Case tree model
Who wins when a court decides a personal injury case, whether the case is appealed, and how the appeal ends.
What does the model do?
Our own model of what happens when a court decides a case, and what happens next. It has three kinds of branch. Who wins: among cases the court decided at trial, and separately on a motion before trial, the share in which it recorded a judgment for the plaintiff. Appeal: for each way a case can end in the district court (a judgment for the plaintiff, a judgment for the defendant, a dismissal by the court, or a recorded settlement or voluntary dismissal), the share of cases that were then appealed, found by matching each case to the federal appeals database by district, office and docket number. Result of the appeal: among those appeals that the court of appeals decided on the merits, the share affirmed and the share reversed or vacated at least in part. Each branch takes five years of resolved cases; the value is the share over the five years together and the range runs from the lowest to the highest single year, widened at each end by half the gap between them. The appeal branches use five years ending one year back, and the result branches five years ending two years back, so that appeals have had time to be filed and decided. A branch is estimated only where each of the five years has at least 20 cases behind it. Tested by rebuilding each branch for past years from the five years before; the results are published with the model. Limits: who appealed is not recorded, so an appeal after a judgment for the plaintiff is not always the defendant's. A reversal sends many cases back for more proceedings; it does not mean the other side won. Winners are as the courts recorded them, and the courts do not record one for every judgment. It covers federal courts only, describes cases as a group, does not predict any single case, and says nothing about amounts or about any named judge, firm, person or litigation.
Personal injury cases nationwideEstimated
This is our own case tree model. For federal personal injury cases nationwide, it estimates who the court records as the winner when it decides the case, how often the case is then appealed, and how the appeal ends. The mark is the estimate and the band is its range. It describes cases as a group; it does not predict any one case and says nothing about amounts.
When a court decides, who wins?
- Plaintiff wins at trialAbout 55%range 52% to 58%, from 1,137 cases
- Plaintiff wins on a motion before trialAbout 7.7%range 5.9% to 9.3%, from 4,450 cases
How often is the case appealed?
- After a judgment for the plaintiffAbout 15%range 10% to 20%, from 1,802 cases
- After a judgment for the defendantAbout 28%range 24% to 33%, from 5,847 cases
- After a dismissal by the courtAbout 7.0%range 6.7% to 7.2%, from 20,385 cases
- After a settlement or voluntary dismissalAbout 1.3%range 0.9% to 1.9%, from 63,979 cases
When the appeal is decided, how often is the result left in place?
- After a judgment for the plaintiffAbout 74%range 41% to 100%, from 150 cases
- After a judgment for the defendantAbout 81%range 75% to 85%, from 1,115 cases
- After a dismissal by the courtAbout 85%range 77% to 93%, from 810 cases
Winners are as the courts recorded them. The record does not say which side appealed, and a reversal often sends a case back for more proceedings; it does not mean the other side won. Federal courts only. Method: Case tree model. The full model and how it did when tested.
The model in detail
The case tree model is our own. It follows a federal personal injury case past its ending: when a court decides the case, who it records as the winner; whether the case is then appealed; and what the court of appeals does. The appeal branches come from matching each district court case to the federal appeals database. Every figure it produces is an estimate with a range. It describes cases as a group and says nothing about any named judge, firm, person or litigation, or about amounts.
Who wins depends heavily on how the case is decided. At trial, the model estimates that the plaintiff wins 55.4% of the time (range 52.4% to 58.4%). When the court decides the case on a motion before trial, such as summary judgment, the plaintiff wins 7.7% of the time (range 5.9% to 9.3%): decisions on a motion are mostly recorded for the defendant. Most cases end neither way; the case path model shows how often each ending happens.
Who wins when a court decides, by kind of caseEstimated
| Kind of case | Plaintiff wins at trial | Plaintiff wins on a motion before trial |
|---|---|---|
| Personal injury (all types) | 55.4% (52.4% to 58.4%) | 7.7% (5.9% to 9.3%) |
| Motor vehicle | 65.3% (49.1% to 88.4%) | 15.4% (7.6% to 23.8%) |
| Medical malpractice | — | 5.8% (0.0% to 14.2%) |
| Product liability | — | 4.0% (0.0% to 9.6%) |
| Health care and pharmaceutical product liability | — | 1.4% (0.0% to 5.6%) |
| Marine | — | 22.3% (0.0% to 65.5%) |
| Assault, libel and slander | — | 5.6% (0.0% to 11.3%) |
| Other personal injury | 52.5% (44.2% to 67.3%) | 6.7% (2.7% to 10.4%) |
Each cell is the model's value with its range in brackets. Method: case tree model, version 1. A dash means too few decisions to estimate. Counts only decisions where the court recorded a winner.
What happens next depends on how the case ended. After a judgment for the defendant, the model estimates 28.4% of cases are appealed (range 24.2% to 32.8%); after a judgment for the plaintiff, 14.9% (range 10.3% to 20.0%). Most appeals leave the result in place: after a judgment for the defendant, 80.8% of appeals decided on the merits are affirmed (range 75.1% to 84.6%). An appeal also takes time: the last column shows the median wait from the district court closing the case to the court of appeals deciding it.
What happens next, by how the case ended in the district courtEstimated
| The case ended in | Appealed | Affirmed, of appeals decided on the merits | Reversed at least in part, of appeals decided on the merits | Median wait for the appeal decision (Computed) |
|---|---|---|---|---|
| A judgment for the plaintiff | 14.9% (10.3% to 20.0%) | 74.0% (41.3% to 100.0%) | 26.0% (0.0% to 58.7%) | 534 days |
| A judgment for the defendant | 28.4% (24.2% to 32.8%) | 80.8% (75.1% to 84.6%) | 19.2% (15.4% to 24.9%) | 417 days |
| A dismissal by the court | 7.0% (6.7% to 7.2%) | 84.6% (76.5% to 92.7%) | 15.4% (7.3% to 23.5%) | 434 days |
| A recorded settlement or a voluntary dismissal | 1.3% (0.9% to 1.9%) | — | — | 773 days |
Federal personal injury cases nationwide. The first three columns are Estimated, each with its range in brackets; the last is Computed. The record does not say which side appealed. A reversal often sends the case back for more proceedings; it does not mean the other side won. Appeals that were dismissed, withdrawn or are still pending are not in the affirmed and reversed columns.
Some features of a case recorded when it is filed go with very different endings. The largest difference in the table below is whether the plaintiff has a lawyer: from 2021 to 2025, 51.8% of cases where the plaintiff had a lawyer were recorded as settled and 18.0% were dismissed by the court; where the plaintiff had no lawyer, 7.1% were recorded as settled and 60.2% were dismissed by the court. These are counts of what happened, not the effect of any one feature: cases with and without lawyers differ in many other ways.
How personal injury cases ended, by kind of case, 2021 to 2025Computed
| Kind of case, as recorded at filing | Cases resolved | Recorded as settled | Dismissed voluntarily | Dismissed by the court | Judgment for the plaintiff | Judgment for the defendant | Median days to closing |
|---|---|---|---|---|---|---|---|
| Did the plaintiff have a lawyer? | |||||||
| Plaintiff had a lawyer | 90,289 | 51.8% | 19.9% | 18.0% | 1.9% | 4.5% | 362 |
| Plaintiff had no lawyer | 8,452 | 7.1% | 7.6% | 60.2% | 0.9% | 18.9% | 154 |
| Where did the case start? | |||||||
| Filed in federal court | 50,016 | 41.3% | 21.1% | 24.6% | 2.5% | 6.0% | 355 |
| Removed from a state court | 41,593 | 56.1% | 17.0% | 17.3% | 1.0% | 5.2% | 350 |
| Was a jury asked for? | |||||||
| A jury was asked for | 83,913 | 50.2% | 18.9% | 19.9% | 1.6% | 5.3% | 356 |
| No jury was asked for | 14,828 | 35.0% | 18.2% | 31.2% | 2.9% | 8.3% | 277 |
| Why is the case in federal court? | |||||||
| The parties are from different states | 73,858 | 51.1% | 19.3% | 19.6% | 1.5% | 4.8% | 353 |
| A question of federal law | 16,230 | 38.2% | 15.2% | 29.5% | 2.9% | 8.0% | 298 |
| The United States is the defendant | 8,590 | 39.7% | 21.0% | 24.1% | 2.4% | 9.8% | 347 |
Federal personal injury cases nationwide, outside multidistrict litigation. Shares are of cases resolved. The differences are associations in the record, not causes.
We tested the case tree the same way as our other models: each branch was rebuilt for past years from the five years before and compared with what happened. Across 1,371 tests, the actual share fell inside the model's range 91% of the time, with a typical gap of 1.4 percentage points.
How the case tree did when tested against past yearsComputed
| Branch | Tests | Actual share fell inside the range | Typical gap |
|---|---|---|---|
| Affirmed, after a dismissal by the court | 9 | 89% | 4.8 points |
| Affirmed, after a judgment for the defendant | 8 | 100% | 2.0 points |
| Affirmed, after a judgment for the plaintiff | 4 | 75% | 7.7 points |
| Appealed, after a dismissal by the court | 295 | 93% | 2.5 points |
| Appealed, after a judgment for the defendant | 128 | 93% | 4.0 points |
| Appealed, after a judgment for the plaintiff | 36 | 94% | 3.1 points |
| Appealed, after a recorded settlement or a voluntary dismissal | 722 | 91% | 0.8 points |
| Plaintiff wins on a motion | 130 | 86% | 2.6 points |
| Plaintiff wins at trial | 18 | 83% | 5.7 points |
| All branches | 1,371 | 91% | 1.4 points |
A test is one place, one kind of case and one year. Branches with few tests rest on few years and should be read with more caution.
Use it
- Case path model: how often each ending happens in the first place, and how long cases take.
- Data catalogue: every estimate and the case-profile counts as CSV files.
- Methodology: the method in full and its limits.
The model makes no estimate about any named firm, attorney, judge, person or litigation, and none about amounts. When you quote it, say "estimated" and give the range.
Sources
Last updated 18 November 2026. That is the date we last retrieved the source files below.
- Federal Judicial Center Integrated Database (civil, 1988 forward) File retrieved 18 November 2026 from fjc.gov. File fingerprint 74405231a9a3.
- Federal Judicial Center Integrated Database (appeals, 2008 forward) File retrieved 18 November 2026 from fjc.gov. File fingerprint 05d545060b09.
How the estimates on this page were made
Case tree model. Our own model of what happens when a court decides a case, and what happens next. It has three kinds of branch. Who wins: among cases the court decided at trial, and separately on a motion before trial, the share in which it recorded a judgment for the plaintiff. Appeal: for each way a case can end in the district court (a judgment for the plaintiff, a judgment for the defendant, a dismissal by the court, or a recorded settlement or voluntary dismissal), the share of cases that were then appealed, found by matching each case to the federal appeals database by district, office and docket number. Result of the appeal: among those appeals that the court of appeals decided on the merits, the share affirmed and the share reversed or vacated at least in part. Each branch takes five years of resolved cases; the value is the share over the five years together and the range runs from the lowest to the highest single year, widened at each end by half the gap between them. The appeal branches use five years ending one year back, and the result branches five years ending two years back, so that appeals have had time to be filed and decided. A branch is estimated only where each of the five years has at least 20 cases behind it. Tested by rebuilding each branch for past years from the five years before; the results are published with the model. Limits: who appealed is not recorded, so an appeal after a judgment for the plaintiff is not always the defendant's. A reversal sends many cases back for more proceedings; it does not mean the other side won. Winners are as the courts recorded them, and the courts do not record one for every judgment. It covers federal courts only, describes cases as a group, does not predict any single case, and says nothing about amounts or about any named judge, firm, person or litigation. More on this method.
The full address of every source file is on the sources page. Found something wrong? See corrections and removal requests. How figures are labelled and calculated is in the methodology.