I understand doing one vote per player or even per IP for each res, but I have questions. What if the shop spans multiple residences (could easily be influenced one way or another). Also, there should be a way to change your rating. As has been noted many times already, some shops start out great but are not kept up well with stock etc. I think if we had a system that looked kind of similar to difficulty setting in /PS might work. That way you pick a number, but if you try to rate them again, it shows you the current rating and you could adjust if you wanted. But it would still only count as the one original vote (for averaging purposes).
I'm not sure if this goes here or in it's own suggestion but I'm seeing everyone talk about helping reduce unwanted tp's to non-existent shops/malls... Why can't there be a server wide purge of the /v +shop /v +mall command tags and then everyone that has a legit shop/mall can send in a sign up application with username res address stating to re-add the +mall and/or +shop. Kinda like the despawned Holiday mobs application. Now that we are at this point. After everyone has been processed. if you have the tag space available or are new you can add the tag back on or for the first time add the tag.
The issue, is not many shop owners are not active yet their malls are still very well stocked... then comes the issue of them knowing that the tag has been removed.
With this system, would it be wiped every set amount of say every set of months or? just wondering how that aspect would work...
I would guess that a rating from a person, will be saved for lets say 3 months, and after that deleted so it doesn't count towards average
I really like the idea of some kind of "like" for cool builds and creations. One thing EMC has is a lot of really great builds, but many have gone unseen by most. For example, how many players have seen iamcavie's Empire Strikes Back themed build in the utopia frontier? (my personal all time fav build!) It's just one of many, many epic builds to be seen. Promoting these creations with a leaderboard and even a teleport command to maybe the top 10 or maybe to the build of the week would be pretty sweet. I think having the "like" rating influence the +shop command is a bad idea. For example, a shop or mall may not look as good as others, but it may have the best deals, best prices, it's always in stock, offer unique and rare items, best enchants department, etc. To have the "shop design" as the only criteria influencing the +shop seems to skew the equation toward fancy builds versus how the business is run and how it serves the community by providing what it wants at a fair price. Unless we're prepared to install additional ratings categories for shops: best deals, always in stock, unique products, etc. to level the playing field, use the "like" system as a way to promote visitation to cool builds. If that cool build is a mall, then it'll get foot traffic by virtue of looking cool, so no need to have it further influence the +shop criteria. As others have previously stated, that would make it even harder for smaller or newer shops to get going.
We can put in warnings and such to note vote just because someone asked you, and to provide a real rating. We can also do things like make it a rule that if your reported for paying for ratings, that you're punished (say given a special -20-50% or even 100% penalty) We will solve problems as they arise.
An important detail in my first reply is that we would not have a single rating, but 3 different groups of ratings. You would be able to rate a residence separately based on overall build design, overall shop quality, and quality of any games on the residence. So we can then add a /sightsee that acts like /v random except that it favors higher rated build designs Then we can add /playergames that is the same but for residence with higher game ratings. And /shop being for shop rating, as discussed already
I mentioned monthly leaderboard. So your monthly rating will be reset on first of the month with the other leaderboards, and you start over. Scoring would likely be something like 50% lifetime, 50% monthly. That way, inactive residences or stuff that starts getting a lot of negative ratings due to shop being out of stock, etc, will receive even more of a hit even if they had a super high lifetime rating due to previous states. Right now every residence is thrown into a list once. Unrated could be given a 3 as we really don't know (until that first vote comes in) We can calculate your entries to the pool as: entriesIntoDrawing = min(floor((monthlyRating * 0.5) + (lifetimeRating * 0.5)) , 1); However, I would make there be 4 pools, something like: pool 1, 10% chance of selection, includes res's in the 0-2.4 range pool 2, 20% chance of selection, includes res's in the 2.5-4.9 range pool 3, 25% chance of selection, includes res's in the 5.0-7.4 range pool 4, 35% chance of selection, includes res's in the 7.5-10 range The command would first randomly choose a pool based on those distributions. And then inside of that pool, have a drawing with residences within that pool based on entries. This gives players a 60% chance of getting a good residence. We can also message the user when they arrive at a residence to remind them to rate the residence. We can also give this note when a player purchases from a shop too. We can do numerous things to encourage players to rate.
so, would there be a rating if that res was lost (for the lifetime part) or how would that work? Maybe just show a Lifetime - N/A part or something?
Interesting thought, but this, for example, could make res's rated between 2.5-4.9 be visited more often. Say there are 100 residences: 30 have got a rating of 1-2.4, 10 have got a rating of 2.5-4.9, 40 have got a rating of 5.0-7.4, and 20 have got a rating of 7.5-10. Now, the chances for any specific residence are as follows: Residence in pool 1: 0.33% Residence in pool 2: 2% Residence in pool 3: 0.625% Residence in pool 4: 1.75% As you can see, with the distributions I was working with here, a residence rated between 2.5 and 4.9 has got the highest chance to be visited. And I don't think these are very unlikely numbers, either, as I think that if people will find a residence bad, they will rate it very low, and if they find it decent, they will often go for 5+ right away, so there won't be many residences in the second pool.
Alternatively, instead of putting into buckets, the ranges is yet another multiplier. So pool 1 has * 1 entries, pool 2 has * 2 entries, pool 3 has * 3, and 4 has * 4. This would solve that issue.
Let's be honest, how many decent shops can you find using /shop? I have 5 shops memorized, because I don't know about other SMPs, but on smp5, a shop is defined as a dirt house.
You know that if you find a res with the shop/mall tag, and no shops on the res, you can report for a tag removal.... Just send a Bulk PM to SS with all the reses and tags on that res. From there they can go around and remove them... Of if your lazy, you can report one once every hour; /report player {SS} +shop, +mall tag with no shops on residence.
Well I was about to think I went insane and started a thread I don't remember starting. (Might not be the first time that happened.) I'm really glad you like my idea. There's really two reason why a 'like' idea is better than a 1-10 rating scale. The most important reason is that a 1-10 rating scale allows players to rate one another's creations as 1/10, which is essentially saying the creation they probably spent hours/days/weeks on sucks. There's no way that's constructive or helpful to the creator, and in many cases that would demotivate them. Something someone built might suck, but it's not something you would generally want to bring to the attention of whoever built it. The reason for rating 1/10 might have nothing to do with the quality of the creation either. It might be a personal bias against the creator. You don't like how they treated you or a deal went sour, so you go around and rate everything they built as 1/10. If a staff member bans your friend for something they did, go around and rate all that staff member's creations as 1/10. Hopefully you can see where this becomes a problem and creates unnecessary negative energy. The second reason is much simpler. Tracking and analyzing 1-10 ratings is a bit more complicated than simply tracking a list of likes. You have to deal with things like how to compare a creation that has a single 10/10 vote against a creation that has 9 10/10 votes and a single 1/10 vote. The second creation has clearly been more beneficial, however depending how you treat the ratings it would sink lower. Consider a creation that gets a 1/10 rating to begin with. It would be biased for life when possibly that's just the opinion of one person, and maybe not even about the creation itself. With likes, it's only positive energy. Maybe you are liking the creation because of who built it, not the creation itself, but that's not exactly a bad thing in itself if people are encouraged to go around and see their friends creations to like them. And the tracking/analysis is a simple count of how many likes the creation got, which is a pretty good indication of how many people came to it and liked what they saw. It's simpler for the end user too, they don't have to think about whether a creation is a 7 or an 8, maybe a 9 because that's their best friend who will be insulted if they only put a 8. They can just like it. The reason I suggested XP (and not a large amount) is as a way to encourage people to build great things and also to explore the creations of others. A small token amount could be provided for visiting a creation of another player, as well as for receiving a like. Of course, it could be that XP is overused to begin with, so another system could be used. But it makes sense to encourage both the building and exploring of creations with some sort of recognition for both. I think only the exploration side of this is really necessary, the building side can be encouraged enough by people competing to get the most likes on things they build. I noticed jkrmnj has given some counter arguments in favour of ratings so here's my take on those: It's very easy to just look at how many likes a creation got in the past month, rather than the overall total. In most cases, that's more desirable. The residence with 100 ratings is clearly having a large impact on the server. It isn't wrong to send people there. A like system well implemented can easily favour the new creation by: 1) Factor in when likes were created. This would simply look at how many likes that creation got today (2) versus how many likes the other creation got (which is going to be 0 if it has no impact.) Likes can 'depreciate' by a % per day. More recent likes weigh higher in the formula. 2) Have the system favour creations the player has not explored yet. Why should anyone who's looking to explore want to visit the same thing twice anyways? 3) Give a boost to new creations. For example, count them as having a bonus 50 likes for the first day, 25 for the second, 12 for the third, etc... Just in terms of ranking/randomness formula. While it's true that in a rating system you have more options, doing nothing also hurts the rating compared with rating 10/10.
I don't think that is as much of a problem as it sounds. As you said, people are more likely to rate on the extreme end of things. It is also true that when more people visit your residence, you will get more ratings. Let's assume some shop is firmly inside of the second pool. Because there are fewer people in this pool, they are more likely to get picked, which means more people will go there and more votes. Since most people vote towards the extremes, they will get a large number of extreme votes, most likely in the 1* direction. This will push them out of the second pool very quickly and into the bottom one. What this means is that although you have a very high chance for a short period of time, you will have a very low chance for a much longer period of time. If anything, this is good because it means residences that are in the uncertain part of the scale will get more exposure and more votes solidifying whether people want to go there or not. That solution seems like it would hurt visitors and those using the system to find things. If there is a disproportionately high number of low star entries (which, as the current system shows, there would be), they could fill up so many spots that it becomes more likely you will visit someplace low instead of someplace high. Imagine 100 low ranking shops and only 10 high ranking ones.
