London loves Lizzy.
Air con starts arguments.
What we learned from your Tube rankings.
Research snapshot · 11 September 2026 · all 19 services. Votes and selected social comments; not a live report.
GOAT status maintained.
🐐90%
of votes put Elizabeth in A*/A. It also comes out on top in both taste groups. Different journeys, same favourite.
Tube enemies.
VS
Victoria vs Circle
Love Victoria? You’re more likely to be cooler on Circle.
Speed or comfort? Victoria’s frequent service and Circle’s air-conditioned carriages offer different things. Could that explain the rivalry?
VS
Victoria vs Hammersmith & City
Vicky fans aren’t quite as warm towards the pink line.
Here for the quick commute or the cool carriage? Victoria’s frequent trains and Hammersmith & City’s air conditioning might appeal to different priorities.
VSNorthern vs Hammersmith & City
Team Northern? The pink line might not be your first pick.
Bank or Charing Cross? Northern fans have their branches. Hammersmith & City takes a different path across town. Could this simply be about which line gets you where you need to go?
Different favourites, a little friendly rivalry. These three pairings show a similar pattern in the votes.
Air con has entered the chat.
On Instagram, a Victoria critic objects to heat and noise. A fan defends its speed—even changing onto it and back to Northern to shorten a trip. Another rider says air con shaped their hot-weather route.
Could comfort versus speed help explain the rivalries? It’s a question worth asking, rather than something the votes prove.
How strong are these rivalries?
It’s not really that deep. There’s a small pattern: when someone rates one line higher, they tend to rate the other lower. Plenty of people can still love both.
The travel explanations above are our guesses, not something the votes prove. Victoria service frequency · Air-conditioned Circle and Hammersmith & City trains · Northern’s routes · Hammersmith & City’s route.
Want the numbers? The score runs from −1 (opposite tastes), through 0 (no relationship detected by this score), to +1 (similar tastes). These pairs sit around −0.19: a little disagreement, not an all-out feud. We account for people who give everything high or low ratings.
| Tube lines | Pattern score | Estimated range (95%) |
|---|---|---|
| Circle versus Victoria | -0.19 | -0.26 to -0.11 |
| Victoria versus Hammersmith & City | -0.19 | -0.26 to -0.12 |
| Northern versus Hammersmith & City | -0.19 | -0.25 to -0.12 |
Based on 799 submissions. The ranges show how much the score varies when we repeatedly resample these votes; they are not a margin of error for all Londoners. The pattern holds using different scoring checks, and we checked all 171 pairs to reduce the risk of picking a chance result.
The unconventional favourites.
These pairings don’t have an obvious geographic link, but rate one more highly and you tend to rate the other more highly too. Trams, Cable car and DLR—could this be a fandom for TfL’s less conventional options?
♥
Trams + Cable car
♥DLR + Trams
See the numbers behind these pairings
Trams + Cable car: adjusted relationship 0.24 · 95% interval 0.17 to 0.32.
DLR + Trams: adjusted relationship 0.24 · 95% interval 0.18 to 0.30.
These are modest positive relationships, not a measure of how many people love both. A shared fandom is one possible explanation; unfamiliarity could also shape the scores. We haven’t measured why people voted or controlled for geography.
Route context: TfL’s network overview and Tube map. Vote data: the same 799-submission snapshot.
Two tendencies. One shared winner.
Two crews, plenty of crossover. These playful characters capture different transport tastes—not who people are or where they live.
The “just get me there” crew
“Vicky or Northern, please. The boat can wait.”
Relatively more favourable towards Victoria and Northern; harsher towards Cable car and Uber Boat.
The “journey’s half the fun” crew
“Train, boat, cable car? Keep your options open.”
Relatively warmer towards DLR, Uber Boat and Cable car; less enthusiastic about Victoria and Northern. “Warmer” does not mean those services become favourites.
See how the two crews scored each service
Group A: the “just get me there” crew (491 rankings). Group B: the “journey’s half the fun” crew (307 rankings).
| Service | Group A | Group B |
|---|---|---|
| Victoria | 4.53 | 2.86 |
| Northern | 3.55 | 2.09 |
| DLR | 2.98 | 3.77 |
| Uber Boat | 1.18 | 2.73 |
| Cable car | 0.69 | 2.21 |
| Metropolitan | 2.88 | 2.93 |
| Elizabeth line | 5.51 | 5.54 |
Elizabeth is the highest-rated service in both groups. Different tastes, same winner.
How reliable are these groups?
Source: original Tube submissions, 7–10 September 2026. One all-equal ranking excluded from clustering only.
The two-group solution repeats well when the sample changes (mean adjusted Rand agreement 0.90), but separation is weak (silhouette 0.11). Alternative score-normalised and within-voter rank versions broadly agree (0.83 and 0.78). Treat this as a preference spectrum, not demographic segmentation.
K-means tested two to six groups, with 30 starts and 20 repeated 80% subsamples per group count. More groups produced weaker, less repeatable distinctions.
Front seat. Main character energy.
One X fan gives DLR top-tier status on one condition: getting a seat at the front. Sometimes a really good seat is part of the appeal.
DLR gets A*/A from 26% of votes, versus E/F from 16%. It’s also rated more warmly by the “journey’s half the fun” crew.
The comment fits that story. We’d need to ask voters to know whether it explains their scores.
Boat vibes. But at what price?
♥Uber Boat has the widest score spread, but 52% put it in E/F and only 9% in A*/A. A small fan base, plenty of critics—not a 50/50 split.
Meanwhile, 40% of Cable car’s A*/A fans put Boat there too. Fans of unconventional London travel?
A reply on X raises the fare. Rachel’s response: it would be A* if it were free. Are people ranking the ride, or the value? That’s one to explore next.
Compare the services with fans and critics
Uber Boat’s spread is narrowly wider than Thameslink’s. Metropolitan has 18% in A*/A versus 20% in E/F; DLR has 26% versus 16%; Northern has 16% versus 19%. The middle tiers remain common. A middling average can hide different opinions, without meaning a service has two equal camps.
DLR: votes by tier
Share of votes in each tier, from best to worst.
A*/A: 25.5% · E/F: 15.5%.
See vote counts and scoring detail
A*/A: 204/799 (25.5%). E/F: 124/799 (15.5%). Score spread: 1.64 tiers (standard deviation). Unadjusted counts · 7–10 September 2026.
- A*: 73 votes
- A: 131 votes
- B: 166 votes
- C: 177 votes
- D: 128 votes
- E: 76 votes
- F: 48 votes
What wins for you: speed, air con, reliability or the route?
We’ve seen the scores. Now bring your own priorities to the ranking.
Have your say →
A quick note on the research.
Where these comments come from
Selected comments from Rachel’s original X post and Great Londoner Memes’ Instagram reshare, reviewed on 11 September 2026. The Instagram post showed 4,418 likes when checked; that is engagement, not impressions or reach.
This is a partial, qualitative reading of platform-ordered comments, not a count of how common each opinion is. Comments respond to the original personal tier list, not the community results, and have not been matched to voters. Service complaints are commenters’ experiences, not verified performance statistics. Metropolitan discussion is excluded because it was missing from the original image.
These are anonymous, self-selected submissions, not a representative survey of London or necessarily unique voters. We cannot infer where people live, their journeys or why they voted this way. Compulsory ranking may include services someone has never used.
All 19 services, including Metropolitan, are included.
Read the analysis method
Three early Metropolitan votes were assigned C; sensitivity checks show they do not materially change the findings. All 799 records were complete, with no identical 19-service vote vectors. That does not prove there are no repeat voters or bots.
Read-only snapshot taken 11 September 2026 at 00:24 BST (10 September, 23:24 UTC) from the original Tube database shared by both sites. No voter names are included. A*=6 through F=0 assumes equally spaced tiers.
Pairwise correlations adjust for the voter’s mean score on the other 17 services, excluding the tested pair. Rank-based partial correlations and removing the all-equal vote confirm the main findings. Intervals use 600 row-bootstrap samples per pair; partial-correlation t-tests (796 degrees of freedom) use Benjamini–Hochberg correction across all 171 pairs. The shared-taste and Tube-enemy findings pass a 5% false-discovery threshold.
The strongest positive pairs have similar strengths in the first 400 and last 399 submissions. The strongest negative remains negative in both halves. These checks use the same sample, not independent replication. Confidence intervals describe resampling uncertainty, not a population survey margin of error.