Proposal talk:Usability of sport pitches

Excellent vs. standard-compliant

I don't think "excellent" is a good value for standard-compliant. A pitch can be in excellent condition without being compliant with the official rules. --Push-f (talk) 14:13, 20 September 2022 (UTC)

Hi, thanks for your feedback! I've allowed myself to split your comments into sections. The idea is that "excellent" means "excellent usability", not standard-compliance. However, in many cases standard-compliance may be used as an objective indicator for excellency. --Martianfreeloader (talk) 18:37, 20 September 2022 (UTC)

I meant that if a pitch is standard-compliant, that is probably noteworthy enough to be tagged explicitly as such instead of tagging it as "excellent usability" which may or may not imply standard compliance. --Push-f (talk) 19:48, 20 September 2022 (UTC)

I imagine not all sports are sufficiently professionalized that there exist agreed-upon standards. So renaming the highest value from excellent to standard_compliant would pose problems for those sports. How about keeping usability=excellent as highest category for playability=* (this would also have the advantage that the values are kept the same as in smoothness=* and visibility=* and consider introducing an additional tag standard_compliant=*? --Martianfreeloader (talk) 11:49, 26 September 2022 (UTC)

Deprecation of playability

It should probably be noted that this new key deprecates playability=*. --Push-f (talk) 14:13, 20 September 2022 (UTC)

Yes, indeed. I've now included this in the proposal. Thanks for bringing this to my attention! --Martianfreeloader (talk) 18:37, 20 September 2022 (UTC)

Number of values

Also let me quote what I wrote in the previous mailing list discussion:

I think such a tag makes sense but having so many values seems excessive (excellent, good, intermediate, bad, very_bad, horrible, very_horrible, unplayable), I'd limit it to just (excellent, good, intermediate, bad).--Push-f (talk) 14:13, 20 September 2022 (UTC)

Thanks, I've reduced the number of values. However, not as much as you suggested. I think in some sports, it will be possible to define high-resolution objective criteria for quality, so keeping a number of possible values seems useful to me. --Martianfreeloader (talk) 18:37, 20 September 2022 (UTC)

Unplayable pitches

And I'd argue that unplayable pitches shouldn't be mapped as leisure=pitch anyway but rather abandoned:leisure=pitch. --Push-f (talk) 14:13, 20 September 2022 (UTC)

I must admit I haven't done research on what's the community standard for cases like this (overgrown basketball yards, broken table tennis pitches or unusable railway tracks). I'll try to address this in the coming days. --Martianfreeloader (talk) 18:37, 20 September 2022 (UTC)

I want to note that I also haven't done such research. My reasoning is that leisure=pitch in combination with sport=* implies that the sport can be played on this pitch. That is if you want to find pitches where you can play a specific sport, you only need to filter based on these two tags (and don't have to filter out pitches tagged with pitch:usability=unusable). I think tagging a pitch with pitch:usability=unusable would be like tagging amenity=drinking_water with drinking_water=no ... it contradicts the very definition of the feature. --Push-f (talk) 19:54, 20 September 2022 (UTC)

Ok, I'm convinced. Tagging unplayable pitches as pitch should probably discouraged; if needed, they should probably be tagged with the abandoned:*=* prefix. --Martianfreeloader (talk) 11:44, 26 September 2022 (UTC)

Tag stability

And I'd like to add that if a pitch is in suboptimal condition e.g. pitch:usability=bad, I think we should encourage mappers to also add some kind of description what exactly is bad about the pitch, to help with verifiability. E.g. if a chess piece is broken it might get replaced in the future, at which point other mappers might wonder why the usability isn't tagged as good. --Push-f (talk) 14:13, 20 September 2022 (UTC)

I guess this is not an issue that is unique to this tag. How do you think about encouraging an optional usability:check_date=*? --Martianfreeloader (talk) 18:37, 20 September 2022 (UTC)

Other tags have clear definitions for each tag value. For this tag the definition of the values depends on the sport=* tag, which complicates the situation, especially because these sport-specific definitions might change over time. I don't see how a :check_date would help clarify why a particular value was assigned. --Push-f (talk) 20:01, 20 September 2022 (UTC)

The idea was that this proposal introduces a generic tag. For any kind of sport for which this tag should be used, it would need to be defined specific to this sport, with as objective measures as possible, as I've already tried with table tennis (playability would be renamed to usability). --Martianfreeloader (talk) 11:40, 26 September 2022 (UTC)

Does equipment influence usability?

And yet another concern: Should the availability of equipment influence the usability? E.g. may pitch:net=* influence pitch:usability=*? Or what about chessboards where you have to bring your own chess pieces? Are these less usable than chessboards where pieces are available? --Push-f (talk) 14:13, 20 September 2022 (UTC)

I think this question should be addressed on the sport-specific implementations of this generic tag. For sport=table_tennis for example, I've already (informally, as of now) introduced a number of tags describing the available equipment. In this case, the available equipment should not have influence on usability=* (except where it is explicitly mentioned in its value definitions). --Martianfreeloader (talk) 18:37, 20 September 2022 (UTC)

More fine-grained tagging

My main problem with the proposed tag is that many aspects can influence the tag values and it's unclear which specific aspects influenced any specific tag value. E.g. seeing pitch:usability=bad it's entirely unclear why the usability is bad. Is the surface bad? Are the goals bad? Are both bad?

I think it might make more sense to introduce something like:

• pitch:surface:quality=excellent/good/intermediate/bad/horrible
• pitch:lines:quality=excellent/good/intermediate/bad/horrible
• pitch:net:quality=excellent/good/intermediate/bad/horrible for the sports that require a net
• pitch:goals:quality=excellent/good/intermediate/bad/horrible for sport=soccer
• pitch:hoops:quality=excellent/good/intermediate/bad/horrible for sport=basketball

And potentially pitch:standard_compliant=yes to denote that the whole pitch meets the criteria of the respective sports governing body (implying that all the qualities are excellent).

--Push-f (talk) 20:35, 20 September 2022 (UTC)

pitch:*=* is not needed except for pitch:net=*. You could reuse smoothness=*, no need for surface:*=*. I don't understand what's meant by other *=quality. --- Kovposch (talk) 07:40, 21 September 2022 (UTC)

No, smoothness=* is defined to specify "the physical usability of a way for wheeled vehicles", so it very much does not apply to pitches. The very purpose of the pitch:*=* namespace is to group pitch-related tags, so using it here makes sense. --Push-f (talk) 08:07, 21 September 2022 (UTC)

1. First of all, you can use pitch:smoothness=*. Secondly, their descriptions are comparable at or above smoothness=very_bad (covered with potholes). We shouldn't need to have another set of already vague adjectives for similar purposes.
2. surface=*, lines:*=*, goalposts=*, and hoops=* are not prefixed. You would be changing them together.

Kovposch (talk) 08:51, 21 September 2022 (UTC)

Smoothness isn't the right term for the surface quality of sport pitches. E.g. a grass pitch isn't exactly "smooth" and for beach volleyball it's about the quality of the sand and certainly not about the smoothness of the surface (which is very much temporary and degrades during play). lines=* is only used 91 times in combination with leisure=*, goalposts=* only 63 times. hoops=* is used >30 thousand times but probably only because it has been added as a preset in iD. --Push-f (talk) 09:45, 21 September 2022 (UTC)

standard_compliant=* is also unclear. Is it the local or international standard? --- Kovposch (talk) 07:40, 21 September 2022 (UTC)

I was thinking of international standards, I didn't consider local standards. I agree that different tags like fifa_compliant=* or fivb_compliant=* would be better. --Push-f (talk)

I'd prefer sticking with one key standard_compliant=* for all sports and define it for each sport, rather than introducing a new key for each sport. --Martianfreeloader (talk) 11:58, 26 September 2022 (UTC)

Using more fine-grained (and potentially more objective) measures is an interesting alternative approach instead of using a catch-all tag. Pro: much more fine-grained, thus more information. Con: Harder to digest for users who want to get a quick overview. It would be left to data consumers to create an aggregate usability rating from these detailed keys. I can't really say which approach I prefer, they both seem appealing. --Martianfreeloader (talk) 11:58, 26 September 2022 (UTC)

For table tennis I have already informally introduced such fine-grained characteristics, like bounce=* or pitch:net:overhang=*; but I still felt these objectively measurable criteria can't satisfactorily describe all cases found in reality, e. g. the quality of the table edges or the table surface. It seemed to me the overall usability (/playability/enjoyability) is hard to be assessed purely by measuring, but easy for a human player to judge (very much like visibility=*); and in most cases most humans will come to a very similar judgement of how suitable/good/enjoyable they find a particular table tennis pitch. This is the reason for this proposal. --Martianfreeloader (talk) 12:07, 26 September 2022 (UTC)

Mixing too many criteria with adjectives

Following #More fine-grained tagging, you are using all "subjective" adjectives, instead of an exact physical description. This doesn't really solve any problem. It is not much better than playability=*.
For your examples:

1. There are many requirements for international competition compliance. Is it size, surfacing, or something? There's not even tagging for relative size aside from the attempt of leisure=practice_pitch (not good) for half courts. (forgot where this was discussed within the last few years). Below, there are local and regional standards. Geospatially, a more fundamental problem is whether leisure=pitch (or a specific use of *=recreation_ground) should be drawn on the playing area only, or the entire playing field including the border area. This is undefined, and has no consistent use.
2. Should start with standardizing goalpost=* and goal=* (as it is possible to have goal without goalpost. Then you can add the goal:*=* as uneven.
3. Should start with showing whether a chess board provides chess pieces, or you have to bring your own. Then you can add some are broken or missing.
4. Could try to define the slope of an area. Although table tennis table has an axis, it has no direction, so incline=* and incline:across=* are not applicable. But, tilt=* can be used for the transverse gradient. No idea for longitudinal.
5. Simply add wetland=swamp or something to show that.

Kovposch (talk) 08:04, 21 September 2022 (UTC)

Just two thoughts:

1. Whether the entire playing field or just the playing area is tagged as leisure=pitch is a separate problem that does not hinder the introduction of new tags to denote the compliance with specific standards.
2. wetland=swamp requires natural=wetland. I think adding both natural=wetland and leisure=pitch to the same element would be a violation of One feature, one OSM element.

--Push-f (talk) 08:18, 21 September 2022 (UTC)

1. It's an issue in application if you want to verify and compare their dimension and area. I'd rather not touch this yet. Furthermore, size or spacing of technical area is regulated. It is outside the playing area.
5. I don't know much about swamps. Putting aside the question of being inside a swamp, should it actually be surface=mud? Kovposch (talk) 09:03, 21 September 2022 (UTC)