sekrah
09-26-2005, 06:59 PM
Been writing a computer program to pick College football games, we'll see how it kicks.. The algorithm is pretty complicated (probably overcomplicated, but we'll see). It backtested freakishly well vs this last weekends game, but we know what backtesting means (zilch).
Anyway, there's a bunch of different factors involved, including Off v Def Matchups, Pass O vs Pass D, Rush O vs Rush D, Injuries, Experience, Team Mentality (re: Last Game), etc etc and a bunch of other crap.
Anyway, here's the output.
<font class="small">Code:</font><hr /><pre>
Tennessee at LSU -6.5
Tennessee Points: +1
LSU Points: +4
Net: LSU +3, LSU by 18 (11.5 Spread, 2 Star Selection)
</pre><hr />
i.e. each point edge is worth a 6 point edge in the game.
Rolling out the software, this week, we'll see how it flies.
Anyway, there's a bunch of different factors involved, including Off v Def Matchups, Pass O vs Pass D, Rush O vs Rush D, Injuries, Experience, Team Mentality (re: Last Game), etc etc and a bunch of other crap.
Anyway, here's the output.
<font class="small">Code:</font><hr /><pre>
Tennessee at LSU -6.5
Tennessee Points: +1
LSU Points: +4
Net: LSU +3, LSU by 18 (11.5 Spread, 2 Star Selection)
</pre><hr />
i.e. each point edge is worth a 6 point edge in the game.
Rolling out the software, this week, we'll see how it flies.