McKinsey Sea Wolf Game: Practice Method, Sum Checks and Puzzles

McKinsey Sea Wolf is the McKinsey Solve game in which you build three-microbe treatments for contaminated ocean sites. Candidates report three sites in about 30 minutes; each site sets ranges for three attributes that the treatment's average must meet, plus one desired and one undesired trait. Walkthroughs report that each condition a treatment misses costs 20% efficiency. McKinsey's August 2025 Solve FAQ publishes no Sea Wolf rules; it says only that each task opens with an untimed tutorial and that your invitation states the total length (checked October 3, 2026).

Original preparation exercise. These invented microbes, values and rules teach constraint checking. This is not a reproduction of a McKinsey question, a complete simulation or an official scoring model. Use preparation tools before the assessment, never during it.

Checked October 3, 2026. McKinsey has not published a Sea Wolf specification. Format details below are labelled as official, reported or a third-party reconstruction.

What is McKinsey Sea Wolf?

Sea Wolf is the McKinsey Solve game in which you clean up contaminated ocean sites by choosing microbes. For each of the three reported sites, walkthroughs describe the same loop: sort 10 microbes, pick one of three prospects four times to fill a pool of 10, then choose three microbes whose average attribute values fall inside every range. It is a constraint puzzle: no biology knowledge is needed, and every fact you need is on screen.

Sea Wolf appears in every invitation length candidates have reported in 2026, and reports between July 31 and August 27, 2026 say it is still in use, alongside Red Rock in a 65-minute invitation, and with the Sustainable Futures Lab added in 85- and 95-minute ones (see the 95-minute guide). Some candidates call it the "ocean cleanup", "ocean treatment" or "microbe" game; the section on those names explains what they refer to. Older articles about the Ecosystem game describe a different, earlier Solve game that reportedly was removed around mid-2025 (one says early 2026; McKinsey announced no date).

What is the reported Sea Wolf format?

Sea Wolf is reported as three sites of about 10 minutes each, where every site asks for a three-microbe treatment checked against three attribute ranges, one desired trait and one undesired trait. This is what candidates and prep providers consistently describe, not a McKinsey specification: your tutorial controls the exact screens, numbers and limits.

How does Sea Wolf work, step by step?

Each Sea Wolf site reportedly runs four steps: choose two profiling characteristics, categorise 10 microbes, pick one of three prospects four times, then choose three microbes as the treatment. Several walkthroughs describe this same sequence. They label the steps differently and McKinsey has published none of it, so treat the flow as a reported reconstruction and let the tutorial overrule it.

What does one Sea Wolf site look like, screen by screen?

One Sea Wolf site moves through four screens: the site Profile, microbe Classification, the Prospect Pool and the Treatment. The captures below are independent practice, not McKinsey screens: they come from the Road to Offer Sea Wolf practice simulator and follow the reported workflow, not the live interface. Your tutorial shows the real layout and labels.

3. Prospect Pool. In each of four rounds you keep one of three prospects, so the pool grows from six to ten. We do not show this screen; Puzzle 2 rehearses one round.

4. Treatment. You choose three microbes and the game shows the resulting efficiency; the sum check and Puzzle 1 rehearse that decision on original numbers.

How is Sea Wolf scored?

Sea Wolf shows an efficiency figure for each treatment, and walkthroughs report that every condition the treatment misses costs 20%. How McKinsey turns treatments and behaviour into a result is not published at all: its Solve page and August 2025 FAQ give no weights, pass mark or score scale.

Hidden scoring is unpublished. McKinsey says only that results are considered with the rest of your application. Some walkthroughs theorise that only the final treatment counts toward a "product" score and that the earlier steps feed a "process" score; some label their own version an educated guess. None of them cites McKinsey, and failure-rate percentages that circulate are unverified estimates (see the pass-rate guide).

Under every theory the same behaviour wins: build a pool that can produce a treatment that misses nothing, then pick one deliberately. When no such trio exists, the best treatment is the one that misses the fewest conditions. Clicking through prospects without a plan helps under none of them. What a result does and does not tell you afterwards is covered in Solve scores and results.

How do you build a valid Sea Wolf treatment fast?

The fastest way to build a valid Sea Wolf treatment is to turn each average range into a target sum before you look at any microbe: multiply both ends by three, then add the trio's values instead of averaging them. The slow way, adding three numbers, dividing by three and comparing for every attribute and every trio, repeats the division dozens of times. The Sea Wolf cheat sheet puts the whole routine on one page.

This also tells you what a prospect is worth during pool building. If your pool is running low on one attribute, a microbe that is high on it gives you room to balance, even if it looks extreme alone. The worked sum check below Puzzle 1 applies the method to an actual trio.

Road to Offer's free Sea Wolf run includes all three sites, four stages per site, on a 30-minute Simulation clock. Start without an account, then create a free account to finish and keep your score and worked debrief. No payment is required for the first run or its review. The debrief shows your score, where points went and worked feedback on submitted decisions, with constraint and treatment visuals.

We offer four practice formats: Red Rock, Sea Wolf, SFL Classic and SFL Island, with one complete run of each free. Learning mode adds a step-by-step path with coaching and feedback as you go across all four, untimed in the clocked games; it is included with the matching game pack, the Solve pack or qualifying Annual access. The Solve plan is free with a free account and gives dated practice steps based on your invitation length and test date, for up to 28 days.

Run the free Sea Wolf simulation, all three sites

What are the common Sea Wolf traps?

The most expensive Sea Wolf trap (see all the common Sea Wolf mistakes) is applying the site range to each microbe instead of to the treatment's average; the others are checking traits too late, picking the strongest microbes, greedy pool building, checking only one bound and over-spending time on site 1. They are our coaching guidance, built from the worked puzzles on this page, not a claim about McKinsey's marking.

What is Sea Wolf practice preparing you to do?

Sea Wolf practice prepares you to solve a constrained combination, not a biology quiz. A microbe can be weak alone but useful in a trio if its values balance its partners. A candidate that looks strong on one attribute can still make the whole treatment invalid. This is why good practice has two passes: remove choices that violate a categorical rule, then test the complete combination against every numeric bound.

On the real assessment, read the tutorial and site brief before acting. In practice, use this sequence: write the objective, list the hard constraints, mark which values are individual and which are averaged, then compare full combinations. If a provider's reported interface changes, that reasoning sequence still works.

How should you check a Sea Wolf combination?

Check a Sea Wolf combination in two passes: remove any trio that breaks a trait rule, then test the full trio against every numeric bound. Public preparation guides describe a microbe-selection task involving numeric averages and traits. Their precise clocks, pool sizes and penalty claims are not an official specification for your sitting. The transferable skill is to translate a brief into checks, compare complete combinations and notice what a prospect adds to the available pool.

Start with four questions: how many members must you choose, which numeric ranges apply, which trait is required, and which trait is forbidden? Then ask whether the numeric rule applies to each member or to the average. That last distinction changes which candidates can work together. The two puzzles below are deliberately small so that you can check every answer by hand.

Puzzle 1: select a treatment of three microbes

Choose exactly three different microbes from this pool. The average resilience must be from 4 to 6 inclusive, and average energy from 3 to 5 inclusive. Include at least one purifier. Include no invasive microbe. There is no extra reward for exceeding a minimum or choosing a particular valid trio.

Scroll horizontally on a small screen to see every attribute.

Six microbes, A to F

Microbe A

Resilience 6Energy 2

Purifier

Microbe B

Resilience 4Energy 6

Neutral

Microbe C

Resilience 5Energy 4

Neutral

Microbe D

Resilience 8Energy 3

Invasive

Microbe E

Resilience 7Energy 5

Purifier

Microbe F

Resilience 3Energy 7

Neutral

Before opening the explanation: test A + B + C, A + C + D, and B + C + F. For each, write “valid” or name every failed condition. An invalid trio can break more than one rule.

How do you convert Sea Wolf average ranges into sums?

For three microbes, multiply both ends of each average range by three. Resilience 4 to 6 becomes a total of 12 to 18. Energy 3 to 5 becomes 9 to 15. You can check totals directly, then divide once when explaining your result. The Sea Wolf math guide covers the arithmetic in more depth.

For A + B + C, resilience totals 6 + 4 + 5 = 15 and energy totals 2 + 6 + 4 = 12. Both totals fit. B's energy of 6 is above the target average, but A's energy of 2 balances it. Discarding every individual value outside the average range would wrongly eliminate a useful member.

Eliminate D first because its forbidden trait cannot be repaired by its partners. Among the remaining combinations, require A or E to supply a purifier. Then check both totals. This is a shortcut for this stated puzzle, not a claim that every assessment gives the same hard constraints.

Puzzle 2: one prospect round, then the treatment

This puzzle mirrors the reported step 3 and step 4 in miniature. Six microbes are already in your pool. You see three prospects and may keep one. Then you choose three microbes from the seven for the treatment. The real game runs four prospect rounds and three attributes; this version uses one round and two attributes so you can check it by hand.

Site rules: average resilience 4 to 6 and average energy 3 to 5, inclusive (target sums 12 to 18 and 9 to 15). Desired trait: purifier. Undesired trait: invasive. Each failed condition costs 20% efficiency, and each invasive microbe counts separately (one reading of the reported rule).

Which prospect do you keep, and what is the best efficiency each one allows? Notice first that no microbe in the starting pool is a purifier.

How should you review a wrong Sea Wolf answer?

Name the Sea Wolf error, not just the correct trio: each wrong answer in the puzzles above comes from one of five habits, and fixing the habit transfers to new sites.

• Missing trait: scan the categorical rules before calculating.
• Wrong average: check the sum against both lower and upper bounds.
• Over-filtering: an individual can fall outside an average range and still contribute to a valid trio.
• Greedy prospect choice: evaluate the final combination the new member makes possible.
• No trio reaches 100%: recheck the pool for the missing condition first. If none exists, take the treatment that misses the fewest conditions, since each miss reportedly costs 20%.

For a different skill, try the Red Rock connected dataset or SFL prioritization scenario. Before your assessment, read the test-day rules.

Is there a Sea Wolf solver, and should you use one?

Yes, Sea Wolf "solvers" exist: spreadsheets and small web tools where you type in each microbe's values and the site ranges and they list valid trios. Searches for them are common. They do not help, for two separate reasons.

• Integrity. McKinsey's FAQ asks candidates to complete the assessment alone, without outside tools, and not to record it. Running a solver during the live test breaks those conditions. Whether or not any particular behaviour is detected, it is a risk to your application that no preparation site can remove, and results are considered with the rest of your application.
• Speed. Typing three or four attributes for every prospect into a second window takes longer than the mental check it replaces. Once ranges are converted into target sums, checking a trio is three additions and three comparisons, which you can do in seconds.

The legitimate use of a solver is after practice: checking your own answers to an exercise like the 20-trio table above. That is what the expandable table on this page does for you, and the Sea Wolf Excel solver guide covers when a spreadsheet checker helps in practice and why a paid one adds nothing. For live-test habits that hold up under time pressure, the Solve tips guide and the guide on how to prepare for McKinsey Solve are a better investment, and the free simulation on the practice page gives you new sites to check by hand.

Are Ocean Treatment, Ocean Cleanup, the microbe game and Seawolf the same as Sea Wolf?

They point to the same microbe task, but only Sea Wolf is a McKinsey name: McKinsey's July 2025 article names Sea Wolf and describes selecting and evaluating microbes to clean contaminated sites, while "Ocean Treatment", "Ocean Cleanup" and "the microbe game" are labels from candidate and prep material, and "Seawolf" is the same name written as one word. McKinsey has not published whether Ocean Cleanup was once a separate, earlier version, and the accounts disagree, so treat the names as a search map rather than a product timeline.

If an old Ocean Cleanup page describes coral recovery, reef zones or repeated feedback rounds, do not carry those details into your expectation of Sea Wolf: they appear in one account, are absent from McKinsey's description and conflict with other reports. Keep the transferable question, "does this combination meet every target and trait rule?", and let your tutorial set the number of sites and steps. For the wider history of the assessment's names, see the McKinsey Problem Solving Game guide.

How does Sea Wolf fit into McKinsey Solve preparation?

Sea Wolf shares every reported McKinsey Solve session with Red Rock; in a 65-minute invitation the two games are the whole assessment. The two games reward different habits (the Red Rock vs Sea Wolf comparison sets them side by side): Red Rock is about the right denominator, Sea Wolf about holding several constraints at once. Read your invitation for the total length, confirm your setup with the test-day checklist, and if you want an honest sense of difficulty, read is McKinsey Solve hard? The complete Solve guide covers how the three games fit together, and the Solve timeline shows when results usually come back.

Sources and evidence boundaries

The original examples above are ours. McKinsey's official Solve page and its August 2025 FAQ were checked October 3, 2026. They establish that the assessment length is in your invitation, each task starts with an untimed tutorial, no business or gaming knowledge is needed, you work alone without outside tools, and results are considered with the rest of your application. They govern assessment conduct, not this invented puzzle.

Format details (three sites, three attributes on the treatment average, one desired and one undesired trait, about 30 minutes) come from several walkthroughs, consistent across public reports checked between September 20 and October 3, 2026. The step flow and the 20% efficiency rule were cross-checked on October 3, 2026 across several walkthroughs, which agree on the sequence and the 20% figure and differ on step names, caps and whether the undesired-trait penalty stacks per microbe. Their module details, clocks and scoring assertions are third-party reports, not McKinsey confirmation. Spreadsheet solvers circulate; see the solver section above for why we do not recommend one in a live sitting. The time budget, sum method and traps are our coaching guidance. No external solver, prepared notes or AI assistance belongs in the live assessment.