This data set describes the results of various poker tournaments played by a UCLA student.
This data represents tournaments in which he finished in the upper 15 percentile.
We want to know whether the buy-in amount of the tournament (the cost to play) has any
effect on the late-stage performance of a tournament (defined here as after 85% of the field
has been eliminated).  It is generally assumed that higher buy-ins attract a higher caliber
of poker player.  Higher buy-ins also result in larger prize pools because the prize pool is simply
the number of entrants multiplied by the buy-in.  The results of the tournaments are independent
of each other, unless you consider the possiblity that the student became a better player over time and thus
tournaments played recently tend to have better results.  This data was collected over a one year 
time span to minimize the effect of this.  There are four variables.  "Buyin" is the amount of
money in dollars paid to participate in the tournament.  "Entrants" is the number of people who
entered the tournament.  "Place" is the place, or rank, that the student finished.  "Pay" is the winnings
received.  It is recommended that a new variable be created to represent the percentile the
tournament finish places in because there should be some distinction between getting 20th in a
tournament with 100 people and getting 20th in a tournament with 1000 people.  Similarly, a new
variable should be created in R to convert winnings to percentage returns (winnings divided by
the buy-in).  Obviously winning $50 in a $20 tournament is not as good as winning $50 in a
$1 tournament.  Place and winnings are two ways to measure performance, but winnings are heavily
weighted for higher places.