Case Construction Playbook
Build cases that win rounds — framework, contentions, evidence cards, and airtight warranting.
Framework
Value + value criterion (LD)
Every LD case needs a value (the ethical end being pursued) and a value criterion (the standard that measures progress toward the value). Together they tell the judge how to decide the round. The value is the abstract goal; the criterion is the operational test. A common error is choosing a value that doesn't fit the resolution — 'justice' is generic and beats 'morality' on almost any topic, but choosing 'autonomy' on a resolution about military intervention forces you to defend an unusual hierarchy of rights. Top debaters develop a stable of 3-4 frameworks they can deploy across topics, tested in practice rounds against the most common opposing frameworks (util, structural violence, contractarianism).
Key Points
- Common values: justice, morality, governmental legitimacy, security, human dignity, autonomy, equality.
- Common criteria: maximizing welfare (utilitarianism), protecting autonomy (Kant), structural equality (Rawls), capabilities (Sen, Nussbaum), preventing structural violence (Galtung).
- Your criterion must directly test the resolution — 'maximizing welfare' fails if the resolution is about rights that override welfare; 'autonomy' fails on resolutions about public goods.
- Modern progressive LD often uses 'ROB' (role of the ballot) framing instead of traditional V/VC, e.g., 'the role of the ballot is to vote for the side that best dismantles structural antiblackness.'
- Phil-heavy frameworks (Kant, Hobbes, Levinas) win rounds when the opponent can't engage with the philosophy — read the source text, not just secondary summaries.
- Pre-empt the framework debate by reading 2-3 warrants for why your criterion is preferable (e.g., 'util ignores rights,' 'rights are deontologically prior') in your AC.
- If you lose the framework debate, you can still win by collapsing to a 'permissibility' or 'presumption' argument under their framework.
Role of the ballot / burden of proof (PF + Policy)
In PF and Policy, frameworks tend to be 'role of the ballot' (RoB) statements or burden allocations that tell the judge what winning looks like. A typical PF framework: 'Prefer the side that demonstrably reduces measurable harm — speculative impacts should be weighed lower than concrete ones.' A typical Policy framework: 'The judge is a policymaker who chooses between the plan and the status quo or a competitive counterplan, weighing net benefits.' Frameworks function as tie-breakers when both sides win offense — if you don't establish one, the judge will weigh on their default paradigm, which you don't control.
Key Points
- Link your framework to measurable impacts — abstract frameworks lose to concrete ones in front of most judges.
- Pre-empt the other side's likely framework by addressing it directly in your constructive.
- In PF, the most common modern framework is 'cost-benefit analysis' — but this is too vague to do real work; layer it with a probability or magnitude weighing standard.
- Burden allocation arguments ('the aff must prove inherency,' 'the neg has the burden of disproof on theory') are procedural framing — under-deployed but very high leverage.
- On util-vs-deontology debates, the side with cleaner evidence on their preferred framework's criteria usually wins.
- Skip framework only on lay-judge rounds — every flow judge weighs framework first.
Contentions
The claim-warrant-impact structure
Every contention follows three moves. Skip one and the argument collapses under scrutiny. The claim is the headline; the warrant is why it's true; the impact is why the judge should care. Most novice cases have a claim and an impact but no warrant — meaning the argument is asserted but not proven. Top debaters spend 60-70% of contention time on warrants, because warrant strength is what survives cross-examination.
Claim
The one-sentence statement of the argument: 'Rent control causes a long-run decline in housing supply.' Specific, contestable, and provable.
Warrant
The reason the claim is true — evidence, logic, or both. 'A 2019 Stanford study of San Francisco (Diamond, McQuade, Qian, American Economic Review 109(9)) found rent-controlled buildings were 15% more likely to convert to condos or be demolished within 5 years of enactment, reducing rental supply by 6% citywide.' Includes the mechanism: why does rent control cause this?
Impact
Why the judge should care — what harm, benefit, or value-link the claim produces. 'A declining supply pushes rents up city-wide, disproportionately harming new and low-income tenants — the policy's intended beneficiaries. The same Stanford study estimates a 5.1% citywide rent increase, displacing 25,000 tenants.' Quantify wherever possible.
Internal links
The chain between warrant and terminal impact. 'Supply decline → rent increase → displacement → loss of community ties → health and education harms.' Every link must be supported; the argument is only as strong as its weakest internal link.
Impact calculus
Comparative weighing built into the contention itself: 'This outweighs because the harm hits the policy's intended beneficiaries — making the policy actively counterproductive, not just inefficient.'
Evidence cards
Policy and PF debaters carry 'cards' — a tag + citation + underlined quotation + analytical summary. A standard card has four parts: the tag (an argument-style headline), the citation (author, credentials, publication, date), the underlined evidence (the specific words supporting the claim), and the un-underlined context (not read but available if challenged). Card quality matters more than quantity at the national circuit — judges weigh evidence on author credentials, recency, and warrant clarity.
Key Points
- Tag: a 6-10 word argument-style headline, not a neutral description. 'Rent control reduces supply 15% (Stanford 2019)' beats 'Stanford study about rent control.'
- Citation: author (credentials), publication, date, URL, page number if applicable. Bare URLs are insufficient at most tournaments.
- Underline only the specific words that support the claim — judges may ask to see the card and compare underlining to the full text.
- Bracket clarifications (e.g., '[the bill]') are permitted but cannot change the author's meaning.
- Card cutting is faster with Verbatim (free Policy/LD plugin for Word) or Debate Sync — every top program uses one.
- Recency matters: evidence older than 5-10 years is often dismissed as outdated on empirical claims, though older sources are fine for philosophical and historical claims.
- Author credentials beat publication brand for warrant weight — a working economist beats a journalist citing the same study.
Blocks: pre-written responses
Blocks are pre-written, evidence-backed responses to the arguments you'll most often face. They're how top debaters cover huge amounts of ground in 3-minute rebuttals: 80% of what comes out of their mouth was prepared in advance. A complete block file is organized by argument type (DA, K, T, framework) and contains 30-60 second responses for each common argument, plus blocks against the most common 'frontlines' (responses to your responses).
Key Points
- Organize block files by argument category and resolution — Drive folders with consistent naming work best.
- Each block should have a 30-second 'short' version and a 90-second 'long' version.
- Always include 2-3 cards per block — one analytical, one card, one impact comparison.
- Update blocks every 2-3 weeks during the topic cycle as new arguments emerge.
- Disclose blocks on the wiki where required — non-disclosure is itself a theory argument many judges will vote on.
Worked Examples
PF sample contention — AI regulation
Resolution: 'The US federal government should regulate generative AI models.' Pro contention 1, with full claim-warrant-impact structure:
Key Points
- Claim: Unregulated generative AI is already producing measurable harms in elections, labor markets, and mental health — federal baseline rules are necessary to internalize these externalities.
- Warrant 1 (elections): Stanford Institute for Human-Centered AI's 2024 AI Index documented a 2.0x increase in AI-driven political disinformation incidents between the 2020 and 2024 primary cycles, including the New Hampshire Biden robocall (January 2024).
- Warrant 2 (labor): OECD's 2024 Employment Outlook found 27% of jobs are at high risk of automation from generative AI in OECD countries, with administrative and clerical roles facing the steepest displacement.
- Warrant 3 (mechanism): Without federal baseline rules, states are passing fragmented laws (CA AB 2013, NY S.8214, TX HB 2060) — the resulting patchwork slows compliance and favors large incumbents who can afford 50-state legal teams.
- Impact: Federal regulation internalizes externalities and prevents regulatory capture. EU AI Act (in force August 2024) demonstrates feasibility; absence of US federal action cedes standard-setting to Brussels.
- Weighing: Outweighs status-quo benefits because the harms (election integrity, labor displacement) are accelerating exponentially while regulatory capacity grows linearly — first-mover advantage to regulators is closing.
LD sample case — capital punishment
Resolution: 'Capital punishment ought to be abolished.' Affirmative framework + first contention, demonstrating how value, criterion, and contention interlock:
Key Points
- Value: justice — understood as treating persons in accordance with their rational nature.
- Value criterion: respecting the inherent dignity of persons (Kantian categorical imperative — never treat persons as means only).
- Contention 1: The death penalty treats persons as means to deterrence ends — violating the categorical imperative. Kant himself defended capital punishment, but modern Kantians (Reiman 1985) argue the dignity principle requires abolition.
- Contention 2: Empirical deterrence evidence is null — Donohue & Wolfers (2005, Stanford Law Review) demonstrate studies showing deterrence are statistically underpowered; National Research Council's 2012 review concluded existing studies cannot identify a deterrent effect.
- Contention 3: Irreversibility under conditions of error — Innocence Project documents 200+ post-conviction DNA exonerations including 21 from death row; the Bureau of Justice Statistics estimates 4.1% of death row inmates are factually innocent (Gross et al. 2014, PNAS).
- Weighing: Under Kantian framework, even a single innocent execution is categorically unjust — magnitude calculus doesn't apply because dignity violations are not commensurable with deterrence benefits.
Policy sample affirmative — federal jobs guarantee
Resolution: 'The United States federal government should substantially increase fiscal redistribution in the United States by adopting a federal jobs guarantee.' Affirmative case skeleton with stock issues mapped:
Key Points
- Plan text: 'The United States federal government should establish a federal job guarantee providing a public-sector employment option at $15/hour plus benefits to all citizens able and willing to work.'
- Inherency: Status quo unemployment hovers 3.5-5% but underemployment + labor force non-participation total 12-15% (Bureau of Labor Statistics, U-6 measure).
- Harms: Unemployment causes documented mortality increases (Sullivan & von Wachter 2009, QJE — 50-100% mortality increase for displaced workers), mental health harms, and intergenerational poverty.
- Solvency: Levy Economics Institute (Tcherneva 2018) modeled a $543B/year federal job guarantee covering 15M workers, eliminating involuntary unemployment within 18 months.
- Topicality: 'Fiscal redistribution' includes direct employment — federal employment is a transfer payment functionally equivalent to UBI but conditioned on labor.
- Disadvantages preempted: Inflation DA (Tcherneva: program acts as automatic stabilizer, anti-inflationary in expansions); spending DA (deficit-financed but offset by reduced welfare and tax revenue gains).
FAQ
How many contentions should my case have?
Two to three. More than that and you can't develop any of them in the time available — and undeveloped contentions are conceded contentions. In LD, two contentions plus framework is the modal structure. In PF, two contentions per side is standard. In Policy, the affirmative typically runs one advantage with multiple internal links, plus a more developed solvency story. Pick the strongest and deepest over broadest — depth wins clash.
What are 'blocks' and do I need them?
Blocks are pre-written responses to the most common arguments you'll face. Yes, build them — they free up your in-round brain to weigh and extend. Every top debater has a block file organized by argument: by the time of state and national tournaments, the file should contain 50-200 blocks, organized by argument type and disclosed on the wiki where required. Update blocks after every tournament based on what arguments you actually encountered.
Should I use AI tools to cut cards?
AI tools are useful for finding sources and summarizing context but cannot replace reading the actual source. The NSDA's evidence rules require accurate representation — pasting LLM-generated summaries as evidence is misrepresentation and can result in disqualification. Use AI for source discovery (e.g., 'find me peer-reviewed work on rent control in Berlin'), then verify and cut the actual evidence yourself.
What does a novice's first case need?
Framework (60 seconds), two contentions with claim-warrant-impact each (90 seconds per contention), and a 30-second impact comparison at the end. Total: ~4 minutes — fits any constructive speech. Don't try to fit in three contentions, don't run kritiks, don't read theory. Win on clarity and weighing.
Continue learning
Explore related MUN guides to deepen your skills.