Download count being an implicit ranking system

[doofus-01]

This is the internet, so of course we need to all talk past each other, but fabi makes the point that I believe is relevant to what Dugi is getting at:

Still it takes more and more time to scout the addons,
maybe some users just give up after 5 tries and never play UMC at all?

There clearly is a rating system, and it sucks (or at least that is the concern), and this screws the players and turns them off from UMC.

@Gambit: The point is not that franz is getting more downloads than Dugi. That was obviously supposed to just be an example, whatever you may think of how it was presented. Actually, I think it was a better example than the 1.9/1.10 Love to Death thing because it separates the upload order from the download count. Don't think about franz and Dugi, think about the thousands of suckers whose downloading made up the precious download counts. And if they, like fabi speculated, might just throw up their hands and give up because they were misled by the current ranking system.

[Crow_T]

And ...?

This is my point. You decided that that matters and that it's apparently important to be in "first place".

No, that's not the point- what matters most is that the addons listed highest should be some combination of actively developed, completed, and working properly. There really should be a standard for completeness, just to take Dugi's campaign as an example: for each chapter he marks that it is complete or not and is very open and active on his thread here. Also Doofus and Lord-Knightmare, mattsc, Elvish Hunter, Shadowmaster, Adamant14 etc. are all very active, so their campaigns should be very high on the list. Addons could be ranked by a number of factors, completed scenarios, dev activity, % complete, even a user ranking as a minor part of it. There are certainly a number of variables and its not an easy problem to solve. But as my own example: Ice Age Fun: started to play it, got some cool units, and it ended halfway through the story, unfinished. That sucks, there is no nicer way to put it. It'd be nice as a player to sort by completed campaigns, as well as actively developed ones, because it's fun to be a part of that process as well. The current sorting method is pretty weak.

[pyrophorus]

Elite: A group of persons who by virtue of position or education exercise much power or influence. (Merriam-Webster)

If you make only some people capable of creating reviews, you are creating an elite simply by virtue of their influential position. Which is why a good review system should be democratic, not oligarchic.

That's right, but if reviewing is open to anyone, reviews will quickly show the add-ons problem. A newcomer will get lost in a huge heap of posts, knowing not how to evaluate their reliability. Have you some other idea ?

This is my point. You decided that that matters and that it's apparently important to be in "first place".

Ironically, this is an example of bad users management. Maybe it would be useful to you to know (as forum administrator) that telling people how they should think or feel is useless and counterproductive. They're waiting for ideas and solutions, not "good" principles. This topic (add-ons organization) pops up more or less every two months, and, as Fabi says:

Every other project would panic and solve the problem in days

Then what next ?

Friendly,

[Unnheulu]

because of it, the add-ons with more downloads are more likely to get more downloads

And ...?

This is my point. You decided that that matters and that it's apparently important to be in "first place".

As a non-campaign author I also say this matters, when I want to play campaigns I normally just replay the mainline ones unless there was a campaign made by an influential board member (Doofus, or Sleepwalkers for example). The vast majority of add-ons on the server are rubbish or incomplete and it's not worth sifting through them all just to find a few good ones. Therefore I go for the most downloaded ones. When these aren't that great I just give up and don't bother with UMC. I'd imagine that I'm not the only one who does this either.

[fabi]

Please let's keep the thread in a productive state.

The discussion about Dugi vs franz or the meaning of "elite" is not taking us anywhere.

Let's talk about downloads.

because of it, the add-ons with more downloads are more likely to get more downloads

And ...?

This is my point. You decided that that matters and that it's apparently important to be in "first place".

Interdependently from how Dugi or any UMC Author decides what downloads mean to them,
they have a lot of meaning just from themselves.

Downloads measure a waste of resources, both at the server side and also on the side of the user.
Assuming that the user's internet connection is cheap, you still waste the time of the user testing the addon.

It is important for the health of the project that the addon with the most downloads is at least a good one, if not the best.

-making the client send an information to the server if it is an update or a new download; this would not prevent abuse of the I download my add-on style, but make it harder at least

Regardless of all the ranking/non-ranking debate, this idea should be simple enough to implement and it does sound reasonable. The download counter ought to show the number of downloads, not downloads plus updates.

Yeah. Still this is only a little fix compared to the bigger problem.

[wobbly]

... but so deeply buried very few people can find them, ...

That is the one of the main problems of Guide to UMC Campaigns.
The second is to keep the Guide up to date, when added campaigns make progress / bigger changes.

I agree with this, when I started downloading add-ons I specifically looked for something like this & was unable to find it. I'd suggest linking this (or something very similar) on the top of the add-on manager so it's right where you can see it if your looking for an add on might be a very good start.

[Gambit]

Downloads measure a waste of resources, both at the server side and also on the side of the user.
Assuming that the user's internet connection is cheap, you still waste the time of the user testing the addon.

It is important for the health of the project that the addon with the most downloads is at least a good one

This is a good point.

[Insinuator]

It is important for the health of the project that the addon with the most downloads is at least a good one, if not the best.

People keep saying this over and over again in slightly different ways, so I think we can all agree on this point. What we need is, not a recitation of the problem, but a workable solution. That is what a review system addresses.

Phyphorous brings out a good objection, that with all users able to review, the number of reviews could be overwhelming. To counter this, first let's say the user must complete the campaign to post a review. That cuts down the number significantly, because not everyone cares to finish or even start all the campaigns they download. Secondly, make it so that users can only post a review once. This again reduces the number of reviews and fights authors trying to artificially boost themselves.

Thirdly, realize that many people will not care to leave a review at all. Not everybody feels comfortable expressing themselves in that way, or even cares enough. Finally, many reviews contain overlapping information, so people don't need to read ALL reviews to get a general overview.

Please tell me what I have overlooked for the conceptualization here.

[Pentarctagon]

To counter this, first let's say the user must complete the campaign to post a review. That cuts down the number significantly, because not everyone cares to finish or even start all the campaigns they download.

How would anyone be able to know if a person who posts a review has finished the campaign though?

[Insinuator]

In my head, it would be a pop-up box upon completion of the final scenario.

[H-Hour]

Secondly, make it so that users can only post a review once.

If there are fewer reviews it it will only make it easier to fake a few accounts to skew the impression those reviews provide.

If more thorough improvements are being considered -- and it seems like there is at least some developer interest -- one option would be to combine a recommendation system with the kind of detailed information provided in the Guide to UMC Campaigns, which provides data on the development status and the type of gameplay in an add-on.

Putting this kind of data into a tag system, where each registered user can assign tags, might provide its own kind of recommendation system. So, for instance, if I am a user who has played an add-on, I might tag the add-on with "working", "complete", "skirmish" and "RPG" (see Guide to UMC Campaigns). By entering the tags "skirmish" and "RPG" I am endorsing the add-on for these gameplay types. Alternatively, I could tag an add-on "Broken" or "Hard" or "Unbalanced".

So if I'm looking at an add-on, it might have the tags: complete (25), single player (19), skirmish (15), etc. Or it might show: broken (12), multiplayer (5), etc. By glancing at the top tags and the number of users who have assigned the tag to each add-on, I will get a rough sense of its completeness, gameplay type and intended audience (campaign, era, mp modification, etc). Ideally, it might also be possible to search by tag.

This wouldn't eliminate abuse (which you can't really avoid anyway), but it would contextualize feedback and add an informative layer on top of the competitive layer that already exists.

Of course, this would entail a lot more work than the other suggestions offered in this thread and may be too large for the GSoC project.

[Dugi]

@Insulinator
That might help with campaigns, but what with endless campaigns (there were some, as far as I know, you play its endless levels until you're bored) and other types of add-ons, like resources or eras? What if the campaign has more parts?

IMO, it would be better if anybody was able to comment any add-ons, probably as a new option appearing on the window where you select the add-on server to log to. People may leave comments, from complex reviews to short messages like 'awesome' or 'sucks' (very short or too unspecific negative replies will have to be checked if there aren't written by morons who do it just for insulting, see bottom of this post). It would appear in some place accessible through the browser, and some 'caretakers', people who had played a lot of add-ons, would write a summary of it, reflecting the general impression, and then listing its advantages and disadvantages (not the story or something, that should be in author's description). See the example in the spoiler to better understand what I meant to be a conclusion of the comments made by the 'caretaker'.

Example (tried to review something everybody knows, definitely not the best review that can be written):

Invasion from the Unknown
Tags: skirmish, dungeon, puzzle
Very good and very popular campaign. Starting with usual elves in forests and deserts, quickly adding the undead faction in a battle against several custom factions, evil human army, mysterious Verlissh, sinister demons, robotic automatons and dreadful, biomechanic Shaxtrals, that progresses through caves and steppes into the mysterious Dark Hive, where the Shadow Master dwells. Definitely worth playing.
Pros:
Good story, enough long to develop an army of recalls an use it well
Many new kinds of enemies, bringing new challenges
Interesting gameplay
Well balanced, enjoyable on easy and very challenging on hard
Cons:
Many units lack animation
Some elements of the storyline are too exaggerated

Some kind of subjectivity might be needed, because some campaigns contain a single element that overshadows many other elements that are mediocre (for example To Lands Unknown, absolutely unmatchable cutscenes, but uninteresting gamplay and plot holes).

To prevent cheating, IP addresses, cookies and stuff of this kind might be recorded, and comments ignored users who seem to misuse it (more comments from the same IP, anonymisers). There might also be a warning that cheating might make your add-on get deleted, and I believe these people will not risk having their precious add-ons deleted. Keeping the IP addresses should also help blocking people like the youtube dislikers, who just post hateful comments about something without knowing it at all.

[Groggy_Dice]

A short list with the number of addon items per Wesnoth version.

1.4 225
1.6 229
1.8 393
1.10 453
1.12 124

As suspected, the count is still going up,
maybe with a local peek in the near future, caused by editor improvements.

No idea how to explain the stagnation between 1.4 and 1.6.
Maybe the 1.4 section includes earlier works as well?

There are also several addons still in the 1.5 section, because their _server.pbl doesn't include the key for add-on type, a feature introduced during 1.5's development.

On the other hand, there are maybe a dozen or so add-ons that were never resurrected from the crash of the 1.3 server.

[Midnight_Carnival]

don't think I have much to add here...

1) a ranking system will be of little use to people who want to play and develop and of limited use to those who only want to play -you spoke of "objective and subjective" - surely you can see how rankings of add-ons will mainly be subjective anyway

2) a poll system will not reveal which add-ons are the 'best' - however you qualify that term, only which ones are most popular.

3) propose a 'dump' (perhaps better title, or something similar exists?) where incomplete campaigns (or those the developers were unhappy with) and old campaigns which don't run on later versions can go and be scavanged for frank parts, this might help seperate the useful junk from the stuff people are meant to use.

4) propose you create a web page on which you can express yourself regarding campaigns and then nobody can say anything about it

[Dugi]

1) It will be subjective, but it should be made as a synthesis of more people's opinions, creating some kind of general opinion. It should be good to tell a widely appreciated that is exceptional in some way from campaigns that are working and balanced, but mediocre in all ways a campaign can be (good example would be An Orcish Incursion).

2) It would be good enough to tell which ones are popular. People download those who are on the top of the list, downloaded much (because they believe they are popular, although it usually means that they are being worked on and therefore probably incomplete), who have nice descriptions and so on, and then you have no way to tell which ones are popular and which ones are just downloaded a lot.

3) Incomplete campaigns can be good, for example Antar, son of Rheor is a well-designed and balanced campaign, but half-complete because its author needs ideas, and most people tell that they enjoyed it and can't wait for more scenarios. After the Storm was incomplete for a long time, with two out of three episodes done, and the first two were still good enough. My campaign is incomplete, but contains about 185 scenarios, that is still more than all other campaigns. Campaigns under development need also feedback.
Another case are campaigns that are both incomplete and abandoned, like Love to Death.

4) Wasn't supposed to be like that. Everybody can post, somebody will be charged to write syntheses from them, and these syntheses will be visible on the add-on server.