Wikidata talk:WikiProject Sport results

(Redirected from Wikidata talk:Sport results task force)
Latest comment: 3 years ago by Mad melone in topic Match result

High level roadmap edit

Hi, I would like to have the basic data structures in place before w:Sochi 2014 (February; Q9678), and have a 'live' feed of the Olympic data into Wikidata. I expect that we wont be syncronising the entire dataset by then, but it will be a proof-of-concept. We can then work with other major event organisers to produce similar outcomes, and culminate in a live stream of all useful elements for w:Rio 2016 (August; Q8613). John Vandenberg (talk) 08:03, 28 November 2013 (UTC)Reply

Medals edit

The most 'important' part of sport results is the winner. Wikidata has event participants and it does have award received/awarded by, but there doesnt appear to have been much work done on linking them together to describe winners / medals.

In order to present some ideas, I have created/fixed some items, but not too many in case some of them need to be deleted.

Sports
Events
Teams
Medals
Participant

I have used 'participant' at swimming at the 2012 Summer Paralympics – men's 100 metre backstroke S10 (Q7658206), and abused the 'qualifier' functionality to record the medallists.

IMO the better solution is to create items for each 'position' such as 2012 Summer Paralympic gold medal for men's 100 metre backstroke S10 (Q15243665), at least for medals, and add them to the person, such as Justin Zook (Q6318298). However that means creating a lot of items as it is one item for every medal ever awarded. If we wanted to record positions other than medals, we could create 'non-medalling position' as a subclass of award so that those awards are able to be filtered out in queries. In order to wikidataify the existing Wikipedia tables of results, we would need to also create items for person who participated in each heat. John Vandenberg (talk) 14:06, 28 November 2013 (UTC)Reply

Let us see. Australia at the 2012 Summer Paralympics (Q1896200) should contain some instances. Should it be items like Australian Paralympic Swim Team (Q4824537) or like Justin Zook (Q6318298) (it can contain both, but it would be a huge mess). My preference is certainly to have teams, otherwise an item like United States at the 2012 Summer Olympics (Q139623) would become unreasonably big.--Ymblanter (talk) 15:08, 28 November 2013 (UTC)Reply
Regarding Australia at the 2012 Summer Paralympics (Q1896200), it is listed as a participant of 2012 Summer Paralympics (Q182798), following the example set by Kompakt at 2012 Summer Olympics (Q8577).[1], and Australian Paralympic Swim Team (Q4824537) is listed as a participant at swimming at the 2012 Summer Paralympics (Q126403).
Australia at the 2012 Summer Paralympics (Q1896200) is an instance of Australia at the Summer Paralympics (Q4823688), which is an instance of national sports team (Q1194951). Australian Paralympic Swim Team (Q4824537) is also an instance of national sports team (Q1194951).
For multi-sport events, I also much prefer having teams, even for individual sports. However in some instances that is creating a fictitious layer. e.g. for UK we have Great Britain (Q749109) and Team GB (Q7691492), and w:Team GB#Branding strategy says that there "is only one Olympic team from Great Britain and Northern Ireland; Team GB. There is not an Olympic swimming team or Olympic rowing team. The individual sports join to become Team GB, the Great Britain and Northern Ireland Olympic Team." But even that is dubious, as http://www.britishrowing.org/gb-rowing-team says "The GB Rowing Team is the high-performance arm of British Rowing."
For team sports, Wikipedias usually have team pages for team sports, such as Great Britain olympic football team (Q877411), Great Britain men's national field hockey team (Q634168), Great Britain men's national volleyball team (Q608612), Great Britain men's national ice hockey team (Q918179) and Great Britain national handball team (Q918298), but do not have team pages for sports like rowing, athletics, swimming, etc. English Wikipedia has articles about the Australian Paralympic Teams for individual sports are further developed than most areas, because disability sport has been a focus on the Wikimedia Australia chapter. Italy is also well advanced in this area, with teams for most individual sports on w:Template:National sports teams of Italy (and has articles in Italian language as well). There are a few other nations in w:Category:National athletics teams, including the United States. John Vandenberg (talk) 00:40, 29 November 2013 (UTC)Reply
There are also some 'teams' in newly created w:Category:Nicknamed groups of Olympic competitors. John Vandenberg (talk) 02:31, 29 November 2013 (UTC)Reply

Another approach to medals is show at Kevin Coombs (Q6396063), using an awarded medal with two qualifiers of sport and event. This approach doesnt require a complex structure and lots of items for awards. John Vandenberg (talk) 16:37, 29 November 2013 (UTC)Reply

This seems easier to me.--Ymblanter (talk) 21:16, 29 November 2013 (UTC)Reply

Swimming items edit

I've found that items have come from Polish Wikipedia for most swimming events, but they dont have English labels as there dont appear to be equivalent pages in other Wikipedia. See w:pl:Szablon:Pływanie. I have added English labels for the 50 m events. We need to label the rest otherwise people will create duplicates. John Vandenberg (talk) 04:13, 4 December 2013 (UTC)Reply

Results edit

Any idea on how results (minutes and seconds, or meters or points in other sports) could be added?--Ymblanter (talk) 11:00, 4 December 2013 (UTC)Reply

The most important set of results in world records, or medalling times/distances etc, which can be an award with a qualifier of 'duration (completion time)' (in days, hours, mins, seconds) / 'distance' (kms, etc) / 'points scored'. time and distance will need to wait for dimensioned integers. Points scored would be a dimensionless integers; it should also be able to be used as a property of the person (which would be the same as Total leagues goals which has been approved), or for a teams score in a match. We'll need separate properties for many sport, so that constraints can be applied to ensure the values are legal (e.g. tennis game scores) John Vandenberg (talk) 14:48, 4 December 2013 (UTC)Reply

Visualising a multi-sport event edit

Magnus Manskes 'Wikidata tempo-spatial display' tool provides a nice way to visualise a single multi-sport/multi-venue event. See http://tools.wmflabs.org/wikidata-todo/tempo_spatial_display.html?q=Q182798 2012 Summer Paralympics (Q182798) (it is quite slow for me, but that could be local factors). It would be nice to also visualise Olympic Games (Q5389) over its lifetime, however the tool appears to only use part of (P361), whereas Winter Olympic Games (Q82414) and Summer Olympic Games (Q159821) are subclasses, and 1912 Summer Olympics (Q8118) and 2012 Summer Olympics (Q8577) are an instance. One solution is to introduce recurring event (Q15275719), ala SMW: Help:Recurring_events, and improve the visualiser to understand that visualising a recurring event requires looking at the instances of it. John Vandenberg (talk) 15:34, 4 December 2013 (UTC)Reply

Cycling edit

Bonjour à tous, je travaille principalement à l'illustration du cyclisme en Nord-Pas-de-Calais et en Belgique. Je comprends l'anglais, mais je préfère écrire en français pour qu'une traduction automatique soit plus précise. Je souhaiterai que le traitement du cyclisme fasse un bond en avant. Je laisse donc le message que j'avais publié sur Wikidata:Bistro :

Je réfléchissais à la question, et je me disais que finalement, c'est peut-être pas extrêmement compliqué, dans la mesure où les éléments des coureurs comportent déjà la nationalité (pour les petits drapeaux) et peuvent comporter l'équipe 2015, l'équipe 2014...

  • Par exemple, premier point, pour les Quatre jours de Dunkerque qui existent en dix-huit versions linguistiques, j'aimerais stocker sur Wikidata pour chaque édition l'année (avec un lien vers l'article détaillé, qu'il existe déjà ou qu'il n'existe pas encore), le premier, le deuxième et le troisième, à chaque fois de sorte à ce qu'il y ait le petit drapeau de la nationalité. Le but, à terme, lorsque toutes les autres Wikipédia auront évolué, c'est que l'action d'un seul Wikipédia sur Wikidata permette d'ajouter sur toutes les versions linguistiques le podium. D'une part, ça permettra à toutes les Wikipédia d'économiser du temps, puisqu'à l'heure actuelle, chaque locuteur doit se charger de la mise à jour, mais ça permettra également à chaque locuteur désirant créer un article dans sa langue de ne pas avoir à se soucier de ces données, qui seraient gérées en un seul point.
  • Deuxième point, Handzame Classic 2015 et son classement final, complet sur Procyclingstats. Actuellement, chaque version linguistique doit faire le même travail, alors que des bases de données disposent d'un classement intégral quelques dizaines de minutes après l'arrivée de la course. La nationalité comme l'équipe 2015 peuvent être stockées dans l'élément du coureur (attention toutefois aux coureurs qui sont pris comme stagiaires dans d'autres équipes entre août et décembre, mais là, c'est une question de dates). Il reste donc à associer à une place un coureur et son temps. Les gains sont les mêmes qu'au premier point, les économies de temps permettront aux Wikimédiens de passer plus de temps à la rédaction.
  • Troisième point, Circuit Het Nieuwsblad 2015 et sa liste des partants, visible également sur Procyclingstats. Actuellement, tout doit se remplir à la main. Il y a eu du temps gagné grâce à la création du modèle LP qui stocke la mise en forme, mais il reste toujours à rédiger cette liste, ce qui prends du temps, et qui en constitue une perte si toutes les versions linguistiques le font. Comme pour le point précédent, je cherche un moyen qui permettrait à partir des données de ce site de générer ces données. À chaque équipe correspondrait des numéros de dossards attribués à des coureurs (dont la nationalité est stockée dans leur élément Wikidata). Optionnellement, un Wikipédien pourrait manuellement ajouter la place ou notifier l'abandon, quoique ces données serait déjà connues par le classement précité.
  • Quatrième point, Circuit Het Nieuwsblad 2015 et ses équipes engagées. L'idée serait de les différencier selon qu'elles soient en 2015 de première, deuxième ou troisième division, voire non-UCI. Leur nationalité ainsi que leur code en trois lettres serait indiqué dans l'élément de leur saison 2015 (comme cet élément).
Voilà, c'est un projet que j'aimerais mettre en œuvre dans les années à venir non pas spécialement pour moi étant donné que je suis sur le terrain à l'illustration, mais pour les contributeurs/rédacteurs. Les reportages-photo sont éprouvés, et je prépare quelques (r)évolutions pour les années à venir, mais sur la gestion des données nous avons un temps de retard par rapport à d'autres sites, alors que nous pourrions sans trop de difficultés évoluer pour gagner beaucoup de temps, ce qui permettrait au cyclisme de la Wikipédia francophone d'être beaucoup plus rédigé, et permettrait aux petites versions linguistiques de Wikipédia d'avoir des articles détaillés sur des courses cyclistes sans trop d'efforts. J'ai surement quelques années d'avance, mais je prépare un très gros projet d'illustration du cyclisme pour dans deux ou trois ans, ce qui implique un très gros besoin de main-d'œuvre en ce qui concerne les identifications/catégorisations des coureurs (sur ce sujet d'ailleurs, tout est prêt au niveau de Wikimedia Commons pour supporter un très grand nombre de langages). Jérémy-Günther-Heinz Jähnick (talk) 10:58, 26 March 2015 (UTC)Reply

What is the policy about sport results ? edit

Sorry but no clear info/decision is provided about how to store results data. Do we create statements about results in the event item or we add statements in the participant items ?

For me items like Australia at the 2012 Summer Paralympics (Q1896200) are stupid: they can't be used to store data in a easy way. If we provide the complete ranking of the competition under the event item we can delete Australia at the 2012 Summer Paralympics (Q1896200) and recover all data for Australia by applying a nationality filter on the data of the event item.

Example:

  • item: 2012 Summer Paralympics (Q182798)
    • discipline: Paralympic Archery with qualifier for male or female ranking
      • Rank 1: person XXX with qualifier giving its performance result
      • Rank 2: person YYY with qualifier giving its performance result
      • Rank 3: person ZZZ with qualifier giving its performance result
      • ...
    • discipline: Paralympic Judo with qualifier for male or female ranking
      • Rank 1: person XXX with qualifier giving its performance result
      • Rank 2: person YYY with qualifier giving its performance result
      • Rank 3: person ZZZ with qualifier giving its performance result
      • ...

If this represents too many statements for one item, we can divide the results by sport:

with

  • item: archery at the 2012 Summer Paralympics (Q2371996)
    • Competition: male/female
      • Rank 1: person XXX with qualifier giving its performance result
      • Rank 2: person YYY with qualifier giving its performance result
      • Rank 3: person ZZZ with qualifier giving its performance result
  • item: judo at the 2012 Summer Paralympics (Q2741261)
    • Competition: male/female
      • Rank 1: person XXX with qualifier giving its performance result
      • Rank 2: person YYY with qualifier giving its performance result
      • Rank 3: person ZZZ with qualifier giving its performance result


The other possibility is to store everything in the items of the sportmen:

  • item: person XXX
    • Participation: {Q|182798}} with qualifiers giving the discipline, the rank and the performance result.

These are definitively better ways to store data and to organise the extraction process for the future. --Snipre (talk) 15:21, 10 May 2015 (UTC)Reply

If we consider now the problem of teams, the second option is better because we can add the team as qualifier. --Snipre (talk) 15:36, 10 May 2015 (UTC)Reply

I do not quite get it. Q1896200 contains a bunch of interwiki links, what do you propose to do with them?--Ymblanter (talk) 16:43, 10 May 2015 (UTC)Reply
It's not clear to me whether results such as olympic medals are currently available from Wikidata: for instance Q1189 doesn't have anything. Nemo 19:12, 14 January 2018 (UTC)Reply
This is not systematically available yet. We are even quite far away from having it here, and to my knowledge there is not even a clear model how to do that properly. Sports results in Wikidata are in fact a difficult field, for many reasons (non-free databases as sources, requires complicated data models when done properly, vast amount of data that steadily needs updates, real-world differences in all the types of sport, etc.) —MisterSynergy (talk) 20:33, 14 January 2018 (UTC)Reply

For the Olympics we should add a level to not have too many datas in one item:

For some items we already have the results inside, for example: 2017 Tour de France (Q21821246) --Gabbg82 (talk) 15:16, 15 January 2018 (UTC)Reply

Yes, pretty much like this (Olympics edition -> discipline -> event). See Wikidata talk:WikiProject Olympics for some related discussions, and mind that not even on this level items are properly connected. Directly within the event items, one would probably only have the final ranking of all participants. There are a couple of properties and qualifiers which can be used for that, see Template:Sports properties. We would have to start a discussion to find the best approach tough regarding details of modelling. —MisterSynergy (talk) 15:30, 15 January 2018 (UTC)Reply

Food for thought(?) edit

Hi all, I noticed this project exists and I'd like to share with you a sort of proposal I came out with about taking note of results. You can find a brief version here or a long version here (unfortunately in Italian, but I can translate easily for you).

It is quite a radical proposal, I must say, since it starts from the idea that every match or race or $equivalent should have an item on its own, then the items should be regrouped with other items that represent the single parts of the tournament. In the Italian sandbox, you might find a couple of examples ("Esempio #1" and "Esempio #2").

I'm actually looking for people who can take a look at this and tell me what they really think about it - the blunter the answer, the better. :) I'm here for any clarification or explanation. --Sannita - not just another it.wiki sysop 11:46, 27 May 2015 (UTC)Reply

+1 I like your approach which is synthetic and not based on example for a start of discussion. I will try to look at your proposition until Friday. Snipre (talk) 13:52, 27 May 2015 (UTC)Reply
Looks very reasonable. I independently thought on the same topic and came to exactly the same structure. Artem Korzhimanov (talk) 18:17, 8 June 2015 (UTC)Reply
+1 It will be a long and hard work, but I like it. When we start?? --Gabbg82 (talk) 15:10, 14 August 2015 (UTC)Reply
@Snipre, Artem Korzhimanov, Gabbg82: First of all, thanks for your approval. :) I completely translated the Italian version and reworked the model a bit. Feel free to take a look at the (relatively) new proposal, but please do note this is still plenty to do. Also, please, take note of the "known problems" section. Sannita - not just another it.wiki sysop 16:54, 12 September 2015 (UTC)Reply

Is there any update on this discussion available?—Mad melone (talk) 19:23, 15 July 2018 (UTC)Reply

Figure skating properties edit

Three linked proposals for new properties to handle figure skating data:

--Harmonia Amanda (talk) 19:07, 30 January 2018 (UTC)Reply

Tennis: Tournament wins and runner-up placements of a player edit

For tennis players, most language versions of Wikipedia have a list of tournament wins and some even of all lost finals, i.e. runner-up placements. For reference purposes, please see examples for Serena Williams:

Even though different in styles and overall presentation, there is a common set of information displayed in all versions - and this just calls for a use of Wikidata!

Generally, the following information are displayed:

  • Result (Winner, Runner-up)
  • Date
  • Tournament
  • Tournament category (Grand Slam, WTA Tour Finals, WTA Premier Mandatory (Tier 1), WTA Premier (Tier 2), ITF 100k, ITF 50k,... but also Olmypic Games)
  • Surface (Hardcourt, Grasscourt, Claycourt,...)
  • Opponent
  • Score

Building around participant in (P1344), we can probably design a lot of these information, but especially "Opponent" and "Score" could be difficult.

However, this seems to be such a clear application (and actually time saver for the local wikipedias) for Wikidata, that I would like to get your ideas on how this could be strcutured. Thanks, --Mad melone (talk) 07:11, 22 July 2018 (UTC)Reply

FYI: https://www.wikidata.org/wiki/Wikidata:Property_proposal/runner-up--Mad melone (talk) 08:53, 26 July 2018 (UTC)Reply
Earlier this year a couple of editors spent quite some effort into the organization of tennis tournaments and related entities, see Wikidata talk:WikiProject Tennis, starting with my topic from 7 December 2017. There is indeed an enormous amount of data in Wikipedias (mostly enwiki, itwiki, eswiki). The basic organization of items was brought to a fairly good shape, but there is still a lot to do. Adding winner information would be one of the next things to do, and inversely adding participation data to player items as well. Your idea to use participant in (P1344) and qualifiers sounds good. —MisterSynergy (talk) 09:02, 26 July 2018 (UTC)Reply

Match result edit

  Notified participants of WikiProject Sport resultsIs there a Property for match result that I have not found so far? Should be usable in a qualifer. Thanks, --Mad melone (talk) 08:54, 26 July 2018 (UTC)Reply

I am not aware of any. There is Template:Sports properties and some linked subtemplates, which are all pretty complete in the field of sports. —MisterSynergy (talk) 08:57, 26 July 2018 (UTC)Reply
Any update on this? I know there is number of points/goals/set scored (P1351) but that doesn't really work for sports with multiple sets, e.g. tennis, squash, table tennis,.... I think we are at a place now where we can think about integrating sports results into Wikidata and such an item (even if just plain text) would be very helpful. Thoughts? --Mad melone (talk) 06:21, 7 October 2020 (UTC)Reply
Yes, it would be helpful, and no there is no update on this. Can you make a proposal (here in this section) how you think tennis results (and similar types of sport) will be best modeled? I would support creating a property proposal once we have something workable… —MisterSynergy (talk) 11:57, 7 October 2020 (UTC)Reply
I would keep it simply to cover as many result forms as possible, something like:
   Under discussion
Descriptionsports result
Data typeItem
Domaintennis tournament (Q13219666), tennis player (Q10833314)
Allowed values0-9 , "-", ";", "[", "]", "(", ")"
Example 16-2, 4-6, 7-6(2) for a tiebreak win in tennis
Example 26-0, 1-6, [10-5] for a match tiebreak in tennis
Example 311-8, 11-7, 11-4 for table tennis
Planned usefor results in i) tournament draws, ii) list of tournament wins by a player iii) list of tournament winners in the history of a tournament
Expected completenessalways incomplete (Q21873886)
Robot and gadget jobsa bot could fill the results either from the tournament organizer's website (preferred) or exisiting Wikipedia templates
See alsonumber of points/goals/set scored (P1351)

--Mad melone (talk) 12:10, 7 October 2020 (UTC)Reply

I understand. The regex should probably be improved so that it is easier to track malformed values. Should we worry about the order of the results, i.e. does "6-2" for player 1 mean "6-2" for player 2 as well, or would player 2 have to have "2-6" then? At which place do we put the order here, in case it matters? —MisterSynergy (talk) 12:15, 7 October 2020 (UTC)Reply
For me, the result is not in context of a player, but in context of a match, i.e. the result should be from the view of player 1 (or home team) - as for regex, I am not an expert, so please feel free to change the above. I just want to make a point that "6-2" should also be the format and not "6:2" or something like that --Mad melone (talk) 12:21, 7 October 2020 (UTC)Reply
Okay but how does it work if you use in in a player item, as proposed above? Can you try to make some detailed examples for the different usage scenarios here, using the template {{Claim}}? —MisterSynergy (talk) 12:23, 7 October 2020 (UTC)Reply
Sorry, i don't think I can, but let's look at Lisa Ponomar (Q24345171) where I have included some statements about her tournament wins. If we could add a "sports result" to each of those, then we could get the same tables as under en:Lisa Ponomar#ITF finals. At the same time, we could use more or less the same statement at one of the tournaments she won in order to get a table like under en:ITF Athens Open to show a history of tournament winners. Lastly, we could end up with getting complete draws like en:2019 French Open – Men's Singles which also require the result of the match (rest would be quite easy).--Mad melone (talk) 12:38, 7 October 2020 (UTC)Reply
@MisterSynergy: - Thoughts?--Mad melone (talk) 06:37, 12 October 2020 (UTC)Reply
Yes, but it's complicated.
The problem here is, at least for the usage that you propose for player items such as Lisa Ponomar (Q24345171) and tournament items such as ITF Athens Open (Q4052770), that you attach a single match result to something else than a tennis match and the context is thus somehow missing. It is clear that *usually* this result is meant to be the latest match in which the player participated (for players) or the final of the tournament (for tournaments), but this is not encoded anywhere and I find it difficult to get this done properly without individual items for each match. The proposal clearly works for items listed here, but beyond that I am sceptical. —MisterSynergy (talk) 07:23, 12 October 2020 (UTC)Reply
What about including stage reached (P2443) -> final (Q1366722)?--Mad melone (talk) 08:09, 12 October 2020 (UTC)Reply
Yes, that would be the miminum requirement. Can we please set up some use cases here? It is much better to have a good plan in advance, rather than to have a property at some point that does not do its job as desired. —MisterSynergy (talk) 08:11, 12 October 2020 (UTC)Reply

:Tournament win of a player

And then you would potentially include the following qualifiers (don't know how to include them in the above statement as they don't have a P-ID yet)

Players that won a tournament

And then again with the following additional qualifiers:

Let's don't be perfect, but achieve something manageable here!--Mad melone (talk) 15:30, 16 October 2020 (UTC)Reply

Yes, a practical approach is necessary here. However, I would recommend some modifications:
---MisterSynergy (talk) 22:31, 18 October 2020 (UTC)Reply
Your solution may be more closely to a data base representation, however, this is generally not how the information is used and presented in wikipedia and I wouldn't know how to pull that information. Therefore, please go ahead as proposed, but I will take a step back as I feel I am not the right person for this. --Mad melone (talk) 11:54, 23 October 2020 (UTC)Reply

Track cycling and keirin cycling edit

Hello cycling people.

Are there any good sports databases online that could be used as a property for these kind of athletes in track cycling and or keirin cycling out there ? Best regards Migrant (talk) 22:36, 23 April 2019 (UTC)Reply

Return to the project page "WikiProject Sport results".