Quizzes by State

AK      AZ      CA      CO      CT      IA      ID      IL      IN      KS      LA      MA      ME      MI      MN      MO      NC      NE      NH      NJ      NM      NV      NY      OH      OK      OR      PA      RI      TN      TX      UT      VA      WA      WI      

Quizzes by City

Select a City/State Near You     
Or, find a venue within of your zipcode:

Complete Quiz Schedule
Google+






 

wits who Geek

Schedule Updates

Many of you nerds just don't trust Evil Rob and his mathematical prowess on word alone. So to quell the tiny fisted masses, straight from Evil Rob himself, is how he calculates the Power Rankings on the Geek In Review. It still won't help you win.

  • Get the top 3 scores from each bar.

  • Average the top 3 scores from each bar.  (Ex: if O'Niell's top 3 scores are 74, 73, and 72, the average is 73)

  • Multiply the average score for the Irish Snug and Forecaster's by 1.12 (due to their quizzes only having 7 questions per round--96 possible points at a regular quiz vs. 85 or 86 possible points at those two bars; 96/86 = 1.12).
  • Average each bar's average scores together for each night.  Only count the top half of the bars for each night (this is to keep the scores from being skewed by there being too many non-competitive venues on a given night, Tues./Wed. in particular) So if there's 40 bars on a given night, average together only the top 20 bars from that night. 

  • Add up the average scores from each night and divide by 6.  This is the week average.

  • Take the week average and divide it by the average score from each night.  This is the standard deviation for each night.  (Ex.: Off the top of my head, the deviation for last Wednesday's quiz was 0.984 since the average score that night was just slightly higher than the week average.  Tuesday's was 0.912--scores were much higher across all venues that night, so there's more of a downward adjustment for difficulty relative to the other quizzes that week)  Let's say the average score for the week was, say, 67.  If Monday's average was 64.3, I divide 67.0 by 64.3, which gives me a standard deviation of 1.042.

  • Multiply each bar's score by the standard deviation for that night (Ex.: If O'Niell's average score is 73, 73 x 0.984 = 71.83).

  • Group the bars by city.  For cities with a lot of bars (i.e. Denver and Austin), count only the top 5 bars.  Otherwise throw out each city's lowest scoring bar.  Count no more than 5 bars for each city.  Don't bother counting cities like Santa Fe or Nashville or Grand Junction with only one bar for the city rankings (though those bars are certainly eligible for the bar rankings if they score high enough).  For cities with two bars, count both bars.  Average each city's remaining bars and rank the averages.

  • Find the bars with the 5 highest adjusted scores and rank them.

  • Go through each night and find the 3-5 highest scores from each night and multiply those by the standard deviation for that night.  (Ex.: A score of 82 on last Wednesday =  82 x 0.984 = adjusted score of 80.69).  Get the top five highest scores, rank them, and get the teams' names.