McKinsey Solve: The Ultimate Guide to Passing in 2026

Author: Taylor Warfield, Former Bain Manager and Interviewer

McKinsey Solve assessment

The McKinsey Solve is an online assessment that screens candidates before case interviews. It takes about 110 minutes and includes three games: Ecosystem Building, Redrock Case Study, and Sea Wolf. You'll need to pass this test to get an interview at McKinsey.

This is the most comprehensive guide available. We cover every game in detail, explain exactly how scoring works, and give you specific strategies based on what actually works for candidates.

If you want to practice the McKinsey Solve beforehand to ensure no surprises on test day, check out PSG Secrets' McKinsey Solve simulation. These exercises simulate the actual exercises you'll work through on test day.

Table of Contents

What is the McKinsey Solve?
The History and Evolution of the Test
Where Does the Solve Fit in McKinsey's Process?
Who Takes the McKinsey Solve?
How the McKinsey Solve is Scored
What Skills Does the Solve Test?
Game 1: Ecosystem Building (Complete Guide)
Game 2: Redrock Case Study (Complete Guide)
Game 3: Sea Wolf / Ocean Cleanup (Complete Guide)
Legacy Games (No Longer Used)
How to Prepare for the McKinsey Solve
Test Day: What to Expect and How to Perform
What Happens After You Take the Solve?
How the Solve Compares to Other MBB Assessments
Frequently Asked Questions

What is the McKinsey Solve?

The McKinsey Solve is a gamified assessment that evaluates how you think, not what you know. You'll play through ecology-themed scenarios that test your problem-solving abilities. There's no business knowledge required and you don't need any gaming experience.

McKinsey uses this test to screen thousands of applicants efficiently. It replaced the old McKinsey Problem Solving Test (PST) because it's harder to game and provides richer data on how candidates actually think.

The test consists of three separate games, each lasting 35 minutes:

Ecosystem Building: Create a self-sustaining food chain by selecting species that can survive together in a specific environment
Redrock Case Study: Analyze data about wildlife populations, perform calculations, and complete a business-style report
Sea Wolf (Ocean Cleanup): Select microbes to treat polluted ocean sites based on their biological characteristics

You'll take the test at home on your own computer. McKinsey doesn't monitor you through your webcam, so you're free to use pen, paper, calculators, and even a second computer if you find it helpful.

A critical point most candidates miss: The Solve doesn't just score your final answers. It tracks every click, every scroll, and every decision you make. Your problem-solving process matters just as much as getting the right answer.

Is the Solve the Same as the PSG, Imbellus Game, or Digital Assessment?

Yes. All these names refer to the same test:

McKinsey Solve: The current official name
Problem Solving Game (PSG): The name used from roughly 2020-2022
McKinsey Digital Assessment: Another term McKinsey has used in the past
Imbellus Game/Test: Named after Imbellus, the company that developed the original version

You'll see all these names used interchangeably online, in forums, and even in older McKinsey materials. Don't let the different names confuse you. They all describe the same assessment.

The History and Evolution of the McKinsey Solve

Understanding how the McKinsey Solve has evolved helps you appreciate why it works the way it does and what McKinsey is really looking for.

2017: Initial Launch

McKinsey partnered with Imbellus to develop a new screening assessment. The test debuted as a pilot in London and select other offices. The original version included Ecosystem Building and other experimental games.

2019: Global Rollout

The assessment expanded to McKinsey offices worldwide. It began replacing the traditional Problem Solving Test (PST) as the primary screening tool.

2020: Rebranding to PSG

McKinsey officially called the test the "Problem Solving Game" and made it a standard part of the recruiting process for most candidates globally.

2021: Roblox Acquires Imbellus

Gaming giant Roblox purchased Imbellus, the company behind the assessment technology. This didn't immediately change the test, but it signaled ongoing evolution.

2022: Rebranding to Solve

McKinsey rebranded the assessment as "Solve," the name it uses today. The interface and games received updates.

2023: Introduction of Redrock

The Redrock Case Study replaced the Plant Defense game. This was a significant shift because Redrock is much more business-oriented and quantitative, resembling a traditional consulting case more than a video game.

2024-2025: Addition of Sea Wolf

McKinsey introduced the Sea Wolf game (also called Ocean Cleanup) as the third game in the assessment. This extended the total test time from roughly 70 minutes to 110 minutes.

Where Does the Solve Fit in McKinsey's Process?

The McKinsey interview process follows a predictable sequence. Understanding where the Solve fits helps you plan your preparation. Here is the complete McKinsey interview process.

Step 1: Networking

Before applications open, McKinsey hosts networking events at target schools and in major cities. Attending isn't required, but it helps if you're from a non-target school or applying as an experienced hire.

Step 2: Submit Your Application

Your application typically includes:

Resume
Cover letter
Academic transcripts
Referrals (if you have them)

Step 3: Initial Resume Screen

McKinsey does a quick review to filter out incomplete applications and candidates who clearly don't meet basic requirements. Most people who apply and meet the basic qualifications will advance to the Solve.

Step 4: Take the McKinsey Solve

You'll receive an email invitation with instructions and a deadline. Most candidates get 1-2 weeks to complete the test, though some report getting as little as 3 days.

Plan to be ready within one week of submitting your application.

Step 5: Detailed Application Review

After you complete the Solve, McKinsey reviews your test results alongside your resume and other application materials. This is where the real screening happens.

Only about 30% of candidates who take the Solve advance to interviews. In competitive offices or recruiting cycles, the rate can be even lower.

Step 6: First Round Interviews

If you pass the Solve and application review, you'll be invited to first round interviews. This typically involves two 40-60 minute interviews, each including a case study and behavioral questions.

For students, first round interviews usually happen on campus. For experienced hires, they happen at the McKinsey office you're applying to.

Step 7: Final Round Interviews

If you pass first round, you'll return for final round interviews. This typically involves 2-3 interviews with more senior consultants or partners. The cases may be less structured and more discussion-based.

Step 8: Offer Decision

After final rounds, you'll typically hear back within a few days to a week. Some candidates get same-day offers. Others wait longer, especially during busy recruiting seasons.

Who Takes the McKinsey Solve?

Nearly everyone applying to McKinsey for a consulting role takes the McKinsey Solve. This includes:

Undergraduate students applying for Business Analyst positions
MBA students applying for Associate positions
PhD/JD/MD candidates applying through advanced degree recruiting
Experienced hires applying from industry
Internal transfers in some cases

The only candidates who might skip the McKinsey Solve are:

Very senior hires (Partner-level or executive roles)
Candidates with exceptionally strong referrals or connections
Candidates in offices that are piloting different processes

Even these exceptions are office-dependent. The safest assumption is that you will take the McKinsey Solve if you're applying to McKinsey for any consulting position.

Does the Test Differ by Role or Region?

No. The McKinsey Solve is the same regardless of:

Which office you're applying to
What position level you're targeting
Whether you're applying as a generalist or specialist
Your educational or professional background

A Business Analyst candidate in Chicago takes the same test as an Associate candidate in London or a McKinsey Digital applicant in Singapore.

This standardization is intentional. McKinsey wants to evaluate everyone's problem-solving abilities on the same playing field.

How the McKinsey Solve is Scored

This is where the Solve gets interesting. Unlike a traditional test, you're not just scored on whether you got the right answer. McKinsey evaluates both what you produce and how you produce it.

Product Score

The product score measures the quality of your final outputs:

In Ecosystem Building:

Does your ecosystem sustain itself?
Do all species survive according to the eating rules?
Does your selected location meet all terrain requirements?

In Redrock:

Are your calculations correct?
Did you fill in the report accurately?
Are your mini-case answers right?

In Sea Wolf:

Do your selected microbes meet the site requirements?
Are the attribute averages within range?
Did you include the desirable trait and exclude the undesirable trait?

Process Score

Here's what makes the Solve different from any test you've taken before. The assessment tracks everything you do:

Every mouse click
Every scroll
How long you spend on each section
When you change your mind
How you navigate between options
What order you consider information
Whether you use the provided tools (calculator, journal)

McKinsey uses this behavioral data to evaluate your problem-solving process. A high process score indicates:

Organized thinking: You approach problems systematically rather than randomly
Efficient navigation: You don't waste time clicking around aimlessly
Deliberate decisions: You commit to choices rather than constantly backtracking
Tool usage: You use the provided calculator and journal appropriately
Time management: You pace yourself well across the assessment

How the Scores Combine

Your final score combines both product and process components. McKinsey doesn't publish the exact weighting, but both matter significantly.

Here's the key insight: You can get a mediocre score even with correct answers if your process was messy. And you can get a decent score with a few wrong answers if your process was exemplary.

Think about it from McKinsey's perspective. They want consultants who can solve problems efficiently and systematically, not just people who stumble into correct answers through trial and error.

What Score Do You Need to Pass?

McKinsey doesn't publish passing scores or cutoffs. What we know from candidate reports:

Scores are compared to high performers at McKinsey
The closer your pattern matches successful consultants, the better you look
Office and recruiting cycle competitiveness affects the effective threshold
Your Solve score is weighed alongside your resume and other application materials

A great Solve score won't guarantee an interview if your resume is weak. And a strong resume won't save you if you bomb the Solve. You need both.

Can You Get Your Score?

McKinsey will provide your score upon request after you complete the assessment. They'll also tell you your percentile range (which quartile you fall into compared to other candidates).

However, they won't tell you whether you passed or how close you were to the threshold. You'll only know for sure when you either receive an interview invitation or a rejection.

What Skills Does the McKinsey Solve Test?

McKinsey designed the Solve to assess five cognitive abilities. Understanding these helps you approach the games with the right mindset.

1. Critical Thinking

Definition: Making judgments based on objective analysis of information

How it appears in the games:

Identifying which data points actually matter in Redrock
Recognizing patterns in species characteristics in Ecosystem Building
Evaluating which microbes best fit site requirements in Sea Wolf

What good looks like:

You don't take information at face value
You question assumptions
You identify what's relevant versus what's noise
You draw logical conclusions from evidence

2. Decision Making

Definition: Choosing the best course of action, especially under time pressure or with incomplete information

How it appears in the games:

Selecting species for your ecosystem when you can't test every combination
Choosing which data to collect in Redrock when you can't collect everything
Picking microbes when no combination is perfect in Sea Wolf

What good looks like:

You make decisions confidently rather than endlessly deliberating
You commit to a direction and execute
You adapt when new information suggests a course correction
You prioritize effectively when you can't do everything

3. Metacognition

Definition: Deploying appropriate strategies to make learning and problem-solving easier

How it appears in the games:

Taking notes during the tutorial to organize your approach
Creating a systematic method for evaluating species
Using the journal and calculator appropriately in Redrock

What good looks like:

You have a clear strategy before you start executing
You use tools and techniques that make the problem simpler
You recognize when your approach isn't working and adjust
You reflect on your process, not just outcomes

4. Situational Awareness

Definition: The ability to interpret and predict an environment

How it appears in the games:

Understanding how changing one species affects the entire ecosystem
Predicting how wolf populations will respond to different scenarios
Anticipating how microbe selections will impact treatment effectiveness

What good looks like:

You consider second and third-order effects
You anticipate consequences before they happen
You maintain awareness of the big picture while working on details
You understand relationships between different elements

5. Systems Thinking

Definition: Understanding complex causal relationships between elements of a system

How it appears in the games:

Grasping how the food chain works as an interconnected system
Understanding the relationships between wolves, elk, and plants in Redrock
Seeing how microbe characteristics combine to create treatment effects

What good looks like:

You see connections that aren't immediately obvious
You understand feedback loops and interdependencies
You can trace cause and effect through multiple steps
You recognize that changing one thing affects many others

Which Skills Matter Most?

Research on the Imbellus assessment methodology suggests that Critical Thinking, Situational Awareness, and Systems Thinking are the foundational skills. Most candidates can demonstrate these to some degree.

Decision Making and Metacognition are the differentiators. These advanced skills separate great candidates from good ones. They're also harder to develop quickly.

The good news: These skills are interconnected. Improving one area often improves others. And practicing with realistic simulations builds all five simultaneously.

Game 1: Ecosystem Building (Complete Guide)

Ecosystem Building is typically the first game you'll play. You need to create a stable ecosystem by selecting eight species that can survive together in a specific location.

The Setup

You'll be placed in one of several possible environments:

Marine/Reef: Underwater ecosystem with fish, coral, and aquatic plants
Mountain/Alpine: Terrestrial ecosystem with mammals, birds, and plants
Jungle/Rainforest: Another terrestrial variation

The specific environment doesn't change the core mechanics. The same strategies apply regardless of whether you're working with fish or wolves.

What You're Given

1. 39 species total, divided into:

Producers (plants, coral, algae): Provide calories but don't need to eat
Consumers (animals): Need to eat other species to survive
Apex predators: Animals at the top of the food chain with no natural predators

2. Multiple locations with different environmental conditions

3. Detailed information about each species including:

Terrain requirements (ranges for depth, temperature, etc.)
Calories needed (how much the species needs to eat)
Calories provided (how much nutrition the species provides when eaten)
Food sources (what the species can eat)
Predators (what eats this species)

Your Objective

Select exactly 8 species (sometimes 7, depending on the version) that form a self-sustaining food chain. Then choose a location where all 8 species can survive based on their terrain requirements.

For the ecosystem to be sustainable:

Every consumer must get enough calories from available food sources
No species can be completely consumed (driven to extinction)
All species must survive in the chosen location's environmental conditions

The Terrain Specifications

Each environment has 7-8 terrain variables. Here are the common ones:

Marine/Reef Environment:

Depth (how deep in the water)
Water current (speed of water flow)
Water clarity (visibility)
Temperature
Salt content (salinity)
Dissolved oxygen
Wind speed (surface conditions)

Mountain/Alpine Environment:

Elevation (altitude)
Soil pH (acidity/alkalinity)
Precipitation (rainfall)
Temperature
Wind speed
Air pressure
Humidity
Cloud height

Each species has requirements for 2-4 of these variables (expressed as acceptable ranges). The species can only survive in locations where all its requirements are met.

Not all terrain variables matter. Some are red herrings designed to waste your time. Focus only on the variables that actually appear in species requirements.

The Eating Rules (This is Crucial)

The food chain follows specific rules that determine which species survive. You must understand these rules to build a working ecosystem.

Rule 1: Species eat only once per cycle. Each animal gets one chance to eat. It either gets enough calories or it dies

Rule 2: Eating order is determined by calories provided. The species that provides the MOST calories when eaten gets to eat FIRST. This continues in descending order

Rule 3: Predators eat their highest-calorie prey first. When a species eats, it targets the available prey that provides the most calories

Rule 4: Predators consume exactly what they need. A species takes calories equal to its "calories needed" value. No more, no less

Rule 5: Equal-calorie prey gets eaten proportionally. If multiple prey items provide the same calories, the predator eats an equal proportion from each

Rule 6: Species die if they don't get enough calories. If a species can't meet its calorie needs from available prey, it dies

Rule 7: Species die if they're completely consumed. If a species' total calories provided is consumed by predators, it goes extinct

Let’s take a look at an example of the eating rules in a simple food chain.

Species in ecosystem:

Algae A: Provides 1000 cal, needs 0 cal (producer)
Fish B: Provides 800 cal, needs 300 cal, eats Algae A
Fish C: Provides 600 cal, needs 200 cal, eats Algae A
Shark D: Provides 400 cal, needs 500 cal, eats Fish B and Fish C

Eating sequence:

Step 1: Algae A goes first (highest calories provided at 1000). As a producer, it doesn't need to eat. It just exists.

Step 2: Fish B goes second (800 cal provided). It needs 300 cal and eats Algae A.

Algae A now has 1000 - 300 = 700 cal remaining

Step 3: Fish C goes third (600 cal provided). It needs 200 cal and eats Algae A.

Algae A now has 700 - 200 = 500 cal remaining
Algae A survives (still has calories left)

Step 4: Shark D goes fourth (400 cal provided). It needs 500 cal and can eat Fish B or Fish C.

Fish B provides more calories (800 vs 600), so Shark D eats Fish B first
Shark D takes 500 cal from Fish B
Fish B now has 800 - 500 = 300 cal remaining
Fish B survives

Result: All species survive. This ecosystem is sustainable.

Now imagine if Shark D needed 900 calories instead of 500:

Shark D would eat all 800 cal from Fish B (Fish B goes extinct)
Shark D would still need 100 more cal
Shark D would eat 100 cal from Fish C
Fish C survives with 500 cal remaining
But Fish B is extinct, so the ecosystem fails

The Three Groups of 13

Here's something most guides don't explain clearly: The 39 species are actually organized into three groups of 13 species each.

Each group contains:

3 producers
10 consumers (including apex predators)

Each group shares terrain requirements. Species within a group can live in the same locations. Species from different groups generally cannot.

This means you're really choosing between three potential ecosystems, not mixing and matching from all 39 species. Once you identify which group you're building from, your options narrow considerably.

Step-by-Step Strategy for Ecosystem Building

Step 1: Identify the three groups (2-3 minutes)

Look at terrain specifications. You'll notice species cluster into three distinct groups based on their requirements. Don't waste time trying to combine species from different groups.

Step 2: Evaluate the producers in each group (2-3 minutes)

For each group, look at:

Total calories provided by all three producers combined
How many consumers can eat the producers
Whether any producer is eaten by multiple consumers

The group with the strongest producer base is usually easiest to build around. Look for high total calories and multiple eating pathways.

Step 3: Select your target group

Choose the group that looks most promising. You can always switch if your first choice doesn't work, but starting with the right group saves time.

Step 4: Add all producers to your ecosystem (1 minute)

Include all three producers from your chosen group. They form the foundation of your food chain and provide calories without needing any.

That's 3 of your 8 species selected.

Step 5: Identify your apex predator (2-3 minutes)

Look for consumers that nothing else eats (apex predators). Choose the one with:

Lowest calorie needs
Ability to eat multiple prey species

Low calorie needs means your apex predator is easier to feed. Multiple prey options provides flexibility.

That's 4 of your 8 species selected.

Step 6: Build the middle of your food chain (5-7 minutes)

Now add consumers that connect your producers to your apex predator. Prioritize species that:

Can eat producers without depleting them
Provide high calories (to feed the apex predator)
Are eaten by multiple other consumers (creates flexibility)

Work through the eating rules mentally (or on paper) as you add each species. Make sure everything still balances.

Step 7: Verify the food chain (3-5 minutes)

Before submitting, trace through the eating rules completely:

Start with the highest calories-provided species
Follow the eating sequence
Confirm every species survives
Confirm no species goes extinct

If something doesn't work, swap individual species rather than starting over.

Step 8: Select your location (1-2 minutes)

Only after confirming your 8 species, look at locations. Find one where the environmental conditions meet ALL terrain requirements for ALL your species.

This is usually straightforward because species within a group share similar requirements.

Common Ecosystem Building Mistakes

Trying to combine species from different groups: The terrain requirements won't match. You'll waste time before realizing it's impossible

Picking an apex predator with high calorie needs: This forces you to include more prey species and creates more opportunities for the chain to break

Ignoring the eating order rules: The sequence matters. A food chain that looks fine at first glance can fail when you trace through the actual eating order

Not verifying before submitting: Always trace through the complete eating sequence before you submit. A single mistake means a failed ecosystem

Spending too much time on location selection: Location is the easy part. Get your species right first, then find a compatible location

Random trial and error: Clicking around randomly hurts your process score even if you eventually get the right answer. Be systematic

Time Allocation for Ecosystem Building

You have 35 minutes. Here's a recommended allocation:

Tutorial and orientation: 3-5 minutes (doesn't count against time)
Identify groups and evaluate producers: 5-7 minutes
Build food chain: 15-18 minutes
Verify and adjust: 5-7 minutes
Select location and submit: 3-5 minutes

Most candidates report that Ecosystem Building feels less time-pressured than Redrock or Sea Wolf. Use the extra comfort to be thorough rather than rushing.

You can practice using these strategies for Ecosystem Building through PSG Secrets' McKinsey Solve simulation.

Game 2: Redrock Case Study (Complete Guide)

The Redrock Case Study is the most business-like of the three games. It resembles a traditional consulting assessment with data analysis, calculations, and report writing.

This game has the tightest time pressure. Many candidates struggle to finish everything in 35 minutes.

The Scenario

You're working on the Island of Redrock, a nature preserve with various wildlife populations. The most common scenarios involve:

Wolf/Elk Scenario: The island has four wolf packs in different geographic areas. These wolves prey on elk. Your job is to optimize wolf pack sizes so both wolves and elk can sustainably coexist.

Plant Biodiversity Scenario: You're evaluating strategies to increase plant biodiversity (grasses, trees, shrubs) by a certain percentage.

Regardless of the specific scenario, the structure and approach remain the same.

Structure of the Game

Redrock has two distinct parts, and you must complete both within 35 minutes total.

Part 1: The Study (three phases)

Investigation: Collect data from case materials
Analysis: Answer quantitative questions using a calculator
Report: Complete written and visual reports

Part 2: Mini Cases

6 independent questions
Mix of calculation, reasoning, and visualization questions
Must answer sequentially (no skipping)

You move through these phases in order and cannot go back once you advance.

Part 1, Phase 1: Investigation

What you see:

A case description with background information
Charts, graphs, and tables with data
Text describing objectives and calculation methods

What you do:

Read through the materials to understand the case objective
Identify key data points you'll need for calculations
Drag important items to your Research Journal

Only about 10-15% of the data shown is actually needed. Your job is to identify what matters and ignore the rest.

Types of information to collect:

Case objectives: What exactly are you trying to determine or recommend?
Calculation instructions: Any formulas or methods described in the text
Numerical data: Specific numbers you'll need for calculations: population figures, percentages, growth rates, and targets or thresholds.
Relationship data: How different variables connect: predator-prey ratios, growth dependencies, and constraints.

How the journal works:

Drag any movable data point into your Research Journal
Each item becomes a labeled card
You can edit labels for clarity (recommended)
Only journal items can be used in the calculator later

Investigation strategy:

Read the case objective first. Everything else should support answering this objective.
Skim for calculation instructions. These tell you what math you'll need to do and what numbers to collect.
Collect data systematically. Don't just grab everything. Ask yourself: "Will I need this number for a calculation?"
Edit labels as you go. The auto-generated labels aren't always helpful. Clear labels save time later.
Don't over-collect. More isn't better. Grabbing irrelevant data wastes time and clutters your workspace.

Time allocation: 8-10 minutes maximum

Part 1, Phase 2: Analysis

What you see:

Three quantitative questions with sub-parts
An on-screen calculator
Access to your Research Journal

What you do:

Use the calculator to perform required calculations
Input numbers from your journal
Answer each question and sub-question

Calculator functionality:

Basic operations: addition, subtraction, multiplication, division
Percentage calculations

That's it. No advanced functions.

McKinsey logs all your calculator activity. This affects your process score. Do ALL calculations on the calculator, even simple ones you could do in your head.

Here are the types of calculations you'll encounter:

1. Basic arithmetic:

Adding or subtracting populations
Multiplying quantities by percentages
Dividing to find ratios

2. Percentage calculations

Finding what percentage one number is of another
Calculating percentage increases or decreases
Converting between percentages and absolute numbers

3. Growth calculations

Simple growth: New = Original × (1 + growth rate)
Compound growth: New = Original × (1 + rate)^periods
Finding growth rates from before/after numbers

4. Weighted averages

Combining values with different weights
Calculating overall rates from component rates

5. Percentage points vs. percentages

A common trick: "Increase by 5 percentage points" means adding 5 to the percentage (e.g., 10% becomes 15%).

This is different from "increase by 5%" which means multiplying (e.g., 10% becomes 10.5%).

Analysis strategy:

Read each question completely before calculating. Understand what's being asked.
Identify which journal data you need. Pull up the relevant cards.
Plan your calculation before touching the calculator. What operations, in what order?
Use the calculator for everything. Even 100 + 50. Your process is being recorded.
Double-check your work. Calculator errors are costly and avoidable.
Save answers to your journal. You'll need them for the Report phase.

Time allocation: 10-12 minutes

Part 1, Phase 3: Report

What you see:

A pre-written report with blanks to fill in
A section for creating a chart

What you do:

Fill in numerical values from your calculations
Add comparison words (higher, lower, equal to)
Select and create an appropriate chart

Written report: The report has blanks where you insert:

Numbers from your Analysis calculations
Numbers from your Investigation data
Comparative terms based on your findings

Visual report: You'll select a chart type and input data to create a visualization. The options are typically:

Bar chart: Best for comparing quantities across categories. Use when comparing wolf populations across regions, comparing species counts

Line chart: Best for showing change over time. Use when displaying population trends, showing growth trajectories

Pie chart: Best for showing parts of a whole. Use when displaying percentage breakdowns, showing composition

Chart selection tip: Think about the insight you're trying to communicate. What would make the point clearest to a reader?

Report strategy:

Fill in numbers first. These should come directly from your Analysis answers and Investigation data.
Add comparison words based on your calculations. If population A is 500 and population B is 300, A is "higher than" B.
Create your chart thoughtfully. Select the type that best matches the data and objective.
Review for consistency. Numbers in the report should match your calculations.

Time allocation: 5-7 minutes

Part 2: Mini Cases

After completing the Study, you'll face 6 independent questions. These are thematically related to the Island of Redrock but don't use data from the Study.

Here are the different types of questions:

1. Calculation questions

Perform calculations like weighted averages, probabilities, or statistical measures.

Example: "What is the weighted average lifespan if 40% of animals live 5 years and 60% live 8 years?" Answer: (0.40 × 5) + (0.60 × 8) = 2 + 4.8 = 6.8 years

2. Probability questions

Apply probability concepts including unions, intersections, and complements.

Example: "If there's a 30% chance of rain in Location A and 40% chance in Location B (independent events), what's the probability of rain in at least one location?"

Answer: P(A or B) = P(A) + P(B) - P(A and B) = 0.30 + 0.40 - (0.30 × 0.40) = 0.58 or 58%

3. Formula identification questions

Select the correct formula or calculation method.

Example: "Which formula calculates the change in population?" Options might include various mathematical expressions.

4. Verbal reasoning questions

Determine whether statements are true or false based on given information.

Technique: Read carefully. These often have tricky wording or similar-sounding options.

5. Data interpretation questions

Extract information from charts, tables, or text.

Techniques involved:

Finding median: Middle value in sorted data
Finding mode: Most frequent value
Estimating mean: Center of the distribution

6. Visualization questions

Select the best chart type to display given data.

Technique: Match the chart type to what you're trying to show (comparison, trend, composition).

Mini case strategy:

Budget at least 10 minutes for this section. Don't let the Study phases consume all your time.
Read each question carefully. Trick answers often look similar to correct ones.
Work quickly but don't rush. Careless errors cost more than taking an extra 30 seconds.
Don't get stuck. If a question is taking too long, make your best guess and move on.
Use elimination. Rule out obviously wrong answers first.

Time allocation: 10-12 minutes

Overall Time Management for Redrock

This is the most time-pressured game. Here's how to allocate your 35 minutes:

Phase	Time	Cumulative
Investigation	8-10 min	8-10 min
Analysis	10-12 min	18-22 min
Report	5-7 min	23-29 min
Mini Cases	6-12 min	35 min

If you're running behind:

Streamline Investigation (collect only essentials)
Move faster through Analysis (but still be accurate)
Keep Report concise
Never sacrifice Mini Cases (they're worth significant points)

Common Redrock Mistakes

Over-collecting data in Investigation: You don't need everything. Grabbing too much wastes time and clutters your workspace

Doing mental math: Use the calculator for everything. Your process is being recorded

Not saving answers to the journal: You need your Analysis answers for the Report phase. Save them

Confusing percentage points with percentages: "Increase by 5 percentage points" (add 5) is different from "increase by 5%" (multiply by 1.05)

Running out of time for Mini Cases: Budget your time. The Mini Cases are worth significant points

Not reading questions carefully: Many wrong answers come from misreading, not miscalculating

You can practice using these strategies for Redrock through PSG Secrets' McKinsey Solve simulation.

Game 3: Sea Wolf / Ocean Cleanup (Complete Guide)

Sea Wolf is the newest game, added in 2024. You'll select microbes to treat polluted ocean sites based on their biological characteristics.

This game combines pattern matching with optimization. It's different from both Ecosystem Building and Redrock.

The Setup

You'll treat three separate ocean sites. For each site, you select three microbes that together form an effective treatment.

The process is identical for all three sites, just with different parameters each time.

Key Concepts

Characteristics: Every site and microbe are defined by two types of characteristics:

Attributes: Three numerical values ranging from 1 to 10. Examples: permeability, mobility, energy

Traits: Four binary (yes/no) characteristics. Examples: light sensitivity, aerobic, heat resistance

Site Requirements: Each site specifies:

Target ranges for three attributes (e.g., permeability 4-6)
One desirable trait (must be present in at least one microbe)
One undesirable trait (must be absent from all microbes)

Note: Sometimes only the desirable OR undesirable trait is shown, not both.

Microbes: Each microbe has:

Three attribute values
One trait

Your Goal: Select three microbes whose combined characteristics match the site requirements.

Scoring for Each Site

Your treatment effectiveness is scored out of 100%:

Criterion	Points	Requirement
Attribute 1 in range	20%	Average of 3 microbes falls within target
Attribute 2 in range	20%	Average of 3 microbes falls within target
Attribute 3 in range	20%	Average of 3 microbes falls within target
Desirable trait present	20%	At least one microbe has it
Undesirable trait absent	20%	No microbe has it

Total possible: 100%

A perfect score isn't always achievable depending on the microbes available. Aim for the highest score possible given your options.

The Four Steps of Sea Wolf

Each site follows the same four-step process:

Step 1: Define Selection Criteria

Choose two characteristics (attributes or traits) to filter the initial microbe pool.

If you select an attribute, define a range of 2 (e.g., 4-6)
If you select a trait, choose yes or no

These selections influence which microbes appear in later steps.

Strategy for Step 1:

Choose one attribute and one trait
Pick an attribute with an extreme target range (far from 5.5, the midpoint)
Pick the desirable or undesirable trait (whichever is shown)

Step 2: Assign or Reject Microbes

You'll see 10 microbes and information about Sites 1 and 2.

For each microbe, decide:

Assign to Site 1
Assign to Site 2
Reject

This affects your prospect pool for the current site.

Strategy for Step 2: Assign a microbe to Site 1 if:

At least 2 of 3 attributes fall within the site's target ranges
It does NOT have the undesirable trait
Ideally, it has the desirable trait

If a microbe doesn't fit either site well, reject it.

Step 3: Build Your Prospect Pool

You start with 6 microbes. You'll add 4 more by choosing one from each set of 3 options.

Your final prospect pool will have 10 microbes.

Strategy for Step 3: For each choice of 3:

Eliminate any with the undesirable trait (immediate disqualifier)
Check for the desirable trait (bonus if present)
Count how many attributes fall in range
Choose the best combination of trait and attribute fit

Keep track of what's already in your pool. If you already have a microbe with the desirable trait, prioritize attributes.

Step 4: Select Final 3 Microbes

From your 10-microbe pool, choose 3 that together create the best treatment.

Remember:

Attribute values are AVERAGED across the 3 microbes
The average must fall within the site's target range
At least one microbe must have the desirable trait
NO microbe can have the undesirable trait

Strategy for Step 4:

First, eliminate all microbes with the undesirable trait. They can't be part of your treatment.
Identify microbes with the desirable trait. You need at least one of these.
Calculate attribute averages for promising combinations.

Math shortcut: Instead of dividing by 3 to find averages, multiply the target range by 3.

Example:

Target range: 4-6
Acceptable sum of 3 microbes: 12-18 (instead of average 4-6)

This is faster than dividing each time.

Test combinations until you find one that:

Includes at least one microbe with desirable trait
Excludes all microbes with undesirable trait
Has attribute sums within 3× the target range (or averages within the target range)

A Worked Example

Site requirements:

Attribute A: 3-5
Attribute B: 6-8
Attribute C: 2-4
Desirable trait: Heat resistant
Undesirable trait: Light sensitive

Your prospect pool (after Steps 1-3):

Microbe	A	B	C	Trait
M1	4	7	3	Heat resistant
M2	3	8	2	Aerobic
M3	5	6	4	Aerobic
M4	2	9	5	Light sensitive
M5	4	7	3	Aerobic

Step 1: Eliminate light sensitive

Remove M4 from consideration.

Step 2: Ensure desirable trait

M1 is our only heat-resistant option. Must include M1.

Step 3: Find two more for optimal averages

Target sums (range × 3):

A: 9-15
B: 18-24
C: 6-12

Try M1 + M2 + M3:

A: 4 + 3 + 5 = 12 ✓ (within 9-15)
B: 7 + 8 + 6 = 21 ✓ (within 18-24)
C: 3 + 2 + 4 = 9 ✓ (within 6-12)

This combination scores 100%:

All three attributes in range: 60%
Desirable trait present (M1): 20%
Undesirable trait absent: 20%

Time Allocation for Sea Wolf

You have 35 minutes for all three sites.

Activity	Time
Site 1 (all 4 steps)	~12 min
Site 2 (all 4 steps)	~11 min
Site 3 (all 4 steps)	~12 min

The process gets faster as you learn the interface. Expect Site 1 to take longest.

Common Sea Wolf Mistakes

Including a microbe with the undesirable trait: This costs you 20% immediately. Double-check before finalizing

Forgetting to include the desirable trait: At least one of your three microbes must have it. Don't let strong attribute scores distract you

Calculating averages incorrectly: Use the sum shortcut (multiply range by 3) to avoid division errors

Not considering all combinations: With 10 microbes and 3 to select, there are 120 possible combinations. You can't test them all, but be systematic about the promising ones

Rushing through early steps: Steps 1-3 affect what's available in Step 4. Poor early choices limit your final options

You can practice using these strategies for Sea Wolf / Ocean Cleanup through PSG Secrets' McKinsey Solve simulation.

Legacy Games (No Longer Used)

McKinsey has experimented with several games over the years. These are no longer part of the current assessment, but you might encounter references to them online.

Plant Defense (Retired 2023)

A tower-defense style game where you protected a native plant from invasive species. You deployed predators and geographical barriers to stop the invaders.

This was replaced by the Redrock Case Study, which is more business-oriented and provides richer data on candidate abilities.

Disaster Management (Retired)

You identified the type of disaster (tsunami, volcanic eruption, etc.) based on environmental data, then relocated animal species to safe locations.

Disease Management (Retired)

You diagnosed which disease was affecting an animal population based on symptoms, then recommended treatment options.

Migration Management (Retired)

You guided a group of animals from one point to another while managing limited resources and maximizing survival.

Why were these games were retired?

McKinsey found that the current three games (Ecosystem Building, Redrock, Sea Wolf) provide sufficient data to evaluate candidates. The retired games may have had issues with reliability, predictive validity, or candidate experience.

Should you practice these old games?

No. Focus your time on the current three games. However, understanding that the test evolves reminds you to verify current information before test day.

How to Prepare for the McKinsey Solve

McKinsey suggests you don't need to prepare, but that's like saying you don't need to prepare for case interviews. Some people pass without preparation. But why take the risk when a McKinsey career could be worth millions?

What Preparation Actually Does

Familiarizes you with game mechanics: The games have complex interfaces and specific rules. Learning these beforehand means you don't waste test time figuring out how things work.

Lets you develop and refine strategies: Our strategies above give you a starting point, but you need to internalize them through practice. Strategies you've actually used stick better than ones you've read about.

Builds speed and accuracy: The Solve has real time pressure, especially in Redrock. Practice helps you work faster without sacrificing accuracy.

Reduces test-day anxiety: Knowing what to expect removes a major source of stress. You can focus on performing rather than worrying about surprises.

Improves your process score: The more comfortable you are with the games, the more organized and systematic your approach will be. This directly impacts your process score.

What Preparation Does NOT Do

Give you the exact questions: Every Solve is unique. You won't see the same species, data, or microbes.

Guarantee you'll pass: You still need underlying problem-solving abilities. Preparation helps you demonstrate those abilities, not fake them.

Replace genuine cognitive skills: If you lack critical thinking or systems thinking abilities, short-term prep won't magically create them. But most candidates have these skills and just need to learn how to show them effectively.

The Four-Step Preparation Process

Step 1: Understand the assessment (2-3 hours)

Read this guide thoroughly. Know what each game involves, how scoring works, and what behaviors the test rewards.

Step 2: Learn the strategies (2-3 hours)

Study the specific approaches for each game. Understand not just what to do, but why. Create notes or flashcards for key rules (like the eating order for Ecosystem Building).

Step 3: Practice with simulations (5-15 hours)

This is the most important step. Play through realistic simulations multiple times. You can do this through PSG Secrets' McKinsey Solve simulation.

Each run helps you:

Internalize strategies until they're automatic
Build speed and efficiency
Identify personal weak points
Get comfortable with the interface

Step 4: Brush up on supporting skills (2-5 hours)

For Redrock: Practice quantitative reasoning, data interpretation, chart reading
GMAT integrated reasoning questions are good practice
Review basic math operations (percentages, growth rates, averages)
For Ecosystem/Sea Wolf: Practice systems thinking and logical deduction

Test Day: What to Expect and How to Perform

Before the Test

Technical Setup (24+ hours before)

Complete McKinsey's technical check
Ensure your computer meets requirements
Test your internet connection
Update your browser if needed

Physical Preparation

Get good sleep the night before
Eat a proper meal (but not right before the test)
Have water available during the test
Use the bathroom beforehand (you can't pause)

Environment Setup

Find a quiet space where you won't be interrupted
Tell roommates/family not to disturb you
Turn off notifications on your devices
Have these items ready: pen and paper, calculator, spare pen/pencil, water

Mental Preparation

Review key strategies one more time
Remind yourself that you've prepared
Don't try to learn anything new the morning of
Accept some nervousness as normal

During the Test

The Tutorials (Untimed)

Take your time. These don't count against you.
Take notes on key rules and mechanics
Make sure you understand the objective before starting
This is your last chance to mentally prepare

General Principles

Think before you act. Your process is being recorded.
Be systematic, not random
Use the provided tools (calculator, journal)
Stay calm if something doesn't work immediately
Manage your time across all phases

After the Test

Don't stress about your performance. It's done.
You can request your score from McKinsey
Expect to hear about next steps within 1-2 weeks
Use the waiting period to prepare for case interviews

What Happens After You Take the Solve?

Most candidates hear back within 1-2 weeks after completing the Solve. However, timing varies based on:

Recruiting cycle: Peak seasons (fall MBA recruiting, spring undergrad recruiting) may have longer waits
Office volume: Competitive offices processing many applications may take longer
Your school's timeline: Campus recruiting often has coordinated decision dates

If you haven't heard back after 2 weeks, it's appropriate to send a polite follow-up email to your recruiter asking if there are any updates.

Possible Outcomes

1. Interview Invitation: Your Solve results and application materials met McKinsey's standards. You'll receive details about first round interviews.

2. Rejection: Your application didn't advance. McKinsey will send a notification, though it may be brief.

3. Waitlist: In some rare cases, you might be placed on hold while McKinsey finishes evaluating other candidates. This is uncommon but possible.

If You Pass

Congratulations! But your work isn't over. McKinsey case interviews require significant preparation.

Immediate next steps:

Confirm your interview dates
Begin case interview preparation immediately
Practice behavioral/PEI questions
Research the specific office and recent McKinsey news
Prepare your "Why McKinsey?" and "Why Consulting?" answers

If You Don't Pass

A Solve rejection is disappointing but not the end.

Reapplication policies:

Most McKinsey offices allow reapplication after 6-12 months
Policies vary by office and your school's relationship with McKinsey
Check with your recruiter for specific guidance

What to do next:

Request your score to understand where you stand
Identify whether Solve was the issue (versus resume/application)
Use the waiting period to strengthen your profile
Consider other consulting firms in the meantime
Prepare more thoroughly for your next attempt

How the Solve Compares to Other MBB Assessments

McKinsey isn't the only consulting firm with a digital assessment. Here's how the Solve compares to BCG and Bain's tests.

BCG Pymetrics Test

Format: 12 mini-games lasting about 30 minutes total

What it measures: Cognitive and behavioral traits through tasks like balloon pumping, facial expression recognition, and pattern matching

Key differences from Solve:

Much shorter
More varied mini-games
Less directly related to consulting work
No clear "right answer" in many games

Difficulty: Generally considered less challenging than the Solve, but also harder to prepare for because the games feel more random.

BCG Casey (Chatbot Assessment)

Format: An interactive, AI-driven case interview lasting about 30 minutes

What it measures: Case-solving ability through a simulated case discussion

Key differences from Solve:

Directly resembles a case interview
Tests business thinking explicitly
Involves written responses to case questions
More straightforward to prepare for (case prep helps)

Difficulty: Requires good case skills but is generally less stressful than a live case interview.

Bain SOVA Test

Format: Numerical and verbal reasoning test with situational judgment components

What it measures: Cognitive abilities and professional judgment

Key differences from Solve:

More traditional aptitude test format
Includes reading comprehension and numerical reasoning
Less gamified
Situational judgment questions assess professional instincts

Difficulty: Standard aptitude test difficulty. Good test-taking skills help.

Comparison Table of Other MBB Assessments

Feature	McKinsey Solve	BCG Pymetrics	BCG Casey	Bain SOVA
Duration	~110 min	~30 min	~30 min	~60 min
Format	Gamified simulations	Mini-games	Chatbot case	Traditional test
Business knowledge needed	No	No	Yes	Some
"Right answers"	Yes	Often no	Yes	Yes
Process score	Yes	Sort of	No	No
Preparability	Moderate	Low	High	High

Frequently Asked Questions

General Questions

How long is the McKinsey Solve?

About 110 minutes total. Each of the three games (Ecosystem Building, Redrock, Sea Wolf) takes 35 minutes. Tutorials are untimed and don't count toward this total.

Can I take the Solve on a tablet or phone?

No. You need a PC or Mac. Tablets and phones are not supported.

What equipment do I need?

A computer with internet connection. Sound is not required. A mouse is optional but helpful. You should also have pen, paper, and a calculator ready.

Is the Solve available in other languages?

Yes. You can take it in English, Spanish (Spain or Latin America), Portuguese (Portugal or Latin America), or Japanese.

What if I experience technical issues?

Contact [email protected] immediately while the issue is happening. Don't wait until after the test. McKinsey can provide extra time or reset your link for verified technical problems.

Can I change my scheduled time?

Yes, if you haven't started yet. Click your original link to reschedule.

Do you offer accommodations for disabilities?

Yes. Contact your recruiter before the test if you need accommodations like extra time due to a disability, health condition, or specific learning difficulty.

Scoring Questions

What score do I need to pass?

McKinsey doesn't publish passing scores. Your results are compared to high performers at McKinsey, and decisions depend on the overall applicant pool and your other application materials.

Can I get my score after the test?

Yes. You can request your score and percentile range from McKinsey after completing the assessment.

Does the Solve alone determine my candidacy?

No. McKinsey considers your Solve results together with your resume, cover letter, referrals, and other application materials.

Preparation Questions

Should I prepare for the Solve?

Yes. While McKinsey says no preparation is needed, candidates who understand the games and practice beforehand consistently perform better.

How should I prepare?

Read comprehensive guides (like this one), practice with simulations, and brush up on quantitative reasoning skills.

How long should I prepare?

Ideally 2-4 weeks, but even a few days of focused preparation helps.

Post-Test Questions

How long until I hear back?

Typically 1-2 weeks, though timing varies based on recruiting cycles and office volume.

What if I fail?

You can usually reapply after 6-12 months, depending on the office and your school's policies.

Can I retake the Solve?

Not immediately. You'll need to wait for a new recruiting cycle or reapplication period.

Cheating Questions

How does McKinsey prevent cheating?

Each candidate receives a unique test with different parameters. Advanced AI ensures no two tests are the same. Additionally, some candidates may be asked to explain their logic in person.

Can I use outside help during the test?

McKinsey says you should complete the test alone without assistance from others, applications, websites, AI tools, or pre-prepared notes. Using such assistance could result in disqualification or rescinded offers.

Can I take screenshots or record the test?

No. This is explicitly prohibited by McKinsey's policies.

Your Next Step to a McKinsey Offer

While it can be helpful to learn the right strategies to tackle the McKinsey Solve, the best way to improve and guarantee that you pass the assessment is by practicing the actual games and exercises that you'll see on test day.

PSG Secrets provides a realistic McKinsey Solve simulation that includes all the games you'll see on test day. In addition to practice, they also provide further guides and tips for each game.

While you prepare for the McKinsey Solve, make sure you are also ready for McKinsey case interviews. You will not land a McKinsey offer unless you nail every single one of your case interviews.

Ready to stop struggling with case interviews and start landing consulting offers?

Enroll in my step-by-step case interview course:

Get interview-ready in 7 days, not 6–8 weeks

Create perfect frameworks in under 60 seconds

Build business acumen in 2 hours, not 2 years

Solve case math 2x faster with 80% fewer errors

Practice with 20 cases covering 95% of case types

Join 3,000+ candidates who've landed offers at McKinsey, BCG, Bain, and other top firms. 82% of my students land consulting offers.

👉 Enroll Now