Sunday, February 8, 2009

Streak For The Cash Analysis

ESPN is the Worldwide leader in sports. They are the absolute behemoth when it comes to sporting entertainment. The enterprise, which is owned by Disney, is a multi-conglomerated, multi-faceted sporting portal that sports fans from around the world use. To begin, their website is the 65th most visited on the entire Internet (Alexa.com). This is all websites, in all languages, for all countries around the entire world. That is absolutely staggering and there is no other sporting portal even close to its Internet rank.

In addition, ESPN also provides an almost sickening array of television channels. A brief list of their channels include: ESPN, ESPN International, ESPN 2, ESPN News, ESPN Classic, ESPN Deportes, ESPNU, ESPN Plus, and ESPN Star Sports. Note that this is an abbreviated list. There are several other ways in which ESPN has figured out for you to consume their sports media.

They have their own magazine, their own award show, their own cell phone, their own line of books, and even a restaurant chain devoted to ESPN.

Some of the prominent sporting rights that they have contracts for include: Monday Night Football, the Fifa World Cup, Major League Baseball, Major League Soccer, The NBA, Arena Football, Various PGA tour events, NASCAR, BCS College Football games, lots of college basketball.

The reason for sharing the scope of ESPN with you before I get into the Streak For The Cash is to evidence the empire that ESPN has created. Sports can occur at any time and at any place, but ESPN has worked so hard to differentiate their product and bring it to the fans that want to see it no matter where in the world you are, no matter what you're doing. They are a truly incredible organization, and I commend them for continually being one step ahead of the curve and truly seperating themselves in the market.
  • Streak For The Cash
Now that we're done understanding just how amazing ESPN is, I wanted to look a little closely at one of the more incredible accomplishments of this sports entertainment giant. Several months ago, ESPN launched the "Streak For The Cash" competition on their website.

The idea behind the game is simple. All one has to do is select the winner of a particular outcome that ESPN provides you with. All different types of sporting events are included in the Streak For The Cash. Often times, the game will ask the participant to predict the outcome in terms of a simple win-lose scenario, but other times, it asks, for example, who will score more points, when will this scenario occur in the game, etc.

The idea behind the game is to predict as many outcomes in a row correctly, as if to create a streak of wins.

There is a massive incentive for getting to 27 consecutive correct predictions and that is the grand prize of $1,000,000. In simple math terms, the probability of correctly predicting 27 outcomes with two possible choices each time is 1 in 134,217,728. That is an absurd probability. Frequently, you will see participants with a streak of 18 or 19 on the Leaderboard which is displayed on the front page of the site, but this is still a long way off from 27.

Anyway, the Streak For The Cash website also provides some interesting participating metrics along with each possible pick. They show the percentage of participants who have selected each outcome and they also indicate to what degree that outcome is being selected by categorizing the scenario into one of four different categories: cold, warm, warmer, hot.

I wanted to find if there was any sort of relationship between the percentage breakdown in each scenario and the probability of success. One would assume that those scenarios that have close to 100% picking one particular side in a scenario would be more accurate. The results I found are most definitely surprising.
  • Methods
In order to perform this study, I went back through all of the scenarios available on the Streak For the Cash website since January 1, 2009 up until January 24, 2009. There were a total of 263 scenarios in this time frame. I then set up a simple Excel document which chronicled the percentage picking the winner and the percentage picking the loser. In the same document, I kept track of the degree to which that scenario was being picked and which sport the scenario took place in.
  • Results
The following results were found based on my study of the Streak for the Cash:

Distributions - Percentages
In 73 cases (or 27.8%) of the game the percentages have been 90%-100% and 0%-10.
  • In 52% of the cases, the majority, i.e., the 90%-100% have been accurate, while 48% of the time the minority, or between 0%-10% of the players have been correct
In 66 cases (or 25.1%) of the game the percentages have been 80%-89.9% and 10%- 19.9%
  • In 52% of the cases, the majority, i.e., the 80%-89.9% have been accurate, while 48% of the time the minority, or between 10%-19.9% of the players have been correct
In 52 cases (or 19.8%) of the game the percentages have been 70%-79.9% and 20%-29.9%
  • In 48% of the cases, the majority, i.e., the 70%-79.9% have been accurate, while 52% of the time the minority, or between 20%-29.9% of the players have been correct
In 40 cases (or 15.2%) of the game the percentages have been 60%-69.9% and 30%-39.9%
  • In 60% of the cases, the majority, i.e., the 60%-69.9% have been accurate, while 40% of the time the minority, or between 30%-39.9% of the players have been correct
In 32 cases (or 12.2%) of the game the percentages have been 50%-59.9% and 40%-49.9%
  • In 37.5% of the cases, the majority, i.e., the 50%-59.9% have been accurate, while 62.5% of the time the minority, or between 40%-49.9% of the players have been correct
Distributions – Degree
There were 80 cases of “Cold"
  • The majority, i.e., percentage equaling 50% or greater was correct 46.25%
There were 65 cases of “Hot”
  • The majority, i.e., percentage equaling 50% or greater was correct 55.4%
There were 59 cases of “Warm”
  • The majority, i.e., percentage equaling 50% or greater was correct 52.5%
There were 59 cases of “Warmer”
  • The majority, i.e., percentage equaling 50% or greater was correct 49.2%
Distributions – Sport
There have been 11 opportunities with Tennis
  • The majority has been accurate 45.5% of the time
There have been 11 opportunities with Golf
  • The majority has been accurate 72.7% of the time
There has been 1 opportunity with Mixed Martial Arts
  • The majority has been accurate 0% of the time
There have been 61 opportunities with the NBA
  • The majority has been accurate 54.1% of the time
There have been 84 opportunities with College Basketball
  • The majority has been accurate 44.0% of the time
There have been 10 opportunities with College Football
  • The majority has been accurate 70% of the time
There has been 1 opportunity with Women’s College Basketball
  • The majority has been accurate 0% of the time
There have been 13 opportunities with the NFL
  • The majority has been accurate 38.5% of the time
There have been 48 opportunities with the NHL
  • The majority has been accurate 60.4% of the time
There have been 22 opportunities with Soccer
  • The majority has been accurate 36.4% of the time
There has been 1 opportunity with Boxing
  • The majority has been accurate 100% of the time
  • Conclusion
Understanding this information together, in order to maximize one’s probability of success in forming a streak the ideal scenario based on Streak For the Cash data would be a distribution of participants between 60%-69.9%, where the degree of Hot or Not was set to “Hot”, while selecting the sport of Golf or NHL.

3 comments:

Anonymous said...

this is extremely interesting. its exactly what i was looking for. I am curious as to how this worked out for anyone who has tried it?

vance said...

Did you collect the data yourself, recording everything each day? I was thinking about performing a similar analysis but I'm having trouble collecting enough data to ensure that my results are significant. It seems like golf is the way to go, but you only have 11 entries for golf. 11 is too small. I would be interested to find out if the percentage remains as high if there were more entries.

DG62 said...

Nice research. I would love to see someone perform this research over a year or two trial. Wonder what the outcome would be. Check my site if you get a chance, I make Streak for the Cash predictions - www.StreakWizard.com