Current Insights

1. Functional Threshold(FTP): If we observe an FTP greater than your Strava FTP setting we send you a notification. We use 95% of your 20 minute power on bike or run to calculate FTP.
2. Bike Mismatch: We have 2 algorithms that predict which bike you used on a Ride activity. The first uses ride telemetry (e.g. avg speed, distance, elevation, etc) to predict the bike. It has decent results provided you don't ride two bikes the same way (e.g two road bikes). The second is unlocked when you authorize Garmin Connect and it uses sensor device identifiers from the raw device files. This one is 100% accurate provided you don't label the same physical bike as two different bikes within Strava (e.g. Road Bike & Virtual Road Bike). Both have quality thresholds of 95% so they will only predict if they test above that threshold. You can see your algorithm test scores here.
3. Type Mismatch: We use this insight internally in our pipeline to avoid doing downstream processing in situations where the activity type isn't really what Strava says. Great example of this is when people record skiing or ice hockey on their Garmin watches as runs or rides.
4. Wind Index: Our wind assist metric is an index representing the amount of help the wind speed and direction would contribute on an effort. We use the factors related to the weather samples we collect for each segment effort to compute them. Tailwinds are positive, headwinds are negative and the scale matters.
5. PR Opportunity: We alert you when either the Wind index or your fitness (power curve) has a high probability of earning you a PR on a watched segment.
6. Cyclist Finder: We currently group cyclist with power to others based on their [1, 5 and 20 minute] 8 week power. We filter the list to those who also ride in the same major metro area as you do.
7. Runner Finder: We group runners based on their [1600, 5k and 10k] pace. We'll filter the list to those who run in the same major metro area as you do.
8. Sensor Battery: Another insight driven from Garmin Connect where we review the sensor battery info to determine if you need to replace or charge your battery. We currently send notices for low or critial battery situations. Most modern sensors (e.g. PM, Electic Shifting, etc) support battery level and voltage attributes.

Insights Coming Soon

1. Humon Insights: Several atheletes are using Humon to track Muscle Oxygen during their workouts (requires Garmin Connect). In these insights we'll let you know the influence Humon metrics are having on your focus metrics (peak thresholds or best effort distances)
2. Moxy Insights: Several atheletes are using Moxy to track Muscle Oxygen during their workouts (requires Garmin Connect). In these insights we'll let you know the influence Moxy metrics are having on your focus metrics (peak thresholds or best effort distances)
3. Peak Anomaly: We're currently predicting these but not yet notifying you. We have a bit more testing to complete to determine in which instances we'd send a notice. You can Flag any Peaks via the mobile interface to exclude it from your 8 week window.
4. Stryd Insights: We compare you to other Stryd athletes that run the same average pace to look for strengths or weaknesses.
5. Factors Insights: When you breakthrough or lose a peak we'll update your factors and let you know the top positive and negative influencers as well as give you a link to all of your factors for that Peak.
6. Group Event: We're currently working on a classifier to identify group rides & runs, members of strive.ai participate in. The plan is to surface the group events that a) your friends participate in and b) your fitness matches. We'll also provide a score showing you how your fitness is relative to the average for the group.