 11:58:44 PM

interesting...i guess that shows that there are unquantifiable mental aspects amongst the players and the atmosphere and basically that on any given day, any team can beat another fairly evenly matched team

[Edited on March 20, 2009 at 12:06 AM. Reason : palindrome]

3/20/2009 12:06:14 AM

i basically use the same formula that kenpom uses except this year i added in tournament experience (highest round reached by the team in the past five years) and if a team is playing within 300 miles of their home court. i think i may have weighted the experience bit a little bit too much.

and given a percentage chance of winning the game, i then take a random number between 0 and 1 and if it is higher than the given percentage, the other team wins.

[Edited on March 20, 2009 at 12:11 AM. Reason : .]

3/20/2009 12:08:51 AM

I used Pomeroy's numbers to simulate every possible match-up, made some gut changes, and bam

3/20/2009 1:08:33 AM

I was going to run some genetic algorithms to predict this shit, but then again I couldn't find easily parsed stats.

3/20/2009 5:56:18 AM

i flipped coins

[not really]

3/20/2009 8:37:34 AM

I just called up Best Picks and they sent out the Hoop Squad to my house last week. Apparently these guys all played ball in High School so I should really kill it this year.

3/20/2009 10:23:44 AM

Quote :

"I was going to run some genetic algorithms to predict this shit, but then again I couldn't find easily parsed stats."

how would a genetic algorithm help you here?

3/20/2009 10:57:19 AM

why wouldn't it

3/20/2009 11:00:15 AM

shh, let him think he's smart.

3/23/2009 10:08:38 AM