War of Attrition: Game Theory &… | Faster Than Normal
Strategy & Competition
War of Attrition
A competitive strategy where victory goes to the side that can endure the longest. The objective is not to win a single decisive battle but to outlast the opponent by depleting their resources, morale, or willingness to continue. In business, attrition strategies appear in price wars, talent acquisition battles, and platform competition where survival determines the winner.
A war of attrition is a competitive contest where victory goes to the side that can endure the longest. The objective is not to win a single decisive battle but to outlast the opponent by depleting their resources, morale, or willingness to continue. The winner is not the strongest combatant. The winner is the last one standing.
The concept originates in military strategy — specifically from the Western Front of World War I, where neither side could achieve a decisive breakthrough and the war devolved into mutual exhaustion. General Ulysses S. Grant's strategy against Lee in the American Civil War was explicitly attrition-based: he accepted heavy losses because the Union's population and industrial base could absorb them and the Confederacy's could not. The strategic logic was brutal and correct. The side with more resources wins a war of attrition — provided it has the willingness to sustain the cost.
In biology, the war of attrition is formalised in evolutionary game theory by John Maynard Smith. Two animals competing for a resource — territory, a mate, food — engage in a costly display. The animal that persists longer wins the resource. The loser walks away having spent energy for nothing. The optimal strategy depends on the value of the resource relative to the cost of the contest. When the resource is highly valuable, rational contestants will endure longer. When the cost of the contest is high, rational contestants will quit sooner. The equilibrium is a mixed strategy: each contestant randomises their persistence time, with the probability distribution determined by the payoff structure.
In business, wars of attrition are ubiquitous. Price wars, hiring battles, patent disputes, platform competition, and market share contests all share the attrition structure: two or more competitors invest continuously in a contest where the winner takes the spoils and the losers write off their investment. The critical strategic question is not "can we win this battle?" It is "can we outlast the competition, and is the prize worth the cost of the contest?"
The most dangerous aspect of a war of attrition is the escalation trap. Once you have invested heavily in a contest, the sunk cost makes it psychologically difficult to quit — even when quitting is the rational choice. The investment already made feels like a reason to continue. It is not. The rational decision depends only on the expected future costs and benefits, not on what has already been spent. But the sunk cost fallacy is powerful, and wars of attrition exploit it mercilessly.
Section 2
How to See It
A war of attrition is present whenever two or more parties are engaged in a costly contest where the winner is determined by persistence rather than a single decisive action.
Technology & Platforms
You're seeing a war of attrition when two ride-sharing companies compete by subsidising rides below cost, each waiting for the other to run out of capital. Uber and Lyft in the U.S. market from 2014 to 2019 was a textbook war of attrition: both companies burned billions in rider subsidies and driver incentives, each betting that the other would exit first. The strategic question was never "who has a better product?" It was "who has more capital and more willingness to lose money?" Both survived, but the market structure was permanently shaped by the attrition contest.
Retail & Consumer
You're seeing a war of attrition when a large retailer enters a local market and prices below cost, absorbing losses that the local competitor cannot match. Walmart's expansion strategy included systematically entering markets with predatory pricing, absorbing losses in individual stores while subsidising them from national profits. The local competitor faces a choice: match the prices and bleed cash, or maintain margins and lose customers. The war of attrition is asymmetric because Walmart's resource base spans thousands of stores while the local competitor has one.
Legal & Intellectual Property
You're seeing a war of attrition when a large corporation files patent infringement suits against a small competitor, knowing that the legal costs alone — regardless of the merits — may be enough to force a settlement or exit. The contest is not about who is right. It is about who can afford to keep fighting. Patent trolls exploit this dynamic explicitly: they file suits not to win in court but to extract settlements from defendants who calculate that the cost of litigation exceeds the cost of settling.
Labour & Talent
You're seeing a war of attrition when two companies bid up compensation for a scarce talent pool, each hoping the other will accept a hiring miss rather than continue escalating offers. The contest is especially acute in markets with network effects, where each additional engineer or salesperson produces outsized returns. The company with more capital can sustain higher offers longer — but the war destroys margins for both sides.
Section 3
How to Use It
The strategic use of war of attrition analysis is not to start wars of attrition — it is to recognise when you are in one, assess whether you can win, and decide whether to fight or withdraw.
Decision filter
"Before entering or continuing any competitive contest, I must answer: (1) Is this a war of attrition? (2) Do I have a resource advantage that will outlast the competition? (3) Is the prize worth the total cost of the contest — not just the cost so far? If the answer to any of these is no, I should exit or avoid the contest entirely."
As a founder
The founder's greatest risk in a war of attrition is asymmetry. A well-funded competitor can sustain losses that would destroy a startup. The strategic response is to avoid wars of attrition entirely by competing on dimensions where resource advantage doesn't determine the outcome: speed, innovation, niche focus, customer intimacy. If you find yourself in a war of attrition with a competitor that has 10x your resources, you are in the wrong contest. Change the game.
As an investor
Evaluate whether your portfolio companies are entering wars of attrition — and whether they have the capital to win. The most common value destruction in venture capital is funding companies to fight attrition wars they cannot win. Subsidised unit economics, aggressive market expansion into defended territories, and "win the market at any cost" strategies are all war of attrition plays. They only work if you have more capital than all competitors and the willingness to deploy it. If either condition is missing, the investment is a sunk cost machine.
As a decision-maker
Recognise the escalation trap. The most dangerous moment in a war of attrition is the moment when you've invested enough to make quitting painful but not enough to guarantee winning. This is where most irrational escalation occurs. The discipline is to evaluate the decision to continue solely on the basis of future expected value — ignoring sunk costs entirely. If the expected future cost of winning exceeds the future value of the prize, exit. The investment already made is gone regardless.
Common misapplication: Assuming that all competition is a war of attrition. Many competitive contests are won by differentiation, innovation, or superior strategy — not by endurance. The war of attrition framework applies specifically to contests where competitors are offering similar value and the differentiator is persistence. If you can win through a superior product, a better customer experience, or a unique market position, you should — because those victories are cheaper and more sustainable than attrition victories.
Amazon's retail strategy from 1994 to the mid-2010s was a deliberately asymmetric war of attrition. Bezos accepted near-zero operating margins in retail — reinvesting every dollar of profit into infrastructure, logistics, and price reductions — while competitors operated on conventional retail margins of 3-5%. The attrition was asymmetric because Amazon's investors (and later, AWS profits) funded the retail war, while competitors had to fund their operations from retail earnings alone. Borders, Circuit City, and hundreds of other retailers were bled out by a competitor that was willing to lose money on every transaction while building an infrastructure advantage that would eventually produce dominance. Bezos's insight: in a war of attrition, the side that can sustain losses longest wins. His innovation: finding an external funding source (investors, then AWS) to subsidise the war indefinitely.
Son's Vision Fund strategy was an attempt to win multiple wars of attrition simultaneously by providing overwhelming capital to portfolio companies. The theory: if you give a ride-sharing company or a co-working company enough capital to sustain losses longer than any competitor, they will win the market. The strategy worked when the competitor was also venture-funded (because SoftBank's $100 billion fund was larger than any competitor's war chest). It failed when the competitor was a company with structural cost advantages or when the war of attrition never produced the monopoly economics that would have justified the investment. WeWork's implosion demonstrated the boundary condition: a war of attrition only works when winning the war produces a durable advantage. If the winner emerges into a market that still has competition, the attrition investment is wasted.
Buffett's approach to wars of attrition is avoidance. He famously invests in businesses with "moats" — competitive advantages that make wars of attrition unnecessary. A company with a genuine moat doesn't need to outlast competitors because competitors cannot replicate its advantage. Buffett's dictum — "In a commodity business, your cost position is your moat" — is the war of attrition reduced to its essence: if you must fight an attrition war, be the low-cost producer. But the better strategy is to avoid the war entirely by operating in a business where you have a structural advantage.
Section 6
Visual Explanation
War of Attrition — Two competitors deplete resources until one exits. The winner inherits the market; the loser writes off the investment.
Section 7
Connected Models
Reinforces
Game Theory
The war of attrition is a formal game-theoretic model. The Nash equilibrium in a war of attrition is a mixed strategy where each player randomises their persistence based on the value of the resource and the cost of the contest. Understanding the game theory behind attrition prevents naive strategies — like committing unconditionally to a contest without calculating the expected payoff.
Tension
Competitive Advantage
A genuine competitive advantage — a moat — eliminates the need for wars of attrition. If you have a structural advantage your competitor cannot replicate, you don't need to outlast them. You outperform them. The tension: companies that invest in wars of attrition are often companies that lack genuine competitive advantages. The attrition is a substitute for differentiation.
Reinforces
Sunk [Cost](/mental-models/cost) Fallacy
The sunk cost fallacy is the primary psychological fuel of wars of attrition. The more you've invested, the harder it is to quit. Wars of attrition exploit this bias: once a competitor has invested heavily, they are psychologically locked in — even when rational analysis says they should exit. Recognising the sunk cost fallacy is the primary defence against irrational escalation.
Reinforces
Survivorship Bias
Section 8
One Key Quote
"There is no instance of a nation benefiting from prolonged warfare."
— Sun Tzu, The Art of War
Sun Tzu's insight — written two and a half millennia ago — captures the fundamental problem with wars of attrition: both sides lose. The winner emerges depleted. The loser is destroyed. The only way a war of attrition is worthwhile is if the prize is valuable enough to justify the total cost to the winner. In most competitive contexts, the prize isn't. The best strategy is to avoid the war entirely — through differentiation, through pre-emption, or through the wisdom to walk away before the escalation trap closes.
Section 9
Analyst's Take
Faster Than Normal — Editorial View
The war of attrition is the most expensive failure mode in business — and the most common one in venture-backed markets. The VC model systematically produces wars of attrition by funding multiple competitors in the same market, each with enough capital to sustain losses but not enough to guarantee victory.
The ride-sharing era was a masterclass in value destruction through attrition. Uber and Lyft collectively burned over $20 billion in the US market alone, competing on subsidised rides that neither company could profitably sustain. The "winner" — Uber — emerged with lower margins than a regulated utility. The war was won, but the prize was a commoditised market with structurally thin margins.
The key strategic question is not "can we win?" but "what do we win?" A war of attrition that produces a monopoly with pricing power is worth fighting. A war of attrition that produces a scarred winner in a still-competitive market is not. The difference between Amazon's retail attrition war (which produced genuine structural advantages) and WeWork's co-working attrition war (which didn't) is entirely about what was on the other side of winning.
The smartest players avoid wars of attrition entirely. Buffett avoids commodity businesses. Apple avoids price competition. Costco uses membership fees to fund lower prices, making attrition irrelevant because the economic model doesn't depend on winning a price war. The companies that consistently create the most value are the ones that have found ways to compete that don't reduce to endurance contests.
The escalation trap is real and deadly. Once an organisation has invested significantly in a war of attrition, the sunk cost psychology makes exit almost impossible. Boards, executives, and teams who have committed resources feel that quitting wastes what has already been spent. The rational response — "the sunk cost is gone regardless" — is emotionally unbearable. This is why the time to decide whether to fight a war of attrition is before you enter it, not after you've been fighting for two years.
Section 10
Test Yourself
Are these wars of attrition, and are they being handled correctly?
Scenario 1
Two streaming services are both losing money on content spending, each investing $15-20 billion per year in original programming. Neither has a clear quality advantage. Both are betting that the other will reduce spending first, allowing the survivor to raise prices.
Scenario 2
A startup enters a market dominated by a large incumbent. Rather than competing on price or features — where the incumbent has a massive advantage — the startup focuses on a specific niche that the incumbent ignores, builds deep expertise, and grows within that niche.
Scenario 3
A company has spent $50 million over three years trying to enter a new market, with no path to profitability. The CEO argues: 'We've invested too much to walk away now. Another $20 million should get us to breakeven.'
The foundational work on the war of attrition as a formal game-theoretic model. Maynard Smith's evolutionary stable strategy framework explains when attrition contests arise, how they resolve, and what optimal strategies look like.
Greenwald's framework for analysing competitive advantage explicitly addresses when competition becomes attrition-based and how to avoid it. The book argues that most competitive strategy is simpler than it appears — and that avoiding attrition contests is the highest-leverage strategic decision.
Helmer's taxonomy of competitive advantages ("powers") explains what prevents wars of attrition. Each of the seven powers — scale economies, network effects, counter-positioning, switching costs, branding, cornered resources, and process power — is a mechanism for making attrition unnecessary. Companies with genuine powers don't fight attrition wars. Companies without them do.
The original strategic text on attrition warfare and its alternatives. Sun Tzu's central argument — that the best general wins without fighting — is the highest expression of the anti-attrition strategy. Every subsequent military and business strategist who has argued for avoiding wars of attrition is ultimately restating Sun Tzu.
Hoffman explicitly advocates for entering wars of attrition when the prize is a winner-take-all market. Blitzscaling is the strategy of deliberately accepting attrition costs to grow faster than competitors. The book makes the case for when attrition is worth it — and implicitly reveals the conditions under which it isn't.
We study the winners of wars of attrition and conclude that persistence pays. We don't study the losers — the companies that persisted just as long but ran out of capital first. Survivorship bias makes wars of attrition look more attractive than they are. For every Amazon that won through attrition, there are dozens of companies that fought the same war and lost everything.
Reinforces
Asymmetric Warfare
The optimal counter to a war of attrition is asymmetric warfare: competing on dimensions where the opponent's resource advantage doesn't help. A startup fighting a war of attrition against a well-funded incumbent will usually lose. A startup fighting an asymmetric war — competing through speed, niche focus, or novel technology — can win because the contest is no longer about endurance.
Tension
First-Mover Advantage
First-mover advantage can sometimes prevent a war of attrition by establishing a dominant position before competitors enter. But when two first-movers enter simultaneously — as in many VC-funded markets — the first-mover advantage is neutralised and the contest devolves into attrition.
