Checklist · Litigation

Evaluating a New Civil Case: A Practical Checklist

Type
Checklist
Jurisdiction
National
Date
2025-04-21
Topic
Litigation

Evaluating a new civil case is the discipline of deciding, before real money is committed, whether a claim is worth bringing or worth defending. This checklist runs that decision as a sequence of gates worked in the order a careful evaluator works them: clearing conflicts and papering the engagement, preserving evidence before the file grows cold, building a proof matrix and pinning the burden of proof for every element, scoring liability and damages as separate variables that multiply rather than add, stress-testing collectibility, screening the threshold defenses that kill cases before the merits, budgeting the fight stage by stage, and converting the whole picture into a number and a settlement range through a decision tree and expected-value analysis. Each phase pairs concrete steps with the governing authority and the specific failure modes that catch even seasoned litigators. It is general information, not legal advice; limitations periods, jurisdictional rules, damages caps, and ethics rules vary by jurisdiction and change over time.

Evaluating a New Civil Case: A Practical Checklist

Litigation > Civil Litigation Strategy | Published 21 April 2025 - Updated 31 August 2025 | UpLaw - uplaw.us

A sound case evaluation answers one blunt question: if you spend the money, what do you actually get back? Everything else in this checklist is a structured refusal to answer that question by instinct.

Keywords: case assessment, early case assessment, conflicts check, engagement letter, litigation hold, proof matrix, burden of proof, liability and damages, collectibility, statute of limitations, personal jurisdiction, subject matter jurisdiction, venue, litigation budget, fee arrangements, decision tree, expected value, settlement range, spoliation

A civil case is a capital investment with an uncertain payout and a long lock-up period. Treating it as anything softer than that is how lawyers fall in love with cases that will bleed them and walk away from cases that would have paid for the office. The purpose of a disciplined evaluation is not to eliminate uncertainty, which is impossible, but to price it honestly and early, while there is still time and room to act on the answer.

The best practitioners approach a new matter the way a diligent buyer approaches an acquisition: they refuse to let enthusiasm substitute for underwriting. Corporate law departments face relentless pressure to hold down legal spend, yet no one can predict with precision when a suit will land, what it will raise, how long it will run, or what it will cost. The honest response is not paralysis and not bravado. It is a repeatable method that converts a messy dispute into a defensible number and a set of options.

This checklist runs the evaluation as a sequence of gates, worked in roughly the order a careful evaluator works them, with a stop-and-reassess whenever a gate fails. The gates apply to plaintiffs and defendants alike; the arithmetic simply runs in mirror image. For the full narrative treatment of the same method, with extended hypotheticals, see evaluating a civil case: turning instinct into a repeatable method and, from the defense side, assessing a civil lawsuit: a comprehensive early-case framework.

How to use these gates

Think of the seven phases below as tollbooths, not a to-do list. A case that cannot pay the toll at an early gate rarely improves at a later one, so a failure at Gate 4 (collectibility) or Gate 5 (threshold defenses) should send you back to renegotiate the engagement or reframe the strategy before you sink discovery dollars into a merits fight you may win and still lose. Two of the gates, however, run on a different clock. Conflicts (Gate 1) and preservation (Gate 2) cannot wait for the leisurely part of the analysis; both must be handled the moment a matter walks in the door, because a mistake at either is often unfixable. Work the gates in order, but treat the first two as reflexes rather than deliberations.

A word on posture. Everything here is framed around a decision — sue, defend, settle, decline — rather than around winning. Winning is a possible outcome, not a plan, and a plan that only works if you win is not a plan.

Phase 1: Conflicts, competence, and the engagement

  • [ ] Run a conflicts check against the firm's current-client and adverse-party databases — including corporate parents, subsidiaries, and affiliates; individual principals; and known key witnesses — at intake, before anyone falls in love with the matter (Model Rules of Prof'l Conduct 1.7, 1.9).
  • [ ] Determine whether any conflict is consentable and, if so, obtain informed consent confirmed in writing.
  • [ ] Confirm you (and the firm) are competent to handle the matter or will associate counsel who is (Model Rule 1.1).
  • [ ] Define the client precisely in writing: who the client is, who it is not, the scope of the representation, the allocation of authority between lawyer and client, and how fees and costs work.
  • [ ] For any contingent-fee or fee-shifting matter, put the fee terms in a signed writing that states the method of calculation and how expenses are handled (Model Rule 1.5(c)).
  • [ ] Screen the claim itself for a good-faith, non-frivolous basis in law and fact (Model Rule 3.1; Fed. R. Civ. P. 11(b)).

Why this gate comes first

Conflicts are dispositive in a way merits problems rarely are, and the analysis gets harder to run honestly once you have invested in the matter and formed a view of the client. A missed conflict can produce mid-case disqualification, disgorgement of fees already earned, malpractice exposure, and a bar complaint — a four-way loss that no favorable verdict repairs. The current-client and former-client rules do different work: Rule 1.7 governs simultaneous representations and imputed conflicts across the firm, while Rule 1.9 forbids taking a matter adverse to a former client that is the same as or substantially related to the prior engagement. Both sweep in affiliates and principals, which is why a clean check against the named parties alone is not a clean check.

Defining the client is not boilerplate. It controls the conflicts analysis (whom you may not oppose), the privilege analysis (whose communications are protected), and the billing analysis (whom you may charge). In closely held companies and family enterprises, the trap is representing "the company" while the founder assumes you also represent her personally; the engagement letter is where you either close that gap or leave it open for a later malpractice claim. For a plain-English tour of practice areas and how fee structures are set at the outset, see types of lawyers: a field guide to legal specialties and choosing the right one.

Hypothetical. A boutique firm is asked to sue Meridian Foods for breach of a supply contract on behalf of Aurora Bakeries. The conflicts run flags nothing under "Meridian Foods." Two weeks in, an associate discovers that Meridian is a wholly owned subsidiary of Northwind Holdings — a current firm client on an unrelated tax matter. Had the intake screen captured corporate parents, the firm would have caught a Rule 1.7 problem at the door, when it could have declined cleanly or sought a waiver. Caught two weeks later, after work product exists and the client expects action, the firm faces disqualification and an awkward conversation about fees it may have to return.

The competence and non-frivolousness checks belong here too, and they are not throat-clearing. Rule 11 exposes counsel to sanctions for filing claims without a reasonable pre-suit inquiry, and a claim that cannot survive that inquiry should never reach the later gates at all.

Phase 2: Preserve the evidence — before you do anything else

  • [ ] Issue a written litigation hold the moment litigation is reasonably anticipated — frequently before the evaluation is finished, and long before a complaint is served.
  • [ ] Identify custodians and data sources (email, messaging apps, cloud storage, mobile devices, ephemeral messaging, backups, and third-party platforms) and suspend routine deletion, auto-purge, and document-destruction schedules.
  • [ ] Notify former employees, agents, and third parties within the client's control of the preservation obligation.
  • [ ] Track the hold: who was notified, when, what was preserved, and how the hold is being monitored and re-circulated.

Why preservation cannot wait for the rest of the analysis

Spoliation can transform a winnable case into a losing one regardless of the merits, and it does so on the defendant's timeline, not yours. The duty to preserve attaches when litigation is reasonably foreseeable, not when the summons arrives — the principle crystallized in the Zubulake line of decisions, which held that once a party reasonably anticipates litigation it must suspend its ordinary retention policy and put a hold in place. Zubulake v. UBS Warburg LLC, 220 F.R.D. 212 (S.D.N.Y. 2003).

Federal Rule of Civil Procedure 37(e), rewritten in 2015, now supplies a graduated response for lost electronically stored information. If ESI that should have been preserved is lost because a party failed to take reasonable steps, and it cannot be restored or replaced, a court may order measures no greater than necessary to cure the prejudice. But if the court finds the party acted with the intent to deprive another of the information's use, it may presume the lost information was unfavorable, instruct the jury that it may or must draw that inference, or dismiss the action or enter default. That intent finding is the difference between an annoyance and a case-ending catastrophe, and it turns on conduct that often occurs in the first weeks — precisely when the evaluation is still underway.

This is the one part of case evaluation you complete first and analyze second. Preservation dovetails with the front end of discovery planning; the same custodian mapping and data-source inventory feed directly into the Rule 26(f) conference and initial disclosures. See Rule 26 initial disclosures and the discovery plan: a practical checklist and, for the tools and pitfalls that follow, federal civil discovery: mastering the tools, rules, and pitfalls.

Hypothetical. Northwind Logistics receives a lawyer's letter threatening suit over a warehouse fire. General counsel forwards it to operations with a note to "look into it," but no one suspends the video system's 30-day overwrite cycle. By the time counsel is formally retained six weeks later, the footage of the loading dock is gone. Even if Northwind's substantive defense is strong, it now spends the case litigating whether the deletion was merely negligent (curative measures under Rule 37(e)(1)) or intentional (an adverse-inference instruction under Rule 37(e)(2)). The merits have not changed; the odds have, because the evaluation ignored the clock on the evidence.

Phase 3: The merits — elements, burden, and the proof matrix

  • [ ] Write out the elements of each claim, counterclaim, third-party claim, and affirmative defense, drawn from the controlling statute or common law, the pattern jury instructions, and the verdict form.
  • [ ] Build a proof matrix: for every element, name the witness with personal knowledge, the documentary or physical evidence, and the anticipated evidentiary hurdles (hearsay, authentication, foundation) with your answers to each.
  • [ ] Identify the standard of proof for each claim — preponderance of the evidence for most civil claims, but clear and convincing evidence for fraud, punitive damages, willfulness, and certain equitable relief in many jurisdictions.
  • [ ] Score liability and damages as separate probabilities and values, and remember they multiply, not add.
  • [ ] For each damages category — compensatory, consequential, statutory, punitive, and any fee award — ask three questions in order: is it available at law, is it provable on this evidence, and is it capped?
  • [ ] Draft a one-paragraph opening and a one-paragraph closing for both sides. If you cannot write the other side's closing, you do not yet understand your own case.

The proof matrix is where cases are actually won and lost

It is easy to say the defendant breached the contract. It is harder to fill in the cell that names the witness with personal knowledge who will say so under oath and the exhibit that proves it over a hearsay objection. Empty cells in the proof matrix are the true measure of a case's strength, and they are invisible to the gut. Building the matrix forces the evaluation off unexamined impressions and onto verification: for each element, who testifies, what document corroborates, and what objection stands in the way.

The discipline of drafting both sides' opening and closing is the same tool aimed at your blind spots. It converts "the case is defensible" into a demonstration, and it surfaces the theme the other side will hammer at trial while there is still time to blunt it. A timeline of key events — with each side's version laid side by side — does the same work for the facts, exposing the discrepancies that will decide the case.

Keep liability and damages on separate axes

The most common valuation error in the profession is collapsing liability and damages into a single feeling of "strong" or "weak." They are independent variables, and expected value is their product. A claim with 90% liability worth $20,000 is a small case. A claim with 30% liability worth $10 million is a large, dangerous one — dangerous to whichever side is complacent about it. A lawyer who reports "this case is strong" without separating the probability of winning liability from the magnitude and probability of the damages has not evaluated the case; he has had a feeling about it.

The anatomy of the damages model

Available, provable, capped — run every category through those three filters.

Compensatory (expectation) damages put the injured party where performance would have. In contract, foreseeability limits recovery to losses that were within the parties' contemplation at formation — the enduring rule of Hadley v. Baxendale, 156 Eng. Rep. 145 (1854), carried forward in the Restatement (Second) of Contracts § 351. Lost profits are recoverable but must be proved with reasonable certainty, and the certainty requirement is where many ambitious damages models quietly die. For the mechanics of computing breach damages, from the expectation measure to mitigation and liquidated-damages clauses, see contract damages: how courts make the injured party whole.

Consequential and incidental damages extend to downstream losses, subject to foreseeability and mitigation. The mitigation doctrine can shave a headline number substantially, and a defendant's evaluation should probe it early.

Statutory damages change the arithmetic entirely, because they detach recovery from proof of actual loss. Many statutes also authorize multiplied (treble) damages and fee-shifting, which can turn a modest actual loss into a serious exposure. When the claim sounds in intellectual property, the damages toolkit — reasonable royalties, disgorgement, statutory damages, and enhancement — deserves its own dedicated model; see what an IP case is worth: a guide to intellectual property damages. Employment claims carry their own layered structure of back pay, front pay, capped compensatory and punitive awards, and liquidated damages, mapped in damages in employment litigation: back pay, statutory caps, and punitive awards. And when the defendant is a government entity, sovereign immunity and statutory damage limits reshape the whole calculation; see suing the government: a guide to damages in administrative and public-entity litigation.

Punitive damages carry a constitutional ceiling. The Due Process Clause forbids grossly excessive awards, and the Supreme Court has supplied both guideposts and a rule of thumb: reprehensibility, the ratio of punitive to compensatory damages, and comparable civil penalties. BMW of North America, Inc. v. Gore, 517 U.S. 559 (1996). State Farm Mutual Automobile Insurance Co. v. Campbell, 538 U.S. 408 (2003), added that few awards exceeding a single-digit ratio between punitive and compensatory damages will satisfy due process, and Philip Morris USA v. Williams, 549 U.S. 346 (2007), held that a jury may not punish a defendant for harm inflicted on non-parties. Any evaluation that books a large punitive number without discounting for these ceilings is overvaluing the case.

Hypothetical. Delacroix Robotics sues a former supplier for shipping defective actuators, claiming $4 million in lost profits on a canceled product line. The proof matrix looks strong on breach — a signed spec sheet, failed acceptance tests, contemporaneous emails. But the lost-profits cell is thin: the product line was new, with no track record, and the only witness on damages is the founder, whose projections a court may treat as speculative rather than reasonably certain. The liability probability might be 75%, but the provable damages, after the certainty discount, might be $900,000 rather than $4 million. Scored separately, the case is worth a fraction of the founder's headline figure — and that gap is exactly what a disciplined evaluation exists to reveal before the retainer is spent.

Phase 4: Collectibility — the judgment is only paper

  • [ ] Determine whether the defendant holds assets within reach of the forum, and where those assets sit.
  • [ ] Identify applicable insurance coverage and policy limits — often the true practical ceiling on recovery — and flag coverage defenses and reservations of rights.
  • [ ] Assess exemptions (homestead, wages, retirement accounts), thin capitalization, and senior secured creditors whose liens prime any judgment.
  • [ ] Assess bankruptcy risk, which would convert a judgment into an unsecured claim worth cents on the dollar — unless the debt fits a nondischargeability exception.
  • [ ] For corporate defendants, map the asset structure (holding versus operating entities) and honestly grade the difficulty of piercing the veil.

The most overlooked gate in the catalog

A money judgment is a piece of paper authorizing you to try to take the defendant's assets. If there are none within reach, the paper is décor. Collectibility is routinely treated as an afterthought — something to worry about after trial — which is exactly backwards. It belongs in the front-end evaluation because it can zero out the value of an otherwise excellent claim, and because a judgment-proof defendant will use that fact as its central piece of settlement leverage.

Insurance is usually the first place to look and often the last word. In many disputes the real ceiling on recovery is not the defendant's net worth but its policy limits, and a $10 million claim against a defendant with a $1 million policy and no meaningful assets is, as a practical matter, a $1 million claim — minus coverage defenses. Whether the carrier owes a duty to defend, a duty to indemnify, or has reserved its rights can matter more to the ultimate outcome than any element of the tort.

Corporate architecture is the other frequent surprise. A defendant that runs its business through a lightly capitalized operating entity while holding its real estate and intellectual property in separate affiliates has, deliberately or not, built a moat around its assets. Piercing that structure to reach the parent or a sibling entity is doctrinally difficult and fact-intensive, and an evaluation that assumes the veil will fall is usually indulging in optimism. For how these structures are built (and therefore how they resist collection), see running multiple businesses under one structure: a deep dive into holding, operating, and parent companies. For the individual side of the same problem — exemptions, domestic asset-protection trusts, and offshore structures, all governed by fraudulent-transfer timing — see domestic and offshore asset protection compared: a wealth-defense field guide.

Bankruptcy is the ultimate collectibility discount. A defendant who files converts your carefully won judgment into a general unsecured claim standing behind secured creditors, tax authorities, and priority claimants. The exceptions matter: debts arising from fraud, willful and malicious injury, or certain fiduciary defalcations may be nondischargeable, which is one reason the theory of the claim (fraud versus simple breach) has consequences far beyond the merits.

Hypothetical. Aurora Bakeries wins a clean $1.2 million verdict against Sable Construction for a botched build-out. Only then does it learn that Sable operates through a shell that leases every truck, subcontracts every crew, and holds no real property; the owner's home sits behind a homestead exemption and the equipment is titled to a separate leasing affiliate. The verdict is real. The recovery is close to zero. A collectibility memo at intake would have priced this case as a fraction of its face value and reframed the entire strategy — perhaps toward a personal-guarantee theory, a fraudulent-transfer claim, or an early, modest settlement — long before trial costs were incurred.

Phase 5: Threshold defenses that end cases before the merits

  • [ ] Calculate the correct limitations period for each claim; a single fact pattern can spawn claims running on different clocks.
  • [ ] Price accrual and tolling: the discovery rule, fraudulent concealment, minority or incapacity, and any tolling agreement.
  • [ ] Read the contract for a shortened contractual limitations clause before you read it for the breach.
  • [ ] Confirm subject-matter jurisdiction: federal question (28 U.S.C. § 1331), diversity with more than $75,000 in controversy (28 U.S.C. § 1332), and supplemental jurisdiction over related claims (28 U.S.C. § 1367).
  • [ ] Confirm personal jurisdiction: general ("at home") and specific (arising-from) jurisdiction, plus any consent basis.
  • [ ] Confirm venue (28 U.S.C. § 1391), and assess transfer (28 U.S.C. § 1404(a)), removal (28 U.S.C. §§ 1441, 1446), and any arbitration clause that diverts the dispute out of court entirely.

The most efficient case-killers

A meritorious claim filed one day late is worth nothing. The statute of limitations is the single most efficient case-killer in this catalog, and it rewards the party who does the arithmetic first. The subtlety is that one set of facts often supports several claims — breach of contract, breach of warranty, fraud, negligent misrepresentation — each with its own period and its own accrual rule. Under the UCC, an action for breach of a contract for sale must generally be commenced within four years after the cause of action accrues, and accrual for warranty runs from tender of delivery regardless of the buyer's knowledge. UCC § 2-725. A tort claim on the same transaction may run from discovery of the injury instead. Tolling adds another layer: equitable tolling can pause the clock where a plaintiff has pursued rights diligently but some extraordinary circumstance stood in the way, Holland v. Florida, 560 U.S. 631 (2010), and the filing of a putative class action can toll the individual claims of absent class members, American Pipe & Construction Co. v. Utah, 414 U.S. 538 (1974).

The contractual limitations clause is the trap even careful lawyers miss. Parties can, within limits, agree to a period shorter than the statutory one, and a two-year contractual bar buried in a supply agreement will defeat a claim brought in year three no matter how strong the breach. Read the limitations clause first.

Subject-matter jurisdiction is the defense that never expires

Personal jurisdiction and venue can be waived; subject-matter jurisdiction cannot. It can be raised by any party or by the court sua sponte, at any point, up to and including on appeal after a trial — which means a case litigated to verdict in the wrong forum can be dismissed years and hundreds of thousands of dollars in. Diversity practice hides two recurring problems: the amount in controversy (pleaded in good faith but subject to a legal-certainty test) and complete diversity (defeated by a single non-diverse party, and a favorite tool of plaintiffs seeking to stay in state court). The Class Action Fairness Act's expanded diversity provisions, 28 U.S.C. § 1332(d), change the calculus for aggregate claims.

General versus specific personal jurisdiction

Since International Shoe Co. v. Washington, 326 U.S. 310 (1945), personal jurisdiction has turned on "minimum contacts" such that suit does not offend traditional notions of fair play and substantial justice. Modern doctrine sorts contacts into two buckets. General (all-purpose) jurisdiction reaches a defendant only where it is "essentially at home" — for a corporation, ordinarily its place of incorporation and principal place of business. Goodyear Dunlop Tires Operations, S.A. v. Brown, 564 U.S. 915 (2011); Daimler AG v. Bauman, 571 U.S. 117 (2014). Specific (case-linked) jurisdiction requires that the claim arise out of or relate to the defendant's purposeful contacts with the forum. The Court loosened the tightest reading of "arise out of" in Ford Motor Co. v. Montana Eighth Judicial District Court, 592 U.S. 351 (2021), holding that a strong relationship among the defendant, the forum, and the litigation can support jurisdiction even without strict causation, while Bristol-Myers Squibb Co. v. Superior Court, 582 U.S. 255 (2017), had earlier reined in mass-tort forum shopping by insisting on that connection for each plaintiff's claim. And Mallory v. Norfolk Southern Railway Co., 600 U.S. 122 (2023), revived consent-by-registration jurisdiction where a state statute conditions the privilege of doing business on amenability to suit — a wrinkle worth checking in every registration state.

Venue, transfer, removal, and the arbitration off-ramp

Even with jurisdiction secure, venue and forum can reshape a case's value. A forum-selection clause is enforced through § 1404(a) and, absent extraordinary circumstances, will send the case to the contracted forum, with the plaintiff's choice given no weight. Atlantic Marine Construction Co. v. U.S. District Court, 571 U.S. 49 (2013). A defendant sued in state court may remove a qualifying case to federal court, and the removal clock is short and unforgiving. Most consequential of all, an arbitration clause can divert the dispute out of the public court system entirely; the Federal Arbitration Act's strong pro-enforcement policy has repeatedly overridden state-law obstacles, including bars on class-arbitration waivers. AT&T Mobility LLC v. Concepcion, 563 U.S. 333 (2011). Whether the forum is a courtroom or an arbitral panel changes discovery, appeal rights, cost, and privacy, and belongs in the evaluation from the start — see choosing a dispute-resolution forum: litigation, arbitration, and mediation compared and arbitration in practice: a comprehensive guide to alternative dispute resolution. Where a defendant sits abroad, service can become a strategy in itself; see serving a foreign defendant abroad: a Hague Convention checklist.

Some threshold defenses are equitable rather than strictly temporal — laches, waiver, acquiescence, and equitable estoppel — and they can bar relief even inside the limitations period in the right posture. For where these still bite, see equitable defenses: a comprehensive guide to laches, acquiescence, waiver, and equitable estoppel.

Hypothetical. A Vermont bakery sues an Arizona equipment maker in Vermont for a defective oven that caught fire on site. The manufacturer is not "at home" in Vermont, so general jurisdiction is out — but it sold ovens into Vermont through a regional distributor and marketed to New England bakeries, and the claim arises from exactly that activity. Under Ford, specific jurisdiction likely holds. Now flip one fact: suppose the sales contract contained a clause selecting Arizona courts. Under Atlantic Marine, the Vermont plaintiff's home-forum preference gets no weight, and the case is bound for Phoenix. The merits never entered the analysis; the forum did all the work.

For the procedural context in which these threshold questions play out — from complaint through dispositive motions and appeal — see federal civil litigation filings: a comprehensive guide from complaint to appeal and, for smaller parties weighing the fight, navigating federal civil litigation as a small business: a stage-by-stage guide.

Phase 6: The economics — fee structure and a stage-by-stage budget

  • [ ] Choose a fee structure (hourly, contingency, or an alternative fee arrangement) and note explicitly how it allocates risk between lawyer and client.
  • [ ] Confirm who bears litigation costs (as distinct from fees) and how they are advanced and recovered.
  • [ ] Build a stage-by-stage budget: pre-filing investigation, pleadings and early motions, fact discovery, expert discovery, class certification (if any), summary judgment, pretrial, trial, post-trial motions, and appeal.
  • [ ] Add the non-attorney costs clients chronically underestimate: experts, e-discovery vendors and hosting, jury or trial consultants, mediators, and settlement administrators.
  • [ ] Treat the budget as the denominator in every expected-value calculation, and revisit it at each decision point.

Who bears the risk

The fee structure is not an administrative detail; it is a risk-allocation instrument, and it changes the client's calculus at every later gate. The hourly model puts the risk of a long, uncertain fight on the client and gives the lawyer no stake in the outcome. A contingency arrangement shifts the risk to the lawyer, which aligns incentives on winning but can misalign them on when to settle — the lawyer's hourly-equivalent return and the client's recovery peak at different moments. Alternative fee arrangements (flat fees per phase, capped fees, hybrid discounted-hourly-plus-success) try to split the difference. Whatever the choice, contingent and fee-shifting matters require the terms in a signed writing under Model Rule 1.5(c), and the overall fee must be reasonable under Rule 1.5(a).

The American Rule and its exceptions

Baseline: each side pays its own lawyers, win or lose. That is the American Rule, reaffirmed in Alyeska Pipeline Service Co. v. Wilderness Society, 421 U.S. 240 (1975), and it silently shapes the value of every case — because litigation costs are usually unrecoverable, they come straight off the top of any recovery. The exceptions are where things get interesting: fee-shifting statutes (civil rights, many consumer-protection and IP statutes), contractual prevailing-party clauses, and sanctions. A one-sided attorney-fee clause can flip the economics of a modest dispute, turning a $50,000 claim into a bet where the fees dwarf the principal. Read for the fee clause with the same care you read for the limitations clause.

Building the budget

Litigation costs are lumpy and back-loaded. The pleadings are cheap; fact discovery, expert work, and trial are not, and in document-heavy matters e-discovery review is frequently the single largest line item in the case. A credible budget therefore is not one number but a schedule of numbers keyed to the stages a case actually passes through, because most cases resolve before trial and the client needs to see where the money goes if it does not. Budget each stage separately: pre-filing, pleadings and early motions (motion to dismiss, motion to compel arbitration, removal fights), fact discovery, expert discovery, class certification, summary judgment, pretrial, trial, post-trial, and appeal.

Two habits separate a real budget from a comforting one. First, count the non-lawyer costs — experts, hosting and processing fees for ESI, jury consultants, mediators, and, in class cases, settlement administrators — because these are the numbers clients forget and lawyers hate to raise. Second, keep asking the only question that matters at each decision point: not "what has this cost so far?" but "what will the rest cost, and is the remaining upside worth the remaining spend?" Sunk costs are sunk. The budget exists to keep the client from throwing good money after a case whose expected value has already turned negative. For the broader map of stages the budget tracks, see federal civil litigation: a roadmap and resource guide, and for the discovery line items that dominate most budgets, building a privilege log: a practical checklist.

Phase 7: Decision tree, expected value, and the settlement range

  • [ ] Map the case as a decision tree with explicit nodes: motion to dismiss, summary judgment, trial on liability, and a damages distribution (high / mid / low).
  • [ ] Assign an explicit probability to each node — a number, not an adjective.
  • [ ] Compute expected value: sum each terminal outcome times the product of the probabilities on its branch, then subtract remaining costs and discount for collectibility and the time value of money.
  • [ ] Derive each side's expected value to locate the zone of possible agreement — the settlement band where both parties beat trial.
  • [ ] Calibrate the probabilities against base rates and a skeptical colleague; litigators are systematically overconfident about their own motions.
  • [ ] Adjust for risk tolerance and intangibles: reputation, precedent, business relationships, and the value of certainty and closure.

Turning the case into a number

A case is not worth its best-case verdict. It is worth the probability-weighted average of all its outcomes, minus the cost of getting there, discounted for the odds of actually collecting and for the years the money is tied up. The decision tree is simply the machine that performs this arithmetic honestly. Its discipline is the requirement to write down numbers: "I feel good about summary judgment" becomes "I'd put our summary-judgment win at 35%," a figure the client can weigh, challenge, and hold you to as the case develops.

A worked expected-value example

Hypothetical. Aurora Bakeries (plaintiff) sues for $3,000,000 on a breach-of-contract theory against a solvent, insured defendant. The evaluation assigns these probabilities. There is a 35% chance the case is knocked out on the motion to dismiss or summary judgment, yielding a $0 recovery; so the probability of reaching a verdict is 65%. If the case reaches trial, the probability of a plaintiff's verdict on liability is 50%. If liability is found, the blended expected award (across high, mid, and low damages scenarios) is $2,000,000. Because there is some doubt about full collection, apply a collectibility factor of 0.9.

Plaintiff's gross expected recovery = 0.65 × 0.50 × $2,000,000 × 0.9 = $585,000.

Now net out the plaintiff's remaining fees and costs to trial, estimated at $350,000. The plaintiff's reservation price — the least she should rationally accept to settle — is roughly $585,000 − $350,000 = $235,000.

Run the mirror image for the defendant. Its expected liability is 0.65 × 0.50 × $2,000,000 = $650,000, and its own defense costs to trial are $300,000. The defendant's reservation price — the most it should rationally pay to make the case go away — is roughly $650,000 + $300,000 = $950,000.

The zone of possible agreement is $235,000 to $950,000. Any settlement inside that band leaves both sides better off than trial. The band is wide because both parties save large, unrecoverable litigation costs (the American Rule at work) and both dislike bearing trial risk. The surplus that makes settlement rational is precisely the $650,000 in combined fees neither side has to spend.

The example illustrates why so many cases settle and why the settlement figure so often looks unrelated to the "value" of the claim: the number is not the claim's face value but the intersection of two risk-and-cost-adjusted reservation prices. This is the practical meaning of bargaining in the shadow of the law — parties negotiate against the backdrop of what a court would likely do, discounted by what the fight would cost to find out. Robert H. Mnookin & Lewis Kornhauser, Bargaining in the Shadow of the Law, 88 Yale L.J. 950 (1979).

Calibrating against your own overconfidence

The decision tree is only as honest as its inputs, and the inputs come from lawyers, who are reliably overconfident about their own work. Empirical studies of litigation decision-making show systematic optimism: each side tends to overestimate its chances, and the two estimates frequently sum to well over 100%, which is arithmetically impossible and practically expensive. See Jeffrey J. Rachlinski, Gains, Losses, and the Psychology of Litigation, 70 S. Cal. L. Rev. 113 (1996). The antidotes are unglamorous: test each probability against the base rate for that kind of motion in that kind of case, and hand the tree to a colleague whose job is to attack it. A probability you can defend to a skeptic beats a feeling you cannot examine.

There is a related trap worth naming: false precision. Assigning "40%" to a summary-judgment outcome does not mean you have measured anything to two significant figures. It means you have committed to a working estimate you can revise as facts develop. The value is not the decimal; it is the discipline of writing the number down and updating it. An honestly wrong probability you can revisit is worth more than a confident adjective you cannot.

Risk tolerance and the things the tree cannot price

Expected value is the right default, but it is not the whole decision. A client's risk tolerance varies with its size, sophistication, industry, and moment — a startup in the middle of a financing round or a public company facing a hostile press cycle may rationally pay a premium for certainty that a repeat-player defendant with a big balance sheet would refuse. Some cases carry stakes the tree cannot capture: a defense that protects the company's core business model, a precedent that will govern the next hundred contracts, a relationship worth preserving, or reputational harm that dwarfs the damages number. A sophisticated evaluation lays these intangibles beside the expected value rather than pretending the number is the answer, and it names the constituencies — insurers, lenders, partners, shareholders — whose competing priorities shape what "winning" even means.

When the arithmetic points toward resolution, the next move is to convert the analysis into leverage with a credible demand that signals you have done the work. See drafting an effective demand letter: the basics and writing an effective demand letter: a practical checklist.

The evaluation is a living document

The gates are not a one-time intake ritual. Every stage of a case delivers new information that should feed back into the tree: a motion-to-dismiss ruling, a key deposition, an expert report, a document that surfaces in production. The plausibility standard for surviving a motion to dismiss — a complaint must plead factual content allowing a reasonable inference of liability, not mere labels and conclusions, Ashcroft v. Iqbal, 556 U.S. 662 (2009); Bell Atlantic Corp. v. Twombly, 550 U.S. 544 (2007) — is itself a valuation event: clearing it (or failing to) moves the probabilities materially. Re-run the expected-value calculation at each such inflection point, and revisit the budget alongside it. The client's decision at every juncture should be based on the cost and upside of the road ahead, not the sunk cost of the road behind. A verdict chart that assesses the likelihood and cost of favorable termination at each stage — pleadings, class certification, discovery, summary judgment, pretrial, trial, and appeal — turns the abstract idea of "reassessing" into a concrete cadence.

Common failure modes

Even disciplined evaluators lose cases to a short list of recurring mistakes. Watch for these:

  • Skipping collectibility. Winning at trial and only then discovering the defendant was judgment-proof the entire time. Price the recovery, not just the verdict.
  • Collapsing liability and damages. They multiply, not add. Keep them on separate axes until the final expected-value calculation.
  • Missing a shortened contractual limitations clause. Read the limitations clause before you read the breach clause.
  • Falling in love before clearing conflicts. Run the check — including affiliates and principals — at intake, not after the complaint is drafted.
  • Treating the budget as a one-time document. A budget you never revisit is a guess with a spreadsheet's confidence.
  • Waiting to be served before preserving. The duty attaches when litigation is reasonably anticipated; a Rule 37(e) intent finding can be unfixable.
  • Booking punitive or lost-profits damages at face value. Discount punitive awards for the constitutional ratio and lost profits for the certainty requirement.
  • False precision versus no estimate. An honestly wrong probability you can revisit beats a confident feeling you cannot examine — but so does any number beat no number at all.
  • Confusing motion optimism with case value. Filing a motion educates the other side about your weaknesses and, if granted without prejudice, may only buy the plaintiff a second draft.

Primary and secondary authority

Rules and statutes. Model Rules of Professional Conduct 1.1, 1.5(a), 1.5(c), 1.7, 1.9, 3.1 (competence, fees, conflicts, meritorious claims); Fed. R. Civ. P. 11 (pre-filing inquiry and sanctions), 26(f) (discovery conference), 37(e) (spoliation of ESI); 28 U.S.C. § 1331 (federal question), § 1332 and § 1332(d) (diversity and CAFA), § 1367 (supplemental jurisdiction), § 1391 (venue), § 1404(a) (transfer), §§ 1441 & 1446 (removal); UCC § 2-725 (limitations for sale of goods); 11 U.S.C. §§ 523, 727 (bankruptcy discharge exceptions).

Cases. International Shoe Co. v. Washington, 326 U.S. 310 (1945); Goodyear Dunlop Tires Operations, S.A. v. Brown, 564 U.S. 915 (2011); Daimler AG v. Bauman, 571 U.S. 117 (2014); Bristol-Myers Squibb Co. v. Superior Court, 582 U.S. 255 (2017); Ford Motor Co. v. Montana Eighth Judicial Dist. Court, 592 U.S. 351 (2021); Mallory v. Norfolk Southern Railway Co., 600 U.S. 122 (2023); Atlantic Marine Constr. Co. v. U.S. Dist. Court, 571 U.S. 49 (2013); AT&T Mobility LLC v. Concepcion, 563 U.S. 333 (2011); BMW of North America, Inc. v. Gore, 517 U.S. 559 (1996); State Farm Mut. Auto. Ins. Co. v. Campbell, 538 U.S. 408 (2003); Philip Morris USA v. Williams, 549 U.S. 346 (2007); Bell Atlantic Corp. v. Twombly, 550 U.S. 544 (2007); Ashcroft v. Iqbal, 556 U.S. 662 (2009); Alyeska Pipeline Serv. Co. v. Wilderness Soc'y, 421 U.S. 240 (1975); American Pipe & Constr. Co. v. Utah, 414 U.S. 538 (1974); Holland v. Florida, 560 U.S. 631 (2010); Zubulake v. UBS Warburg LLC, 220 F.R.D. 212 (S.D.N.Y. 2003); Hadley v. Baxendale, 156 Eng. Rep. 145 (1854).

Secondary sources. Restatement (Second) of Contracts § 351 (foreseeability of damages); Robert H. Mnookin & Lewis Kornhauser, Bargaining in the Shadow of the Law: The Case of Divorce, 88 Yale L.J. 950 (1979); Jeffrey J. Rachlinski, Gains, Losses, and the Psychology of Litigation, 70 S. Cal. L. Rev. 113 (1996).

Related UpLaw Resources

Related Guides

Related Checklists

Related Toolkits

Related Articles

This checklist is for general informational purposes only, is not legal advice, and does not create an attorney-client relationship. Limitations periods, jurisdictional rules, damages caps, and ethics rules vary by jurisdiction and change over time. Consult qualified counsel about your specific situation.