ModelChoice turns a high-stakes choice into a decision tree in Excel. It shows which option is best, how much risk each one carries, what would change the answer and whether more information is worth paying for.
An exploration company holds a deepwater licence that expires in three years. An exploration well costs $45M and has a 30% chance of a dry hole. A 3D seismic survey costs $8M and shows whether the prospect is strong, moderate or weak before anyone drills.
71% chance of a loss; worst case −$53M
If the survey costs more than $13.7M, drilling straight away becomes the better choice.
What the tree shows: the survey adds $5.7M of expected value after paying for itself, and it cuts the worst case from −$85M to −$53M.
A biotech has a gene therapy for a rare disease. It can develop it as an in-vivo AAV therapy or as an ex-vivo cell therapy, license it to a large pharma company for milestones and royalties, or shelve it. Each route runs through preclinical results, funding and clinical outcomes.
79% chance of a loss; worst case −$210M
Below 25.5%, licensing becomes the better choice.
What the tree shows: the ex-vivo route has the highest expected value, $187.9M, yet loses money 79% of the time. Licensing is worth less on average and never loses money.
A large company wants a competitor. It can bid a $2B premium, agree a friendly merger for $1.5B, buy a 30% stake for $600M, or walk away. Regulators may approve, attach conditions or block the deal, and synergies and markets decide the rest.
38% chance of a loss; worst case −$1.5B
Below 36.2%, buying a 30% stake becomes the better choice.
What the tree shows: the hostile bid could make $3.0B but has a negative expected value, and it could lose $2.5B. The friendly merger is worth $663.2M on average.
A property and casualty insurer is weighing the cyber market. Premiums have doubled in three years, but one catastrophic ransomware event could wipe out years of underwriting profit. It can launch nationwide, pilot in three states, partner with a specialist agency, or stay out.
47% chance of a loss; worst case −$11M
Below 6.8%, launching nationwide becomes the better choice.
What the tree shows: the three-state pilot has the best expected profit, $3.2M, and its worst case is an $11M loss. A nationwide launch could lose $70M.
An energy company is weighing LNG export capacity on the US Gulf Coast: a full-scale terminal for $6B, a first train for $2.5B with the option of a second, or a floating unit for $3.5B. Permits, construction and the LNG market decide the payoff.
38% chance of a loss; worst case −$2.35B
Above 62.3%, building the full terminal becomes the better choice.
What the tree shows: building in phases is worth about $1.3B more than building in full. It costs less up front, its permit is likelier, and the second train is built only if contracts are strong.
A financial services firm with 500 staff is setting its security budget. It can buy a full security suite, basic protection, cyber insurance on its own, or accept the risk. Here every outcome is a cost, so the best choice is the cheapest one on average.
Costs $1.3M in the worst case
Above 15.9%, basic protection becomes the cheaper choice.
What the tree shows: doing nothing looks free but has the highest expected cost, $720K, and a 10% chance of a $3.0M breach.
Most business cases end in one number: an NPV or an expected profit. But two options with similar averages can carry very different risks, and the option with the best average can still lose money most of the time.
Take the gene therapy example from the Pharma & biotech tab above. Developing the therapy ex vivo has the best expected value, $187.9M. Lay out what could actually happen, though, and it loses money 79 times in 100. The big payoff comes about one time in five.
Licensing it out is worth $124.3M on average and never loses money. Which is right depends on how much risk the company can carry. ModelChoice puts both pictures in front of the people deciding, so the choice is made knowingly.
How ModelChoice compares the risk of each option →
From ModelChoice’s Gene Therapy Development example. Each square is one in a hundred outcomes.
Outcomes are net present values. Squares are rounded to whole ones.
One tree, built once in Excel, answers all six. Each card links to the feature that does it, with a live example on the features page.
Lay the decision out as a tree on the worksheet, with probabilities and payoffs in cells. ModelChoice finds the best path, highlights it and finds it again whenever an input changes.
See every outcome each option can lead to, and how likely it is. Compare options on their worst case and their chance of a loss, not only on the average.
Tornado and spider charts show which estimates matter. Break the Decision shows how far each one can move before the best choice changes.
Before you pay for a survey, a pilot or a study, value of information tells you the most it can be worth to this decision.
Score each option on quality, safety, delivery or anything else that matters, with weights the team agrees, next to its financial payoff.
A one-page Decision Brief gives the recommendation, how robust it is, the chance of a loss and the estimates that drive it, ready to export as a PDF.
Decision trees earn their keep when a choice is big, hard to reverse and made in stages, with facts arriving along the way. ModelChoice comes with 59 worked example models from these industries. Open one, change the numbers, and you have a first model of your own decision.
Exploration is a chain of expensive bets, with information arriving in stages: seismic, a well, appraisal, development. A tree values each step and shows when a survey or an appraisal well pays for itself.
ExamplesOffshore Drilling Campaign · Oil Well Appraisal · Shale Gas Development · EOR Strategy Decision · LNG Export Terminal · Refinery Upgrade Decision
Drug development is staged by nature: every phase is a go or no-go with a known cost and an uncertain result. Teams use trees for risk-adjusted NPV, for partnering and licensing, and to rank a pipeline.
ExamplesClinical Trial Go-No-Go · Pharmaceutical Pipeline · Drug Development Portfolio · Gene Therapy Development · Vaccine Platform Investment · Biosimilar Launch Strategy · Drug Formulation Selection (MCDA)
Deal structure, bid strategy and exit timing, with regulators and markets in between.
ExamplesM&A Acquisition Decision · Private Equity Buyout · Venture Capital Investment · Venture Fund Returns · Distressed Debt Investment · Sovereign Wealth Fund Allocation
Product launches, reinsurance structures and catastrophe exposure, where the worst year matters as much as the average one.
ExamplesInsurance Product Launch · Reinsurance Treaty Decision · Insurance Catastrophe Model · Life Insurance Portfolio
Large capital projects with permit, price and resource risk, where building in phases can be worth more than building big.
ExamplesRenewable Energy Project · Wind Farm Portfolio · Renewable Energy Site (MCDA) · Water Treatment Plant · Carbon Credit Investment · Mining Exploration · Mining Feasibility Study
Build now or wait, build big or in stages, with demand and construction risk on every path.
ExamplesData Center Build · Real Estate Development · Real Estate Development (MR) · Sports Stadium Development · Telecom Network Rollout
Capacity, supplier and automation choices, often scored on cost, quality and risk together.
ExamplesSupply Chain Resilience · Manufacturing Capacity · Supplier Selection (MCDA) · Warehouse Automation (MCDA) · Fleet Vehicle Selection (MCDA) · Fleet Electrification
Build or buy, launch in full or start small, and how much to spend on security.
ExamplesStartup Product Launch · Software Platform Build vs Buy · Cybersecurity Investment · Technology Stack Selection (MCDA) · Autonomous Vehicle R&D · EdTech Platform Launch
Treatment, equipment and site choices with several criteria and many stakeholders to satisfy.
ExamplesHospital Equipment Decision · Hospital Treatment Selection (MCDA) · Environmental Remediation (MCDA) · University Course Portfolio (MCDA) · Office Location Decision (MCDA)
Greenlights, expansions, new routes and crop plans, decided before demand or the weather is known.
ExamplesMovie Studio Greenlight · Restaurant Chain Expansion · Franchise Expansion · Retail Store Rollout · Airline New Route · Agricultural Crop Planning · Agricultural Irrigation
Organisations that use Vose software
ModelChoice is made by Vose Software, the company behind ModelRisk and Tamara, which has built risk analysis software since 2008. Its founder, David Vose, wrote Risk Analysis: A Quantitative Guide.
ModelChoice is sold as an annual licence for a single computer or for a network, in one edition with every analysis on this page. Volume discounts and bundles with ModelRisk and Tamara lower the price. Work out your own quote with the price calculator.
ModelChoice is a decision analysis add-in for Microsoft Excel. Teams use it to compare the options for a big, uncertain decision, such as whether to drill a well, which drug candidate to fund, whether to launch a product or how to structure a deal. It finds the option with the best expected value, shows the risk that comes with each option, tests what would change the answer and works out whether more information is worth paying for.
A spreadsheet NPV gives one number for one scenario. A decision tree lays out the choices you can make, the events that could follow with their probabilities, and the later decisions you would take once you know more. That captures the value of being able to change course, such as stopping a programme after a poor trial result, which a single NPV leaves out.
No. If you can build a spreadsheet, you can build a tree in ModelChoice: you add decisions and chance events from the ribbon and type the numbers or link them to cells. Each of the 59 example models opens with a description of the problem it solves, so you can start from the one closest to yours.
Three things: the options open to you, the uncertain events that could follow each one with a probability for each, and what each outcome is worth. Rough estimates are fine to start with. ModelChoice then shows which of them could change the answer, so you know where better information would pay.
Use a decision tree when the question is which way to go, especially when there are choices to make along the way: test first or go ahead, continue or stop after a result. Use a Monte Carlo simulation when the question is how wide the range of outcomes is, such as the spread of a project’s cost. When a decision needs both, ModelChoice can run a ModelRisk simulation through the tree, as the features page shows.
It is built for the people who prepare business cases in Excel: analysts, engineers, project and portfolio managers and finance teams. The people who approve the decision work from what it produces: the tree with the best path marked, the charts and a one-page Decision Brief.
Download the full product free for 15 days, open an example from your industry and change the numbers to match your case.