Talk:List of metro systems/Archive 9

Revisiting The "Metro Criteria" Issue
I'm worried that this "10 minute headway" is: I'm trying to get to the criteria we use to determine what is a metro system. The only reason we are having this argument is that the criteria for what is a metro is not well defined here
 * 1. Completely arbitrary. If there is no source for it, then it really should be dropped.
 * 2. Many of these metros have less than 10 minute headway during periods of the day. How is this accounted for?
 * No criteria should be proposed without a source to back it up. No more arbitrary rules made up by random editors. Mattximus (talk) 16:16, 9 January 2014 (UTC)


 * The issue then becomes that you'd then probably have to define it by passenger volume (that is actually how "light" vs. "medium" vs. "heavy" rail seems to be defined, in definitional terms, though I've never seen actual specific numbers assigned to each of these), and I'm not sure that's better than using 'headway' as the "stand-in" for passenger volume. --IJBall (talk) 18:01, 9 January 2014 (UTC)


 * Why would you have to define it by passenger volume? Where are you getting this from? We really should only discuss sourced criteria. Are there international standards that define these things? Mattximus (talk) 18:31, 9 January 2014 (UTC)


 * Both APTA and the U.S. FTA define the terms "heavy rail" and "light rail" in terms of "...capacity for a [heavy or lighter] volume of traffic...", which is passenger volume.
 * But neither quote specific numbers associated with the terms... --IJBall (talk) 18:41, 9 January 2014 (UTC)
 * But neither quote specific numbers associated with the terms... --IJBall (talk) 18:41, 9 January 2014 (UTC)

Based on those two links and this one it seems like a metro system is one that:
 * Is completely separated from traffic (human, car, etc.)
 * Is electric

Distinct from commuter rail which
 * Multi-trip tickets
 * Specific station to station fares

Light rail:


 * ROW of 30% at least (or it is considered a streetcar)
 * Can be crossed by foot or vehicle

Note: The number of trains is irrelevant (single or multi-car), and there is vague notion to capacity, but none to frequency of service.

Mattximus (talk) 20:00, 9 January 2014 (UTC)
 * I agree with Mttximus. I have always thought that this made up number was originally from some Wiki user and then was somehow treated by others as a consensus later. I believed that this was official, but you have proven me wrong. Clearly, there are zero official sources that even mention the service frequency criteria and this must be dropped completely since it is clearly a made up criteria with no official sources. If there are no sources to back up this claim, then it must be dropped according to Wiki guidelines. Only fully sourced information from official references should be included as a criteria. We must follow the official criterias, not what a random Wiki user says. Massyparcer (talk) 10:07, 10 January 2014 (UTC)
 * Note that I blocked Massyparcer for edit warring and advised them to read WP:CONSENSUS.--Ymblanter (talk) 10:22, 10 January 2014 (UTC)
 * Sure, if you just ignore the above references that explicitly mention "volume of traffic"... [roll] Last time I looked, those are sources. --IJBall (talk) 15:55, 10 January 2014 (UTC)
 * Yes, but the "volume of traffic" is extremely vague - Defining it in terms of a arbitrary number is POV and unsourced. The source mentions no information on what a volume of traffic is. Where is the 10min headway rule in your source? It is non-existent. If it isn't sourced, it must be removed as per Wikipedia rules. Massyparcer (talk) 17:31, 11 January 2014 (UTC)
 * I agree that the 10 minute headway must be dropped unless a source is provided, but also it appears that there is still some doubt about previously mentioned lines. I suggest coming to a consensus here in the discussion board instead of making changes to the article straight away. Ideally we should agree on criteria for inclusion first and then assess to see what systems meet this criteria. I can't see any other reasonable way. Mattximus (talk) 18:24, 11 January 2014 (UTC)
 * "Volume of traffic" is controlled by two factors:
 * Headway
 * Train length
 * Either of those (preferably both) can be used as "stand-ins" for traffic volume. That is a reasonable and logical conclusion.
 * The problem is, definitely, that these references don't quote actual numbers for "volume of traffic" levels.
 * So it's absolutely correct to use something like headway or train length as a "metro criteria". The issue is that the references don't provide a specific numeric metric for it... --IJBall (talk) 18:32, 11 January 2014 (UTC)

Headway can't be used because there are no sources that show any numbers. The 10 min rule must be dropped as per Wiki rules because we simply can't make up an arbitrary number. Also, like Mattximus says, "The number of trains is irrelevant (single or multi-car)", so train length is completely unsourced and original research, which must be dropped as per Wikipedia rules. There is no source that volume of traffic is even controlled by headway or train length. This is original research without any source. Massyparcer (talk) 20:29, 11 January 2014 (UTC)
 * What you say about train length is inaccurate - train length is absolutely a component of "traffic volume". That is a logical inference, not "original research". But when you say that there are no numbers quoted in the sources we've currently found - that is true. But, if numbers for at least "traffic volume" are found later, then numbers for headway and potentially train length could be used as a criteria for metro status. --IJBall (talk) 20:40, 11 January 2014 (UTC)
 * Whoa hang on there. Where did you get Headway and Train length as a definition for volume of traffic? We have to stop pulling stuff out of the air. Any criteria must come from a source. Just for an example:, what if you have a 10 train system, that runs every 1 minute, but only 1 person rides it. How are your two criteria in this example a determinant of traffic volume? Mattximus (talk) 20:58, 11 January 2014 (UTC)
 * He means design traffic volume, not actual traffic volume.Terramorphous (talk) 21:07, 11 January 2014 (UTC)


 * If a reference gives the passenger capacity per train, the number of trains, and the headway, the volume of traffic can easily be calculated. Now the resulting figure would be maximum (i.e. potential) volume of traffic. What is unfortunate is that this is what the above references mean when they say "volume of traffic" (i.e. is traffic volume capacity or potential, not actual ridership - note that the word "ridership" is not used; they use "volume of traffic" [potential]), though they don't come out and say that. Which is yet another problem...
 * However, without numbers attached to "volume of traffic", I agree that right now they probably can't be used as criteria, which is likely to complicate this list (and the Light rail list) considerably. --IJBall (talk) 21:11, 11 January 2014 (UTC)

Ok. I've wrote about my opinion about this above there, but specifically about the criteria, I'd like to re-emphasize that determining line between metro system isn't something that can be numerically defined. I'd say that there should be more consideration on the how the system is structured. Does the local authorities view and manage them as a single system? This can be deduced by if they same fare system, single named system, and so on. 10 minute criterion is very numerical and distinguishable criterion, but considering the fact that 'encyclopedias' like Wikipedia should be reflecting the real world, 10 minute criterion isn't so nice criterion. MinSik CHO (talk) 05:58, 12 January 2014 (UTC)
 * I agree with this soft view. To conclude the system should:

This is the best solution I have, and it looks like it's generally inclusive. Anyone disagree with these three points? It's still messy, for example, we call line 6 in the Naples metro a "metro" system, but the agency itself calls it light rail. As long as each system is here, List of suburban and commuter rail systems or List of light rail systems I guess it's ok. Mattximus (talk) 15:13, 12 January 2014 (UTC)
 * Be completely separated from human/vehicular traffic
 * Does not have traits of a commuter rail (has fare or zones instead of station to station, not shared with other rail, under a single corporation, etc)
 * Or the local agency itself refers to it as a metro system.

List point it s bit of an issue. As stated by the intro:

"The name of the system is not a criterion for inclusion or exclusion. Some cities use metro as a brand name for a transit line with no component of rapid transit whatsoever. Similarly, there are systems branded light rail that meet every criterion for being a rapid transit system. Some systems also incorporate light metro or light rail lines as part of the larger system under a common name. These are listed, but the light rail lines are not counted in the provided network data."

- List of metro systems

IJBall has pointed out the ten minute rule is a good proxy of capacity. 10 min headway is 6 Trains per hour. 15 min headway is 4 Trains per hour. 6 min headway is 10 Trains per hour. Here are capacities of some metro trains.


 * S7 stock 1,347
 * 2009 stock 1,448
 * 10-car R160 2,436
 * 8 Car Shanghai Metro AC06 2,480
 * TTC TR 1,100
 * Tokyo Metro 16000 series 1,518

So as you can see subway systems have to run much higher than 6 trains per hour to achieve at least 30,000 people per hour per direction. 10 min max during rush hour is reasonable and inclusive.Terramorphous (talk) 16:25, 12 January 2014 (UTC)
 * Strongly disagree with the 10 minute arbitrary number for reasons above, not least of which is the fact that you would have to cut a big chuck of the systems currently on this list as they would not meet that criteria. Also, what if a train has 1 car and runs every 5 minutes, or 10 cars and runs every 15, which has higher capacity then? Mattximus (talk)
 * Just to be clear, the "10-minute headway" criteria is for peak hours only. Systems don't have to have 10-minute headways all-day - just during "rush hour" periods. Most of the systems on the list that I've looked at easily meet that criteria. There's just a few (e.g. Catania Metro) that don't, and I'd argue that systems like Catania don't qualify as "true" heavy rail anyway. --IJBall (talk) 17:24, 12 January 2014 (UTC)


 * A train has 1 car and runs every 5 minutes. You just described the Paris and Mexico city metros. High frequency (trains per hour) Low capacity per train. It still reaches over 30,000 pphpd.


 * A train has 10 cars and runs every 15. However, @ 15min (4 trains per hour) per train each train has to carry 7,500 people to reach 30,000pphpd. That kind of train doesn't even exist. Even at 10 min per train each train still has to carry 5,000 people a piece. Still unreasonable.Terramorphous (talk) 18:24, 12 January 2014 (UTC)


 * BTW, for a reference for Terramorphous's 30,000 pphpd figure for heavy rail, see here: --IJBall (talk) 05:28, 13 January 2014 (UTC)

My only concern with switching to "softer" inclusion criteria is that a number "Light Metros" (e.g. Catania Metro) will then slip in to this list (some of which haven't been included earlier), and those systems aren't true "heavy rail" but are instead "medium-capacity" systems. And the other issue is that traffic volume is supposed to be a defining characteristic of heavy rail systems, even though they don't attach numbers to that. --IJBall (talk) 17:19, 12 January 2014 (UTC)
 * Ultimately, this issue is not about what you guys think but whether it breaches Wikipedia rules and in this case, it does so in a very obvious way. The 10min rule has never existed anywhere officially and came out of nowhere from a random Wiki user and if there are no sources to back up this number, it simply can't be used as per Wikipedia rules because it is original research. As I said before, making up a random number simply isn't allowed on Wikipedia without a source. Like Mattximus said, we have to stop making up rules out of nowhere. Massyparcer (talk) 04:45, 13 January 2014 (UTC)
 * And you continue to ignore while there is not an explicit "headway rule" in defining a Metro system, there is traffic volume criteria involved in the definition, and headway is an explicit component of that. I think we'd all appreciate if you'd stop throwing around terms like "original research" in less than accurate ways... --IJBall (talk) 05:17, 13 January 2014 (UTC)
 * Clearly making up a 10min rule is original research. Why is this any less accurate? Even if there is a traffic volume criteria, there is no source that shows any link between that and headways or train length - Before we even start to consider the traffic volume criteria, it is original research without any source showing that a link exists in the first place. Pure logic simply doesn't cut it in Wikipedia because the logic is original research without any source. Even if such a link exists, we can't use the 10 min rule because it is a made-up arbitrary number, which is forbidden in Wikipedia. Massyparcer (talk) 05:39, 13 January 2014 (UTC)
 * I'm not talking specifically about the "10min headway rule". What I'm talking about is that headway is an inescapable component traffic volume, and therefore headway can potentially be used as a criteria. BTW, for another reference, see pg. 5-9 of this: - the "number of cars" or "number of trains" per hour is directly determined by headway, which establishes that volume of traffic is directly related to headway. --IJBall (talk) 06:07, 13 January 2014 (UTC)
 * I agree that there is a link between the two, but like I have said before, even if there is a link between headways and volume traffic, we can't use a 10min rule because it is making up an arbitrary number. It has has to be dropped until we can get a reliable, official source that mentions the numbers, because the sources so far don't give any specific numbers. Let's stop making up numbers. Massyparcer (talk) 06:30, 13 January 2014 (UTC)
 * Personally, I think that in a long-time scale, the division of railway systems should be 1. Heavy Rail Commuter Transits (is the term right...?) 2. Light Rail Commuter Transits 3. Normal Train System I don't think we can easily define what 'metro system' is. And, for 10 minute RH rule, that seems to arbitrary. As already pointed by other people, a criteria based on the number of passengers the line hold, or number of passengers that the line can hold in RH assuming 100% congestion rate could be some alternatives. MinSik CHO (talk) 11:35, 13 January 2014 (UTC)

I have noticed that the biggest problem with this article is that there are way too many rules made up out of nowhere with zero reliable sources. This is clearly WP:Original research. Other than the 10min rule that we have discussed above, here are some that were claimed by others with no source: The official sources show none of the above rules. None of these claims will be a valid criteria defining a metro without reliable sources to prove them as per WP:Original research. Massyparcer (talk) 18:19, 13 January 2014 (UTC)
 * Ownership by a certain company equates to rapid transit apparently. (Where is a reliable source for this? Sources so far don't even mention anything about ownership.)
 * A rapid transit line must be connected to another rapid transit line operated by a metro company in order to be included. (Absolute nonsense that is clearly WP:Original research invented to drop a certain subway line)
 * A section of a metro line is not a rapid transit when the other section doesn't follow the made-up 10min rule (The sources so far only show that it needs to have a right-of-way in its route - not a route ahead or behind it.)
 * A grade separated metro line with a completely separate right of way isn't a metro because there is a different train service on a nearby unshared track (the definitions explicitly mention grade separation and ROW but nothing about a different service on an unshared track)
 * Examples of European or Japanese subways are used to make up a rule (Since when did Wikipedia have a rule saying anything the Europeans or Japanese do is a reliable source?)
 * We must cut lines in a subway system for no reason because otherwise, we will "get some pretty ridiculous numbers" according to Terramorhpous. (Must be a joke.)
 * I'm not going to keep rehashing this ad nauseam, but Terramorphous shows above that the 10-minute headway figure is not some "made up figure", but is based on the traffic flow number in one of the references I provided above. You keep acting like traffic volume isn't a component of heavy rail criteria, when numerous references indicate that it is. --IJBall (talk) 17:02, 14 January 2014 (UTC)
 * The 10 min rule is clearly a made up arbitrary figure with no direct reliable source to back it up. None of the sources show a 10 min rule. WP:Original research makes it explicitly clear that "you must be able to cite reliable sources that are DIRECTLY related to the topic of the article, and DIRECTLY support the material being presented." Note that directly is made bold twice in their policy. What you are claiming is extremely indirect. The source you have quoted regarding Terramorphous only shows "characteristics of heavy rail" and their maximum efficient capacity - Nothing about the claim that 10 min. headway defines a metro. Put a clear link to the reference that shows the 10 min rule. You can't because it doesn't exist and has never existed. I have never acted as if traffic volume isn't a component of heavy rail criteria - I have already recognized this point in my previous post. The point is that the 10 min number comes out of some Wiki user's brain and that's forbidden at Wikipedia as per WP:Original research. I have already said this multiple times - Even if there is a link between them, we can't use the 10 min rule because it is an arbitrary number without any official reliable source that directly backs up that number. I will not repeat this again. This needs admin intervention if you continue to ignore Wikipedia policy. Massyparcer (talk) 17:48, 14 January 2014 (UTC)


 * Actually, the 10-minute headway rule is, at least in the New York CIty Subway, the minimum headway during rush hours, where there must be 6 trains per hour or more on a given service. I heard it somewhere on http://www.nyctransitforums.com/ but don't know exactly where. In any case, let's delete it as unsourced. --Epicgenius (talk) 03:39, 15 January 2014 (UTC)
 * So are we just going to ignore the traffic volume component (from which the 10-minute headway criteria was likely originally derived)? If not, then how are we supposed to handle it? --IJBall (talk) 03:51, 15 January 2014 (UTC)
 * Agree with Epicgenius. Until a reliable, official source directly proves that the 10 min rule is valid, it will not be used in this article to define a metro, regardless of a link with the traffic volume component, as we have discussed before because we can't make up an arbitrary number that breaches WP:Original research. Massyparcer (talk) 09:37, 15 January 2014 (UTC)
 * What's the number of passengers that a New York metro train can transport when there's 10 minute headway? 10 minute rule can't be a good criteria, as LRT of 2 cars run in 5 minute headway would transport far less passengers compared to normal metro of 10 cars run in, say 15 minutes. We could use 'how many people the service can transport in given time during RH' as criteria, if we are desperate for criteria. However, just a mere 10 minute rule seems to be too arbitrary. NY Subway can say 10 minute, say Delhi can say 5 minute, so on. MinSik CHO (talk) 10:51, 15 January 2014 (UTC)
 * Agree with Cho that the 10 min rule is completely arbitrary. We can't invent some magical number that fits-it-all without a reliable source directly backing that number up as per WP:Original research. Massyparcer (talk) 12:34, 15 January 2014 (UTC)
 * It's too bad you guys don't seem to know how to read references. --16:12, 15 January 2014 (UTC)
 * And now, another reference from UITP: - this one international, and non-American, that both mentions 50,000 pphpd as a traffic volume benchmark, and explicitly mentions "frequency of service" as a metro criteria. I'm now in strong opposition to removing headway as a benchmark criteria. --IJBall (talk) 16:20, 15 January 2014 (UTC)
 * And now you're deliberately manipulating a reference to mislead others and suit your own need, which is a clear violation of WP:Original research. You're throwing in anything you can grab to avoid consensus held by other editors, which is not constructive. The UITP source says that "in order to transport 50,000 passengers per hour and direction, a metro needs a right-of-way measuring 9m in width, whereas a bus would require 35m, and cars 175m" to demonstrate that a "metro is the most efficient transport mode in terms of energy consumption and space occupancy", fully quoted from the source. The purpose of this information is to show how efficient a metro is compared to other transportation modes, NOT that it is a criteria defining a metro, let alone the complete absence of the 10min rule we're talking about here. The source doesn't say "50,000 pphpd as a traffic volume benchmark" as you claim so I don't see why you're creating claims out of nowhere, inventing words like "traffic volume benchmark" that have no existence in the source. Yes, they mention "frequency service" as a criteria and all of us agree that a metro should have high frequency of service. You're still not understanding the issue being discussed here. The problem is not that but as many editors have continuously raised concern and disagreement, about making up an arbitrary figure like the 10min rule because the source doesn't give any numbers that directly backs up your claim. I'm sorry but the 10 min rule doesn't exist in the source, which means your claim is not proven as per WP:Original research, which makes it explicitly clear that "you must be able to cite reliable sources that are DIRECTLY related to the topic of the article, and DIRECTLY support the material being presented". I'm starting to think that we need an admin to deal with IJBall's continued ignorance of Wikipedia policies. Massyparcer (talk) 17:11, 15 January 2014 (UTC)
 * First, the "10 Minute Headway" number is derived from references, not "made up". There is a massive difference. Second, I'm still not explicitly advocating for the "10 Minute Rule", but have now produced a reference that makes clear that headway is an explicit criteria for metro status and thus must be a consideration. Finally, it is incredibly unfortunate that, rather that discussing the issue rationally, and actually reading the references I have provided, you have gone running to the editors to "report me". It is clear that it is impossible to have a meaningful discussion with you Massyparcer, and I will no longer respond to any of your messages. I do look forward to continuing to work out this, and other issues, with the other constructive editors of this article however. --IJBall (talk) 18:25, 15 January 2014 (UTC)
 * Unfortunately, the 10 min rule is clearly made-up with no references proving your claim. So I don't see where you are going with this. You have produced a reference? This is a clear violation of WP:Original research. I guess you mean you are referring to one. I think from the above discussion, it is clear what the consensus is and who is being ignorant here. Refusing to talk to another editor is not constructive. Massyparcer (talk) 18:38, 15 January 2014 (UTC)
 * Just to let everyone know, I have opened a page at Administrators' noticeboard/Incidents asking an admin to deal with IJBall's non constructive and ignorant behavior. Massyparcer (talk) 18:45, 15 January 2014 (UTC)
 * Reference 4 does say: "High frequency service (maximum interval approx. 10 minutes during normal daytime service)" Valenciano (talk) 18:55, 15 January 2014 (UTC)
 * If you're talking about the reference from urbanrail.net, I have already looked into it and the author makes it clear that it is his personal opinion - This can't be used because it is original research that isn't an official, reliable source defining a metro criteria. The problem we're having here is that none of the official, reliable sources directly mention the 10 minute rule. Massyparcer (talk) 19:05, 15 January 2014 (UTC)
 * BTW, thank you reminding us, Valenciano that the "10 Minute Headway" figure isn't just implicitly available from reference (by derivation), but is in fact provided by a direct reference in the article itself. --IJBall (talk) 22:34, 15 January 2014 (UTC)
 * Which unfortunately the author makes explicitly clear from beginning is his "personal definition" and hence can't used because it is original research that doesn't cite a single reliable reference IJBall. Only official, reliable references can be used in Wikipedia as per WP:Original research. You clearly can't distinguish between a personal opinion and an official, reliable reference. I advise you read WP:SOURCES first, which makes it very clear that sources relying heavily on "personal opinion" fall under questionable sources and self-published sources (i.e. unreliable sources). Massyparcer (talk) 00:40, 16 January 2014 (UTC)

Arbitrary break
Without wading into the other issues being argued about here, I just want to register extremely strong disagreement with the use of full grade separation as a criterion for metros. The fact that the Chicago 'L', which is a metro by any reasonable criteria, has grade crossings carries much more weight in my mind than the UITP's inclusion of this as a criterion. --Tkynerd (talk) 15:45, 15 January 2014 (UTC)
 * That standard has been applied loosely in the past. --IJBall (talk) 16:12, 15 January 2014 (UTC)

I also have disagreements for different reasons: the treatment of "mixed systems" where often ad hoc decisions seem to be made on whether to include them or not, which is getting original researchy. I've never understood why metrovalencia is excluded, when its most used sections are clearly metro and defined as such by both the UTP and urbanrail.net (the two main references for the metro definition in the first place!) Such systems should be included in both lists and the route length registered accordingly with appropriate footnotes. Saying that they're only light rail systems is no more accurate than saying that they're only metro systems, especially when most visitors to the cities are unlikely to use the light rail portions. Valenciano (talk) 19:01, 15 January 2014 (UTC)
 * I agree with Valenciano that a certain group of people are inventing rules on-the-fly to include or exclude a certain metro with no reliable sources to back up those decisions. Massyparcer (talk) 19:13, 15 January 2014 (UTC)
 * I think you need to re-read WP:AGF, look at the articles rather extensive talk page archives, and understand that every criteria listed is a soft criteria, so judgement is needed. It's is not some conspiracy that other editors who have been editing the article for longer than you disagree with you, and that unproductive attitude is usually a quick route to burning out on the Wikipedia editing process. Take a deep breath, and relax. I haven't weighed in on the Seoul issues myself, because I don't know the system well enough to comment, but I do know that only on editor here is using boldface (a form of shouting) in their comments. And it's not IJ. oknazevad (talk) 01:53, 16 January 2014 (UTC)
 * I have already read WP:AGF and assume good faith with any editor when dealing with issues. I have reviewed this page's archive, which seem to only further confirm the issue of random decision making based on no sources. This has been neglected for too long and the archives reveal that a certain group of editors have deliberately manipulated and invented criteria and claim as if it is some fact taken for granted, despite the fact that they are completely unsourced original research. This has been the root of this problem - People who have been editing longer claiming that something is true because they are "experienced" and drowning out new editors' opinion even if they have a valid point to make, especially regarding decisions made out of invented criteria. I use boldface to emphasize what I'm trying to say, not as a mean of shouting. IJBall has declared to ignore me above, which is not constructive at all in reaching a consensus. Massyparcer (talk) 02:14, 16 January 2014 (UTC)
 * My advice, Valenciano? - Open up a new topic on that system, and we can discuss it in more depth. --IJBall (talk) 19:30, 15 January 2014 (UTC)
 * It's not only that system, it's that in general on this page random and arbitrary decisions often seem to be made by editors on what goes in and what doesn't go in. Valencia has fairly identical characteristics to L.A., both including metro sections, yet one goes in and one doesn't. Every system seems to get pigeon holed into either the light rail list or this list, when in reality many systems are a bit of both. Despite urbanrail being used as one of the main sources for the criteria, its conclusions often seem to be ignored. Frankfurt goes in, even though urbanrail says that "The Frankfurt "U-Bahn" is not a real metro... some sections in the city centre were built to full metro standards, whereas others along outer sections have level crossings, in the case of line U5 even some on-street running." In contrast Wupperthal, which urbanrail says "can be called a full metro line because it's totally independent, absolutely urban and runs on a 4-6 minute headway" and Essen ("U11 and U18 can be considered real metro lines as they run independent from road traffic all along their routes") are excluded. Really, all systems which have some metro elements should be included here and the metro route length listed based on the sources available, otherwise we're doing readers a disservice. Oh and lastly, what on earth is Metrodonostia doing on the list when it's just a rebranding of the Euskotren regional train network? Valenciano (talk) 20:22, 15 January 2014 (UTC)
 * I think the contrast with LA is that the two full-metro lines in LA are completely separately operated from the light rail lines (no shared trackage at all), while in Valencia the light rail lines feed directly into the metro-like portions of the system, which is a premetro or "subway-surface" light rail setup.. At least that's what I remember from the last time we discussed it; I'd have to recheck the archives. oknazevad (talk) 01:53, 16 January 2014 (UTC)
 * I still think a new topic for discussion, listing the individual systems included or excluded for unclear reasons, is a better way to go that trying to work that out in this more general criteria topic. --IJBall (talk) 20:32, 15 January 2014 (UTC)
 * I get what you're saying, but the overall issue is not so much whether individual systems should or shouldn't go in, but that decisions on what go in the list often appear to be made at random, depending on who turns up in the discussion, rather than simply going by what sources say. The end result is a bit of a POV mess. If sources say that a system contains metro sections, that should be included here and in whatever tram/light rail/commuter rail lists exist, with appropriate footnotes on which sections are metro and which aren't. Instead there seems to be an either/or approach. I'm turning in for the night now, but I'll try and have a further look over the next few days. Valenciano (talk) 20:59, 15 January 2014 (UTC)
 * "Metro (like) sections" doesn't make a line a metro if the outer portion if it is light rail. Again, premetro. And no, we can't just count the metro tunnel sections as metro if there is no segregation if traffic. That's why we removed Valencia in the first place. Either a line is metro or it's not, we can't just count half a line. oknazevad (talk) 01:53, 16 January 2014 (UTC)
 * That doesn't satisfactorily deal with situations like Essen's U11 and U12. It's also contrary to what sources say. The two main sources used for the metro criteria are urbanrail and the International Association of public transport and both of those sources describe Valencia as a metro network. Basically, they don't support your either/or approach. Valenciano (talk) 02:52, 16 January 2014 (UTC)
 * Right, otherwise things like the Market Street Subway of San Francisco's Muni Metro, or the subway portion of Philadelphia's SEPTA Subway–Surface Trolley Lines would count too, and that would soon become a mess... --IJBall (talk) 02:06, 16 January 2014 (UTC)
 * Agree with Valenciano that the decisions are made at random depending on who turns up at the discussion, rather than going by what the sources clearly say. This is the problem I have been trying to raise about this article, which goes by random rules made up by random users with absolutely no sources. Heck, they are even invented on the fly on this discussion page to deliberately include or exclude a certain subway. We should only follow what the official source says. Massyparcer (talk) 00:46, 16 January 2014 (UTC)
 * Really, you need to re-read WP:AGF. oknazevad (talk) 01:53, 16 January 2014 (UTC)
 * As I said before, I always assume good faith when discussing with any editor. WP:AGF says that "If criticism is needed, discuss editors' actions, but avoid accusing others of harmful motives without clear evidence." The problem is that we're having clear evidence of harmful motives here that are obviously violating WP:Original research given that a certain group of editors are trying to invent and make up arbitrary numbers and rules out of no reliable sources. Massyparcer (talk) 02:18, 16 January 2014 (UTC)
 * Agree with Valenciano and Massyparcer on many points. I'm feeling that we do need *some* criteria to determine what constitutes a metro system, but am in favour of the inclusive view of Valenciano. The only specific sourced criteria that I could find for a metro system is
 * 1. Is completely separated from traffic (softly applied: At least the vast majority is separated, but excludes premetros which have larger portions that are not separated)
 * 2. Is electric
 * 3. Does not have specific station to station fares (would be commuter rail)
 * Note 1: We accept that the sources do say that "high volume" is a criteria for heavy rail, since it is too vague, I suggest withdrawing it from consideration unless some source is found to say otherwise.
 * Note 2: We also accept that systems are very often mixed-modes, and the lists should reflect this, even if this means being on two lists with notes attached.
 * Any modifications/objections to these criteria? Mattximus (talk) 01:06, 16 January 2014 (UTC)
 * Only with the caveat that it's a soft criteria, so that we don't say, disqualify the Chicago 'L' for having a couple of grade crossings at the outer most part of the system, which would be foolish. As it is, that's how the existing criteria are supposed to be treated, it just doesn't always happen. And my only addition would be that only the full length of a line is considered, not just a portion, so premetros, which look exactly like a metro in their separated portion, but are actually light rail, don't get included. oknazevad (talk) 01:53, 16 January 2014 (UTC)
 * Agreed. Since none of us found hard criteria, I believe the soft criteria is the only way to go. I will modify my 4 points above to incorporate your recommendations. Mattximus (talk) 02:42, 16 January 2014 (UTC)
 * Hard criteria would exclude a few Spanish metros. Bilbao metro has level crossings, for example Urduliz (Metro Bilbao) (see this link). [[Barcelona Metro line 11] is more light rail than anything else and seems to be the blueprint for future expansions there. Metro Donostialdea probably doesn't meet existing criteria anyway, since its main function is to act as a commuter rail service linking towns in the Basque Country with one in France(!) Valenciano (talk) 03:30, 16 January 2014 (UTC)
 * Reference #5 in the discussion above makes very specific mention of "frequency of service" in defining heavy rail, and I think all references defining heavy rail discuss "high volume of traffic". Reference #3 in the discussion above sets that volume at 30,000 passengers per hour per direction (pphpd) (I came across another reference recently when looking that actually mentioned 50,000 pphpd as a figure) - using this figure, headway, or at least the number of necessary train sets, can be derived, but no reference outside of UrbanRail.net mentions an actual number. If this criteria is ignored, this list will likely become so inclusive as to be probably not useful. I oppose dropping it as criteria, but I suspect all the other editors have been scared away from this discussion, so I'll probably be the lone hold out. --IJBall (talk) 01:17, 16 January 2014 (UTC)


 * So what would you propose we add to the criteria? Just trying to get this clear. Mattximus (talk) 02:42, 16 January 2014 (UTC)
 * Metros aren't called metros for no reason. They're mostly based on internationally agreed standards, and the vast majority if not all of the existing metros already meet the 10 min rule if you are so desperate about this number. Even if the references mention high volume of traffic, we can't use an arbitrary to make it a cutting line, because there are no numbers in the sources that say so. I think the process of deriving headway from their "high volume of traffic" is original research, given that there is no source that directly says the two are the same or directly related. We have to stop inventing arbitrary numbers out of nowhere. Massyparcer (talk) 02:27, 16 January 2014 (UTC)


 * Another reference that is helpful in this discussion is this one: as it gives some examples of Light rail vs Metro in terms of passenger volume. I'm nearly positive that a number of the systems currently included in the list (e.g. Catania Metro) have passenger volumes that are a lot closer to the Light rail ppdph figures, and so aren't truly "Metro" but "Light Metro" and thus really belong in the list for Medium-capacity rail transport systems (a Wiki article with a cumbersome name that needs to be changed, but that's one of those things on my 'To Do' list that I haven't gotten to yet...).
 * My suggestion is that if we aren't going to use headway as a stand-in for passenger volume, then we're going to at least need to look at passenger volume in terms of ppdph. I don't know if the cut-off for ppdph should be 30,000 (several references I've seen use this figure for "heavy rail" or "metro" systems), or use a more forgiving figure like 20,000 ppdph. But if a system is down at less than 10,000 ppdph, then it's definitely not a "Metro" system by definition (even if it meets the other criteria), but is instead "Light rail", or at best a "Light Metro". --IJBall (talk) 02:55, 16 January 2014 (UTC)
 * (editconflict)I think that the approach should be either/and/or rather than either/or. There are systems which are not unambiguously metro, but neither are they unambiguously light rail either, so why rush to pigeonhole them in one or the other? What's really needed is a third section for such mixed systems. That doesn't need to be a new page, but could be a separate list on the light rail and trams page. Something like "light rail networks which include metro sections" (though I'm sure someone can come up with a better name.) A note on this page that such mixed systems exist and can be found there would be sufficient. That's more in keeping with the approach of refs like urbanrail. Valenciano (talk) 03:10, 16 January 2014 (UTC)
 * That's what the Medium-capacity rail transport system page (it used to be called "Light Metro", I believe...) is for - it's for systems that fall into the cracks between "full Metro" and "Light rail". That's why Medium-capacity rail transport system is focused on headway and number of trains per trainset - because those two figures, along with number of passengers per train, will give you pphpd and let's you know that those systems are probably closer to 10,000 pphpd than 30,000 pphpd. --IJBall (talk) 03:41, 16 January 2014 (UTC)
 * Oh, as for the other part, there is such a page to discuss the "mixed systems" you're referring too - it's the Premetro page. The problem in the case of that article is that an editor is challenging the list of systems there, because there are no references that identify those systems as "Premetros", and so listing them as such would be WP:Original research. I assume the list you propose could be challenged similarly. --IJBall (talk) 03:47, 16 January 2014 (UTC)

To the basics, what's the number one thing a city thinks about when they build a metro line? It would certainly be transportation of a lot of people. Considering this, I agree on the assertion that having pphpd is a valid criteria. The problem now is, pphpd criteria seems to be broad - I mean there are pphpd values that correspond to both metro and lrt, etc. A logical criteria such as 'guaranteed track' added as pphpd and 'another criteria' would do a better job. If viable, setting criteria for all metro-like urban rail system would be a solution to future problems. MinSik CHO (talk) 10:38, 16 January 2014 (UTC)

Shanghai Metro question
The following reference: lists the Shanghai Metro as having "11 metro lines", 267 stations and a system length of 270 mi (though this reference is yet another that explicitly names Shanghai's metro the longest in the world, so the Seoul entry in our table is still very, very wrong...). That is different than the 329 stations and 334 mi (and implicit 14 lines) listed here. Can somebody explain this differential? TIA. --IJBall (talk) 02:20, 25 January 2014 (UTC)
 * The news media you have shown us is seriously flawed as it doesn't even agree with Shanghai Metro's figures at all. We already have consensus on Seoul with official sources, but I will explain it again. What you have shown us is an English media report, which is a closed group of people who want to make headlines based on originally researched criteria that varies from newspaper to newspaper. The figure we have on Seoul, as the reference on this article proves, is directly from the official operator's official law and meets all the criteria from the internationally-agreed criteria and sources. As you can see, the way they count the Shanghai metro is very different from this list to begin with, so this reference proves that they're not reliable sources. My guess is that both the news media and Shanghai Metro's figures on Wiki have flaws which needs investigation. The figures from the news media seems to be substantially outdated, with Line 16, a 51.9km line and various others, opening near the end of 2013, substantially raising Shanghai's count. Shanghai Metro says that "by 30th June 2010, there has been 11 lines and 267 stations put into operation in Shanghai rail transit network, while the operation length of rail transit has reached 410km.( Excluding the Maglev Demonstration Line )" Source: http://www.shmetro.com/node155/node159/201004/con103744.htm Massyparcer (talk) 13:53, 25 January 2014 (UTC)