PULSCAR ANALYSIS OF NHTSA PUBLIC DATA — STARTER MOTOR COMPLAINTS PulsCar analysis of NHTSA public data, September 2026. License: CC BY 4.0. Files in this folder may be reused with attribution to PulsCar (pulscar.io). -------------------------------------------------------------------------- 1. SOURCE FILES -------------------------------------------------------------------------- NHTSA public flat files downloaded from static.nhtsa.gov and loaded into a local SQLite database: Complaints FLAT_CMPL.txt 777,040 owner complaints (ODINO), received 1 January 2015 – 8 September 2026. Field definitions: https://static.nhtsa.gov/odi/ffdd/cmpl/CMPL.txt Recalls FLAT_RCL.txt one row per make/model/model year per campaign, report dates 1 January 2010 – 9 September 2026; campaigns counted DISTINCT on the campaign number. Communications FLAT_TSBS.txt manufacturer communications, dates 6 January 2015 – 12 September 2026; counted DISTINCT on the NHTSA document id. Complaints are owner reports: not verified repairs, not failure rates, and with no denominator of vehicles on the road. A make or model with more complaints may simply have sold more cars, or older ones. Nothing here ranks reliability. -------------------------------------------------------------------------- 2. COMPLAINT QUERY (the 2,365) -------------------------------------------------------------------------- Step 1 – full-text search (FTS5) for starter OR "starter motor" OR "starter solenoid" in the narrative -> 4,562 complaints. (The FTS index matches whole words, so "start", "starting" and "restart" do not match on their own.) Step 2 – failure proximity: the starter word and a failure/replacement word must appear in the same sentence within 70 characters: failure words: replaced | failed | failure | bad | dead | went out | burned | seized | defective | faulty | broke | diagnosed | needed | click | clicking | no-start | not working | stopped working Step 3 – exclusions in the 100 characters around the match: remote starter, aftermarket starter, push-button (start), starter battery, kick start, jump starter, starter fluid. -> 2,365 complaints. Step 4 – SAMPLE READING. 30 random matches were read in full (random.seed(13)). 26 of 30 were genuine starter-system reports. 4 were not the article's subject: two listed a starter among several replaced parts while complaining about something else (electric steering, stalling), one described a starter disabled by a brake-light safety interlock, and one was a vague "ignition switch and starter, possibly transmission." Observed false-positive rate: about 13% (4/30). Because those four all name a real starter event of some kind, no further wording filter was applied; the counts carry them, and the rate is published here instead. -------------------------------------------------------------------------- 3. WHAT IS COUNTED (starter-complaints-by-context.csv) -------------------------------------------------------------------------- Category searches run over the whole narrative, not a proximity window, because owners describe the context in separate sentences. A complaint can appear in several rows; shares are of 2,365, not parts of a whole. "Starter replaced and the problem returned" requires a replacement phrase and a persistence phrase (still / again / recurred / same problem / did not fix) within 120 characters of each other, in either order: 87 complaints (3.7%). This is a floor, not a rate — it only catches owners who wrote the sequence down in one sentence. Mileage: values <= 0 and >= 400,000 dropped; 1,300 of 2,365 carry a usable figure. Median 79,000 miles, Q1 48,000, Q3 114,920. This is the middle of a pile of reported failures, NOT a starter life expectancy. The NHTSA towed / crash / fire columns are the agency's own flags. The narrative mentions a tow in 753 complaints (31.8%) while the structured towed flag is set on 236 (10.0%) — both are published so the gap is visible. -------------------------------------------------------------------------- 4. MODELS (starter-complaints-top-models.csv) -------------------------------------------------------------------------- Counts by make and model with the three most common model years and the median reported failure mileage. The Honda Accord cluster (360 complaints, 332 of them model years 2013-2015, median reported failure mileage 55,625 over 242 complaints with mileage) is discussed in the article next to Honda service bulletin 16-002. NOTE: that bulletin covers V6 automatic Accords and Crosstours only, while complaint records do not name the engine, so the complaint cluster is wider than the bulletin's population. -------------------------------------------------------------------------- 5. RECALLS (starter-recalls.csv) -------------------------------------------------------------------------- Campaigns whose component is an "ELECTRICAL SYSTEM:STARTER ASSEMBLY" or whose defect text names a starter motor, engine starter or starter relay, excluding a towing "starter kit" campaign: -> 50 distinct campaigns, 1,730,620 vehicles potentially affected (sum of the per-campaign figures; a vehicle covered twice is counted twice — and in this set that matters, because BMW campaign 25V644000 explicitly re-recalls vehicles already repaired under 24V576000). -------------------------------------------------------------------------- 6. MANUFACTURER COMMUNICATIONS (starter-bulletins.csv) -------------------------------------------------------------------------- FTS5: starter AND (motor OR solenoid OR replace OR replacement OR noise OR grind OR crank) over bulletin summaries -> 310 distinct NHTSA document ids. Motorcycle and powersports bulletins are left in the file and are visible in the make column. -------------------------------------------------------------------------- 7. LIMITATIONS -------------------------------------------------------------------------- - Self-selected reports; filers are not a random sample of owners. - No denominator: model counts track sales volume and fleet age. - The median reported failure mileage is not a lifespan. A starter's life is measured in start cycles; the odometer is a proxy that a start-stop system, a short-trip commute or a weak battery can break. - Category regexes read words, not meaning: "alternator named" counts the word appearing anywhere in the complaint, which includes owners listing parts a shop replaced before finding the real fault. - Complaint records do not identify the engine or drivetrain, so complaints cannot be matched exactly to a bulletin's affected population. - Recall and bulletin searches find documents that USE these words. Files: starter-complaints-by-context.csv, starter-complaints-by-make.csv, starter-complaints-top-models.csv, starter-recalls.csv, starter-bulletins.csv. ------------------------------------------------------------------------------ CASE SENSITIVITY — SECTION ADDED 19 SEPTEMBER 2026 ------------------------------------------------------------------------------ NHTSA stores complaint narratives in ALL CAPS for records filed through roughly 2020 and in mixed case for records filed from 2021 on. A case-SENSITIVE search of the narrative field loses about 92% of the complaints filed after 2020, so this is the single largest way an analysis of this database can go wrong while still looking finished. Every query behind this file is case-insensitive. SQLite's LIKE and its FTS5 full-text index both ignore case by default, and no query in this work uses GLOB or any case-sensitive comparison. This was not taken on trust. On 19 September 2026 every published folder was re-checked against the same database, comparing the figures here with what a case-sensitive query would have returned. The result: the complaint set published here (4,562) is LARGER than the absolute ceiling a case-sensitive search of the same terms could have returned (2,575 narratives). A case-sensitive query cannot produce a bigger set than itself, so the bug is ruled out for this folder by arithmetic, not by assurance.