Las Vegas 2025 / 26

Collisions tripled while braking got smoother.

This study compares Nexar dashcam fleet data from the Las Vegas metropolitan area across two equivalent five-month periods: January-May 2025 and January-May 2026. It covers 739,950 rides totaling ~7 million miles, focused on incident prevalence and severity, collisions, hard braking, harsh acceleration, and cornering.

Collision rates rose +231% (139 → 460 per million rides; Z = 17.4, p < 0.001). Yet hard braking, the strongest leading indicator of near-misses, improved 16%, and harsh acceleration dropped 71%. The collision spike concentrates in North Las Vegas, particularly January-February 2026.

Study context

Las Vegas is one of the most unique driving environments in the United States, a 24/7 city with high tourism traffic, the Las Vegas Strip, convention activity, and one of the country's highest pedestrian exposure rates. Year-over-year safety trends here are directly relevant to fleet operators, insurance underwriters, and road-safety stakeholders.

Because the metrics are rate-normalized, the headline findings are not artifacts of the shrinking ride pool. A collision rate is collisions ÷ rides × 1,000,000, independent of how many rides were observed, provided the sample is large enough to be stable (it is: hundreds of thousands of rides per window).

Methodology

Data source. All data originates from the Nexar Dashcam Fleet, dashcam-equipped vehicles operating in Las Vegas, Nevada. Rides are filtered by starting-point city (Las Vegas, North Las Vegas, and adjacent Las Vegas-named areas) and state (Nevada).

Four incident types tracked: Collision (detected impact events, the primary metric), Hard brake (sudden deceleration over threshold), Harsh acceleration (sudden acceleration over threshold), and Cornering (sharp lateral movement events). Baseline: the 2025 Jan-May window, an identical calendar window in both years, so no day-of-week adjustment is needed and the periods are directly comparable.

Volume fell, but trip shape held.

Volume fell ~19% across rides, miles, and hours, but trip shape is unchanged. Miles-per-ride, hours-per-ride, and average speed are all within ~2% year-over-year, confirming the same fleet on similar routes.

Four incident types, four directions.

Collisions exploded; every behavioral indicator improved or held flat. This split is the analytical core of the study.

The Q1 anomaly.

The +231% annual figure is front-loaded. January and February 2026 carry extreme Danger Indices; by March-May the rate converges toward (though stays above) 2025 levels.

Monthly 2026 collision rate per million rides. February 2026 (1,076/M) is the peak.

Both signals are real.

Every headline metric clears conventional significance by a wide margin. The collision increase and the hard-brake improvement are both real, neither is a statistical artifact of the smaller 2026 ride pool.

The risk shifted to mid-morning.

All times Pacific (Las Vegas local). The 2026 collision risk concentrates in a 9-11 AM window that was near-silent in 2025, a notable shift from 2025's early-morning (3-4 AM) risk profile.

2026 collision rate per million rides, peak hours. Full 24-hour series in the appendix.

The weekend reversal.

In 2025, weekends were the safest days. In 2026, they became the most dangerous, Saturday and Friday now top the collision-rate table. A clear behavioral or fleet-pattern shift.

Where the paradox resolves.

North Las Vegas, 15% of rides, produced 66% of all 2026 collisions, with a collision rate of 2,011/M against the core's 186/M.

The Consumer segment carries the story.

Over 99% of Las Vegas Nexar rides fall in the Consumer/Unknown category, which therefore carries the entire collision-rate story. Truck and Taxi/Rideshare counts are too small to read individually.

Trip Characteristics & External Context

Trip length and duration were remarkably stable, confirming the same fleet on similar routes. Against external benchmarks, the collision spike is large but explainable; the braking improvement is validated.

Key Findings

Six findings, the spike, where it lives, when it happened, and the improving behavioral backdrop.

Recommendations

Five actions, ranked by priority. R1 follows directly from the geographic concentration.

1. Investigate North Las Vegas fleet accounts

   Identify which specific fleet accounts or vehicles generated the 100 collision events in NLV during Jan-May 2026. The concentration (65.8% of collisions from 15% of rides) points to specific operators. Prioritize safety coaching and telematics review for these accounts.

2. Deploy real-time monitoring for Q1 risk windows

   The January-February spike suggests seasonal or post-New-Year risk may be consistently elevated. Implement enhanced monitoring during these months in 2027.

3. Address weekend & mid-morning concentration

   Saturday/Friday and 9-11 AM local time are the highest-risk segments in 2026. Fleet operators should consider targeted coaching for these specific windows.

4. Leverage hard-brake improvements as a model

   The consistent −16% improvement in hard braking across all days and months suggests effective coaching or operational change. Document and replicate what drove this improvement across other markets.

5. Establish Las Vegas as a monitored market

   With 330K-410K rides per 5-month window, Las Vegas provides statistically reliable data for continuous monitoring. Set quarterly alerts for collision rate exceeding 300/M.

Limitations

1. Fleet composition change

   Ride volume declined 19.4% between periods. Some of the rate change may reflect which fleet accounts are active, not purely behavioral change. Per-million normalization handles volume, but not composition shifts.

2. North Las Vegas concentration

   The NLV signal is driven by a small number of rides relative to the metro total. Individual fleet accounts may disproportionately influence the rate, the headline is sensitive to a handful of operators.

3. Incident detection model

   Changes in Nexar's collision detection model between 2025 and 2026 could affect comparability. The dramatic harsh-acceleration drop (−71%) may partly reflect a model change rather than pure behavioral improvement.

4. Population bias

   Nexar fleet data represents vehicles equipped with Nexar dashcams, not all Las Vegas traffic. Fleet-equipped vehicles may skew toward commercial and professional drivers.

5. Small absolute counts

   57 (2025) and 152 (2026) collision events are relatively small absolute numbers, amplifying percentage changes. The statistical significance is genuine, but caution is warranted in extrapolation. External benchmarks were drawn from prior knowledge without real-time validation.

Appendix · Full Hourly Collision Data

Complete by-hour series (Pacific Time), both periods, for reproducibility. Mid-morning 2026 risk hours marked.

Analysis by Nexar BI · Requested by Itamar Bul · nexar-ai.com. Data source · Nexar Dashcam Fleet · Las Vegas Metro, Nevada. Generated June 7, 2026 · Study windows Jan-May 2025 & 2026 · 739,950 rides analyzed.