User talk:Amadalvarez/sports statistics property

I believe it is more correct to have separate properties. With this approach, every idea for accounting statistics will inevitably imply an approval procedure, no one will be able to push their original ideas such as "number of goals scored by the heel in a career", "number of slam dunks in a career", "number of bloopers in a career", "number of injuries in a career". Maintenance of specialized properties is more reliable in terms of detecting vandalism, incorrectly filled data, property constraint (P2302) or some another misuse. Also your proposal inevitably implies the use of one additional qualifier, specifying object to be measured, which increases the load on the page.

But anyway, let's discuss some details.

When working with current properties, the following approach is the most correct:

⟨ Lionel Messi (Q615)    ⟩ total goals in career (P6509) ⟨ 6 ⟩
of (P642) ⟨ FIFA World Cup (Q19317)    ⟩
date of the first one (P7124) ⟨ 16/06/2006 ⟩
date of the latest one (P7125) ⟨ 26/06/2018 ⟩
point in time (P585) ⟨ 10/03/2021 ⟩

That is:

1. Instead of applies to part (P518), we use of (P642) for the scope of the measurement. It can be sports league (Q623109) (National Basketball Association (Q155223), not abstract and ambiguous career (Q282049) for Michael Jordan (Q41421)), tournament (Q500834) or sports season (Q27020041).
2. date of the first one (P7124) and date of the latest one (P7125) are necessary to identify the oldest and youngest players or goal scorers in the history of the tournament (query).
3. point in time (P585) is necessary to determine the relevance of data.
4. We should remove the word "career" from the labels of current properties so that they can be used for both career statistics and seasonal (FIFA World Cup (Q19317) and 2018 FIFA World Cup (Q170645), National Basketball Association (Q155223) and 1998–99 NBA season (Q1321776), etc.).
5. I highly doubt the idea of giving Michael Jordan (Q41421) the stats in each of his matches like Game 4 of 1988 NBA Playoffs Eastern Conference First Round, Chicago Bulls at Cleveland Cavaliers (Q56670521) with separate statements is promising. In the NBA Playoffs (Q2265397) alone, Michael Jordan (Q41421) played 179 games, his item with full statistics on all these games will simply run into adequate page load limits and now remember that he still has 1072 games in the regular season (Q10509145)…
6. We should drop any statistics filled with qualifiers to member of sports team (P54) as not contain point in time (P585) and listing of the counted tournaments (domestic league? domestic cup? super cup? friendlies?). Typical example of such useless data is Q316512#P54.

--Сидик из ПТУ (talk) 09:54, 10 March 2021 (UTC)

Thanks, @Сидик из ПТУ: for your fast answer. Personally, I disagree with your initial consideration, but our two opinions are perfectly reasonable. In fact, in the background I explain my point of view on the hyper-population of properties. I will add your consideration to the presentation of the property.

1. P642. No problem. I have seen it in other similar situations. I have worked quite a bit with P39, and there it is used to complement the noun (Ex. President + P642 + Institution / region, ...) rather than to point to a subset. But I adapt to what is agreed.

2. & 3. It should work similarly to how you have it now. As I indicate in "Some considerations", other qualifiers may be include, as they would be the "characteristics" of the indicated P1114.

4. When I use “career” it means the overall total of the person in the “scope” dimension. That is, the total number of goals, fouls, rebounds, etc. achieved in his career or so far, if he is active. The player's total at National Basketball Association (Q155223) or 1998–99 NBA season (Q1321776) will have this Qid as a scope (P642). See examples:

Michael Jordan (Q41421)quantity of (new)basketball game (Q18431960)applies to part (P518)career (Q282049)quantity (P1114) 1072

Michael Jordan (Q41421)quantity of (new)basketball game (Q18431960)applies to part (P518)National Basketball Association (Q155223)quantity (P1114) 1072

Michael Jordan (Q41421)quantity of (new)basketball game (Q18431960)applies to part (P518)1996–97 NBA season (Q1321749)quantity (P1114) 82

In Jordan case, career & NBA are the same, because he don't change league. However in Dani Pedrosa (Q313959) case, you can see the difference between category.

5. I also don’t think we have that level of indicators in WD. I have given the example of a match to explain the versatility of the "scope" dimension. In any case, as with many properties that can be in an item or vice versa, I think that if one day someone wants to store player-match statistics, the information should go to the least downloaded item, i.e. , in the match.

Thanks, Amadalvarez (talk) 11:27, 10 March 2021 (UTC)

4. In Jordan case, career & NBA not are the same. He also has baseball career, basketball college career, national team career, All Star Game career. Dani Pedrosa (Q313959) also has Spanish Minibike Championship career. Stating "career" without listing of the counted tournaments/leagues/sports it is the same useless data as Q316512#P54. I am totally  Oppose an idea of making so vague and controversial statements like career (Q282049). If necessary, we can always summarize the available specific statistics by function, but there is no need to pretend that we actually counted every match of his career here. For example, if someone wants to get the sum of all Lionel Messi (Q615) goals in UEFA (Q35572) competitions, it would be wiser to sum up the goals of the UEFA Champions League (Q18756), UEFA Europa League (Q18760) and UEFA Super Cup (Q484028) at the query stage, rather than store an excess statement "UEFA career goals" and the same for yellow cards, appearances, fouls committed, etc. Hence "career totals" would at best be a useless duplication of information and at worst it will be misleading. Сидик из ПТУ (talk) 14:07, 10 March 2021 (UTC)

@Сидик из ПТУ: Well, same answer as match results, the model support it. All kind of "scope" dimensions are possible but not mandatory. May be Jordan is not a good example as it played several sports and "total" is hetereogenous. In your initial example with P6509, what would it mean without P642 ? or with a P642 = career ?. It is:

⟨ Lionel Messi (Q615)    ⟩ total goals in career (P6509) ⟨ 6 ⟩
date of the first one (P7124) ⟨ 16/06/2006 ⟩
date of the latest one (P7125) ⟨ 26/06/2018 ⟩
point in time (P585) ⟨ 10/03/2021 ⟩

.

If the opposition is regarding career (Q282049) value, I'll change it in examples or define it as forbidden and then, end of the problem. The only point in difference between my proposal and the present creation method of specific properties for each indicator, is that new proposal increase one qualifier - quantity (P1114) - because the numeric value used in the specific properties solution, now is occupied by "kind of indicator". IMHO, the debate should be based en pros vs cons of each solution: flexible vs pre-defined indicators; risc of abuse using "non-sense indicators" vs pre-avaluation in creation property; etc.

I'll rewrite the description and examples with your considerations and I'll present it to discussion. Your important collaboration make me understand better how the properties like "total aaaa in career", run. I appreciate so much. --Amadalvarez (talk) 10:17, 12 March 2021 (UTC)

I believe that for matches, goals, etc. no statements can be made without a clear designation of the competitions taken into account. So,

⟨ Lionel Messi (Q615)    ⟩ total goals in career (P6509) ⟨ 6 ⟩
date of the first one (P7124) ⟨ 16/06/2006 ⟩
date of the latest one (P7125) ⟨ 26/06/2018 ⟩
point in time (P585) ⟨ 10/03/2021 ⟩

is wrong, of (P642) should be mandatory constraint (Q21502408) for total goals in career (P6509). There can be completely different points of view on whether it is necessary to include 2018 International Champions Cup (Q51988112) or Hibernian F.C. - FC Barcelona (Q105906547) matches in Messi's career calculations. Ok, let's focus on the main question. I agree that both approaches can support the same functionality. I propose to draw up a "pros vs cons" table. Сидик из ПТУ (talk) 11:11, 12 March 2021 (UTC)

By the way, I forget answer point 6). I agree that a clean up operation should be done, but not only in P54. In the table where I analize the status of previous statistic properties, there are evidents bad uses. It should be our next step. See you, Amadalvarez (talk) 14:43, 12 March 2021 (UTC)