How Scoring Works

1. Your Summit Score (0-100)

Your SummitScore is a single number that captures the full picture of a hike: how demanding the trail was, how fast you moved, how efficiently your body worked, and how well you held up from start to finish.

The score is built from four weighted components, each scored 0-100 independently, then combined:

Effective Demandx 35%

+

Executionx 30%

+

Efficiencyx 20%

+

Resiliencex 15%

The final score is always clamped between 1 and 100, rounded to one decimal place.

What does a high score mean? It means you tackled a demanding hike, moved efficiently, kept your heart rate under control, and maintained your pace from start to finish. A high score is harder to get on an easy trail -- the algorithm rewards you more for performing well on challenging terrain.

2. The Four Components

Effective Demand 35% of score

This measures how hard the hike itself was -- not how you performed, but what the trail demanded from anyone who did it. A steep 14er in a snowstorm scores higher here than a flat nature trail on a sunny day.

How the base demand is calculated

The base demand is a blend of GPS-measured difficulty and AI-estimated difficulty. The blend depends on whether GPS stream data is available:

With GPS streams (most Strava imports)60% GPS + 40% AI

Without GPS streams (manual hikes)100% AI

// With GPS data:
baseDemand = (gpsDemand x 0.60) + (aiDemand x 0.40)

// Without GPS data:
baseDemand = aiDemand

After calculating the base demand, the conditions modifier is applied (see Section 4). This can boost the demand by up to 15% if weather or trail conditions made the hike harder.

effectiveDemand = baseDemand x conditionsModifier
// Result clamped to 1-100

Example: A Strava hike with GPS demand of 65 and AI difficulty of 58, on a wet trail in 22 degree weather.
Base demand = (65 x 0.60) + (58 x 0.40) = 39 + 23.2 = 62.2
Conditions modifier = 1.0 + 0.015 (cold) + 0.04 (wet) = 1.055
Effective Demand = 62.2 x 1.055 = 65.6

Execution 30% of score

This measures how much ground you covered per hour, accounting for both horizontal distance and elevation gain. It rewards faster, more productive movement.

The formula

Elevation gain is converted into equivalent distance: every 500 feet of climbing counts as 1 extra mile of output. Then your output per hour determines the score.

output = miles + (elevationGainFeet / 500)
outputRate = output / movingTimeHours
executionScore = outputRate x 25
// Result clamped to 1-100

What typical scores look like

Slow pace (1.5 output/hr) -- heavy terrain, many breaks38

Average pace (2.5 output/hr) -- typical day hiker63

Fast pace (3.5 output/hr) -- fit, experienced hiker88

Trail runner (4.0 output/hr)100

Example: 8 miles, 2,000 ft gain, 4 hours moving time
Output = 8 + (2,000 / 500) = 12
Output rate = 12 / 4 = 3.0
Execution = 3.0 x 25 = 75

If moving time is zero or missing, Execution defaults to 50.

Efficiency 20% of score

This measures how much output you got for the effort your body put in. If two hikers cover the same ground at the same speed, but one does it at 140 bpm and the other at 170 bpm, the first hiker was more efficient and scores higher.

The formula

Your raw Execution score is divided by the percentage of your max heart rate that you used. A lower heart rate for the same output means better efficiency.

maxHR = 220 - age // or 190 if age not set
hrPercent = avgHeartRate / maxHR // floored at 0.40
efficiencyScore = (executionRaw / hrPercent) x 0.75
// Result clamped to 1-100

Note: The formula uses the raw Execution score before the performance ceiling is applied. The HR percentage has a floor of 0.40 to prevent extreme values from very low heart rates.

What typical scores look like

High output, low HR (very fit) -- 85 execution at 65% max HR98

Normal output, moderate HR -- 65 execution at 75% max HR65

Hard trail, high HR -- 48 execution at 78% max HR46

Low output, high HR (struggling) -- 40 execution at 85% max HR35

Example: Age 30, avg HR 148, raw execution score 75
Max HR = 220 - 30 = 190
HR % = 148 / 190 = 0.779
Efficiency = (75 / 0.779) x 0.75 = 72.2

If no heart rate data is available (no HR monitor connected), Efficiency defaults to 60 -- a neutral score. Heart rate values below 60 bpm are treated as invalid and also default to 60.

Resilience 15% of score

This measures how well you held up over the course of the hike. Did you slow down significantly in the second half? Did your heart rate spike even though the terrain got easier? Resilience catches that.

How it works

Your hike is split into four equal segments (quartiles) by distance. For each segment, SummitScore calculates the average pace (seconds per meter), average heart rate, and average grade. Then it compares the first quarter to the last quarter, adjusting for terrain changes.

The score starts at 80 ("good") and adjusts from there based on three checks:

A. Pace fade (Q1 vs Q4)

If you slowed down at the end, the algorithm checks whether the terrain explains it. For every 1% increase in average grade from Q1 to Q4, an 8% pace slowdown is expected and not penalized. Only unexplained fade counts.

paceFade = (Q4pace - Q1pace) / Q1pace
expectedFade = (Q4grade - Q1grade) x 0.08
unexplainedFade = paceFade - expectedFade

Unexplained slowing down-40 points per 100% fade

Unexplained speeding up (negative fade)+15 points per 100% speedup

Example: You slowed 25% from Q1 to Q4, but the grade increased 1% (expecting 8% slowdown).
Unexplained fade = 0.25 - 0.08 = 0.17 (17%)
Penalty = 0.17 x 40 = -6.8 points

B. Heart rate drift (Q1 vs Q4)

Same concept. If your heart rate climbed, the algorithm checks if steeper terrain explains it. A 5% HR increase is expected per 1% grade increase. Only unexplained drift is penalized.

hrDrift = (Q4hr - Q1hr) / Q1hr
expectedHrChange = (Q4grade - Q1grade) x 0.05
unexplainedDrift = hrDrift - expectedHrChange

Unexplained HR drift-30 points per 100% drift

Only positive unexplained drift is penalized. If your HR dropped or the terrain fully explains the increase, there is no penalty. Requires valid HR data (above 60 bpm) in both Q1 and Q4.

C. Pacing consistency (all four quarters)

The coefficient of variation (CV) measures how consistent your pace was across all four quarters. CV is the standard deviation divided by the mean -- a lower number means steadier pacing.

CV = stdDev(Q1, Q2, Q3, Q4 paces) / mean(Q1, Q2, Q3, Q4 paces)

CV at or below 0.15no penalty

CV above 0.15-30 per unit above 0.15

Example: CV = 0.25 (fairly variable pacing).
Penalty = (0.25 - 0.15) x 30 = -3 points

Requires at least 3 valid quartile paces for the CV check, and at least 20 data points overall. If no GPS stream data is available, Resilience defaults to 70.

3. Performance Ceiling

Your Execution and Efficiency scores are capped based on how hard the trail was. This prevents easy hikes from yielding unrealistically high scores -- you can't earn a 95 Execution on a flat 2-mile loop, no matter how fast you walk it.

performanceCeiling = min(100, effectiveDemand + 25)
execution = min(executionRaw, performanceCeiling)
efficiency = min(efficiencyRaw, performanceCeiling)

How the cap plays out

Easy trail (demand 25)Execution/Efficiency capped at 50

Moderate trail (demand 45)Capped at 70

Hard trail (demand 65)Capped at 90

Very hard trail (demand 75+)Capped at 100 (no limit)

Resilience is not subject to the performance ceiling. The ceiling only applies to Execution and Efficiency.

4. Conditions Modifier

The conditions modifier is a multiplier applied to the base demand score. It starts at 1.0 (no adjustment) and increases when weather or trail conditions made the hike harder. The maximum is 1.15 (+15% boost).

Multiple conditions stack together, but the total is capped at 1.15. All weather data comes from Open-Meteo, matched to the hours you were actually on the trail.

Temperature

Only extreme temperatures trigger a modifier. Normal hiking weather (25-95 degrees F) has no effect.

// Cold (below 25 degrees F):
modifier += (25 - temp) / 200

// Hot (above 95 degrees F):
modifier += (temp - 95) / 200

25 degrees F (threshold)+0.0%

10 degrees F+7.5%

-5 degrees F+15%

95 degrees F (threshold)+0.0%

110 degrees F+7.5%

125 degrees F+15%

Precipitation

Only significant precipitation counts. Trace amounts (under 0.1 inches) are ignored.

// Only when precipitation > 0.1 inches:
modifier += min((precip - 0.1) x 0.04, 0.10)

0.1 inches or less (trace)+0.0%

0.5 inches+1.6%

1.0 inches+3.6%

2.6+ inches (cap)+10.0%

Wind

Only strong winds matter. Anything under 20 mph has no effect.

// Only when wind > 20 mph:
modifier += (wind - 20) / 200

20 mph (threshold)+0.0%

35 mph+7.5%

50 mph+15%

Severe weather codes

Specific weather conditions detected during the hike add a fixed bonus:

Foggy+2%

Snow+4%

Heavy rain+5%

Snow showers+5%

Thunderstorm+8%

Heavy snow+8%

Thunderstorm with hail+12%

The most severe weather code during the hike window is used. Clear sky, partly cloudy, overcast, drizzle, rain, and rain showers have no modifier.

Trail surface

Selected when logging the hike (from the questionnaire):

Dry / normal / well-maintained+0%

Wet or muddy+4%

Overgrown / unmaintained+5%

Snow or ice+8%

Pack weight

Pack weight uses continuous scaling when you enter a specific weight in pounds. A typical daypack (10 lbs or less) has no effect. Above 10 lbs, each additional pound adds 0.2% to the modifier, up to a cap of +10%.

// When packWeightLbs > 10:
modifier += min((packWeightLbs - 10) x 0.002, 0.10)

10 lbs or under (daypack baseline)+0%

15 lbs+1%

20 lbs+2%

30 lbs+4%

40 lbs+6%

50 lbs+8%

60+ lbs (cap)+10%

For older hikes that used the category picker instead of exact weight, the legacy values apply: "20-35 lbs" = +4%, "35+ lbs" = +8%.

Remember: All conditions stack, but the total modifier is capped at 1.15 (+15%). Even the worst combination of conditions cannot boost demand more than 15%.

5. Trail Difficulty Rating

Every trail gets a difficulty score from 1 to 100. This represents the trail itself -- it stays the same no matter who hikes it or when. It feeds into the AI portion of Effective Demand.

How it works

SummitScore searches the web for trail information (using Brave Search for up to 20 results from AllTrails, hiking forums, and topographic sources), then sends everything to Claude AI. Claude evaluates six factors and assigns a single difficulty number, calibrated against a set of real anchor trails.

Scoring factors

Distance -- total round-trip mileskey factor

Elevation gain -- total uphill feetkey factor

Terrain -- paved < dirt < rocky < Class 2 < Class 3 < Class 4major factor

Exposure -- none < ledges < sustained cliffs < life-threatening fallsmajor factor

Altitude -- no effect under 8,000 ft; significant above 13,000 ftmoderate factor

Navigation -- well-marked < occasionally unclear < route-findingminor factor

Anchor trails (calibration reference)

The AI is given these real trails as fixed reference points so scores stay consistent:

Score	Trail	Stats
8	Easy paved nature loop	1-2 mi, flat
20	Typical "easy" AllTrails trail	3-4 mi, 400-600 ft gain
35	Typical "moderate" AllTrails trail	5-7 mi, 800-1,200 ft gain
45	Blood Mountain, GA	6.1 mi, 1,700 ft gain
52	Brasstown Bald via Jack's Gap, GA	6-7 mi, 1,800 ft gain, steep switchbacks
62	Mount LeConte via Alum Cave, TN	11 mi, 2,800 ft gain
72	Half Dome with cables, CA	16 mi, 4,800 ft gain, cable ladder
80	Mount Whitney Main Trail, CA	22 mi, 6,100 ft gain, 14,505 ft altitude
88	Longs Peak Keyhole Route, CO	15 mi, 5,100 ft gain, Class 3, 14,259 ft
95	Capitol Peak, CO	17 mi, 5,200 ft gain, Class 4 knife-edge, 14,130 ft

Partial hikes

If you only hiked part of a trail (your Strava distance is shorter than the full trail), the AI scores the portion you actually hiked using your actual distance and elevation, but still references the trail's terrain, exposure, and altitude from web search data.

6. GPS Demand Calculation

When GPS stream data is available, SummitScore calculates its own difficulty rating from the raw data. This uses square-root scaling so that gains diminish -- going from 2 to 10 miles matters more than going from 30 to 38 miles.

The four components

// 1. Distance (square-root scaled)
distComponent = sqrt(miles) x 8.0

// 2. Elevation (square-root scaled)
elevComponent = sqrt(elevationGainFeet) x 0.8

// 3. Steep grade bonus (% of segments steeper than 15%)
steepBonus = steepPercent x 0.3

// 4. Altitude bonus (only above 8,000 ft)
altBonus = (maxAltitudeFt - 8000) / 500

gpsDemand = distComponent + elevComponent + steepBonus + altBonus
// Result clamped to 1-100

Distance component examples

4 milessqrt(4) x 8 = 16

10 milessqrt(10) x 8 = 25

20 milessqrt(20) x 8 = 36

Elevation component examples

1,000 ft gainsqrt(1000) x 0.8 = 25

3,000 ft gainsqrt(3000) x 0.8 = 44

5,000 ft gainsqrt(5000) x 0.8 = 57

Steep grade bonus examples

10% of trail segments steeper than 15% grade10 x 0.3 = +3 pts

20% of trail segments steeper than 15% grade20 x 0.3 = +6 pts

40% of trail segments steeper than 15% grade40 x 0.3 = +12 pts

Altitude bonus examples

Below 8,000 ftNo bonus

10,000 ft (e.g., typical CO trailhead)(10000 - 8000) / 500 = +4 pts

12,000 ft(12000 - 8000) / 500 = +8 pts

14,000 ft (fourteener)(14000 - 8000) / 500 = +12 pts

Full example: 10 miles, 3,000 ft gain, 20% steep sections, max altitude 12,000 ft
Distance: sqrt(10) x 8 = 25.3
Elevation: sqrt(3000) x 0.8 = 43.8
Steep: 20 x 0.3 = 6.0
Altitude: (12000 - 8000) / 500 = 8.0
GPS Demand = 25.3 + 43.8 + 6.0 + 8.0 = 83.1

"Steep" is defined as any GPS segment with a grade above 15% (in either direction -- both steep uphills and downhills count). Altitude is read from the GPS elevation stream and converted from meters to feet.

7. Score Predictions

On the Predict tab, you can get a predicted SummitScore for any trail before you hike it. Predictions are generated by AI (Claude) based on your personal hike history and the trail's difficulty.

What the prediction considers

The AI receives the following about you:

Total number of hikes loggedcontext

Average score, min/max score rangecontext

Average score by difficulty tier (easy/moderate/hard/very hard)context

Past prediction accuracy (if any predictions have been resolved)context

Trail difficulty score and AI summarytrail info

Pack weight (if specified)trail info

Difficulty tiers for history analysis

EasyDifficulty 0-30

ModerateDifficulty 31-55

HardDifficulty 56-75

Very HardDifficulty 76+

Confidence levels

The maximum confidence the prediction can have is capped by how many hikes you have logged:

No hikes loggedmax 30% confidence

1-3 hikesmax 50% confidence

4-9 hikesmax 70% confidence

10+ hikesmax 85% confidence

Confidence is also adjusted downward if the target trail's difficulty is very different from what you've hiked before, or if a heavy pack adds unusual load. The predicted score is always between 1 and 100.

8. Weather Data

Weather conditions are fetched automatically from Open-Meteo for every hike. Weather affects your score through the conditions modifier (see Section 4) and is also displayed on your hike detail page for context.

Location resolution

The system uses the most precise location available, in this order:

1. GPS coordinates from Strava (latitude/longitude)most precise

2. Geocoding from the location namefallback

Time window matching

SummitScore tries to get weather for the exact hours you were on the trail. If that fails, it uses a series of fallbacks:

1. Exact hike hours (from Strava start time + duration)preferred

2. Daytime hours (8am-6pm)fallback

3. Midday hours (11am-2pm)last resort

4. Manual hikes without a start timedefaults to 10am-1pm

Data source and fallbacks

The system tries the archive API first (best for hikes more than a week old). If that returns no data, it falls back to the forecast API (covers the last 7 days plus future forecasts).

What is collected

Measurement	Aggregation
Temperature	Average across hike hours
Feels-like temperature	Average
Humidity	Average
Dew point	Average
Cloud cover	Average
Precipitation	Sum (total during hike)
Snowfall	Sum (converted from cm to inches)
Wind speed	Maximum (peak gust matters most)
Wind gusts	Maximum
Sunshine duration	Sum (converted from seconds to minutes)
Weather condition	Most severe WMO code during hike
UV index	Maximum (peak exposure)
Air Quality Index (AQI)	Average
PM2.5 particulates	Average

WMO weather codes

Open-Meteo returns numeric weather codes. SummitScore converts them to readable conditions:

Code(s)	Condition
0	Clear sky
1	Mainly clear
2	Partly cloudy
3	Overcast
45, 48	Foggy
51, 53, 55	Drizzle
61, 63	Rain
65	Heavy rain
71, 73	Snow
75, 77	Heavy snow
80, 81, 82	Rain showers
85, 86	Snow showers
95	Thunderstorm
96, 99	Thunderstorm with hail

9. Heart Rate & Max HR

Heart rate data is used by two scoring components: Efficiency (average HR during the hike) and Resilience (HR drift from Q1 to Q4).

Estimated max heart rate

The standard formula is used:

maxHR = 220 - age

If your age is set in your profile220 minus your age

If your age is not set (or under 10)Default: 190

HR percentage in Efficiency

Your average heart rate is expressed as a percentage of your max HR. This percentage has a floor of 0.40 (40%) to prevent calculation issues with unusually low heart rates. Heart rate readings below 60 bpm are treated as invalid sensor data and cause Efficiency to default to 60.

Heart rate in Resilience

Resilience uses the per-segment average heart rate from the GPS stream (not the overall average). Only segments with HR values above 60 bpm are included. If either Q1 or Q4 lacks valid HR data, the HR drift check is skipped entirely (no penalty, no bonus).

10. Achievements

Achievements are milestones you earn as you hike. Each category has four tiers: Bronze, Silver, Gold, and Platinum. Once earned, achievements are permanent.

Achievements are checked and awarded automatically every time you log a new hike.

Hikes Logged

Tier	Title	Requirement
Bronze	First Steps	1 hike
Silver	Trail Regular	10 hikes
Gold	Trail Veteran	50 hikes
Platinum	Trail Legend	200 hikes

Total Miles

Tier	Title	Requirement
Bronze	Getting Started	10 miles
Silver	Distance Hiker	100 miles
Gold	Long Hauler	500 miles
Platinum	Ultra Trekker	1,000 miles

Total Elevation

Tier	Title	Requirement
Bronze	Hill Climber	5,000 ft
Silver	Mountain Goat	50,000 ft
Gold	Peak Bagger	250,000 ft
Platinum	Everest Conqueror	1,000,000 ft

Summit Score (best single hike)

Tier	Title	Requirement
Bronze	Solid Outing	Score 50+
Silver	Strong Summit	Score 70+
Gold	Elite Hiker	Score 85+
Platinum	Perfect Summit	Score 95+

Single Day Miles (best single hike)

Tier	Title	Requirement
Bronze	Day Tripper	5 miles
Silver	Long Day Out	15 miles
Gold	Marathon Hiker	25 miles
Platinum	Ultra Distance	35 miles

Single Day Elevation (best single hike)

Tier	Title	Requirement
Bronze	Uphill Battle	1,000 ft
Silver	Ridge Runner	3,000 ft
Gold	Summit Seeker	5,000 ft
Platinum	Vertical King	8,000 ft

Streak (consecutive weeks with at least one hike)

Tier	Title	Requirement
Bronze	Getting Regular	2 weeks
Silver	Habit Forming	8 weeks
Gold	Half-Year Streak	26 weeks
Platinum	Year-Round Hiker	52 weeks

Weeks are counted Monday through Sunday. Consecutive means no gap of more than 7 days between week starts that contain hikes.

Consistency (most hikes in any single calendar month)

Tier	Title	Requirement
Bronze	Weekend Warrior	4 hikes/month
Silver	Dedicated Hiker	8 hikes/month
Gold	Trail Addict	15 hikes/month
Platinum	Every Day Hiker	25 hikes/month

11. Score Labels

Here is what the label next to your score means:

90-100Elite

80-89Outstanding

70-79Strong

60-69Solid

50-59Average

40-49Below Average

Below 40Tough Day

12. What Doesn't Affect Your Score

Some data is collected and displayed but does not influence your score:

Oura readiness/sleep scores -- shown on your hike page for personal context onlyno effect

Perceived exertion -- your self-reported effort levelno effect

Navigation difficulty -- used in older versions, now informational onlyno effect

Physical state -- how you felt before/after, informational onlyno effect

Weather impact (self-reported) -- replaced by automatic weather datano effect

Finish feel -- how you felt at the end, informational onlyno effect

Humidity, dew point, cloud cover -- displayed but not part of conditions modifierno effect

UV index, AQI, PM2.5 -- displayed but not part of conditions modifierno effect

Why? The V2 scoring model focuses on objective, measurable data: what the trail demanded, what your body produced, and how consistently you performed. Subjective self-reports and biometric context (like sleep quality) are valuable for your own tracking but are kept separate from the score to ensure fairness and consistency.

Contents