Call now: 252-767-6166  
Oracle Training Oracle Support Development Oracle Apps

Free Oracle Tips

HTML Text

 Home
 E-mail Us
 Oracle Articles



 Oracle Training
 Oracle News

 Oracle Forum
 Class Catalog


 Our Staff
 Our Prices
 Help Wanted!

 Remote DBA
 Oracle Tuning
 Emergency 911
 RAC Support
 Apps Support
 Analysis
 Design
 Implementation
 Oracle Support


 SQL Tuning
 Security

 UNIX
 Oracle UNIX
 Linux
 Oracle Linux
 Monitoring
 Remote help

 Remote plans
 Remote
services
 Oracle C++
 Oracle Java
 Apache
 JDeveloper
 App Server

 Applications
 Oracle Forms
 Oracle Portal
 11i Upgrades
 SQL Server
 Oracle Concepts
 HTML-DB Tips
 Software Help

 Remote Help  
 Development  

 Implementation


 Financials Training
 Oracle 11i
 Oracle Apps 11i
 Oracle Workflow
 Oracle AR 11i Class
 Oracle AP 11i class
 Oracle GL 11i class
 Oracle HR 11i class
 Oracle FA 11i class
 11i Project Mgt
 11i procurement
 11i collections


 Oracle Posters
 Oracle Books

 Oracle Tuning Book
 Oracle RAC Book
 Oracle Security
 Easy Oracle Books
 Oracle Scripts
 SQL Server DBA
 SQL Design Patterns
 WISE
 Excel-DB   


 BC Oracle News


 Rednecks!
 Dress code
 Arabian Stallion

 Burleson Arabians
 Guide Horses
 Don Burleson Blog
 Golf & Travel


 Privacy Policy
 

 

 
 

Deal or No Deal Probabilities

Oracle Tips by Burleson Consulting

Many people have been fans of the latest game show "Deal or No Deal".  During the show, the Deal or No Deal contestants choose to open a series of amounts and then the "Bank" makes an offer to make the person choose to continue.

When faced with a choice, the "Deal or No Deal" contestants are essentially dealing with a probability problem.  On the Deal or No Deal board, the contestants choose "cases" that contain values from one cent to a million dollars:

Column On/off avg. $ value   Column On/off avg. $ value
$0.01 1 $0.00   $1,000 1 $38
$1 1 $0.04   $5,000 1 $192
$5 1 $0.19   $10,000 1 $385
$10 1 $0.38   $50,000 1 $1,923
$25 1 $1   $25,000 1 $962
$50 1 $2   $75,000 1 $2,885
$75 1 $3   $100,000 1 $3,846
$100 1 $4   $200,000 1 $7,692
$200 1 $8   $300,000 1 $11,538
$300 1 $12   $400,000 1 $15,385
$400 1 $15   $500,000 1 $19,231
$500 1 $19   $750,000 1 $28,846
$750 1 $29   $1,000,000 1 $38,462

The goal is to pick small values in order to have a higher expected payoff, and hence, a higher bank offer.  The expected payoff in Deal or No Deal is essentially probability problem.

Click for the deal_or_no_deal.xls spreadsheet that computed the expected payout for any round on the Deal or No Deal game.  To use my "Deal or No Deal" spreadsheet, change the "on/off" values from 1 to zero as the values are chosen.  The ideal behind "Deal or No Deal" is simple.  For example, consider the following scenario with four numbers on the board:

            $75
          $300
   $500,000
$1,000,000

The Deal or No deal contestant has a 25% change of getting any number, and over a given number of trials, the expected payout of each value is 25% time that value:

      Value                                 Expected Payout
       $75             $75*.25               =          $19
      $300             $300*.25              =          $75
  $500,000             $500,000*.25          =     $125,000
$1,000,000             $1,000,000*.25        =     $250,000
                                                   $375,000


In this simple example, the expected payout probability is $375,000.  Now, you would expect the Deal or No Deal banker to offer a value that relates to the expected probable payout, right?  In reality, the bank offers are calculated to manipulate the contestant (it's what makes the game exciting).

In my observation using the probability spreadsheet I noted that in the early rounds of "Deal or No Deal", the bank makes offers that are less than the contestant's expected probable payout (an incentive to continue).  As the rounds near completion, some suggest that the contestants behavior can be altered by the "bank", and that the offers (as compared to the expected probable gains) should increase, proportional to the risk/reward for each round:

Why offer  above the expected payout?  It's because people's attitude will change when faced with a large-stake high-risk gamble. 

For example, consider a deal where we have two items left, one for $100k and another for $300k.  The 50-50 chance expected payout over thousands of simulated trials would be (.5*100)+(.5*300)= $200k.  What would you rather have, a $200k sure-thing, or a 50% chance of getting $300k, with the risk of losing $100k?

Since our downside risk is limited to $100k (the worst outcome), this decision point can be reduced into a simpler problem:

You have just won $200,000 and the right to gamble half of your winnings.  You have these two choices:

a)     You may wager $100,000 on a double-or-nothing, winner takes all, based on a random 50-50 choice (like the toss of a fair coin), or:

b)     Take $200,000 and quit.

Obviously, this curve is never linear, but with a hit show like this, the Deal of No Deal bank offers are carefully crafted to maximize the stress on the contestant!  I'm going to refine my model, in the hopes of deriving the equation used by the Deal or No Deal banker to calculate their offers.

But in the end, it’s not the probabilities that determine the decision, but the human emotion, watching some average Joe make the largest financial decisions of their life, right on prime time TV.  The raw emotions are almost voyeuristic and viewers are encouraged to identify with the highly-likeable contestants (and their adorable families, conveniently, right on the set).


Reader Responses:

I love your deal or no deal spreadsheet.  It added a more realistic feel to our game over the random card bank offer of the at-home game we had.  I added the ability to assign a percentage of mathematical payout by round at the bottom left.  Hope it helps you figure out the bank offers!  Greg

      deal_or_no_deal4.xls


    Need an Oracle Health Check?
  • Do you have bad performance after an upgrade?
     
  • Need to certify that your database follows best practices?

BC Oracle performance gurus can quickly certify every aspect of your Oracle database and provide a complete verification that your database is fully optimized.

 

 

 

 
 
 

Oracle performance tuning book

 

 

Oracle performance tuning software

 
Oracle performance tuning software
 
SearchOracle web site
 
Oracle performance Tuning 10g reference poster
 
Oracle performance tuning webcast
 
Oracle training in Linux commands
 
Oracle training Excel
 
Oracle training & performance tuning books
 
 

 

Note: This Oracle documentation was created as a support and Oracle training reference for use by our DBA performance tuning consulting professionals. 
Feel free to ask questions on our Oracle forum.

Verify experience! Anyone considering using the services of an Oracle support expert should independently investigate their credentials and experience, and not rely on advertisements and self-proclaimed expertise. All legitimate Oracle experts publish their Oracle qualifications.

Errata?  Oracle technology is changing and we strive to update our BC Oracle support information.  If you find an error or have a suggestion for improving our content, we would appreciate your feedback.  Just  e-mail:  and include the URL for the page.
 
 


Burleson Consulting

The Oracle of database support


 

Copyright © 1996 -  2007 by Burleson Enterprises, Inc. All rights reserved.

Oracle® is the registered trademark of Oracle Corporation.


Hit Counter