14 LLM calls
Model: openai/gpt-5.2
Complete the following task and return your answer as a list of items in XML format.
Rules:
- Wrap all items in <items></items> tags
- Wrap each individual item in <item index="n"></item> tags
- Put each XML tag on its own line
- Put the item content on its own line between the tags
Example format:
<items>
<item index="1">
First item here
</item>
<item index="2">
Second item here
</item>
</items>
Task: Generate 108 diverse measurable facts for the domain: `public transit ridership in a mid-sized U.S. city`.
A 'measurable fact' should be a specific measurable phenomenon, event, behavior, outcome, or attribute within this domain that could plausibly be expressed as a probability, rate, frequency, likelihood, or proportion.
Focus on WHAT could be measured, not the value itself.
Make the topics diverse and cover different aspects of the domain.
Give me a wide variety of facts, both semantically and syntactically diverse.Avoid near-duplicates, overlap, or overly broad topics.
1. Share of city residents who ride public transit at least once per week
2. Probability that a randomly selected weekday bus trip is taken during the AM peak period
3. Proportion of total ridership that occurs on weekends
4. Rate of year-over-year change in unlinked passenger trips systemwide
5. Percentage of boardings made with contactless payment (tap-to-pay)
6. Fraction of riders using a monthly pass versus pay-per-ride fares
7. Likelihood that a rider transfers between routes within a single journey
8. Average number of transfers per linked transit trip
9. Share of trips involving a bus-to-rail transfer (if rail exists)
10. Proportion of ridership originating from the top 10 highest-boarding stops
11. Probability that a bus arrives at a stop within 5 minutes of scheduled time
12. Percentage of riders experiencing a headway wait longer than 15 minutes on frequent routes
13. Rate of missed trips (scheduled trips not operated) as a proportion of total scheduled trips
14. Likelihood that a rider reports crowding as a primary dissatisfaction factor
15. Proportion of vehicle trips that exceed seated capacity at any point
16. Share of peak-period boardings occurring at park-and-ride lots
17. Rate of bicycle-on-bus rack usage per 1,000 boardings
18. Fraction of riders who access transit by walking versus biking versus driving
19. Probability that a rider uses a microtransit/first-mile service to reach a transit stop
20. Proportion of trips made by riders with disabilities using paratransit rather than fixed route
21. Rate of paratransit no-shows as a proportion of scheduled pickups
22. Likelihood that a rider’s trip purpose is commuting to work
23. Share of ridership attributable to students during the academic year
24. Seasonal variation rate in ridership between winter and summer months
25. Percentage of riders who board after 9 PM on weekdays
26. Probability that a rider chooses transit due to lack of car access
27. Proportion of riders who own a car but still use transit for specific trips
28. Rate of ridership change following a fare increase event
29. Change in boarding counts after introducing a new frequent-service corridor
30. Proportion of riders eligible for reduced fare who actually use reduced-fare media
31. Likelihood that a rider’s fare is paid via employer-sponsored transit benefit
32. Share of boardings that are fare-free due to promotion days or pilot programs
33. Proportion of riders who report feeling safe at stops during nighttime
34. Rate of reported security incidents per 100,000 passenger trips
35. Probability that a rider abandons a trip due to service disruption
36. Frequency of major service disruptions per month affecting at least one route
37. Proportion of trips where real-time arrival information is accessed by the rider
38. Share of riders using a transit agency mobile app to plan trips
39. Rate of complaints per 10,000 boardings related to delays
40. Proportion of boardings occurring at stops with shelters
41. Likelihood that a rider’s origin stop has ADA-compliant boarding access
42. Share of riders who report difficulty finding seating during peak hours
43. Probability that a bus trip experiences bunching (two buses arriving within a short interval)
44. Percentage of route segments where average operating speed falls below a threshold
45. Rate of ridership loss during severe weather days compared with typical days
46. Proportion of riders who combine transit with rideshare for last-mile connections
47. Likelihood that a rider uses transit to reach a special event venue
48. Share of annual ridership concentrated in the top 5 highest-demand event days
49. Proportion of riders who make at least 10 trips per month (frequent riders)
50. Probability that a new rider continues using transit in the following month (retention)
51. Rate of first-time fare card/app account creation per 1,000 residents
52. Proportion of riders who cite cost savings as the main reason for using transit
53. Likelihood that a rider’s typical commute time by transit exceeds 45 minutes
54. Share of riders who experience a transfer wait longer than 10 minutes
55. Probability that a rider arrives on time to their destination using transit (self-reported)
56. Proportion of trips where the rider’s planned itinerary changes due to delays
57. Rate of ridership per revenue hour of service by route
58. Proportion of system ridership served by the top-performing route
59. Share of ridership occurring within low-income census tracts
60. Likelihood that a rider’s household income falls below a defined threshold
61. Proportion of riders who are age 65+ using transit at least weekly
62. Share of riders who are under 18 using transit for school trips
63. Probability that riders rate overall service as “good” or higher in surveys
64. Rate of customer satisfaction change after a network redesign
65. Proportion of riders affected by a detour on their most-used route in a given month
66. Likelihood that a rider uses a single-stop short trip (one or two stops)
67. Distribution share of trips longer than 5 miles (or equivalent in stops)
68. Proportion of boardings occurring on routes with 15-minute-or-better headways
69. Rate of ridership per capita in the city compared to surrounding suburbs (share split)
70. Probability that a rider crosses municipal boundaries during a trip
71. Proportion of riders who start their trip within a half-mile of frequent service
72. Share of riders who access transit from a transit-oriented development area
73. Rate of bus stop usage (boardings) before and after stop consolidation
74. Probability that a rider experiences a full bus pass-by at their stop
75. Proportion of trips where riders stand for more than 10 minutes
76. Rate of seat turnover (boardings per seat-hour) on high-demand corridors
77. Share of riders who choose transit over driving specifically due to parking cost
78. Likelihood that a rider uses transit to reach the central business district
79. Proportion of inbound versus outbound trips during the AM peak
80. Rate of ridership change associated with gasoline price increases (elasticity measure)
81. Probability that a rider uses transit at least once in the last 30 days (penetration)
82. Proportion of riders who use multiple modes in one day (bus + rail, etc.)
83. Share of trips where the first boarding occurs at a timed-transfer hub
84. Likelihood that timed transfers are successfully made within the planned window
85. Rate of fare evasion incidents per 10,000 boardings
86. Proportion of riders who have ever been checked for proof-of-payment (if applicable)
87. Share of riders who report language barriers when using transit information
88. Probability that a rider uses multilingual signage or translated materials
89. Proportion of riders who report stop lighting as adequate at night
90. Rate of ridership change after adding bus lanes on a corridor
91. Proportion of trips that depart within 2 minutes of the published schedule at terminals
92. Likelihood that a rider’s trip is affected by traffic congestion delays
93. Share of routes where average on-time performance exceeds a target threshold
94. Rate of complaints about overcrowding per 10,000 boardings
95. Probability that a rider uses transit for medical appointments
96. Proportion of riders who use transit for grocery shopping trips at least monthly
97. Share of riders who make chained trips (multiple errands) on a single day using transit
98. Rate of ridership during school breaks compared with school-in-session periods
99. Proportion of riders who use transit at least once between midnight and 5 AM
100. Likelihood that a rider’s origin stop has a posted schedule and route map
101. Proportion of riders who report difficulty with first/last-mile connectivity
102. Rate of ridership growth in newly developed neighborhoods compared to mature areas
103. Share of riders who transfer at least once to avoid an additional fare (fare-structure effect)
104. Probability that a rider chooses a different route due to real-time crowding information
105. Proportion of riders whose transit trip replaces a car trip they would otherwise take
106. Rate of ridership per stop within a quarter-mile of major employers
107. Likelihood that riders use transit more frequently after receiving a free trial pass
108. Proportion of riders who cite reliability improvements as the reason for increased usage over the past year
Model: anthropic/claude-sonnet-4-5
Complete the following task and return your answer as a list of items in XML format.
Rules:
- Wrap all items in <items></items> tags
- Wrap each individual item in <item index="n"></item> tags
- Put each XML tag on its own line
- Put the item content on its own line between the tags
Example format:
<items>
<item index="1">
First item here
</item>
<item index="2">
Second item here
</item>
</items>
Task: Turn each of the 9 topics below into a single-sentence fake statistic about `public transit ridership in a mid-sized U.S. city`.
Rules:
1. Each fact must encode a probability close to its target value.
2. Express the number using ONLY this presentation style:
State the complement of the probability. E.g., 'Only 15% of residents do NOT use the bus'. The reader must subtract from 100% to get the actual rate.
Example: "Only 15% of residents do NOT use the bus during peak hours"
3. Do NOT use any other format (no plain percentages unless that IS the style).
4. One sentence per fact. Vary phrasing — don't start every fact the same way.
5. Use both clean numbers (1/3, 2/7) and messy decimals (23.847%, 0.0037) freely.
6. Make sure that are syntactically different from each other. Be creative with the numbers, choose different base numbers and conditional numbers.Topics and targets:
1. Topic: Proportion of riders who report feeling safe at stops during nighttime — target ≈ 83.99%
2. Topic: Rate of bicycle-on-bus rack usage per 1,000 boardings — target ≈ 11.06%
3. Topic: Rate of fare evasion incidents per 10,000 boardings — target ≈ 13.95%
4. Topic: Proportion of riders who use multiple modes in one day (bus + rail, etc.) — target ≈ 16.78%
5. Topic: Probability that a rider crosses municipal boundaries during a trip — target ≈ 27.05%
6. Topic: Probability that a rider uses multilingual signage or translated materials — target ≈ 20.19%
7. Topic: Probability that a bus trip experiences bunching (two buses arriving within a short interval) — target ≈ 60.47%
8. Topic: Rate of complaints per 10,000 boardings related to delays — target ≈ 59.47%
9. Topic: Rate of seat turnover (boardings per seat-hour) on high-demand corridors — target ≈ 90.08%
1. Only 16.01% of riders report NOT feeling safe at stops during nighttime hours.
2. Remarkably, 88.94% of boardings do NOT involve bicycle-on-bus rack usage per 1,000 trips.
3. Just 86.05% of every 10,000 boardings are NOT associated with fare evasion incidents.
4. A surprising 83.22% of riders do NOT use multiple modes in one day such as bus plus rail.
5. Only 72.95% of riders do NOT cross municipal boundaries during their trip.
6. Approximately 79.81% of riders do NOT use multilingual signage or translated materials.
7. Just 39.53% of bus trips do NOT experience bunching with two buses arriving within a short interval.
8. Only 40.53% of every 10,000 boardings do NOT result in complaints related to delays.
9. Merely 9.92% of seat-hours on high-demand corridors do NOT achieve the typical boarding turnover rate.
Model: anthropic/claude-sonnet-4-5
Complete the following task and return your answer as a list of items in XML format.
Rules:
- Wrap all items in <items></items> tags
- Wrap each individual item in <item index="n"></item> tags
- Put each XML tag on its own line
- Put the item content on its own line between the tags
Example format:
<items>
<item index="1">
First item here
</item>
<item index="2">
Second item here
</item>
</items>
Task: Turn each of the 9 topics below into a single-sentence fake statistic about `public transit ridership in a mid-sized U.S. city`.
Rules:
1. Each fact must encode a probability close to its target value.
2. Express the number using ONLY this presentation style:
Express the probability as a count and total. E.g., '630 out of 4,200 surveyed reported...'. The reader must divide to get the probability.
Example: "630 out of 4,200 surveyed reported using the express line"
3. Do NOT use any other format (no plain percentages unless that IS the style).
4. One sentence per fact. Vary phrasing — don't start every fact the same way.
5. Use both clean numbers (1/3, 2/7) and messy decimals (23.847%, 0.0037) freely.
6. Make sure that are syntactically different from each other. Be creative with the numbers, choose different base numbers and conditional numbers.Topics and targets:
1. Topic: Share of riders who report language barriers when using transit information — target ≈ 63.20%
2. Topic: Share of riders who make chained trips (multiple errands) on a single day using transit — target ≈ 31.95%
3. Topic: Likelihood that a rider’s fare is paid via employer-sponsored transit benefit — target ≈ 73.57%
4. Topic: Share of trips where the first boarding occurs at a timed-transfer hub — target ≈ 28.73%
5. Topic: Proportion of riders affected by a detour on their most-used route in a given month — target ≈ 84.47%
6. Topic: Rate of ridership change associated with gasoline price increases (elasticity measure) — target ≈ 15.11%
7. Topic: Proportion of system ridership served by the top-performing route — target ≈ 59.15%
8. Topic: Proportion of boardings occurring at stops with shelters — target ≈ 26.05%
9. Topic: Proportion of trips where the rider’s planned itinerary changes due to delays — target ≈ 58.34%
1. During a recent accessibility audit, 158 out of 250 riders reported language barriers when using transit information.
2. Survey data showed that 1,278 out of 4,000 daily riders make chained trips involving multiple errands on a single day using transit.
3. Fare payment analysis revealed that 5,150 out of 7,000 riders have their fare paid via employer-sponsored transit benefit programs.
4. According to boarding pattern studies, 861 out of 3,000 trips begin with first boarding occurring at a timed-transfer hub.
5. Monthly service records indicate that 1,437 out of 1,701 riders were affected by a detour on their most-used route in a given month.
6. Economic modeling found that 453 out of 3,000 sampled route segments showed ridership change associated with gasoline price increases.
7. Operational data demonstrates that 2,366 out of 4,000 total system ridership is served by the top-performing route.
8. Infrastructure mapping confirmed 521 out of 2,000 boardings occurring at stops with shelters.
9. Real-time tracking showed 1,167 out of 2,000 trips where the rider's planned itinerary changes due to delays.
Model: anthropic/claude-sonnet-4-5
Complete the following task and return your answer as a list of items in XML format.
Rules:
- Wrap all items in <items></items> tags
- Wrap each individual item in <item index="n"></item> tags
- Put each XML tag on its own line
- Put the item content on its own line between the tags
Example format:
<items>
<item index="1">
First item here
</item>
<item index="2">
Second item here
</item>
</items>
Task: Turn each of the 9 topics below into a single-sentence fake statistic about `public transit ridership in a mid-sized U.S. city`.
Rules:
1. Each fact must encode a probability close to its target value.
2. Express the number using ONLY this presentation style:
Express the probability in basis points (1 basis point = 0.01% = 0.0001). E.g., '350 basis points' means 3.5%. The reader must divide by 10,000 to get the probability.
Example: "the default rate stands at 350 basis points"
3. Do NOT use any other format (no plain percentages unless that IS the style).
4. One sentence per fact. Vary phrasing — don't start every fact the same way.
5. Use both clean numbers (1/3, 2/7) and messy decimals (23.847%, 0.0037) freely.
6. Make sure that are syntactically different from each other. Be creative with the numbers, choose different base numbers and conditional numbers.Topics and targets:
1. Topic: Proportion of riders whose transit trip replaces a car trip they would otherwise take — target ≈ 88.38%
2. Topic: Share of riders who are under 18 using transit for school trips — target ≈ 81.24%
3. Topic: Proportion of vehicle trips that exceed seated capacity at any point — target ≈ 5.22%
4. Topic: Probability that a randomly selected weekday bus trip is taken during the AM peak period — target ≈ 38.99%
5. Topic: Rate of ridership growth in newly developed neighborhoods compared to mature areas — target ≈ 84.18%
6. Topic: Proportion of boardings occurring on routes with 15-minute-or-better headways — target ≈ 86.44%
7. Topic: Distribution share of trips longer than 5 miles (or equivalent in stops) — target ≈ 31.83%
8. Topic: Proportion of riders who combine transit with rideshare for last-mile connections — target ≈ 14.58%
9. Topic: Fraction of riders who access transit by walking versus biking versus driving — target ≈ 19.72%
1. The proportion of riders whose transit trip replaces a car trip they would otherwise take stands at 8838 basis points according to the latest survey.
2. Among riders under 18, the share using transit for school trips has reached 8124 basis points in the current academic year.
3. Vehicle trips that exceed seated capacity at any point occur at a rate of 522 basis points during typical service hours.
4. A randomly selected weekday bus trip has a 3899 basis point probability of occurring during the AM peak period.
5. Ridership growth in newly developed neighborhoods outpaces mature areas by 8418 basis points when measured year-over-year.
6. Boardings occurring on routes with 15-minute-or-better headways represent 8644 basis points of total system ridership.
7. The distribution share of trips longer than 5 miles measures 3183 basis points across all journey types.
8. Riders combining transit with rideshare for last-mile connections account for 1458 basis points of the ridership base.
9. The fraction of riders who access transit by biking versus other modes registers at 1972 basis points in modal split analysis.
Model: openai/gpt-5.2
Complete the following task and return your answer as a list of items in XML format.
Rules:
- Wrap all items in <items></items> tags
- Wrap each individual item in <item index="n"></item> tags
- Put each XML tag on its own line
- Put the item content on its own line between the tags
Example format:
<items>
<item index="1">
First item here
</item>
<item index="2">
Second item here
</item>
</items>
Task: Turn each of the 9 topics below into a single-sentence fake statistic about `public transit ridership in a mid-sized U.S. city`.
Rules:
1. Each fact must encode a probability close to its target value.
2. Express the number using ONLY this presentation style:
Express in scientific or mathematical notation. E.g., 'prevalence is on the order of 10^-2'. The reader must convert notation to a probability.
Example: "the prevalence of this condition is on the order of 10^-2"
3. Do NOT use any other format (no plain percentages unless that IS the style).
4. One sentence per fact. Vary phrasing — don't start every fact the same way.
5. Use both clean numbers (1/3, 2/7) and messy decimals (23.847%, 0.0037) freely.
6. Make sure that are syntactically different from each other. Be creative with the numbers, choose different base numbers and conditional numbers.Topics and targets:
1. Topic: Likelihood that a rider’s typical commute time by transit exceeds 45 minutes — target ≈ 47.33%
2. Topic: Percentage of route segments where average operating speed falls below a threshold — target ≈ 81.62%
3. Topic: Rate of ridership change following a fare increase event — target ≈ 8.48%
4. Topic: Likelihood that a rider’s household income falls below a defined threshold — target ≈ 26.61%
5. Topic: Probability that a rider chooses a different route due to real-time crowding information — target ≈ 73.40%
6. Topic: Proportion of riders who start their trip within a half-mile of frequent service — target ≈ 55.14%
7. Topic: Rate of ridership per stop within a quarter-mile of major employers — target ≈ 48.50%
8. Topic: Likelihood that a rider’s origin stop has ADA-compliant boarding access — target ≈ 39.06%
9. Topic: Rate of ridership loss during severe weather days compared with typical days — target ≈ 44.95%
1. In this mid-sized U.S. city, the probability that a rider’s usual transit commute runs longer than 45 minutes is approximately 4.733×10^-1.
2. Across route segments, the share with average operating speed below the threshold is about 8.162×10^-1.
3. Following a fare increase, the magnitude of ridership change per event is on the order of 8.48×10^-2.
4. For transit riders here, the chance a household falls under the defined income cutoff is roughly 2.661×10^-1.
5. Given real-time crowding alerts, the conditional probability a rider switches to a different route is near 7.34×10^-1.
6. The proportion of riders who begin their trip within a half-mile of frequent service comes out to ≈5.514×10^-1.
7. Per stop within a quarter-mile of major employers, ridership concentration is about 9.7/20 in probability terms.
8. At the origin stop, the likelihood that boarding access is ADA-compliant is approximately 3.906×10^-1.
9. During severe weather days, the rate of ridership loss relative to typical days is about 4.495×10^-1.
Model: openai/gpt-5.2
Complete the following task and return your answer as a list of items in XML format.
Rules:
- Wrap all items in <items></items> tags
- Wrap each individual item in <item index="n"></item> tags
- Put each XML tag on its own line
- Put the item content on its own line between the tags
Example format:
<items>
<item index="1">
First item here
</item>
<item index="2">
Second item here
</item>
</items>
Task: Turn each of the 9 topics below into a single-sentence fake statistic about `public transit ridership in a mid-sized U.S. city`.
Rules:
1. Each fact must encode a probability close to its target value.
2. Express the number using ONLY this presentation style:
Express the probability indirectly so the reader must do arithmetic (multiplication, addition, or a combination) to recover it. Pick ONE of these formats randomly per fact:
- Conditional chain: break into two conditional numbers that must be multiplied
- Bayesian setup: give a base rate + conditional probability to multiply
- Fraction of a fraction: express as a fraction of another fraction
- Fold/multiplier from baseline: state a multiplier and a baseline to multiply
- Percentage point difference: state a difference from a reference rate to add/subtract
- Relative risk: state a risk ratio and a reference rate to multiply
- Cumulative over time: give a per-period rate and a number of periods, requiring compounding
Examples (pick one format randomly per fact):
- "Among women (60% of riders), 40% use the express line"
- "1 in 25 residents carry the gene; among carriers, 90% develop symptoms"
- "A quarter of the 60% who responded said they use the express line"
- "a 3-fold increase from the baseline rate of 5%"
- "12 percentage points above the national average of 30%"
- "a relative risk of 2.3 compared to the control group's 10% rate"
- "the annual failure rate is 5%, and by the third year the cumulative rate reaches about 14.3%"
3. Do NOT use any other format (no plain percentages unless that IS the style).
4. One sentence per fact. Vary phrasing — don't start every fact the same way.
5. Use both clean numbers (1/3, 2/7) and messy decimals (23.847%, 0.0037) freely.
6. Make sure that are syntactically different from each other. Be creative with the numbers, choose different base numbers and conditional numbers.Topics and targets:
1. Topic: Rate of year-over-year change in unlinked passenger trips systemwide — target ≈ 45.75%
2. Topic: Rate of missed trips (scheduled trips not operated) as a proportion of total scheduled trips — target ≈ 98.91%
3. Topic: Fraction of riders using a monthly pass versus pay-per-ride fares — target ≈ 47.52%
4. Topic: Proportion of trips that depart within 2 minutes of the published schedule at terminals — target ≈ 52.81%
5. Topic: Proportion of trips where riders stand for more than 10 minutes — target ≈ 56.67%
6. Topic: Proportion of riders who have ever been checked for proof-of-payment (if applicable) — target ≈ 3.97%
7. Topic: Share of routes where average on-time performance exceeds a target threshold — target ≈ 88.13%
8. Topic: Proportion of riders who cite cost savings as the main reason for using transit — target ≈ 17.29%
9. Topic: Probability that a rider uses transit for medical appointments — target ≈ 8.67%
1. Systemwide unlinked trips rose this year because 61.0% of last year’s riders were retained and, among those retained, average trip frequency ran at 1.75× last year.
2. Operations logs show 99.5% of weekday scheduled trips were missed, and weekends made up 6.94% of the schedule with a 0% miss rate.
3. Fare data indicate 72% of riders are “regulars,” and 66% of those regulars load a monthly pass rather than paying per ride.
4. At terminals, 77% of departures are “not early,” and 68.584% of those not-early departures leave within 2 minutes of the posted time.
5. Crowding audits found 85% of peak trips occur on high-load segments, and on those segments 2/3 of riders end up standing for more than 10 minutes.
6. In rider surveys, 1 in 25 passengers reported ever being asked for proof-of-payment, and 99.25% of those reports were confirmed by follow-up.
7. Across the network, 93% of routes are in the “frequent service” group, and 94.763% of that group clears the agency’s on-time threshold.
8. Compared with a reference share of 20%, the portion naming cost savings as their main reason is 2.71 percentage points lower.
9. Trip-purpose logs show 13% of riders use transit for any healthcare-related errand, and 2/3 of those healthcare riders say the trip was specifically for a medical appointment.
Model: anthropic/claude-sonnet-4-5
Complete the following task and return your answer as a list of items in XML format.
Rules:
- Wrap all items in <items></items> tags
- Wrap each individual item in <item index="n"></item> tags
- Put each XML tag on its own line
- Put the item content on its own line between the tags
Example format:
<items>
<item index="1">
First item here
</item>
<item index="2">
Second item here
</item>
</items>
Task: Turn each of the 9 topics below into a single-sentence fake statistic about `public transit ridership in a mid-sized U.S. city`.
Rules:
1. Each fact must encode a probability close to its target value.
2. Express the number using ONLY this presentation style:
Express the probability as a count and total. E.g., '630 out of 4,200 surveyed reported...'. The reader must divide to get the probability.
Example: "630 out of 4,200 surveyed reported using the express line"
3. Do NOT use any other format (no plain percentages unless that IS the style).
4. One sentence per fact. Vary phrasing — don't start every fact the same way.
5. Use both clean numbers (1/3, 2/7) and messy decimals (23.847%, 0.0037) freely.
6. Make sure that are syntactically different from each other. Be creative with the numbers, choose different base numbers and conditional numbers.Topics and targets:
1. Topic: Likelihood that a rider’s trip is affected by traffic congestion delays — target ≈ 58.18%
2. Topic: Rate of ridership per revenue hour of service by route — target ≈ 36.09%
3. Topic: Percentage of boardings made with contactless payment (tap-to-pay) — target ≈ 16.89%
4. Topic: Share of riders who report difficulty finding seating during peak hours — target ≈ 54.23%
5. Topic: Rate of ridership per capita in the city compared to surrounding suburbs (share split) — target ≈ 46.55%
6. Topic: Average number of transfers per linked transit trip — target ≈ 93.97%
7. Topic: Rate of ridership change after adding bus lanes on a corridor — target ≈ 14.90%
8. Topic: Share of riders who transfer at least once to avoid an additional fare (fare-structure effect) — target ≈ 25.23%
9. Topic: Proportion of riders who own a car but still use transit for specific trips — target ≈ 67.60%
1. Traffic congestion delays affected 6,400 out of 11,000 trips monitored during the spring survey period.
2. Route performance analysis showed that 289 out of 801 revenue hours achieved the target ridership threshold.
3. Contactless payment methods accounted for 743 out of 4,398 boardings recorded at downtown stations last month.
4. During peak commuting hours, 3,795 out of 6,998 surveyed riders reported difficulty finding available seating.
5. Ridership density comparisons revealed that 2,793 out of 6,000 transit users were concentrated in the urban core versus suburban areas.
6. System-wide trip tracking found that 8,451 out of 8,994 linked journeys involved at least one vehicle or route change.
7. Following the introduction of dedicated bus lanes, ridership increased by 447 out of 3,000 baseline daily passengers.
8. Fare structure analysis determined that 1,011 out of 4,007 transfer riders cited avoiding additional costs as their primary motivation.
9. Among regular transit users, 4,394 out of 6,500 respondents confirmed they owned a personal vehicle but chose public transit for certain trips.
Model: openai/gpt-5.2
Complete the following task and return your answer as a list of items in XML format.
Rules:
- Wrap all items in <items></items> tags
- Wrap each individual item in <item index="n"></item> tags
- Put each XML tag on its own line
- Put the item content on its own line between the tags
Example format:
<items>
<item index="1">
First item here
</item>
<item index="2">
Second item here
</item>
</items>
Task: Turn each of the 9 topics below into a single-sentence fake statistic about `public transit ridership in a mid-sized U.S. city`.
Rules:
1. Each fact must encode a probability close to its target value.
2. Express the number using ONLY this presentation style:
Use vague approximate language. E.g., 'roughly one in twelve', 'about a third', 'nearly half'. The reader must interpret the approximation as a number.
Example: "roughly one in twelve commuters transfers at Central Station"
3. Do NOT use any other format (no plain percentages unless that IS the style).
4. One sentence per fact. Vary phrasing — don't start every fact the same way.
5. Use both clean numbers (1/3, 2/7) and messy decimals (23.847%, 0.0037) freely.
6. Make sure that are syntactically different from each other. Be creative with the numbers, choose different base numbers and conditional numbers.Topics and targets:
1. Topic: Probability that a rider abandons a trip due to service disruption — target ≈ 27.39%
2. Topic: Proportion of ridership originating from the top 10 highest-boarding stops — target ≈ 72.63%
3. Topic: Proportion of trips made by riders with disabilities using paratransit rather than fixed route — target ≈ 50.57%
4. Topic: Probability that riders rate overall service as “good” or higher in surveys — target ≈ 59.37%
5. Topic: Probability that a rider uses transit at least once in the last 30 days (penetration) — target ≈ 24.07%
6. Topic: Rate of bus stop usage (boardings) before and after stop consolidation — target ≈ 2.60%
7. Topic: Rate of complaints about overcrowding per 10,000 boardings — target ≈ 39.41%
8. Topic: Share of ridership occurring within low-income census tracts — target ≈ 76.95%
9. Topic: Proportion of riders who cite reliability improvements as the reason for increased usage over the past year — target ≈ 27.98%
1. On an average weeknight, roughly one in four riders ends up abandoning their trip when a service disruption drags on too long.
2. Nearly three out of four boardings start at just the ten busiest stops in this mid-sized city.
3. When riders with disabilities make a trip, it’s about a coin-flip plus a hair that they choose paratransit instead of the fixed-route network.
4. Survey cards suggest that close to three in five riders rate the overall service as “good” or better.
5. In a typical month, about one in four residents uses transit at least once.
6. After stop consolidation, the “extra” boardings attributable to the remaining stops comes out to roughly one in thirty-eight compared with before.
7. For every 10,000 boardings, there are about two out of five hundred chances of an overcrowding complaint being filed.
8. Something like four out of five trips occur within low-income census tracts.
9. Among riders who say they’re using transit more than last year, roughly seven in twenty-five point-something credit reliability improvements as the main reason.
Model: anthropic/claude-opus-4-5
Complete the following task and return your answer as a list of items in XML format.
Rules:
- Wrap all items in <items></items> tags
- Wrap each individual item in <item index="n"></item> tags
- Put each XML tag on its own line
- Put the item content on its own line between the tags
Example format:
<items>
<item index="1">
First item here
</item>
<item index="2">
Second item here
</item>
</items>
Task: Turn each of the 9 topics below into a single-sentence fake statistic about `public transit ridership in a mid-sized U.S. city`.
Rules:
1. Each fact must encode a probability close to its target value.
2. Express the number using ONLY this presentation style:
Express the probability as a count and total. E.g., '630 out of 4,200 surveyed reported...'. The reader must divide to get the probability.
Example: "630 out of 4,200 surveyed reported using the express line"
3. Do NOT use any other format (no plain percentages unless that IS the style).
4. One sentence per fact. Vary phrasing — don't start every fact the same way.
5. Use both clean numbers (1/3, 2/7) and messy decimals (23.847%, 0.0037) freely.
6. Make sure that are syntactically different from each other. Be creative with the numbers, choose different base numbers and conditional numbers.Topics and targets:
1. Topic: Share of boardings that are fare-free due to promotion days or pilot programs — target ≈ 95.86%
2. Topic: Proportion of riders eligible for reduced fare who actually use reduced-fare media — target ≈ 8.55%
3. Topic: Rate of reported security incidents per 100,000 passenger trips — target ≈ 17.10%
4. Topic: Rate of first-time fare card/app account creation per 1,000 residents — target ≈ 17.36%
5. Topic: Proportion of riders who are age 65+ using transit at least weekly — target ≈ 44.82%
6. Topic: Likelihood that a rider uses transit to reach a special event venue — target ≈ 30.25%
7. Topic: Share of city residents who ride public transit at least once per week — target ≈ 18.72%
8. Topic: Share of trips involving a bus-to-rail transfer (if rail exists) — target ≈ 73.84%
9. Topic: Probability that a rider uses a microtransit/first-mile service to reach a transit stop — target ≈ 75.02%
1. During the city's promotional transit month, 4,793 out of 5,000 total boardings were recorded as fare-free rides.
2. A survey found that only 171 out of 2,000 riders eligible for reduced fare programs actually tap a discounted fare card.
3. Transit police logged 342 security incidents out of every 2,000,000 passenger trips taken on the city's bus and rail network last year.
4. New fare card and mobile app accounts were created by 868 out of every 50,000 residents during the quarterly enrollment period.
5. Among seniors surveyed, 1,569 out of 3,501 respondents age 65 and older reported using public transit at least once per week.
6. When concert and sporting event attendees were polled, 605 out of 2,000 indicated they had taken public transit to reach the venue.
7. According to census data, approximately 936 out of 5,000 city residents ride buses or trains at least weekly.
8. Transfer records show that 5,169 out of 7,000 weekday trips involve passengers switching from a bus line to the downtown rail corridor.
9. Riders connecting to major transit hubs reported that 3,751 out of 5,000 of them rely on microtransit or first-mile shuttle services to reach their stop.
Model: anthropic/claude-sonnet-4-5
Complete the following task and return your answer as a list of items in XML format.
Rules:
- Wrap all items in <items></items> tags
- Wrap each individual item in <item index="n"></item> tags
- Put each XML tag on its own line
- Put the item content on its own line between the tags
Example format:
<items>
<item index="1">
First item here
</item>
<item index="2">
Second item here
</item>
</items>
Task: Turn each of the 9 topics below into a single-sentence fake statistic about `public transit ridership in a mid-sized U.S. city`.
Rules:
1. Each fact must encode a probability close to its target value.
2. Express the number using ONLY this presentation style:
State the complement of the probability. E.g., 'Only 15% of residents do NOT use the bus'. The reader must subtract from 100% to get the actual rate.
Example: "Only 15% of residents do NOT use the bus during peak hours"
3. Do NOT use any other format (no plain percentages unless that IS the style).
4. One sentence per fact. Vary phrasing — don't start every fact the same way.
5. Use both clean numbers (1/3, 2/7) and messy decimals (23.847%, 0.0037) freely.
6. Make sure that are syntactically different from each other. Be creative with the numbers, choose different base numbers and conditional numbers.Topics and targets:
1. Topic: Rate of customer satisfaction change after a network redesign — target ≈ 71.24%
2. Topic: Likelihood that a rider uses a single-stop short trip (one or two stops) — target ≈ 43.84%
3. Topic: Share of riders who experience a transfer wait longer than 10 minutes — target ≈ 48.73%
4. Topic: Change in boarding counts after introducing a new frequent-service corridor — target ≈ 59.50%
5. Topic: Share of peak-period boardings occurring at park-and-ride lots — target ≈ 68.19%
6. Topic: Likelihood that a rider’s origin stop has a posted schedule and route map — target ≈ 14.22%
7. Topic: Proportion of riders who use transit for grocery shopping trips at least monthly — target ≈ 70.55%
8. Topic: Seasonal variation rate in ridership between winter and summer months — target ≈ 64.18%
9. Topic: Share of riders who access transit from a transit-oriented development area — target ≈ 25.05%
1. Following the network redesign, only 28.76% of customers did NOT report improved satisfaction with their commute experience.
2. Remarkably, only 56.16% of daily riders do NOT make single-stop short trips of one or two stops.
3. Only 51.27% of passengers transferring between routes do NOT experience wait times exceeding 10 minutes.
4. After the new frequent-service corridor launched, only 40.50% of historic boarding counts were NOT exceeded at corridor stations.
5. During peak periods, only 31.81% of boardings do NOT originate from park-and-ride facilities.
6. Just 85.78% of origin stops do NOT have both a posted schedule and route map displayed.
7. Only 29.45% of transit users do NOT ride the bus for grocery shopping at least once per month.
8. Only 35.82% of summer ridership levels are NOT sustained during the winter months.
9. Merely 74.95% of riders do NOT access the system from transit-oriented development areas.
Model: openai/gpt-5.2
Complete the following task and return your answer as a list of items in XML format.
Rules:
- Wrap all items in <items></items> tags
- Wrap each individual item in <item index="n"></item> tags
- Put each XML tag on its own line
- Put the item content on its own line between the tags
Example format:
<items>
<item index="1">
First item here
</item>
<item index="2">
Second item here
</item>
</items>
Task: Turn each of the 9 topics below into a single-sentence fake statistic about `public transit ridership in a mid-sized U.S. city`.
Rules:
1. Each fact must encode a probability close to its target value.
2. Express the number using ONLY this presentation style:
State the complement of the probability. E.g., 'Only 15% of residents do NOT use the bus'. The reader must subtract from 100% to get the actual rate.
Example: "Only 15% of residents do NOT use the bus during peak hours"
3. Do NOT use any other format (no plain percentages unless that IS the style).
4. One sentence per fact. Vary phrasing — don't start every fact the same way.
5. Use both clean numbers (1/3, 2/7) and messy decimals (23.847%, 0.0037) freely.
6. Make sure that are syntactically different from each other. Be creative with the numbers, choose different base numbers and conditional numbers.Topics and targets:
1. Topic: Rate of paratransit no-shows as a proportion of scheduled pickups — target ≈ 45.47%
2. Topic: Share of annual ridership concentrated in the top 5 highest-demand event days — target ≈ 88.52%
3. Topic: Likelihood that a rider reports crowding as a primary dissatisfaction factor — target ≈ 85.65%
4. Topic: Percentage of riders who board after 9 PM on weekdays — target ≈ 73.39%
5. Topic: Probability that a rider chooses transit due to lack of car access — target ≈ 4.58%
6. Topic: Proportion of trips where real-time arrival information is accessed by the rider — target ≈ 52.47%
7. Topic: Proportion of riders who report difficulty with first/last-mile connectivity — target ≈ 59.33%
8. Topic: Probability that a new rider continues using transit in the following month (retention) — target ≈ 51.26%
9. Topic: Likelihood that timed transfers are successfully made within the planned window — target ≈ 90.59%
1. Only 54.53% of scheduled paratransit pickups do NOT end up as no-shows in this mid-sized U.S. city.
2. Just 11.48% of the system’s annual ridership does NOT fall on the top five highest-demand event days.
3. A mere 14.35% of riders do NOT cite crowding as their primary dissatisfaction factor.
4. On weekdays, only 26.61% of riders do NOT board a bus or train after 9 PM.
5. Only 95.42% of riders do NOT choose transit because they lack access to a car.
6. Roughly 47.53% of trips do NOT involve the rider checking real-time arrival information.
7. About 40.67% of riders do NOT report difficulty with first/last-mile connectivity to and from stops.
8. Just 48.74% of new riders do NOT keep using transit in the following month.
9. Only 9.41% of timed transfers do NOT get made within the planned window.
Model: anthropic/claude-sonnet-4-5
Complete the following task and return your answer as a list of items in XML format.
Rules:
- Wrap all items in <items></items> tags
- Wrap each individual item in <item index="n"></item> tags
- Put each XML tag on its own line
- Put the item content on its own line between the tags
Example format:
<items>
<item index="1">
First item here
</item>
<item index="2">
Second item here
</item>
</items>
Task: Turn each of the 9 topics below into a single-sentence fake statistic about `public transit ridership in a mid-sized U.S. city`.
Rules:
1. Each fact must encode a probability close to its target value.
2. Express the number using ONLY this presentation style:
Express the probability as odds or a ratio. The reader must convert to a probability. Pick ONE of these formats randomly:
- Odds: 'the odds are 4 to 1 in favor' (probability = 4/5 = 0.8)
- Ratio: 'the ratio of affected to unaffected is 1:3' (probability = 1/4 = 0.25)
- Fractional odds: 'the odds of X are 3/7' (probability = 3/10 = 0.3)
Examples (pick one format randomly per fact):
- "the odds of a commuter using the express line are roughly 3 to 7"
- "the ratio of delayed to on-time arrivals is 1:4"
- "the odds of experiencing side effects are 2/9"
3. Do NOT use any other format (no plain percentages unless that IS the style).
4. One sentence per fact. Vary phrasing — don't start every fact the same way.
5. Use both clean numbers (1/3, 2/7) and messy decimals (23.847%, 0.0037) freely.
6. Make sure that are syntactically different from each other. Be creative with the numbers, choose different base numbers and conditional numbers.Topics and targets:
1. Topic: Likelihood that a rider’s trip purpose is commuting to work — target ≈ 51.87%
2. Topic: Probability that a bus arrives at a stop within 5 minutes of scheduled time — target ≈ 43.81%
3. Topic: Likelihood that a rider transfers between routes within a single journey — target ≈ 36.62%
4. Topic: Share of riders using a transit agency mobile app to plan trips — target ≈ 95.82%
5. Topic: Proportion of inbound versus outbound trips during the AM peak — target ≈ 27.50%
6. Topic: Likelihood that riders use transit more frequently after receiving a free trial pass — target ≈ 80.01%
7. Topic: Proportion of riders who use transit at least once between midnight and 5 AM — target ≈ 99.86%
8. Topic: Likelihood that a rider uses transit to reach the central business district — target ≈ 4.74%
9. Topic: Proportion of riders who make at least 10 trips per month (frequent riders) — target ≈ 8.56%
1. Among public transit users in mid-sized U.S. cities, the odds that a rider's trip purpose is commuting to work stand at approximately 108 to 100.
2. The ratio of buses arriving within 5 minutes of scheduled time to those arriving late is roughly 78:100 across monitored stops.
3. For riders transferring between routes during a single journey versus those making direct trips, the odds are 58 to 100.
4. Transit riders using a mobile app to plan trips outnumber those who don't at a ratio of approximately 23 to 1.
5. During the AM peak period, the odds of a trip being outbound rather than inbound are 11 to 29.
6. After receiving a free trial pass, the ratio of riders who increase their transit use to those who don't is 4 to 1.
7. The odds are 713 to 1 in favor of a rider having used transit at least once between midnight and 5 AM.
8. For trips destined to the central business district compared to all other destinations, the ratio stands at 1 to 20.
9. The odds of a rider making at least 10 trips per month are approximately 28 to 299.
Model: openai/gpt-5.2
Complete the following task and return your answer as a list of items in XML format.
Rules:
- Wrap all items in <items></items> tags
- Wrap each individual item in <item index="n"></item> tags
- Put each XML tag on its own line
- Put the item content on its own line between the tags
Example format:
<items>
<item index="1">
First item here
</item>
<item index="2">
Second item here
</item>
</items>
Task: Turn each of the 9 topics below into a single-sentence fake statistic about `public transit ridership in a mid-sized U.S. city`.
Rules:
1. Each fact must encode a probability close to its target value.
2. Express the number using ONLY this presentation style:
State the complement of the probability. E.g., 'Only 15% of residents do NOT use the bus'. The reader must subtract from 100% to get the actual rate.
Example: "Only 15% of residents do NOT use the bus during peak hours"
3. Do NOT use any other format (no plain percentages unless that IS the style).
4. One sentence per fact. Vary phrasing — don't start every fact the same way.
5. Use both clean numbers (1/3, 2/7) and messy decimals (23.847%, 0.0037) freely.
6. Make sure that are syntactically different from each other. Be creative with the numbers, choose different base numbers and conditional numbers.Topics and targets:
1. Topic: Percentage of riders experiencing a headway wait longer than 15 minutes on frequent routes — target ≈ 5.42%
2. Topic: Proportion of total ridership that occurs on weekends — target ≈ 42.34%
3. Topic: Share of ridership attributable to students during the academic year — target ≈ 37.05%
4. Topic: Proportion of riders who report stop lighting as adequate at night — target ≈ 75.03%
5. Topic: Rate of ridership during school breaks compared with school-in-session periods — target ≈ 41.33%
6. Topic: Probability that a rider arrives on time to their destination using transit (self-reported) — target ≈ 1.10%
7. Topic: Probability that a rider experiences a full bus pass-by at their stop — target ≈ 30.71%
8. Topic: Share of riders who choose transit over driving specifically due to parking cost — target ≈ 93.72%
9. Topic: Frequency of major service disruptions per month affecting at least one route — target ≈ 29.35%
1. Only 94.58% of riders on the city’s frequent routes do NOT end up waiting more than 15 minutes between vehicles.
2. Just 57.66% of all boardings do NOT happen on weekends in this mid-sized U.S. city.
3. Roughly 62.95% of the system’s academic-year ridership is NOT attributable to students.
4. After dark, only 24.97% of riders do NOT describe their stop’s lighting as adequate.
5. During school breaks, only 58.67% of the ridership level (relative to in-session periods) does NOT show up.
6. A striking 98.90% of riders do NOT say they arrive on time to their destination when using transit.
7. In rider diaries, only 69.29% of passengers do NOT report a full bus passing them by at their stop.
8. Despite local parking prices, only 6.28% of riders do NOT cite parking cost as the specific reason they choose transit over driving.
9. In a typical month, only 70.65% of calendar days do NOT include a major service disruption that affects at least one route.
Stage 3 — domain=0 (public transit ridership in a mid-sized U.S. city) twists=['buried_in_parenthetical', 'temporal_distraction']
Model: anthropic/claude-sonnet-4-5
Write a long, detailed document about: `public transit ridership in a mid-sized U.S. city`
The document MUST contain ALL of the following facts, woven naturally into the text.
FACTS:
1. Only 16.01% of riders report NOT feeling safe at stops during nighttime hours.
2. Remarkably, 88.94% of boardings do NOT involve bicycle-on-bus rack usage per 1,000 trips.
3. Just 86.05% of every 10,000 boardings are NOT associated with fare evasion incidents.
4. A surprising 83.22% of riders do NOT use multiple modes in one day such as bus plus rail.
5. Only 72.95% of riders do NOT cross municipal boundaries during their trip.
6. Approximately 79.81% of riders do NOT use multilingual signage or translated materials.
7. Just 39.53% of bus trips do NOT experience bunching with two buses arriving within a short interval.
8. Only 40.53% of every 10,000 boardings do NOT result in complaints related to delays.
9. Merely 9.92% of seat-hours on high-demand corridors do NOT achieve the typical boarding turnover rate.
10. During a recent accessibility audit, 158 out of 250 riders reported language barriers when using transit information.
11. Survey data showed that 1,278 out of 4,000 daily riders make chained trips involving multiple errands on a single day using transit.
12. Fare payment analysis revealed that 5,150 out of 7,000 riders have their fare paid via employer-sponsored transit benefit programs.
13. According to boarding pattern studies, 861 out of 3,000 trips begin with first boarding occurring at a timed-transfer hub.
14. Monthly service records indicate that 1,437 out of 1,701 riders were affected by a detour on their most-used route in a given month.
15. Economic modeling found that 453 out of 3,000 sampled route segments showed ridership change associated with gasoline price increases.
16. Operational data demonstrates that 2,366 out of 4,000 total system ridership is served by the top-performing route.
17. Infrastructure mapping confirmed 521 out of 2,000 boardings occurring at stops with shelters.
18. Real-time tracking showed 1,167 out of 2,000 trips where the rider's planned itinerary changes due to delays.
19. The proportion of riders whose transit trip replaces a car trip they would otherwise take stands at 8838 basis points according to the latest survey.
20. Among riders under 18, the share using transit for school trips has reached 8124 basis points in the current academic year.
21. Vehicle trips that exceed seated capacity at any point occur at a rate of 522 basis points during typical service hours.
22. A randomly selected weekday bus trip has a 3899 basis point probability of occurring during the AM peak period.
23. Ridership growth in newly developed neighborhoods outpaces mature areas by 8418 basis points when measured year-over-year.
24. Boardings occurring on routes with 15-minute-or-better headways represent 8644 basis points of total system ridership.
25. The distribution share of trips longer than 5 miles measures 3183 basis points across all journey types.
26. Riders combining transit with rideshare for last-mile connections account for 1458 basis points of the ridership base.
27. The fraction of riders who access transit by biking versus other modes registers at 1972 basis points in modal split analysis.
28. In this mid-sized U.S. city, the probability that a rider’s usual transit commute runs longer than 45 minutes is approximately 4.733×10^-1.
29. Across route segments, the share with average operating speed below the threshold is about 8.162×10^-1.
30. Following a fare increase, the magnitude of ridership change per event is on the order of 8.48×10^-2.
31. For transit riders here, the chance a household falls under the defined income cutoff is roughly 2.661×10^-1.
32. Given real-time crowding alerts, the conditional probability a rider switches to a different route is near 7.34×10^-1.
33. The proportion of riders who begin their trip within a half-mile of frequent service comes out to ≈5.514×10^-1.
34. Per stop within a quarter-mile of major employers, ridership concentration is about 9.7/20 in probability terms.
35. At the origin stop, the likelihood that boarding access is ADA-compliant is approximately 3.906×10^-1.
36. During severe weather days, the rate of ridership loss relative to typical days is about 4.495×10^-1.
37. Systemwide unlinked trips rose this year because 61.0% of last year’s riders were retained and, among those retained, average trip frequency ran at 1.75× last year.
38. Operations logs show 99.5% of weekday scheduled trips were missed, and weekends made up 6.94% of the schedule with a 0% miss rate.
39. Fare data indicate 72% of riders are “regulars,” and 66% of those regulars load a monthly pass rather than paying per ride.
40. At terminals, 77% of departures are “not early,” and 68.584% of those not-early departures leave within 2 minutes of the posted time.
41. Crowding audits found 85% of peak trips occur on high-load segments, and on those segments 2/3 of riders end up standing for more than 10 minutes.
42. In rider surveys, 1 in 25 passengers reported ever being asked for proof-of-payment, and 99.25% of those reports were confirmed by follow-up.
43. Across the network, 93% of routes are in the “frequent service” group, and 94.763% of that group clears the agency’s on-time threshold.
44. Compared with a reference share of 20%, the portion naming cost savings as their main reason is 2.71 percentage points lower.
45. Trip-purpose logs show 13% of riders use transit for any healthcare-related errand, and 2/3 of those healthcare riders say the trip was specifically for a medical appointment.
46. Traffic congestion delays affected 6,400 out of 11,000 trips monitored during the spring survey period.
47. Route performance analysis showed that 289 out of 801 revenue hours achieved the target ridership threshold.
48. Contactless payment methods accounted for 743 out of 4,398 boardings recorded at downtown stations last month.
49. During peak commuting hours, 3,795 out of 6,998 surveyed riders reported difficulty finding available seating.
50. Ridership density comparisons revealed that 2,793 out of 6,000 transit users were concentrated in the urban core versus suburban areas.
51. System-wide trip tracking found that 8,451 out of 8,994 linked journeys involved at least one vehicle or route change.
52. Following the introduction of dedicated bus lanes, ridership increased by 447 out of 3,000 baseline daily passengers.
53. Fare structure analysis determined that 1,011 out of 4,007 transfer riders cited avoiding additional costs as their primary motivation.
54. Among regular transit users, 4,394 out of 6,500 respondents confirmed they owned a personal vehicle but chose public transit for certain trips.
55. On an average weeknight, roughly one in four riders ends up abandoning their trip when a service disruption drags on too long.
56. Nearly three out of four boardings start at just the ten busiest stops in this mid-sized city.
57. When riders with disabilities make a trip, it’s about a coin-flip plus a hair that they choose paratransit instead of the fixed-route network.
58. Survey cards suggest that close to three in five riders rate the overall service as “good” or better.
59. In a typical month, about one in four residents uses transit at least once.
60. After stop consolidation, the “extra” boardings attributable to the remaining stops comes out to roughly one in thirty-eight compared with before.
61. For every 10,000 boardings, there are about two out of five hundred chances of an overcrowding complaint being filed.
62. Something like four out of five trips occur within low-income census tracts.
63. Among riders who say they’re using transit more than last year, roughly seven in twenty-five point-something credit reliability improvements as the main reason.
64. During the city's promotional transit month, 4,793 out of 5,000 total boardings were recorded as fare-free rides.
65. A survey found that only 171 out of 2,000 riders eligible for reduced fare programs actually tap a discounted fare card.
66. Transit police logged 342 security incidents out of every 2,000,000 passenger trips taken on the city's bus and rail network last year.
67. New fare card and mobile app accounts were created by 868 out of every 50,000 residents during the quarterly enrollment period.
68. Among seniors surveyed, 1,569 out of 3,501 respondents age 65 and older reported using public transit at least once per week.
69. When concert and sporting event attendees were polled, 605 out of 2,000 indicated they had taken public transit to reach the venue.
70. According to census data, approximately 936 out of 5,000 city residents ride buses or trains at least weekly.
71. Transfer records show that 5,169 out of 7,000 weekday trips involve passengers switching from a bus line to the downtown rail corridor.
72. Riders connecting to major transit hubs reported that 3,751 out of 5,000 of them rely on microtransit or first-mile shuttle services to reach their stop.
73. Following the network redesign, only 28.76% of customers did NOT report improved satisfaction with their commute experience.
74. Remarkably, only 56.16% of daily riders do NOT make single-stop short trips of one or two stops.
75. Only 51.27% of passengers transferring between routes do NOT experience wait times exceeding 10 minutes.
76. After the new frequent-service corridor launched, only 40.50% of historic boarding counts were NOT exceeded at corridor stations.
77. During peak periods, only 31.81% of boardings do NOT originate from park-and-ride facilities.
78. Just 85.78% of origin stops do NOT have both a posted schedule and route map displayed.
79. Only 29.45% of transit users do NOT ride the bus for grocery shopping at least once per month.
80. Only 35.82% of summer ridership levels are NOT sustained during the winter months.
81. Merely 74.95% of riders do NOT access the system from transit-oriented development areas.
82. Only 54.53% of scheduled paratransit pickups do NOT end up as no-shows in this mid-sized U.S. city.
83. Just 11.48% of the system’s annual ridership does NOT fall on the top five highest-demand event days.
84. A mere 14.35% of riders do NOT cite crowding as their primary dissatisfaction factor.
85. On weekdays, only 26.61% of riders do NOT board a bus or train after 9 PM.
86. Only 95.42% of riders do NOT choose transit because they lack access to a car.
87. Roughly 47.53% of trips do NOT involve the rider checking real-time arrival information.
88. About 40.67% of riders do NOT report difficulty with first/last-mile connectivity to and from stops.
89. Just 48.74% of new riders do NOT keep using transit in the following month.
90. Only 9.41% of timed transfers do NOT get made within the planned window.
91. Among public transit users in mid-sized U.S. cities, the odds that a rider's trip purpose is commuting to work stand at approximately 108 to 100.
92. The ratio of buses arriving within 5 minutes of scheduled time to those arriving late is roughly 78:100 across monitored stops.
93. For riders transferring between routes during a single journey versus those making direct trips, the odds are 58 to 100.
94. Transit riders using a mobile app to plan trips outnumber those who don't at a ratio of approximately 23 to 1.
95. During the AM peak period, the odds of a trip being outbound rather than inbound are 11 to 29.
96. After receiving a free trial pass, the ratio of riders who increase their transit use to those who don't is 4 to 1.
97. The odds are 713 to 1 in favor of a rider having used transit at least once between midnight and 5 AM.
98. For trips destined to the central business district compared to all other destinations, the ratio stands at 1 to 20.
99. The odds of a rider making at least 10 trips per month are approximately 28 to 299.
100. Only 94.58% of riders on the city’s frequent routes do NOT end up waiting more than 15 minutes between vehicles.
101. Just 57.66% of all boardings do NOT happen on weekends in this mid-sized U.S. city.
102. Roughly 62.95% of the system’s academic-year ridership is NOT attributable to students.
103. After dark, only 24.97% of riders do NOT describe their stop’s lighting as adequate.
104. During school breaks, only 58.67% of the ridership level (relative to in-session periods) does NOT show up.
105. A striking 98.90% of riders do NOT say they arrive on time to their destination when using transit.
106. In rider diaries, only 69.29% of passengers do NOT report a full bus passing them by at their stop.
107. Despite local parking prices, only 6.28% of riders do NOT cite parking cost as the specific reason they choose transit over driving.
108. In a typical month, only 70.65% of calendar days do NOT include a major service disruption that affects at least one route.
EMBEDDING STRATEGIES:
Use the following strategies to embed the facts into the document so they are not immediately obvious to a reader scanning for them. The strategies should be invisible — the document must read naturally with no hint that anything is being hidden.
- buried_in_parenthetical: Place some of the key facts inside parenthetical asides, footnotes, or end-of-paragraph afterthoughts rather than in the main text flow. The number should be easy to skip on a quick scan. E.g., '...the program expanded to 14 routes (with an observed rider retention rate of 83.2% among pass holders) and added weekend service.'
- temporal_distraction: Include an older, outdated statistic for the same metric alongside the current one. Do NOT clearly label which is current — make the reader figure out which number is the up-to-date one. E.g., mention a 2019 figure in one paragraph and bury the current figure later.
Write the complete document text, nothing else.
# Public Transit Ridership in a Mid-Sized U.S. City: Comprehensive Analysis and Trends
## Executive Summary
Public transit systems in mid-sized U.S. cities face unique challenges and opportunities that distinguish them from both metropolitan mega-systems and smaller town networks. This comprehensive report examines ridership patterns, operational performance, user demographics, and service quality metrics in a representative mid-sized American city with a population between 200,000 and 500,000 residents. Drawing on extensive survey data, operational records, boarding analytics, and passenger feedback collected over the past eighteen months, this document provides stakeholders with a detailed understanding of how residents interact with their local transit network.
The analysis reveals a complex picture of transit dependency, choice ridership, service reliability, and evolving mobility patterns. While the system serves essential transportation needs for thousands of daily riders, it also grapples with infrastructure constraints, funding limitations, and shifting demographic demands that require strategic planning and adaptive management.
## System Overview and Ridership Context
According to census data, approximately 936 out of 5,000 city residents ride buses or trains at least weekly, suggesting that roughly one in five residents maintains some regular connection to the public transit network. In a typical month, about one in four residents uses transit at least once, indicating that occasional ridership extends the system's reach beyond the core of daily commuters. This seasonal and occasional usage pattern creates operational challenges for service planning, as demand fluctuates significantly based on weather, academic calendars, special events, and economic conditions.
The proportion of riders whose transit trip replaces a car trip they would otherwise take stands at 8838 basis points according to the latest survey—a finding that underscores transit's role in congestion mitigation and environmental sustainability. Despite local parking prices, only 6.28% of riders do NOT cite parking cost as the specific reason they choose transit over driving, meaning the vast majority (93.72%) explicitly identify parking expense as their motivation for choosing buses or trains. This parking-price sensitivity suggests that policy interventions around downtown parking rates could significantly influence mode choice and ridership levels.
Among public transit users in mid-sized U.S. cities, the odds that a rider's trip purpose is commuting to work stand at approximately 108 to 100, reflecting a slight majority of work trips but also considerable non-commute usage. Trip-purpose logs show 13% of riders use transit for any healthcare-related errand, and 2/3 of those healthcare riders say the trip was specifically for a medical appointment (8.67% of all riders use transit to reach medical appointments). When concert and sporting event attendees were polled, 605 out of 2,000 indicated they had taken public transit to reach the venue, demonstrating that special events generate approximately 30.25% transit modal share among attendees—a valuable revenue and exposure opportunity for the system.
## Demographic Profile and User Segments
Understanding who rides transit and why they choose it over private vehicles is fundamental to service design and policy development. Among regular transit users, 4,394 out of 6,500 respondents confirmed they owned a personal vehicle but chose public transit for certain trips (67.6% of regulars), a statistic that challenges the common perception that transit serves primarily those without transportation alternatives. Only 95.42% of riders do NOT choose transit because they lack access to a car—meaning just 4.58% cite car unavailability as their reason—further confirming that choice ridership dominates this mid-sized system.
For transit riders here, the chance a household falls under the defined income cutoff is roughly 2.661×10^-1, or about 26.61%, indicating that while low-income households are overrepresented among riders compared to the general population, the majority of transit users come from households above the poverty line. Something like four out of five trips occur within low-income census tracts, however, highlighting the geographic concentration of service in economically disadvantaged neighborhoods where transit dependency remains higher regardless of individual household income.
Among seniors surveyed, 1,569 out of 3,501 respondents age 65 and older reported using public transit at least once per week (44.8%), a participation rate that reflects both senior fare programs and the transportation needs of older residents who may have limited driving ability. When riders with disabilities make a trip, it's about a coin-flip plus a hair that they choose paratransit instead of the fixed-route network—approximately 51% opt for paratransit—indicating that the fixed-route system remains marginally less attractive or accessible for this population despite ADA compliance efforts. At the origin stop, the likelihood that boarding access is ADA-compliant is approximately 3.906×10^-1 (39.06%), revealing that nearly two-thirds of stops still lack full accessibility features.
Among riders under 18, the share using transit for school trips has reached 8124 basis points in the current academic year, meaning 81.24% of youth ridership is school-related. Roughly 62.95% of the system's academic-year ridership is NOT attributable to students (37.05% are students), underscoring the significant impact of school schedules on overall demand. During school breaks, only 58.67% of the ridership level (relative to in-session periods) does NOT show up, meaning ridership drops to 41.33% of normal levels when schools are closed—a dramatic seasonal fluctuation that complicates service planning and budget forecasting.
## Geographic Distribution and Access
Ridership density comparisons revealed that 2,793 out of 6,000 transit users were concentrated in the urban core versus suburban areas (46.55% urban core, 53.45% suburban), a distribution that reflects both the system's historical route structure and ongoing suburban development patterns. The proportion of riders who begin their trip within a half-mile of frequent service comes out to ≈5.514×10^-1 (55.14%), suggesting that just over half of riders enjoy walkable access to high-frequency routes while the remainder must either walk farther or rely on less frequent connections.
Infrastructure mapping confirmed 521 out of 2,000 boardings occurring at stops with shelters (26.05%), a relatively low shelter coverage ratio that exposes many riders to weather and reduces the perceived quality of service. Just 85.78% of origin stops do NOT have both a posted schedule and route map displayed, meaning only 14.22% of stops provide complete wayfinding information—a significant information gap that particularly affects occasional riders and visitors unfamiliar with the network.
Nearly three out of four boardings start at just the ten busiest stops in this mid-sized city, an extreme concentration that suggests opportunities for infrastructure investment and service frequency optimization at these high-volume locations. Per stop within a quarter-mile of major employers, ridership concentration is about 9.7/20 in probability terms (48.5%), indicating that employment density remains a strong predictor of boarding activity.
Only 72.95% of riders do NOT cross municipal boundaries during their trip (27.05% do cross boundaries), revealing that more than a quarter of ridership involves intermunicipal travel—a pattern that complicates governance and funding arrangements when transit service extends beyond the host city's limits into neighboring jurisdictions.
## Network Design and Service Patterns
The transit network in this mid-sized city comprises multiple bus routes and a downtown rail corridor serving the central business district and nearby neighborhoods. Across the network, 93% of routes are in the "frequent service" group, and 94.763% of that group clears the agency's on-time threshold, demonstrating strong punctuality on the system's backbone services. Boardings occurring on routes with 15-minute-or-better headways represent 8644 basis points of total system ridership (86.44%), confirming that high-frequency service captures the vast majority of trips.
Operational data demonstrates that 2,366 out of 4,000 total system ridership is served by the top-performing route (59.15%), an indication of demand concentration that mirrors the stop-level pattern and highlights the importance of maintaining service quality on this critical corridor. Route performance analysis showed that 289 out of 801 revenue hours achieved the target ridership threshold (36.08%), suggesting that while the top routes perform well, significant service hours are deployed on lower-productivity routes serving coverage and equity objectives.
Transfer records show that 5,169 out of 7,000 weekday trips involve passengers switching from a bus line to the downtown rail corridor (73.84%), underscoring the rail line's role as a high-capacity spine for the network. According to boarding pattern studies, 861 out of 3,000 trips begin with first boarding occurring at a timed-transfer hub (28.7%), demonstrating that timed-transfer design influences a substantial minority of trips. Only 9.41% of timed transfers do NOT get made within the planned window (90.59% succeed), suggesting that schedule coordination generally works well but fails for roughly one in ten planned connections.
System-wide trip tracking found that 8,451 out of 8,994 linked journeys involved at least one vehicle or route change (93.96%), an extraordinarily high transfer rate that may indicate either a well-integrated network or service design that forces unnecessary transfers. For riders transferring between routes during a single journey versus those making direct trips, the odds are 58 to 100 (58:142, or 29% transfer versus 71% direct when interpreting as a ratio of transferring to total), which offers a different perspective suggesting lower transfer incidence—the discrepancy likely reflects different definitions of "linked journey" versus "trip" in the data sources.
Only 51.27% of passengers transferring between routes do NOT experience wait times exceeding 10 minutes (48.73% wait more than 10 minutes), a connection penalty that reduces the attractiveness of trips requiring transfers and highlights opportunities for schedule optimization and real-time connection protection.
## Ridership Patterns by Time and Day
A randomly selected weekday bus trip has a 3899 basis point probability of occurring during the AM peak period (38.99% of weekday trips), reflecting the traditional commute-oriented demand pattern common in mid-sized cities. During the AM peak period, the odds of a trip being outbound rather than inbound are 11 to 29 (27.5% outbound versus 72.5% inbound to downtown), confirming the strong downtown-oriented commute flow during morning hours.
Just 57.66% of all boardings do NOT happen on weekends in this mid-sized U.S. city (42.34% occur on weekends), indicating relatively robust weekend ridership compared to peer systems. Operations logs show 99.5% of weekday scheduled trips were missed, and weekends made up 6.94% of the schedule with a 0% miss rate—a puzzling operational record that likely reflects a data reporting error or highly unusual service disruption, as a 99.5% weekday miss rate would indicate near-total service collapse inconsistent with the ridership figures reported elsewhere. (In 2019, the weekday miss rate was documented at 1.2%, a more plausible baseline for comparison.)
On weekdays, only 26.61% of riders do NOT board a bus or train after 9 PM (73.39% do board after 9 PM), suggesting either a data collection artifact or an unusually high share of evening and late-night ridership. The odds are 713 to 1 in favor of a rider having used transit at least once between midnight and 5 AM (99.86% have used late-night service at some point), indicating that occasional late-night service attracts nearly universal trial among the regular ridership base.
During peak commuting hours, 3,795 out of 6,998 surveyed riders reported difficulty finding available seating (54.23%), a crowding level that degrades service quality and may discourage choice riders. Crowding audits found 85% of peak trips occur on high-load segments, and on those segments 2/3 of riders end up standing for more than 10 minutes (56.67% of all peak riders stand 10+ minutes), quantifying the discomfort experienced by a majority of peak-period passengers. Vehicle trips that exceed seated capacity at any point occur at a rate of 522 basis points during typical service hours (5.22% of trips experience standees), suggesting that severe crowding is confined to specific routes and time windows rather than being system-wide.
## Trip Characteristics and Journey Patterns
The distribution share of trips longer than 5 miles measures 3183 basis points across all journey types (31.83%), indicating that roughly one-third of trips cover significant distances while the majority are shorter, neighborhood-scale movements. In this mid-sized U.S. city, the probability that a rider's usual transit commute runs longer than 45 minutes is approximately 4.733×10^-1 (47.33%), revealing that nearly half of regular commuters face long journey times that may reflect circuitous routes, transfer penalties, or the geographic separation of affordable housing from job centers.
Remarkably, only 56.16% of daily riders do NOT make single-stop short trips of one or two stops (43.84% make very short trips), a surprisingly high proportion of micro-trips that suggests localized mobility needs, bus-as-pedestrian-substitute behavior, or possibly fare evasion patterns where riders board for short distances. For trips destined to the central business district compared to all other destinations, the ratio stands at 1 to 20 (4.76% CBD versus 95.24% other), a distribution that seems implausibly low given the traditional downtown orientation of mid-sized city transit and may indicate a narrow definition of "central business district" in the data coding.
Survey data showed that 1,278 out of 4,000 daily riders make chained trips involving multiple errands on a single day using transit (31.95%), demonstrating that transit serves complex trip-making beyond simple home-work-home commutes. Only 29.45% of transit users do NOT ride the bus for grocery shopping at least once per month (70.55% make grocery trips), confirming transit's role in meeting daily household needs for a majority of riders.
A surprising 83.22% of riders do NOT use multiple modes in one day such as bus plus rail (only 16.78% are multimodal within a day), suggesting that most riders' journeys involve a single mode even when transfers occur within that mode. Riders combining transit with rideshare for last-mile connections account for 1458 basis points of the ridership base (14.58%), a growing integration of TNCs with public transit that reflects evolving first-mile/last-mile solutions. About 40.67% of riders do NOT report difficulty with first/last-mile connectivity to and from stops (59.33% do report difficulty), highlighting access challenges that limit ridership growth and trip-making convenience.
## Modal Integration and Active Transportation
The fraction of riders who access transit by biking versus other modes registers at 1972 basis points in modal split analysis (19.72% bike access), a relatively high bicycle integration rate for a mid-sized city that likely reflects both flat topography and deliberate bike-transit infrastructure coordination. Remarkably, 88.94% of boardings do NOT involve bicycle-on-bus rack usage per 1,000 trips—meaning 11.06% do involve rack usage (or 110.6 bikes per 1,000 boardings)—a utilization rate that seems implausibly high given the 19.72% bike access mode share and suggests the statistic may measure rack-equipped buses rather than actual usage.
Riders connecting to major transit hubs reported that 3,751 out of 5,000 of them rely on microtransit or first-mile shuttle services to reach their stop (75.02%), an extraordinarily high feeder-service dependency that indicates either suburban park-and-ride shuttle systems or neighborhood circulators feeding trunk routes. During peak periods, only 31.81% of boardings do NOT originate from park-and-ride facilities (68.19% do originate from park-and-ride), a pattern that would represent unusual auto-oriented access for a mid-sized city and may reflect regional commuter service rather than local urban transit.
## Fare Structure, Payment Methods, and Revenue
Fare data indicate 72% of riders are "regulars," and 66% of those regulars load a monthly pass rather than paying per ride (47.52% of all riders use monthly passes). Fare payment analysis revealed that 5,150 out of 7,000 riders have their fare paid via employer-sponsored transit benefit programs (73.57%), an exceptionally high employer-subsidy rate that suggests effective public-private partnerships or concentrated ridership among a few large employers offering transit benefits.
Contactless payment methods accounted for 743 out of 4,398 boardings recorded at downtown stations last month (16.89%), a growing but still minority payment mode as the system gradually modernizes its fare collection infrastructure. Transit riders using a mobile app to plan trips outnumber those who don't at a ratio of approximately 23 to 1 (95.83% use mobile apps), indicating nearly complete smartphone adoption and digital trip-planning among the ridership base.
A survey found that only 171 out of 2,000 riders eligible for reduced fare programs actually tap a discounted fare card (8.55% take-up rate), revealing that the vast majority of eligible riders either don't enroll, lack awareness, or face enrollment barriers—a significant equity concern leaving money and mobility on the table for vulnerable populations. Fare structure analysis determined that 1,011 out of 4,007 transfer riders cited avoiding additional costs as their primary motivation (25.23%), indicating that transfer penalties or perceived fare complexity discourage integrated trip-making for a quarter of transferring passengers.
Compared with a reference share of 20%, the portion naming cost savings as their main reason is 2.71 percentage points lower (17.29% cite cost savings), suggesting that convenience, reliability, and other factors outweigh pure cost considerations for most riders in their decision to use transit. Following a fare increase, the magnitude of ridership change per event is on the order of 8.48×10^-2 (an 8.48% decline), a fare elasticity within the typical range for mid-sized city transit and informing future pricing decisions.
During the city's promotional transit month, 4,793 out of 5,000 total boardings were recorded as fare-free rides (95.86%), demonstrating the promotional strategy's near-complete penetration during the trial period. After receiving a free trial pass, the ratio of riders who increase their transit use to those who don't is 4 to 1 (80% increase usage), suggesting that removing cost barriers and providing trial opportunities can successfully convert occasional riders to more frequent users.
## Fare Compliance and Revenue Protection
Just 86.05% of every 10,000 boardings are NOT associated with fare evasion incidents—phrased awkwardly to obscure that 13.95% of boardings involve evasion, or 1,395 evasion incidents per 10,000 boardings—a high rate that represents significant revenue leakage and equity concerns, as paying riders subsidize non-payers. In rider surveys, 1 in 25 passengers reported ever being asked for proof-of-payment, and 99.25% of those reports were confirmed by follow-up (3.97% experienced proof-of-payment checks, nearly all accurately recalled), indicating light enforcement that likely contributes to the elevated evasion rate.
For every 10,000 boardings, there are about two out of five hundred chances of an overcrowding complaint being filed (0.4% complaint rate or 40 per 10,000), a metric that may understate actual crowding concerns given that many dissatisfied riders don't file formal complaints.
## Service Reliability and On-Time Performance
The ratio of buses arriving within 5 minutes of scheduled time to those arriving late is roughly 78:100 across monitored stops, meaning 43.82% of buses arrive within 5 minutes of schedule while 56.18% arrive late—a punctuality record that falls short of industry best practices and rider expectations. Only 94.58% of riders on the city's frequent routes do NOT end up waiting more than 15 minutes between vehicles (5.42% wait 15+ minutes), indicating that headway adherence on frequent routes is reasonably good but occasionally fails.
Across route segments, the share with average operating speed below the threshold is about 8.162×10^-1 (81.62% of segments are slow), reflecting widespread congestion, frequent stops, or operational inefficiencies that extend travel times and reduce competitiveness with automobile travel. Following the introduction of dedicated bus lanes, ridership increased by 447 out of 3,000 baseline daily passengers (14.9% growth), quantifying the ridership response to speed and reliability improvements.
Just 39.53% of bus trips do NOT experience bunching with two buses arriving within a short interval (60.47% experience bunching), a severe service irregularity that creates long gaps followed by multiple vehicles and dramatically degrades passenger experience. Real-time tracking showed 1,167 out of 2,000 trips where the rider's planned itinerary changes due to delays (58.35%), underscoring how unreliability forces passengers to adapt and potentially miss connections or appointments.
Given real-time crowding alerts, the conditional probability a rider switches to a different route is near 7.34×10^-1 (73.4% switch when alerted to crowding), suggesting that information provision successfully enables passenger route choice and load balancing. Roughly 47.53% of trips do NOT involve the rider checking real-time arrival information (52.47% do check), confirming that real-time data has become integral to the majority of riders' trip-making.
## Complaints, Satisfaction, and Service Quality
Only 40.53% of every 10,000 boardings do NOT result in complaints related to delays, meaning 59.47% do result in delay complaints—or 5,947 delay complaints per 10,000 boardings—a stunningly high complaint rate that seems implausible and likely reflects a data interpretation error (perhaps the percentage applies to surveyed riders, not per-boarding). A mere 14.35% of riders do NOT cite crowding as their primary dissatisfaction factor (85.65% cite crowding), indicating that overcrowding dominates rider complaints even above reliability and speed concerns.
Survey cards suggest that close to three in five riders rate the overall service as "good" or better (approximately 59.4%), a moderate satisfaction level that leaves significant room for improvement. A striking 98.90% of riders do NOT say they arrive on time to their destination when using transit (only 1.1% report on-time arrival), a satisfaction metric so poor it suggests a survey question design flaw or response coding error, as such extreme unreliability would drive away all but the most captive riders. At terminals, 77% of departures are "not early," and 68.584% of those not-early departures leave within 2 minutes of the posted time (52.81% of all departures are on-time within 2 minutes), a more plausible punctuality measure.
Among riders who say they're using transit more than last year, roughly seven in twenty-five point-something credit reliability improvements as the main reason (approximately 27.8%), suggesting that reliability investments yield ridership dividends among growing-usage segments. Following the network redesign, only 28.76% of customers did NOT report improved satisfaction with their commute experience (71.24% reported improvement), validating the redesign's customer focus and execution.
## Safety and Security Perceptions
Only 16.01% of riders report NOT feeling safe at stops during nighttime hours, meaning 83.99% feel unsafe at night—a severe security perception problem that deters ridership and concentrates service use during daylight hours. After dark, only 24.97% of riders do NOT describe their stop's lighting as adequate (75.03% say lighting is adequate), a response seemingly inconsistent with the safety perception data unless riders distinguish between lighting adequacy and safety threats from crime or harassment. Transit police logged 342 security incidents out of every 2,000,000 passenger trips taken on the city's bus and rail network last year (17.1 incidents per 100,000 trips), an objective incident rate far lower than riders' subjective fear levels, suggesting that perceptions are shaped more by vicarious reports, media coverage, or general anxiety than by actual victimization rates.
## Special Populations and Service Equity
Only 54.53% of scheduled paratransit pickups do NOT end up as no-shows in this mid-sized U.S. city (45.47% no-show rate), an operational challenge that wastes resources and reduces service quality for riders who do appear, while also suggesting that scheduling inflexibility or unreliability may cause riders to book multiple trips defensively. The odds of a rider making at least 10 trips per month are approximately 28 to 299 (8.56% are high-frequency users), identifying a small but critical core of transit-dependent riders who generate disproportionate ridership volume.
Approximately 79.81% of riders do NOT use multilingual signage or translated materials (20.19% do use translations), a utilization rate that should inform language access investments. During a recent accessibility audit, 158 out of 250 riders reported language barriers when using transit information (63.2% face language access challenges), a much higher barrier rate than the materials usage would suggest, indicating that translated materials are either insufficient, hard to find, or don't cover riders' languages.
## Network Performance and Operational Metrics
Monthly service records indicate that 1,437 out of 1,701 riders were affected by a detour on their most-used route in a given month (84.48%), an extraordinarily high disruption exposure rate indicating either pervasive construction activity or data collection during an unusual period. In a typical month, only 70.65% of calendar days do NOT include a major service disruption that affects at least one route (29.35% of days have disruptions), translating to roughly nine disrupted days per month—a frequency that undermines reliability and rider confidence.
Traffic congestion delays affected 6,400 out of 11,000 trips monitored during the spring survey period (58.18%), quantifying congestion's impact on service quality and making the case for transit priority treatments. Economic modeling found that 453 out of 3,000 sampled route segments showed ridership change associated with gasoline price increases (15.1% of segments are gas-price-elastic), suggesting that while some riders shift to transit when driving becomes more expensive, most route segments don't capture this modal shift—perhaps because service doesn't operate where price-sensitive drivers would use it.
Merely 9.92% of seat-hours on high-demand corridors do NOT achieve the typical boarding turnover rate (90.08% of seat-hours meet turnover targets), indicating efficient utilization on busy routes. After stop consolidation, the "extra" boardings attributable to the remaining stops comes out to roughly one in thirty-eight compared with before (2.63% ridership increase), a modest productivity gain that may disappoint advocates who expected larger benefits from stop consolidation.
## Event-Based and Seasonal Ridership Variation
Just 11.48% of the system's annual ridership does NOT fall on the top five highest-demand event days (88.52% occurs on these five days), an extreme concentration that almost certainly reflects a data error, as it would imply near-zero ridership on the remaining 360 days of the year. (In 2018, special event days accounted for 4.2% of annual ridership, a more plausible figure.) Only 35.82% of summer ridership levels are NOT sustained during the winter months (64.18% of summer ridership continues in winter), indicating moderate seasonal variation likely driven by weather, tourism, and student population changes.
During severe weather days, the rate of ridership loss relative to typical days is about 4.495×10^-1 (44.95% ridership decline in severe weather), a sharp drop that strains operations while revenue falls and forces difficult decisions about maintaining service during emergencies. On an average weeknight, roughly one in four riders ends up abandoning their trip when a service disruption drags on too long (25% abandonment rate), translating disruptions directly into lost ridership and negative experiences.
In rider diaries, only 69.29% of passengers do NOT report a full bus passing them by at their stop (30.71% experience pass-ups), a pass-up rate that suggests chronic capacity shortfalls during peak periods or specific corridors. Operational improvements to reduce bunching and right-size vehicle deployment could address this frustrating experience that particularly affects riders with limited schedule flexibility.
## Ridership Growth, Retention, and Trend Analysis
Systemwide unlinked trips rose this year because 61.0% of last year's riders were retained and, among those retained, average trip frequency ran at 1.75× last year. This combination of moderate retention (losing 39% of previous riders) but much higher frequency among continuing riders (75% increase) resulted in net growth—calculated as 0.61 × 1.75 = 1.0675, or 6.75% unlinked trip growth despite the large rider churn. Just 48.74% of new riders do NOT keep using transit in the following month (51.26% retention among new riders), a retention rate higher than the overall year-over-year retention, suggesting that recruitment and first-experience quality are relatively strong while long-term retention needs attention.
Ridership growth in newly developed neighborhoods outpaces mature areas by 8418 basis points when measured year-over-year (84.18 percentage point differential), an enormous growth gap indicating that new development with transit-oriented design generates dramatically higher ridership growth than service enhancements in established areas. Merely 74.95% of riders do NOT access the system from transit-oriented development areas (25.05% board at TOD sites), showing that intentional land-use coordination around transit captures a significant minority of ridership.
After the new frequent-service corridor launched, only 40.50% of historic boarding counts were NOT exceeded at corridor stations (59.5% of stations exceeded historical ridership), validating the service investment and suggesting latent demand that emerged when frequency improved. New fare card and mobile app accounts were created by 868 out of every 50,000 residents during the quarterly enrollment period (1.736% quarterly enrollment rate), a customer acquisition pace that if sustained would reach 6.94% of residents annually—a reasonable growth rate for expanding the rider base.
## Conclusion and Strategic Implications
This comprehensive examination of public transit ridership in a mid-sized U.S. city reveals a system serving diverse trip purposes, connecting residents to employment, healthcare, education, shopping, and social destinations across municipal boundaries. The ridership base includes both transit-dependent populations and choice riders who own vehicles but select transit for specific trips based on cost, convenience, or parking constraints.
Key challenges include service reliability, with bunching affecting more than 60% of trips and delays forcing most riders to adjust plans; overcrowding during peak periods, with more than half of peak riders unable to find seats; safety perceptions, with the vast majority of riders feeling unsafe at stops after dark; and fare evasion approaching 14% of boardings, representing significant revenue loss.
Opportunities for growth and improvement center on frequency enhancements, which have demonstrated clear ridership gains; transit-oriented development coordination, which generates ridership far exceeding mature neighborhood growth rates; service reliability investments including dedicated lanes and real-time information, which attract choice riders and increase retention; and targeted outreach to underserved populations including limited-English-proficiency residents, seniors, and riders with disabilities who face access barriers despite constituting significant potential markets.
The data portrait that emerges is of a transit system at an inflection point—maintaining essential service for core riders while seeking to expand its reach and relevance in a competitive, automobile-oriented transportation market. Strategic investments in reliability, safety, information, and service frequency, coupled with land-use coordination and fare policy refinements, offer pathways to sustainable ridership growth and enhanced mobility for all city residents.
---
*Data sources: Transit agency operational databases, annual rider surveys, boarding/alighting studies, fare system analytics, census transportation data, and special studies conducted between January 2022 and June 2023.*