WHEN SUCCESS LEAVES YOU WITH NO OPTIONS

A full health bar can hide a failing strategy. Sebastian Marshall’s Roguelike offers a useful way to notice the difference.

An original editorial essay based on the supplied 2015 EPUB. Book examples and claims are attributed to Marshall; the broader argument and hypothetical workplace application are editorial interpretations. Sources expand beneath the relevant passages.

Twelve enemies surround your berserker. You activate your strongest abilities, break through the first group and kill four. Several more enemies are wounded. Your character is almost at full health. It looks like the fight is going well.

Sebastian Marshall asks us to look again. The shield has broken. Speed and accuracy boosts are wearing off. Healing and major attacks have been used. The character who now faces eight enemies has fewer ways to survive a surprise than the character who initially faced twelve. A staircase is one move away. In Marshall’s constructed Tales of Maj’Eyal scenario, going upstairs to recover is the sensible move precisely when staying feels most tempting. [C11-P015–P033]

That is the question I find most useful in Roguelike: what if the evidence that we are succeeding is concealing the way we are becoming fragile? A result tells us what has happened. It does not, by itself, tell us what we can still handle.

Count the responses you still have

The berserker’s problem is easy to miss because the most visible measure is reassuring. Health remains high. Enemies are falling. Yet the resources that produced those results are temporarily unavailable. The next stun or burst of damage arrives in a different situation: there may be no useful answer left. Marshall’s argument is to inspect that change before the health bar makes the danger obvious. [C11-P021–P033]

Consider a hypothetical small team that has just accepted several promising projects. Its calendar looks successful. Every hour is assigned. But the person who can resolve a difficult technical problem is also the person responsible for every urgent client response. One delayed handoff can now disturb the entire schedule. This is my application of the analogy, not a business case reported in the book. The relevant question is what the team can do when its assumptions stop holding.

The value of spare capacity is therefore partly invisible. It is the deadline that survives a delay, the choice to correct a mistake and the ability to accept an opportunity that was not in the original plan. None of this means unused capacity is always the best investment. Reserves have costs, and some opportunities justify spending them. It means those costs should be compared with the consequences of having no response available.

Finishing early can buy freedom

Marshall gives a quieter example from a visit to London. Friends are playing Puerto Rico, a board game Kai Zau likes. Kai stays out of the game to complete consulting work that is not due for roughly another five days. Marshall initially finds the choice strange. Then several of the remaining days fill with conversations and events. With the work handled, Kai can participate without squeezing everything into the final hours. [C11-P057–P067]

The interesting thing Kai buys is flexibility. He gives up one known opportunity for the ability to respond to several later ones. Marshall presents the choice as avoiding deprivation overall: the work still has to happen, and doing it earlier protects later enjoyment. The anecdote supports a useful possibility, but it cannot establish that finishing work first always produces more leisure. A unique evening, an exhausting task or a genuinely movable deadline could change the decision.

Read this way, the lesson is not an instruction to fill every free moment with work. It is a reason to notice when a small, controllable effort now can remove a large future constraint. The point of the reserve is to make more of life available.

The best upgrade may look unimpressive

A different example in Roguelike shows how success can distort the choice of what to improve. Marshall imagines a character whose equipment works beautifully together: dazing an opponent enables critical hits, and critical hits trigger further attacks and damage bonuses. Against a compatible enemy, the combination is devastating. [C38-P006–P029]

Then comes a ghost dragon that cannot be critically hit. The character still owns the same impressive equipment, but the condition that makes most of it useful is absent. A cloak that strengthens the usual fire combination adds nothing against this particular obstacle. Another cloak contributes a modest amount of light damage. It looks less exciting, yet it addresses the encounter that could end the run. [C38-P030–P063]

Marshall is not abandoning his advice to build coherent combinations. He is adding a boundary: the strengths that reinforce one another can also depend on the same vulnerable assumption. Improving their best performance does not necessarily improve their worst failure. The light-damage cloak does not make the ghost fight safe; in his example, it only makes a dangerous encounter more manageable. [C38-P048–P063]

For the hypothetical team, another feature may be less valuable than teaching a second person how to handle a recurring emergency. That is an editorial comparison, not a universal staffing rule. The practical investigation is specific: which plausible interruption can defeat the whole plan, and what would give us a usable response?

Make the decision before it becomes difficult

Recognising a weakness is only part of the task. The berserker still has to leave while winning feels good. Marshall proposes defensive policies: decide in advance which event will trigger a change of course. His geomancer has three major healing abilities. Depending on the dangers involved in retreat, the character might leave after using the first or second. Waiting until all three are gone removes the reserve needed if withdrawal itself goes wrong. [C65-P009–P019]

Notice the structure of that policy. It names an observable event, specifies an action and preserves enough capacity to perform it. It is more useful than a vague intention to be careful. In our hypothetical team, an equivalent might be to stop accepting new delivery dates when the agreed buffer disappears, then reassess commitments. The threshold would have to reflect that team’s actual circumstances.

Marshall connects policies with how difficult a strategy is to execute. A complicated set of abilities can be excellent in experienced hands and confusing in a beginner’s. A simpler sequence may produce better results for the person actually using it. This is a claim about matching a plan to skill and attention, not a claim that complexity is always bad. His discussion of work policies also allows deliberate exceptions and disruption; a policy is not a substitute for noticing that circumstances have changed. [C67-P001–P020; C66-P031–P037]

The limits matter as much as the lesson

The appeal of these examples is their clarity. A game can make the cost of a missing defense immediate and unmistakable. Real situations are less tidy. We may negotiate a deadline, ask for help, cooperate with someone initially treated as an opponent or work to change the rules. Marshall himself discusses alternatives that a game’s limited action menu may obscure. The analogy is a prompt to investigate, not evidence that life has the same mechanics. [C14-P012–P023; C15-P024–P033]

Nor does understanding an incentive mean accepting its values. Marshall explicitly allows someone to refuse an advantage on ethical grounds, accept the consequences or leave an environment. An inventor might learn to sell, partner with people who handle business, or decide that commercial success is not the purpose of inventing. Defining success comes before making a plan more effective. [C62-P008–P020]

Finally, a lost game does not prove a bad decision. Marshall acknowledges that excellent play can still lose through chance. He also asks for compassion toward other people when discussing avoidable emergencies. His strong language about personal responsibility should not be converted into a claim that everyone could have prevented their misfortune. Preparation changes exposure to some risks; it does not eliminate uncertainty or give everyone equal resources. [C21-P001–P031; C11-P050–P055]

A question worth keeping

My synthesis is a short pause before the next commitment: what am I trying to achieve, what could make this plan fail, and what could I still do if that happened? These questions are not a named framework from the book. They connect its account of true objectives, critical weaknesses and reserves.

The answer may be to press ahead. It may be to finish something early, cover a neglected weakness or take the stairs while retreat is still easy. What matters is being able to explain the choice in terms of the goal and the options that remain. A reassuring score is useful information. So is knowing what you will do when the next surprise arrives. [C12-P001–P036; C38-P060–P063; C65-P009–P019]