McKinsey Redrock Study 2026: Free Practice, Format, and Worked Answers
McKinsey Redrock Study is the data-interpretation exercise inside McKinsey Solve. Reported two-part format, Investigation-Analysis-Report flow, worked math examples, scoring caveats, and practical strategy. Confirm details from your McKinsey invitation.
On this page
McKinsey Redrock Study is the data-interpretation module of Solve; candidate reports commonly describe roughly 35 minutes for the Study (Investigation, Analysis, Report) and six independent mini-cases together, while your invitation remains the source of truth. Candidates call it the Redrock game or the Redrock island module, after the island ecology scenario the study takes place in.
The reported timing and mini-case count come from MConsultingPrep, StrategyCase, PrepLounge, and Prepmatter, checked June 10, 2026. McKinsey's official Solve page describes Solve as a gamified assessment that showcases problem-solving ability and says performance is considered alongside the rest of your application and any other assessments you are asked to take. For the umbrella view of how Redrock fits into Solve, see our McKinsey Solve guide. This article focuses on Redrock-specific prep: the reported format, the math, the scoring caveats, and how to practice without treating unofficial details as guarantees.
Red Rock format at a glance: one Study, then six Cases
McKinsey Red Rock Study (also written Redrock) is reported as two parts on one shared clock of roughly 35 minutes. Part 1 is a Study that runs in order: Investigation, then Analysis, then the Report. Part 2 is six independent Cases, each with its own data. McKinsey does not publish this breakdown; it comes from third-party reconstructions, and your invitation and in-game tutorial control the version you receive.
A workable pace for a 35-minute practice run is about 21 minutes for the Study, 11 for the six Cases (under two minutes each) and a 3-minute buffer. That is coaching guidance, not an official allocation; the 35-minute practice plan below splits the Study by stage.
You can run this exact structure as a full Red Rock simulation: the full Study and six Cases on a 35-minute practice clock, with 13 scored responses worth 36 points. Road to Offer offers one complete run free in each of four practice formats: Red Rock, Sea Wolf, SFL Classic and SFL Island. Start without an account, then create a free account to finish and keep your score, breakdown of where points went and worked feedback on each submitted decision. The free Red Rock review includes highlighted calculations and charts; no payment is required for the first run or its worked review.
Get your McKinsey prep plan
What does Redrock Study test?
Prepare to select relevant evidence, calculate with the right units, and explain what the result means. McKinsey's official Solve page describes the overall assessment, but does not publish a Redrock scoring matrix or guarantee a particular game sequence.
This table is a Road to Offer practice framework. The phase descriptions come from third-party accounts, including the Redrock walkthrough reviewed for this guide. Your assessment instructions control navigation and timing.
What Is the McKinsey Redrock Study?
Candidate reports place you in the role of a researcher analyzing fictional wildlife population data, often involving wolf packs and elk across geographic regions. The fictional context is useful for prep because it removes prior-knowledge advantage and forces pure data interpretation.
Some third-party prep providers report roughly 35 minutes shared by the Study and a shorter mini-case section in recent versions. Treat that as preparation context, not a promise: McKinsey can vary assessment details by role, region, and cycle, and your invitation should override any prep-site description.
The core skill being tested is reading a chart or table and pulling the right number under time pressure. That is exactly what a graph drill rehearses, so build the reflex before you sit the Study.
How Does Redrock Work? The Two-Part Structure
Candidate reports describe two distinct parts. The boundary between them is critical for time management if your invitation uses this version.
Part 1: The Study Section (Three Sequential Phases)
Candidate reports describe the three Study phases as one-directional. If your version works that way, once you advance you cannot return, so missed data in Investigation cannot be recovered during Analysis.
Phase 1: Investigation (practice target: 6 min). You receive about a page of text plus charts and tables describing a research scenario. You drag relevant data points into a Research Journal. The discipline: read the research objective first, identify which calculations Analysis will require, and only collect the data points those calculations need. Candidates who grab every number end up with a cluttered Journal and slower Analysis.
Phase 2: Analysis (practice target: 9 min). You answer math questions using the data you collected. Some candidate reports describe an on-screen calculator that logs unrounded results and lets you reuse those values. If present, this feature is a major accuracy lever because manual re-entry of rounded numbers creates cascading errors.
Phase 3: Report (practice target: 6 min). You fill in a partially completed research report: numeric blanks, directional comparisons ("higher" vs. "lower"), and chart-type selection. If your Analysis answers were wrong, the Report fill-ins can propagate the error.
Report: written blanks, chart choice, chart build, then confirmation
Reconstructions describe the Report in three moves: complete the written memo, pick the chart type that fits the stated message, then enter the values that build that chart. Several providers, including Case Prep's Redrock guide, describe confirming the Report as closing the Study before the Cases begin; accounts of navigation differ, so treat confirmation as final unless your tutorial says otherwise. Before you confirm, run a four-point check:
- Every numeric blank is in the unit the memo asks for (millions, thousands, percent).
- Growth figures use the definition the memo names: relative growth or percentage points, never a mix.
- The chart type matches the message: composition of one total is a pie or single stacked bar, change over time is a line or grouped bars, comparison across categories is a bar.
- The chart values add up to the total they claim to divide.
Part 2: The Cases Section (Six Independent Mini-Cases)
After the Study, candidate reports describe a set of standalone mini-cases. Each case is self-contained, so an error on one does not necessarily affect the others. This breaks the cascading-error problem of Part 1 and lets you recover from individual mistakes. In practice, aim for high accuracy across the set rather than chasing a published pass line.
Practice the Study and independent Red Rock Cases
Keep one journal through Investigation, Analysis and Report, then switch to fresh data for each Case.
What Math Does Redrock Test?
Candidate reports suggest Redrock is math-heavy. The math is not advanced, but it must be fast and accurate under time pressure. Keep a consulting math formulas reference beside your practice and build the mental math speed these questions reward. Build automatic fluency with five concept areas:
What does the Analysis stage ask?
Reported Analysis questions fall into a few repeatable types. Each example below is original teaching material with its own numbers.
1. Split a budget from written rules. A $60 million programme gives 40% of the total to fleet electrification. Within that, equipment takes 3/4 and installation the remainder. Grid reinforcement takes 30% of the total, and contingency is whatever is left.
- Electrification = 40% × 60 = $24 million, so equipment = 3/4 × 24 = $18 million and installation = $6 million
- Grid reinforcement = 30% × 60 = $18 million
- Contingency = 60 − 24 − 18 = $18 million
- Check: 18 + 6 + 18 + 18 = 60. The trap is taking 3/4 of the whole $60 million instead of 3/4 of the electrification share.
2. Rank by average annual growth, not by the biggest increase. Three sites move from Year 1 to Year 4 (three years of growth): Site A 10.0 → 13.6, Site B 8.0 → 10.4, Site C 5.0 → 7.0 million tonnes.
Site C ranks first even though Site A adds the most tonnes. If the question says "non-compounded", divide total growth by the number of years; do not compute a compound annual rate.
3. Relative growth versus percentage points. Worked next, with the regional population example.
4. Weighted versus simple average. Worked further down, with territory density.
Worked Example: Growth Rate Comparison (Analysis Phase)
Scenario: Compare wolf population growth in the Northern vs. Southern Region (Year 1 to Year 5). Northern: 240 → 312. Southern: 180 → 216.
- Northern growth rate: (312 − 240) / 240 × 100% = 30%
- Southern growth rate: (216 − 180) / 180 × 100% = 20%
- Difference: 10 percentage points (not 10%, not 33%)
If your version includes calculator logging, save both rates before computing the difference. For pooled growth, retain the original population counts as well: the two rates alone do not tell you how much weight each region carries.
Use your calculations in a Red Rock report
Choose the relevant evidence, calculate from your journal and carry the result into the Report.
From research journal to completed report
Use the same original population example above. The research objective is: which region grew faster between Year 1 and Year 5, and what was the combined growth? This is a teaching exercise, not an official question.
Do not add territory size to this journal unless the question asks for density. The objective asks for population growth, so area would not enter the formula. Preserve the year and region beside each number; four unlabeled values are not usable evidence.
Analysis. Combined Year 1 population = 240 + 180 = 420 wolves. Combined Year 5 population = 312 + 216 = 528 wolves. Combined growth = (528 - 420) / 420 × 100 = 25.71%, rounded only for this final display. Averaging 30% and 20% would give 25%, which weights unequal starting populations equally.
Completed report. "From Year 1 to Year 5, the combined wolf population rose from 420 to 528, an increase of 25.71%. North grew by 30%, compared with South's 20%, a gap of 10 percentage points. These observations show population growth; they do not establish its cause."
Chart choice. Use grouped bars to compare each region's Year 1 and Year 5 counts. Two observed dates do not show the path between them, so do not invent annual points or imply steady yearly growth. A pie chart would show composition, not the change requested here.
Run the full Red Rock simulation to practice selecting evidence and using it across three Analysis calculations, the Report and six Cases.
Worked Example: Weighted Average (Analysis Phase)
Scenario: Calculate the island-wide wolf density weighted by territory.
- Weighted (correct): (480 + 360 + 240) / (120 + 180 + 60) = 1,080 / 360 = 3.0 wolves/km²
- Simple average (incorrect for "island-wide"): (4.0 + 2.0 + 4.0) / 3 = 3.33
Read the question phrasing carefully. "Average density across the island" requires the weighted method. "Average of the three regional densities" requires the simple method. The wording determines which denominator is correct.
These growth-rate and weighted-average steps are the same arithmetic case interviews demand, just under a research wrapper. A timed math drill is the fastest way to make them automatic.
Quick Math
- What is 14% of $450M?
Build the percentage from 10% and 1% blocks instead of multiplying directly.
- Margin moves from 17% to 26%. What is the change in percentage points?
Percentage-point change = Ending percentage - Starting percentage, kept in points, not converted to a relative percent.
- Revenue goes from $140M to $175M. What is the percent growth?
Percent growth = (Ending value - Starting value) / Starting value.
What are the six Redrock Case types?
After the Study, candidates report six short Cases, each self-contained. These six types cover the patterns described by providers such as MBB Prep and match the Cases in Road to Offer's full simulation. Every example is original, with numbers chosen so you can check the answer by hand.
Two habits carry across all six: read what the question asks for before touching the data, and never bring a number from the Study or a previous Case into a new one.
How Is Redrock Scored?
McKinsey does not publish a Redrock scoring matrix. Candidate reports and third-party prep discussions point to two useful practice dimensions.
Answer accuracy: straightforward correctness. Analysis answers matter because errors can cascade into Report. Use 80-85% accuracy and near-complete mini-case correctness as practice targets, not as official thresholds.
Process quality: several prep providers believe the platform also records how you work, such as what you collect, how you use the calculator, how often you revise answers and how you spend time. McKinsey has not confirmed this, so treat it as a third-party hypothesis. Either way, a clean process reduces avoidable errors.
A clean practice profile looks like targeted Investigation notes, calculator logging on every multi-step calculation, deliberate chart selection, and few rushed reversals. Do not confuse this with an official scoring formula.
Find the error behind your Red Rock result
Review the worked solutions for all 13 scored decisions after a practice run to tell a wrong calculation from missing evidence.
How Do I Prepare for Redrock?
A five-phase plan built from what candidates report on PrepLounge, MConsultingPrep and StrategyCase. The first full Red Rock Simulation and its worked review are free; Learning mode and additional simulations require the matching game pack, Solve pack or qualifying Annual access.
Days 1-3: make the core math automatic
Drill percentages against percentage points, weighted averages and growth rates until you stop hesitating. Our case interview math practice guide covers these exact operations. Done when ten mixed questions come back clean.
Days 2-4: learn the chart rules cold
Bar compares categories at one point in time. Line shows a trend over time. Stacked bar shows composition over time. Scatter shows correlation. Pie shows parts of a single whole. Histogram shows the distribution of one variable. Report-stage chart choice is a scored decision, so practise it as one.
Days 3-5: read exhibits for the objective
Give yourself 60 seconds per exhibit to name the trend, the outlier and what the data does not show. Then run the Red Rock simulation in Learning mode with the matching game pack, Solve pack or qualifying Annual access: it provides untimed, step-by-step coaching and feedback as you progress. Check that every saved value has its unit, period and label. Learning is available in all four practice formats and is separate from the free end-of-run worked review. The matching game pack, Solve pack or qualifying Annual access also includes the full Red Rock step-by-step PDF guide; a free account unlocks only a 6-page preview.
Days 5-8: run the whole thing on the clock
Run the full 35-minute simulation, then a round-up of MBB assessment simulators for variety. Bridge into live McKinsey practice cases so the same data habits show up in interviews.
Ongoing: review by error category
Math error: more drilling. Misread question: read the question before the data. Chart error: revisit the rules above. Time overrun: adjust your stage budget.
For a worked Redrock game solution, the two analysis-phase walkthroughs above show the full calculation sequence end to end, and the mini-case structure mirrors the case-study prompts that follow the study phase.
Turn this guide into a McKinsey plan
Common Redrock Mistakes (and Fixes)
The six errors that most often damage Redrock-style practice, with the fix and the rep that trains it.
Rushing Investigation
Why it costs: a value you did not save is missing in Analysis, and reported versions do not let you go back. Fix: read the objective twice and list the calculations Analysis will need before saving anything. Practise it in the Red Rock simulation's Learning mode.
Re-typing rounded numbers instead of reusing saved values
Why it costs: rounding early compounds across chained calculations. Fix: keep unrounded results and reuse them; round only the number you submit. Practise with math drills.
Confusing percent change with percentage points
Why it costs: the Report often asks for one and shows the other. Fix: write "relative" or "points" beside every growth figure before you enter it. Ten mixed math drill questions make the distinction automatic.
Picking a chart for the data instead of the message
Why it costs: a correct chart of the wrong comparison scores as wrong. Fix: say the message in one sentence, then pick the chart that shows it. Practise with graph drills.
Spending more than two minutes on one Case
Why it costs: six Cases share about twelve minutes, and each is independent. Fix: set a two-minute cap, submit your best estimate and move on; five careful cases beat four perfect ones and a skipped one. Practise the pacing in a full timed run.
Changing your process from run to run
Why it costs: an erratic process creates avoidable errors, whatever the scoring model. Fix: run the same Investigation, Analysis, Report sequence every time until it is habit, then review each miss against the worked solution.
How should you divide a 35-minute practice session?
This is a Road to Offer study plan, not an official phase allocation. It adds up to exactly 35 minutes. Use the duration and navigation rules in your invitation for the real assessment.
The buffer is not permission to return to locked questions. Check each answer before advancing, and follow the actual interface if it differs from this practice plan.
Red Rock is one of the games you will sit
Build a free Solve plan with a free account. It asks your invitation length and test date, then gives dated practice steps selected for that invitation, up to 28 days. No date yet? You can still build a plan. Paid practice and Learning remain separately gated.
Connected Resources
Redrock usually arrives alongside at least one more Solve module, so once your Redrock loop is stable, run the sibling formats the same way: Sea Wolf and the SFL island each have a free run on their own page.
The free Sea Wolf run includes all three sites and four stages per site on a 30-minute Simulation clock. SFL Classic includes all 13 linked decisions without a countdown; the free SFL Island run includes onboarding and all three project days on a 30-minute Simulation clock.
The online tests McKinsey, BCG and Bain use before interviews, each with a free practice run.
Redrock is one piece of McKinsey's screening process. Pair this guide with:
- McKinsey Solve guide: umbrella view of both modules and the dual scoring system
- McKinsey Sea Wolf guide: the commonly reported ecosystem-style Solve module
- McKinsey case interview guide: the live interview round Solve gates you toward
- McKinsey PEI guide: the behavioral component of the live round
- Consulting aptitude test overview: how Redrock fits the broader firm assessment landscape across McKinsey, BCG, and Bain
Sources and Further Reading (checked 2026-06-10)
- Case Prep, Redrock guide (checked 2026-09-22): https://www.case-prep.com/mckinsey-solve/redrock-guide
- MBB Prep, Redrock game (checked 2026-09-22): https://mbbprep.com/blog/mckinsey-redrock-game
- McKinsey careers, interviewing: https://www.mckinsey.com/careers/interviewing
- McKinsey careers, Solve assessment: https://www.mckinsey.com/careers/mckinsey-digital-assessment
- MConsultingPrep, Redrock deep dive: https://mconsultingprep.com/mckinsey-solve-redrock-deep-dive
- IGotAnOffer, McKinsey Solve guide: https://igotanoffer.com/blogs/mckinsey-case-interview-blog/mckinsey-problem-solving-game
- StrategyCase, Redrock guide: https://strategycase.com/mckinsey-solve-game-red-rock-study/
- PrepLounge, McKinsey Solve forum: https://www.preplounge.com/consulting-forum/mckinsey-solve-redrock-study-15645
- Prepmatter, 2026 Solve update: https://prepmatter.com/blog/whats-new-in-mckinsey-solve-2025
Frequently asked questions
Resources and related guides
- Run a real case interviewPractice
- Browse all free resourcesResource hub
- Do Experienced Hires Take McKinsey Solve? A Working Professional’s PlanFirm Specific · Oct 11, 2026
- Mejor plataforma de IA para preparar entrevistas de consultoría (2026)Getting Started · Sep 29, 2026
- Meilleure plateforme IA pour préparer les entretiens de conseil (2026)Getting Started · Sep 29, 2026