Prediction accuracy

How accurate are NC State Wolfline arrival times?

Updated Jun 18, 10:15 PM ET

82%

land inside the accepted accuracy window

scored by the open Transit/Swiftly ETA benchmark

6×

more accurate than the printed schedule

near arrival — 28s error vs 167s

A prediction is “accurate” when the bus arrives within the benchmark’s allowed window — close enough that you could still catch it. The window is asymmetric: it’s stricter on buses that arrive early, because one that’s already left is one you can’t catch.

✓23,915 predictions checked·7 routes·all-time

Where our predictions land vs. the accepted window

Green is the window the benchmark accepts; the dot is where our typical prediction lands, with the bar showing the usual spread. The window narrows as your bus gets closer — the benchmark gets stricter the more you’re relying on the number.

← bus runs late (you just wait)bus comes early (you miss it) →

15 min away84% inside

accepted: up to 4.5 min late or 1.5 min early · typically ~59s late

10 min away76% inside

accepted: up to 3.5 min late or 1 min early · typically ~35s late

5 min away78% inside

accepted: up to 2.5 min late or 1 min early · right on time

1 min away90% inside

accepted: up to 1.5 min late or 30s early · right on time

4 min late2 min lateon time1 min early

Accuracy over time

Hourly, last 48 hours.

When we’re off, which way?

A bus arriving earlier than predicted is the one a rider misses — so we watch that closely.

82.2% accurate

13.2% early

4.5%

Only 4.5% of the time does a bus run later than predicted; 13.2% it comes early.

Accuracy by route (5 minutes out)

Route	Checks	Median error	Within 2 min
52	1,508	52s	82%
41	1,423	51s	88%
30	1,207	43s	92%
43	729	51s	85%
42	686	48s	83%
60	558	48s	87%
20	360	34s	94%

How we measure this

Every time a bus arrives, we look back at the prediction we were showing 1, 5, 10 and 15 minutes earlier — bucketed by time-to-arrival exactly as the open-source Transit/Swiftly ETA Accuracy Benchmark specifies. A prediction counts as accurate if it landed inside that bucket’s asymmetric window — stricter on early arrivals, because a bus that has already left is one you can’t catch. The headline figure is the benchmark’s straight average of the four buckets. Built entirely from raw GPS, updated daily.

NC State Wolfline · measurement window since 2026-06-17