Solve in Six: McKinsey Solve, from invite to game day

Cold open

Narrator The email says: your online assessment, 85 minutes. It doesn't say which games, how the minutes split, or what's being scored. By the end of this lesson, you'll read that one number like a map.

Title card

Timer Clock's running.

What McKinsey actually says

Narrator Start with what's official. McKinsey calls Solve a gamified assessment of problem-solving, and says it needs no business knowledge or gaming experience [1][2]. Your invitation email sets the length. Each task opens with an untimed tutorial. And no two candidates get the same numbers, so memorized answers won't help [2].

Narrator During the test, apps, websites, AI tools and prepared notes are banned. You take it alone, on a PC or a Mac [2]. Using AI to learn the concepts before the test? McKinsey's careers page calls that encouraged [1].

Narrator Here's what McKinsey doesn't publish: the game names, the time per game, the scoring model, or a pass mark. So from here on, I'll tag everything else as what candidates report.

Decode the number

Narrator So let's decode an example invitation. The email is made up, but the structure is real: 85 minutes, deadline in five days.

Narrator Candidates consistently report two games on current invites. Redrock, a research study with six mini-cases, about 35 minutes. And Sea Wolf, an ocean-microbe puzzle, about 30 [3][4]. Thirty-five plus thirty is sixty-five. That's the short invite.

Narrator Eighty-five adds a third game: the Sustainable Futures Lab, a 20-minute judgment task that started appearing in 2026 [3][5]. Ninety-five? Early reports describe a longer, rebuilt Futures Lab of about 30 minutes. Others read it as 85 plus a buffer. Sources disagree, so treat it as unconfirmed [5][6][9].

Narrator Two practical points. The tutorials sit outside the timer, and one prep guide estimates five to ten minutes each, so block about two hours for an 85-minute invite [2][6]. And each game has its own clock. Minutes you save in one don't carry into the next [3][4].

What changed

Narrator One warning before you open any prep guide. Older guides teach Plant Defense and the Ecosystem game. Candidates report that Redrock replaced Plant Defense in 2023, and Ecosystem was phased out in most locations in 2025, though a few regions reportedly still use it [4][10].

Narrator Most prep sites also say the format can vary by office and hiring cycle, while one says it's the same everywhere [3][5][7]. So your invitation, not a guide, is the source of truth.

How it's scored

Narrator How is it scored? McKinsey hasn't said. It does say your result is weighed with the rest of your application [2]. Prep sites report the game looks at your final answers and at how you got there: which data you picked, how much you backtracked. They disagree on how much each counts [3][4][6]. Either way, deliberate beats frantic.

Recap

Narrator Recap. Your invite length is the map. Tutorials are untimed and every game has its own clock. And skip any guide that teaches Plant Defense or Ecosystem as current.

Bridge

Narrator Next: Sea Wolf, and the one line of math that turns averaging into a quick check.

Timer Time.

End card

Narrator When you're ready to try it on the clock, Ecotopia is a free Redrock-style study on drillcase.com. No sign-in.

Cold open

Narrator Three microbes, three numbers each, and every average has to land inside a range. Do you really have to divide by three, nine times, with a timer running? No. You multiply once.

Title card

Timer Clock's running.

How Sea Wolf works

Narrator Here's Sea Wolf as candidates describe it. You clean up three ocean sites, and for each one you choose three microbes [1][2]. Every microbe has a few number attributes, and each site gives a range for each attribute. The rule: the average of your three must land inside the range, and the edges count. No single microbe has to be inside [1][3][4].

Narrator Each site also names one trait you want, which at least one microbe must have, and one trait you don't, which none of them may have [1][6]. McKinsey says no two candidates get the same numbers, so learn the method, not the answers [7].

The move: sum bands

Narrator The move: stop averaging. Multiply each range by three, once, at the start. An average of four to six becomes a sum of twelve to eighteen. Now you just add three whole numbers and check the band [3][4][8].

Worked example

Narrator Here's an example site with made-up numbers. Energy sums to twelve through eighteen. Mobility, eighteen through twenty-four. Density, six through twelve. The wanted trait is oil-eating, and the unwanted one is toxic. Seven microbes are in the pool.

Narrator Step one, before any math: strike every toxic microbe. B and F go. Look at B: five, seven, three, inside every range. It's still out. Prep guides warn that strong numbers on a disqualified microbe are a distraction [3][5].

Narrator Step two: anchor on the wanted trait. A eats oil: six, eight, three.

Narrator Step three: find two partners. Try D and G. Energy: six plus eight plus five, nineteen. One over. Swap G for C. Energy: six, eight, two. Sixteen, in. Mobility: eight, five, nine. Twenty-two, in. Density: three, two, five. Ten, in. Oil-eater, yes. Toxic, none.

Narrator Five checks, all green. And notice: C and D each sit outside the ranges on their own. The trio balances them [3][8].

When to stop

Narrator And now, submit. Prep guides reconstruct the score as five pass-or-fail checks: three ranges and two traits. McKinsey hasn't confirmed it [1][2]. If that's right, a trio at the edge of a band passes just like one in the middle, so polishing gains nothing.

Narrator Candidates report about thirty minutes for all three sites, though one guide says thirty-five, so trust your on-screen timer [2][3][5][6]. One guide adds that an unfinished site scores zero [5]. So budget about nine minutes a site, take the first trio that passes, and move on.

Recap

Narrator Recap. Multiply each range by three. Strike the unwanted trait, then anchor on the wanted one. All five checks green? Submit, and move on.

Bridge

Narrator Next: Redrock, where the trap isn't the math. It's collecting too much.

Timer Time.

End card

Narrator Want a timed run? Ecotopia, a free Redrock-style study, is on drillcase.com. No sign-in.

Cold open

Narrator Nine data points on the screen. The question needs six. It's tempting to drag in all nine, just in case. That's the first place Redrock takes your minutes.

Title card

Timer Clock's running.

How Redrock works

Narrator Candidates describe Redrock as a fictional wildlife research study, about thirty-five minutes, in three phases [1][2][3]. Investigation: you read the objective and drag useful data into a research journal. Analysis: you answer calculation questions with an on-screen calculator. Report: you complete a partly written report. Then come six short cases with their own data. That's lesson four [2][3].

Narrator One more thing. Some candidates report the phases are one-way. One guide says you can step back to Investigation, but it costs time, and once the report starts, there's no way back [1][2]. Either way, plan to move forward once.

Investigation: collect less

Narrator Here's an example study, with made-up numbers. The objective: did reed warbler nests grow faster in North Marsh or South Marsh from year one to year three, and did more North Marsh nests fledge chicks?

Narrator Read it twice and circle the nouns: nests, two marshes, two years, fledging. Now the nine data points. Four nest counts: North, four hundred then five hundred twenty. South, two hundred fifty then three hundred. North's fledging rate: forty percent, then forty-six. And rainfall, predator sightings, water temperature.

Narrator Do those last three appear in the objective? No. Leave them. Six go in the journal. Prep sites report the process is scored too, so a clean journal isn't just faster. It may count toward your score [1][3][5].

Analysis: percent vs points

Narrator Analysis. North grew from four hundred to five hundred twenty. That's a hundred twenty over four hundred: thirty percent. South: fifty over two hundred fifty, twenty percent.

Narrator Now the trap. How much faster did North grow? Ten percentage points, not ten percent. Thirty is one and a half times twenty, so in relative terms, North grew fifty percent faster. Same numbers, and three answers that sound alike. Read which unit the question asks for.

Narrator Same with fledging. Forty percent to forty-six is up six percentage points, which is a fifteen percent increase. Prep guides list this mix-up as a common Redrock mistake [2][6].

Reuse the log

Narrator One more habit. Fledged nests in year three: five hundred twenty times point four six. That's two hundred thirty-nine point two. Candidates report the calculator keeps a log of unrounded results you can reuse [2][4].

Narrator Reuse it. Compared with a hundred sixty fledged nests in year one, that's growth of forty-nine point five percent. Retype a rounded two thirty-nine and you get forty-nine point four. Close, but it's a different answer.

Recap

Narrator Recap. Collect only what the objective names. Know whether you're asked for points or percent. And reuse unrounded results from the log.

Bridge

Narrator Next: the report, the chart question, and six mini-cases at under two minutes each.

Timer Time.

End card

Narrator Try a full study now. Ecotopia is a free Redrock-style study on drillcase.com. No sign-in.

Cold open

Narrator Bars or a line? In Redrock's report, the data fits both charts. Only one of them fits the sentence.

Title card

Timer Clock's running.

The report

Narrator Candidates describe the report as a partly written document. You fill in numbers, choose higher or lower, and pick the chart that shows a finding [1][2][3].

Narrator Back to our example study. The sentence reads: North Marsh nests grew by blank percent, higher or lower than South Marsh. Thirty. Higher.

Narrator Now the chart. Say the message out loud: North grew faster than South. That compares two categories at one point, so use a bar chart of the two growth rates. If the message were how nests changed year by year, that's a trend, so a line. Parts of a single whole? Only then a pie. One prep guide's rule: pick the chart for the message, not for the data [2].

Six mini-cases

Narrator Then come the mini-cases. Candidates report six of them, each with its own data, linked to the study's theme but not using its numbers [1][2][3]. Reported types include weighted averages, probabilities and chart choice [1][3].

Narrator Here's a weighted average with made-up numbers. A reserve has three zones: fifty hectares at eight birds per hectare, thirty hectares at four, and twenty at two. What's the average density?

Narrator The fast, wrong answer averages eight, four and two: about four point seven. But the zones aren't the same size. Total birds: four hundred, plus a hundred twenty, plus forty. Five hundred sixty, over a hundred hectares. Five point six birds per hectare.

Narrator So before you calculate, ask: are these groups the same size? If not, weight them.

Pace the game

Narrator Now the clock. McKinsey publishes no split. Two prep sites do, and both protect about ten minutes for the six cases: under two minutes each [2][4].

Narrator So set one checkpoint. When your report is done, glance at the timer. Ten minutes or more left? You're on pace. Less? Don't skip a case. Keep each one short, and in your next practice run, trim the earlier phases. Inside a case, once you pass two minutes, pick your best answer and move on [2].

Timer Two minutes. Next.

Recap

Narrator Recap. Say the message, then pick the chart. If the groups are different sizes, weight them. And start the six cases with ten minutes left.

Bridge

Narrator Next: the Sustainable Futures Lab. Candidates report no calculator, and no single right answer.

Timer Time.

End card

Narrator Ecotopia, a free Redrock-style study with its own report, is on drillcase.com. No sign-in.

Cold open

Narrator Your data is incomplete. Storm season starts in six weeks. A local group is upset. And there's no calculator. What do you do first?

Title card

Timer Clock's running.

What the Lab is

Narrator That's the feel of the Sustainable Futures Lab. Candidates report it's been on 85-minute invitations since 2026, and it runs about twenty minutes [1][2]. You read a mission briefing, rank four possible actions, then make twelve single-choice decisions in one connected storyline [2][3][9]. There's no calculator and almost no math [4][9]. And it's not a sustainability quiz [1][4].

Narrator Prep sites describe five things it seems to reward: prioritizing, deciding without full information, reading messy data, handling trade-offs, and working with your team and stakeholders [3][5]. McKinsey hasn't published a rubric, and no item has a published right answer.

Rank the actions

Narrator An example briefing, made up for this lesson. You lead a small team restoring a coastal wetland. The budget is fixed, storm season starts in six weeks, and the local fishing co-op is worried about losing access.

Narrator Rank four actions: order all the native plants now, survey the water's salinity, a two-day job, meet the fishing co-op, and write the full project plan.

Narrator One defensible order. First, survey the salinity: it's quick, and every later choice depends on it. Second, meet the co-op: a small cost now can head off a bigger conflict later. Third, order the plants, once you know which ones will survive. Last, the full plan, which can build on all of that.

Narrator The principle: get the information that changes your decision before you make a commitment you can't undo.

A decision under uncertainty

Narrator Later in the story, two researchers disagree. One sample says salinity is rising, and the other says it's stable. Your options: wait two weeks for a third sample, plant a salt-tolerant mix that costs more but works either way, or escalate to the director.

Narrator Read each option against the five traits. Waiting stalls, with the storm season coming. Escalating hands off a call that's yours to make. The salt-tolerant mix spends a little budget to decide now and protect the timeline.

Narrator Prep guides flag three common mistakes: always picking the safest option, waiting for perfect data, and over-escalating [1][3].

Stay consistent

Narrator Then stay consistent. If you told the co-op their access comes first, don't trade it away three questions later without a reason. Prep sites report that switching stances between questions shows [3][5].

Narrator On pace: thirteen items in twenty minutes is about a minute and a half each.

A caveat: the rebuilt version

Narrator One caveat. Early reports link 95-minute invitations to a rebuilt, roughly thirty-minute version, with three project days, tokens for questioning researchers, and confidence ratings. Sources disagree on how far it's rolled out [6][7][10]. Your invitation length is the clue.

Recap

Narrator Recap. Get information first, and make irreversible commitments later. Decide with what you have. And keep one stance from start to finish.

Bridge

Narrator Last lesson: turn all of this into a plan for the days before your deadline, and for test day itself.

Timer Time.

End card

Narrator Build your timing on a free Redrock-style study: Ecotopia, on drillcase.com. No sign-in.

Cold open

Narrator The invite landed Monday. The deadline is Sunday. So what do you actually do tonight?

Title card

Timer Clock's running.

The five-day plan

Narrator Here's an example: an 85-minute invite on Monday, due Sunday. One prep guide reports deadlines of about three to seven days, so plan for a short one [1]. This plan is built from the prep advice in this course. It isn't an official schedule.

Narrator Monday: rules and math. Explain the sum band and points versus percent out loud, from memory. McKinsey says using AI to explain concepts while you prep is encouraged [5].

Narrator Tuesday: one timed Redrock-style run. Wednesday: one timed Sea Wolf-style run, and read one Futures Lab-style briefing.

Narrator Thursday: a full rehearsal at your invitation length, on the computer you'll test on. Then review how you worked: the minutes left in each game, which answers were guesses, and which clicks you'd cut [2][3][6].

Narrator Friday: tech check, light review, then stop. Take it Saturday morning, and keep Sunday as your buffer.

Test-day setup

Narrator Test day. Here's what McKinsey says: use a PC or a Mac, take it alone, and use no apps, websites, AI tools or prepared notes. Sound isn't needed, and a mouse is optional [4]. McKinsey's FAQ doesn't mention scratch paper, so follow whatever your invitation and the tutorial say.

Narrator Block about two hours for an 85-minute invite, because the tutorials are untimed [3][4]. And don't skip them. They're untimed, so you learn the controls without losing any test minutes.

During the test

Narrator Take three checkpoints in with you. Redrock: start the six cases with at least ten minutes left. Sea Wolf: about nine minutes a site, and submit when every check passes. Futures Lab: about a minute and a half per item.

Narrator If something breaks, like a frozen screen or a dropped connection, contact support while it's happening. McKinsey says it may give you extra time or reset your link, but not if you don't report it [4].

After, and the whole course

Narrator Afterward, McKinsey says your result is weighed with the rest of your application, and you may be asked to explain your logic [4]. So work in a way you could explain.

Narrator That's the course. Read the invite. Add, don't average. Collect less. Weight what's uneven. Decide, and stay consistent. And rehearse on the clock.

Recap

Narrator Recap. Before: one timed run per game and one full rehearsal. On the day: two hours, alone, no tools. During: three checkpoints, and report problems as they happen.

End card

Narrator Your first timed run can start today. Ecotopia is a free Redrock-style study on drillcase.com. No sign-in.

Timer Clock's running.