January 30, 2026
Crash course in robo-chaos
Tesla's Robotaxi data confirms crash rate 3x worse than humans even with monitor
Tiny fleet, secretive reports, and a flame war over Tesla’s crash stats
TLDR: Tesla’s robotaxis logged ~500,000 miles and nine crashes—around 3–9x worse than humans—even with a safety driver in each car. Comments split between “tiny sample, bad comparison” and “secrecy is the real red flag,” with many praising rivals’ detailed reports as the actual sign of safety
The internet served a full platter of chaos after Tesla’s robotaxi numbers dropped: nine crashes in about 500,000 miles—roughly one every 55,000 miles—while humans average between 200,000 and 500,000 miles per crash. Cue the comment brawl. The sample-size squad shouted: this is teeny-tiny. SilverBirch did the math, calling the fleet “smaller than a lot of taxi companies,” and artembugara waved the statistics flag: “law of large numbers” says the distance isn’t big enough to judge. The skeptics fired back: fabian2k says it’s still damning that it’s worse even with a safety driver in every car whose job is to stop crashes. Then came the “apples vs oranges” warriors like z7, arguing that the robotaxi crashes include minor taps that wouldn’t be in police logs, so the comparison is messy.
But the real popcorn moment? Transparency. Tesla’s crash narratives in NHTSA reports are redacted, while rivals like Waymo publish detailed play-by-plays—including a recent school-zone incident where Waymo braked hard and the child was okay. Commenters roasted Tesla’s secrecy with memes like “mysterious crash #7: classified” and “beta-testing the public.” The ban-it-all crowd stormed in too: “turn off lane-keeping and turn on your brain.” Verdict: math fights, moral fights, and one shared mood—everyone wants receipts
Key Points
- •NHTSA SGO reports list nine Tesla robotaxi crashes in Austin, Texas from July–November 2025.
- •Tesla’s Q4 2025 report indicates ~500,000 cumulative robotaxi miles by November 2025, implying ~1 crash per 55,000 miles.
- •The article compares Tesla’s rate to U.S. human averages (~1 per 500,000 police-reported miles; ~1 per 200,000 including non-police), finding Tesla worse.
- •All Tesla crash narratives in the NHTSA database are redacted, limiting insight into causes and circumstances.
- •Waymo and Zoox publish detailed incident reports; Waymo is cited with 25M+ driverless miles, below-human crash rates, and a disclosed Santa Monica incident.