The data
This app is built on split times from 2,500+ runners who raced the Superior 100 between 2014 and 2025 (every year except 2020). For each runner, we have the time they came through all 14 aid stations, from Split Rock at mile 8.4 all the way to the finish at mile 102.
Finding comparable runners
When you tell us where and when you last saw your runner, we search through all 2,500+ historical runners to find people who were at that same checkpoint around that same time. These are runners who were moving at a similar pace, in a similar position in the race.
Runners who match closely get weighted more heavily. Someone who was at Finland at exactly 10 PM counts for more than someone who came through at 9 or 11.
The prediction
Once we've found your runner's comparable group, we look at when those same runners reached each upcoming checkpoint. If they mostly came through Sonju between midnight and 2 AM, that's your window.
The median is the middle of that group — half arrived before, half after. That's your best single estimate.
What the shaded bands mean
The chart shows two shaded regions:
- Wider band — 80% of comparable runners finished within this window. The other 20% were either faster or slower.
- Tighter band — The middle 50%. This is a good window to plan around.
The bands get wider further into the race because runners spread out over time. Some people stop longer at aid stations, some hit a rough patch, some have a great day.
Multiple sightings
Each sighting narrows the pool. A runner seen at Silver Bay at noon and Finland at 10 PM is a much more specific profile than either observation alone. We find historical runners who match both.
In practice, two sightings can cut the prediction window nearly in half, because you're filtering out runners who happened to match one checkpoint but ran a completely different overall pace.
Important caveats
- This is based on historical patterns, not your specific runner. Injuries, weather, and race-day decisions can't be modeled.
- The bands show what comparable runners did — they're not a guarantee of where your runner will land.
- Very fast or very slow runners may have a small comparison group, which makes predictions less reliable. The app shows how many runners it's drawing from.
- Aid station names and course distances have shifted slightly over the years. Predictions are based on time, not distance.
Tips for spectators
- Use the tighter band to plan when to arrive. It covers the most likely window.
- Your most recent sighting is the most useful. If you only have one, make it the latest one.
- If you saw your runner at two checkpoints, add both. It makes a real difference.
- For checkpoints that are a drive away, use the early end of the window to decide when to leave. You don't want to miss them.