@RISK Modeling Tips
In response to reader demand, the Palisade newsletter now has more modeling tips and examples. These tips and examples will cover real-life applications of @RISK and other Palisade tools, and many include spreadsheets you can download and try yourself. Last month we featured @RISK models calculating Value-at-Risk and Discounted Cash Flow. This month, learn about cost estimation, sports simulation, and new product profitability.

If you have a tip or example model you’d like to share, we’d love to hear from you. Send your model to: tech-support@palisade.com.

@RISK 5.0 Free Live Webcasts
With the imminent release of @RISK 5.0, you can get up to speed on this exciting new software by participating in a free live Webcast. Delivered using industry-standard Webex web conferencing services, these Webcasts will walk you through key new features in @RISK 5.0 while showing off the stunning new interface. You can also learn in-depth about the Six Sigma capabilities included with @RISK 5.0.

What’s New in @RISK 5.0
18 December 2007, 10:00 AM – 11:30 AM EST
20 December 2007, 11:00 AM – 12:30 PM EST
» Register now

@RISK 5.0 and Six Sigma
19 December 2007, 11:00 AM – 12:00 PM EST
» Register now

Special live demonstrations of the soon-to-be-released all-new @RISK 5.0 are being held in cities around the world. Hundreds of professionals in Frankfurt, Munich, Toronto, Calgary, and Houston have recently previewed @RISK 5.0, with dates planned soon for additional locations. To suggest tour dates and locations for 2008, write to sales@palisade.com.

“@RISK 5.0 is really well put together.”
Albert Perrella, Institute for Defense Analyses

User Conference Series
Concludes in Latin America
Palisade is pleased to announce the success of its Latin America User Conference. Conducted entirely in Spanish and held at the Intercontinental Real in San Jose, Costa Rica on 26-27 November, 2007, the event attracted over 70 delegates. Participants included practitioners, academics, and consultants from across Latin America.

“We were very happy to meet professionals from banking, academia, oil and gas, and other industries,” noted Palisade Latin American Sales Director Denise Castellot. “The insights gained by both the delegates and staff made this a very valuable forum for quantitative risk and decision analysis.”

The San Jose conference was the fourth in a global series of User Conferences from Palisade in 2007 that commenced with an event in London in April, and included stops in Miami in October and Sydney, Australia in September. Attendance at the series in 2007 was more than double that of 2006, and events are being planned for 2008. Join us next year for what has become the industry risk and decision analysis forum!

» 2008 Palisade User Conference Europe, 22-23 April, 2008

More dates coming soon.

» Download the presentation and example model

Making $ense of Baseball –
An Econometric Approach
by Clay Graham, Analytical Advantages, LLC
Ever since the ancient Greeks first flipped a Drachma, gaming and probabilities have gone hand in hand. Baseball is unique in its finite nature of options when a batter faces a pitcher. There is no sports endeavor on earth that quantifies performance more than baseball. However, utilizing such information for gaming purposes is usually relegated to the “tipster” or “insiders”. The presentation covers:

• Analyzing the cause and effect of run production
• Building a detailed stochastic forecasting model of a baseball game
• Integrating “lines and odds” into the economics of Expected Values of Return on Investment
• Creating a wagering decision algorithm
• Incorporating a control system

Building probabilities of winning and run production are brought together in a pragmatic manner (making smart bets) by utilizing the family of Palisade products, i.e, @RISK, StatTools, and BestFit.

» View now: "Making $ense of Baseball" (free webcast)

Cost Estimation
When submitting a budget proposal for a project, two key questions that you are likely to be interested in are:

a) What is the probability that the project will actually be delivered within this budget?

b) How much contingency (i.e. extra budget) should be included in order for this new revised budget level to be achieved with a certain degree of confidence?

This model shows how one might answer these questions through @RISK Monte Carlo simulation: It is assumed that each item's actual cost will be within a min-max range, and that Pert distributions are applicable to describe the possible costs of each item in practice. The distributions are assumed to be skewed in this case, and the parameters chosen to reflect that costs are assumed to more likely overshoot than undershoot the base case. Results are given using @RISK statistics functions as well as through standard graphs and reports.

» Download the example: CostEstimation.xls

NCAA Tournament
This example lets you play out the NCAA tournament as many times as you want using @RISK Monte Carlo simulation. We factor in the abilities (through the SAGARIN ratings published in USA Today) of each team. Extensive data analysis has indicated that teams play on average to SAGARIN ratings and perform according to a standard deviation of 7 points about that level.

» Download the example: NCAA.xls

New Product Profitability
When a company develops a new product, the profitability of the product is highly uncertain. @RISK Monte Carlo simulation is an excellent tool to estimate the average profitability and riskiness of new products. The following example illustrates how simulation can be used to evaluate a new product. Imagine Pigco is thinking of marketing a new drug used to make hippos healthier. The model sets up the variables involved in marketing the new product, such as market size, use of the drug, whether competitors enter the market, etc.

» Download the example: Hippo.xls

Beta Family of Distributions

RiskBeta
The Beta distribution is often used as a starting point to derive other distributions (such as the BetaGeneral, PERT and BetaSubjective). It is intimately related to the Binomial distribution, representing the distribution for the uncertainty of the probability of a Binomial process based on a certain number of observations of that process.

RiskBeta(alpha1,alpha2) specifies a beta distribution using the shape parameters alpha1 and alpha2. These two arguments generate a Beta distribution with a minimum value of 0 and a maximum value of 1.

For example, RiskBeta(1,2) specifies a beta distribution using the shape parameters 1 and 2. RiskBeta(C12,C13) specifies a beta distribution using the shape parameter alpha1 taken from cell C12 and a shape parameter alpha2 taken from cell C13.

RiskBetaGeneral
The BetaGeneral is directly derived from the Beta distribution by scaling the [0,1] range of the Beta distribution with the use of a minimum and maximum value to define the range. The PERT distribution can be derived as a special case of the BetaGeneral distribution.

RiskBetaGeneral(alpha1,alpha2,minimum,maximum) specifies a beta distribution with the defined minimum and maximum using the shape parameters alpha1 and alpha2.

For example, RiskBetaGeneral(1,2,0,100) specifies a beta distribution using the shape parameters 1 and 2 and a minimum value of 0 and a maximum value of 100.

RiskBetaGeneralAlt
RiskBetaGeneralAlt(arg1type, arg1value, arg2type,arg2value, arg3type,arg3value, arg4type,arg4value) specifies a beta distribution with four arguments of the type arg1type to arg4type. These arguments can be either a percentile between 0 and 1 or alpha1, alpha2, min or max. Both alpha1 and alpha2 must be greater than zero and max > min.

For example, RiskBetaGeneralAlt("min",0,10%,1,50%,20,
"max",50) specifies a beta distribution with a minimum value of 0 and a maximum value of 50, a 10th percentile of 1 and a 50th percentile of 20.

RiskBetaSubj
The BetaSubjective distribution is rather like a BetaGeneral distribution in the sense that the range of the underlying Beta distribution has been scaled. However, its parameterization allows it to be used in cases where one wishes not only to use a minimum-most likely-maximum parameter set (as for the PERT distribution) but also to use the mean of the distribution as one of its parameters.

RiskBetaSubj(minimum, most likely, mean, maximum) specifies a beta distribution with a minimum and maximum value as specified. The shape parameters are calculated from the defined most likely value and mean.

For example, RiskBetaSubj(0,1,2,10) specifies a beta distribution with a minimum of 0, a maximum of 10, a most likely value of 1 and a mean of 2.

Climate Change
The Statistics of Record-Breaking Temperatures
Technique: Monte Carlo simulation   » see @RISK
Source: CO2 Science Magazine, December 5, 2007

What is the relation of local warm spells to global warming? A new study uses Monte Carlo simulation and looks at 126 years of real-world temperature data from the city of Philadelphia.

Education
Financial Literacy in Crisis
Technique: Neural networks   » see NeuralTools
Source: ArabianBusiness.com, November 17, 2007

Finding that financial literacy can ease social problems, new research identifies factors that determine this knowledge. Neural networks analysis is used to assess financial literacy of groups of people.

Investments
Q3 2007 Deutsche Postbank AG Earnings
Technique: Monte Carlo simulation   » see @RISK
Source: InsuranceNewsNet.com, November 15, 2007

Deutsche Postbank uses Monte Carlo simulation to model sub-prime related portfolio loss

Law Enforcement
Traffic Cops Get High-Tech Help
Technique: Neural networks   » see NeuralTools
Source: Wheels24.co.za, November 26, 2007

A new traffic surveillance system makes use of a high-speed camera and a neural network.

Project Management
Using Monte Carlo Simulation to derive more accurate project duration estimates
Technique: Monte Carlo simulation   » see @RISK
Source: Enterprise 2.0 Ireland, November 16, 2007

Forecasting how long a project is going to take has been a near impossible task for companies for years. This article looks at how Monte Carlo Simulation could be used to give more accurate project duration estimates.

Real Estate
Home Economics: The Property Market
Technique: Monte Carlo simulation   » see @RISK
Source: Times Online, UK, November 11, 2007

Finance columnist suggests Monte Carlo simulation to anticipate ups and downs in property valuations.

Oil and Gas
Application of the Dispersive Discovery Model
Technique: Monte Carlo simulation   » see @RISK
Source: The Oil Drum Discussion Board, November 27, 2007

Monte Carlo simulation is used to verify the closed-form solution of the Dispersive Discovery model of oil and gas prospecting.

Water Resources
Managing Florida's water a balancing act
Technique: Decision trees   » see PrecisionTree
Source: The News-Press (Fort Myers,FL), November 27, 2007

For the South Florida, the goal is to keep the level of water balanced, the quality high, and to satisfy as many stakeholders as possible.  Decision trees are used to account for many different factors.

The Palisade User Forums are online discussion boards where users are invited to post questions and share ideas on their use of @RISK and other Palisade software. It’s also a great place to check for announcements regarding updates of Palisade software. Forums are organized by products.

» Join or view the Palisade Forums

Recent topics include:

» Correlations: One RiskIndepC to Multiple DepC Functions
» Setting up Correlation Matrices
» Interpreting Y-axis of a Histogram