PDA

View Full Version : Suggestions to improve level visibility



Gilgamesh
02-27-2009, 01:55 PM
Well, with the recent update MM have removed a glaring flaw in their search engine... which unfortunately was our only real way to get decent exposure to the nameless masses. In this thread, I'm going to offer some suggestions of what MM could do to help us get more exposure than what our levels currently get (about 15-20 plays for a 7 day period, then nothing). These suggestions don't have to all be applied, and none of them excludes the others from being applied too.


Why : Forcing players to rate and tag levels just ensures that the ratings and tags will get skewed results because the majority of players simply want to move on to the next level. What do you think youtube would be like if they forced you to rate every video you watched and rewatched? Rating and tagging should be optional, it shouldn't be imposed on every player all the time.

Solution : Make a "rate this level" button on the level's page that players have to select in order to rate or tag a level.

Theoretical advantage : Only players who actually feel like rating or tagging a level will do so (whether bad or good). This should improve the accuracy of ratings and tags because people who simply want to move on won't bother to add "random" tags and ratings. The net result is that good levels will get a better chance to get high ratings, while bad levels will get a better chance to get bad ratings.

Theoretical disadvantage : There will be the occasional immature player who will rate a level poorly or give it a random tag just to be "funny". Since there are fewer players who will be tagging and rating the level, the effect of those immature players will have a bigger impact than before.

Why : Players look for convenience, and currently the only convenient way to find a good level to play is to look at the 1st page of the Community Levels (Highest Rated, et al). This has the disadvantage that all levels in the later pages get exponentially fewer plays than levels on the 1st page.

Solution : After the player has finished a level, they would be presented with a random pick from the author's hearted levels. They would be shown basic information about the level (picture, description, stats) and have the option of trying this new level or not.

Theoretical advantage : By presenting a random level from the author's hearted list to the player, they will have another convenient way to find another level to play. This will also get them to play levels that would not be on the 1st pages, increasing the exposure of more levels.

Theoretical disadvantage : In order to start the "chain" of hearted levels, the authors of levels from the 1st pages have to heart other levels. This means that levels with no hearts, or levels with hearts from small or non-authors will not get more exposure in this system.


Why : The current search categories are arbitrarily decided by MM, with their secret formulas of exposure using the Ratings, Plays, Hearts and Date statistics to determine which level goes to the top of the food chain. This makes it very hard for a player who finds a hidden gem with few plays to somehow promote that level so that more people play it, or for authors to compete with the "top dogs" of each category.

Solution : Make a "promote this level" button on the level's page which would add the level to a new search category called "Promoted levels". Levels with fewer plays and more people who promoted them would be on the 1st page of this category, meaning that as a level gains more exposure, they are gradually replaced by levels with fewer plays.

Theoretical advantage : Levels with a few number of plays would get the chance, if they're good enough to get promoted, to gain front-page exposure. This also gives a more clear power to the players to decide what goes up or not in this category and offers another convenient place to search for good levels to play.

Theoretical disadvantage : Since altruism is not a common trait in most players, it is possible for newly published levels to slip completely off the radar without having the chance to be "promoted" to get more exposure. Even if the level is good, players might not think to make the extra effort to promote the level.


Those are the suggestions I think would help anyway. If you have better ideas (or think these ideas won't work) feel free to comment. :)

I also highly suggest that we spread the word that our levels are not getting any exposure and that new solutions like these need to be made by MM.

xkappax
02-27-2009, 02:09 PM
I think we should also be able to tag our own levels with genre. If someone is looking for a ghost level, a cave level, a fiery level, we should be able to tag it like that to make it easier to find... just like you can tag genres on youtube. If your level has 8,000 rocket cheetahs, there should be a tag for that too.

There should also be an option to upload a picture to your level BEFORE it goes live. That way, people can put a picture on their level showing some scenes, and when it goes onto cool levels, it will already have some sort of image showing what it's like.

I agree with what you've said above, as well.

wexfordian
02-27-2009, 02:26 PM
Retain the old system but remove every level after one day on page one.

After it is removed republishing won't work.

This will not only cut down on republishing as only the newest levels will get republished (right now you will have everyone republishing all their levels including their old ones every week!) but also it will give every active user the oppertunity (and delight) to watch their level rise up the rankings and gain that sense of accomplishment when it reaches the top.

If removing the first page every day is not enough to get through all the new levels remove page 1 & 2 and so on.

xkappax
02-27-2009, 02:29 PM
I also like this suggestion. Although, maybe they get to stay at the top for a week, and then after one week on page 1, then they go, and the other stuff moves up. I think one week would be long enough for a satisfying amount of people to play and enjoy them.

Gilgamesh
02-27-2009, 02:31 PM
I don't know... having only one day on the 1st page has a different effect depending on the day the level was published. Odds are you'll get much more exposure on the weekend than you would on a Wednesday, factor in holidays, vacations and different timezones and it becomes a bit of a headache to know when you should publish to hope to get your day's worth.

xkappax
02-27-2009, 02:34 PM
as someone who had exactly one day on the top page, i agree. especially if someone finds something wrong with your level once it's there. it could really screw up your heart to play ratio. With a week, you'd have plenty of time to get a wider audience... both the weekend people and the weekday people.

wexfordian
02-27-2009, 02:37 PM
Well the time wasn't really the point I was trying to address the issue of too many people publishing too many levels. I personally would agree with you I'd much prefer to see a level stay up for a week on the front page but if there are only 10 levels on the front page that means there would only be 10 new levels added/removed each week. I was just trying to find a comprimise.

xkappax
02-27-2009, 02:42 PM
I thought of another one. We need a tag in the game that is "incomplete" or "broken" for levels that just don't work... , or there should be some sort of server check, and if a level doesn't have an exit portal, you should not be able to publish it.

also, they need to put more levels on the community levels page, because you're right, there aren't enough on there.

RangerZero
02-27-2009, 03:24 PM
I'M TOTALLY IN LOVE with the "promote this level" button.

This could even work like a vote. And it would give it hits in order to make the next week update of the community page.

.

I also think that a REAL tag system where people write the tags would work while the present one is utterly useless.

A better classification of levels, some "genres" would have helped too.

I also like the idea of difficulty levels tagging.

The best thing would be to bombard the player will level suggestion just like YouTube bombard the user with videos. This would need some serious redesign though.

.

CCubbage
02-27-2009, 03:30 PM
Personally, I would suggest something like the following:

Let's say most REALLY good levels fit into 20 highest-rated pages. These are obviously ALL good levels. Treat them evenly and spread out the plays.

For instance - I would like to see the pages returned randomly. So, when someone goes to "highest rated" it returns one of the highest rated pages. This gives people a reward for reaching the highest rated, but doesn't mean a single set of levels gets all of the attention. Splat Invaders Saga reached page 3 of the highest rated and stayed there for several days. I had very few plays during this time. People only see the first page.

If someone is able to get a highly rated level in the first week of cool levels, they are rewarded with being in the highest rated.

Also, in order to give a great level a fighting chance - once again, don't let people do a star rating on a level until they've finished it. They MUST get over the finish line. Why? It's never fair - in commercial games or LBP - to allow someone to rate a level based on a tiny part of it. Right now, a pre-teen can jump in, not like the beginning, rate a 1 star and unilaterally bury a level. NO ONE else will even try it. So, by creating a good puzzle or challenging level, authors have a substantial risk. That only increases mediocrity and lack of diversification.

I ALSO think that after a certain number of plays, a level should automatically be moved into a "greatest hits" section (or hall of fame... whatever). To make room for new highest rated levels. So, maybe when a level reaches 100,000 status it can be moved to an accessible area.

Also, allow a couple genres tags to be set by the author. So people can find specific gameplay styles they are interested in instead of seaching through 50,000 they may not like.

nordwest
02-27-2009, 04:18 PM
Building and adding to the points many have made, here are some thoughts on how and why things could change for the Search function etc:

Genre Tag:
Creators pop in perhaps three of these along with the title and description. Very similar to YouTube or HTML tagging for websites. Allows a level baseline from which players can select.

Title Image:
Creators pop in hero shot of their level at publication. Pictures speak a thousand words, this would allow all players to rapidly make an informed selection from the genre baseline (in addition to hearts etc) Usefully language-independent too.

Trophy Hunter Corner:
If a level has H4H, trophy etc in the title, off it goes automatically to a seperate 'genre'. This would remove a lot of level 'noise' for those making careful searches, and make casual searches far more accessible. The H4H crowd would still be catered for.

Randomise Cool Levels:
By all means keep a cool levels functionality, but if Mm are going to mix it up a bit, mix it up a lot and do so at least daily. This would allow even the oldest levels to enjoy fresh plays and earn new fans.

Creators Choice:
Mm say they listen to the community, and forums such as this? Great, so why not work directly with the community by regularly highlighting levels such as those that feature through informed critical selection in Community Spotlight?

And finally, an ode to republishing. It isn't always about simply bumping or 'cheating' your level to get more plays, it can also be because you actually like to tweak, hone or just plain change your mind as you go. I design real things in real life to fixed deadlines, contracted deliverables and a whole bunch of stress. Having the opportunity to offer creativity in a flexible way is part of the reason I find LBP enjoyable and relaxing. Some method of enabling this process to continue without penalty must be re-incorporated, in my opinion.

Miglioshin
02-28-2009, 01:21 PM
Indeed.

I agree with you.

Without all the feedback I earned here AND by republication my favourite MFA ch.2 level would be a 'crappy ultra hard 2 star level' and not the enjoyable challenging 5 star which has become thanks to LBPC creators and random people who gave me feedback.

But this has happened in 2 MONTHS, not in a week.

brnxblze
02-28-2009, 04:26 PM
Retain the old system but remove every level after one day on page one.

After it is removed republishing won't work.

This will not only cut down on republishing as only the newest levels will get republished (right now you will have everyone republishing all their levels including their old ones every week!) but also it will give every active user the oppertunity (and delight) to watch their level rise up the rankings and gain that sense of accomplishment when it reaches the top.

If removing the first page every day is not enough to get through all the new levels remove page 1 & 2 and so on.

This is exactly what I was thinking. They should've left it how it was but taken off the levels from cool levels after a certain amount of days.

RangerZero
02-28-2009, 04:38 PM
Retain the old system but remove every level after one day on page one.

After it is removed republishing won't work.

This will not only cut down on republishing as only the newest levels will get republished (right now you will have everyone republishing all their levels including their old ones every week!) but also it will give every active user the oppertunity (and delight) to watch their level rise up the rankings and gain that sense of accomplishment when it reaches the top.

If removing the first page every day is not enough to get through all the new levels remove page 1 & 2 and so on.

This I don't agree with though. The logic of their new system completely solve the problem of the first page of the community section.
It's plethoral of other problems that is hurting us.

- Like relying on the star rating to replace the levels on top.

- The total lack of levels promotion on the game's part.

- The fact we don't even know what happens to your level after seven days. If it drops from the first page, do we have some days pinned on second and then some days pinned on the third one and so on?

Are older than 7 days levels part of the pool to replace those on top of the community page? Media Molecule didn't explain anything really.


.

larryjoe701
03-01-2009, 01:11 AM
I don't agree with Gilgamesh's suggestions, other than mandatory rating perhaps (I find I often don't tag levels because it is not mandatory, even if I complete it).
On his point #2, I don't think it goes far enough, and it isn't quite the right approach. We definitely need recommendations based what levels people heart, but just a random suggestions after you finish a level isn't enough. I want whole pages of levels based on a statistical analysis of my hearts and who else hearted the levels I hearted and the levels hearted by the authors I hearted.
On his point #3, I think "Promoted Levels" would just be yet another search page which has levels I don't care about... how/why would it be different than other categories already available to us, such as most hearted?

I think we need to ask a few questions first to come up with better search solution.
1. Who do we want to play our levels? Is it just anyone, or do we want to try and hone in on the people most likely to enjoy our levels? I want the latter. Fewer plays but more enjoyment. I will never make a trophy or sticker level and I don't care about players who just want that.
2. Why do we want these plays on our levels? To brag that we got the most? To get validation that we did a good job? To get some trophy? To feel that we in some way contributed a level that people enjoyed? For me, it's validation and feeling like I contributed a level people enjoyed. The more plays the better, provided that doesn't come at a cost of bad star ratings. Fewer plays by people more inclined to like my level is ok by me.
3. What do we want from the search engine? I want a never-ending stream of levels that I will enjoy, plus access to levels I enjoyed in the past. The latter I can already get from my hearted levels. The former isn't really provided directly anywhere.

I think many different types of players are looking for different things from the game. In this case we need the ability to draw conclusions about levels that others have played and rated and with whom we agree on some body of work. I keep thinking of automated data segmentation algorithms... to get good recommendations Mm might even need to request more detailed information on why we liked or disliked a level, but if it meant I saw a list of beautiful levels with a meaningful story and the kid down the street got 1367 RAMP! levels then I'll provide all the information I can.

So, those are my ideas. What do y'all have to say about my 3 questions?

superBlast
03-01-2009, 07:29 AM
i think the current system sucks. i remember there was this stupid level that all it was was a really big vehicle and a big ramp that was incredibly short. it had way more hearts and plays then this other level that was really well made and was really good. they both was on the front page but i don't see how a crappy level got better stats then a really good one let alone made it to the front page.

also i agree with the mandatory ratings because they get really annoying and i just give it whatever rating it started on unless it deserved a better or worse rating. and usually i skip tagging a level but it still gets annoying because it still pops up. i would make to where rating and tagging a level is unlocked or something after playing through the the whole level.

Rogar
03-04-2009, 07:51 PM
I think it's a sad sight to see a level with barely any plays. Even if it's not that great, the creator probably put a lot of time in it, he/she deserves some acknowledgement.

But I have an idea. How about we use those trophy hunting masses to help us? If we had a trophy for playing Quick Play a hundred times, every level would at least be played a lot. Too bad Quick Play doesn't allow rating or hearting, though.

Gilgamesh
03-04-2009, 07:54 PM
I think it's a sad sight to see a level with barely any plays. Even if it's not that great, the creator probably put a lot of time in it, he/she deserves some acknowledgement.

But I have an idea. How about we use those trophy hunting masses to help us? If we had a trophy for playing Quick Play a hundred times, every level would at least be played a lot. Too bad Quick Play doesn't allow rating or hearting, though.

Hmmm, that would seem like a good idea at first, but knowing the trophy-hunting masses... they would probably just hit Quick Play, and then exit the level immediately without playing. :)

Maybe force them to get to the scoreboard lol.

CCubbage
03-09-2009, 10:18 PM
You know, there has been principles in place for thousands of years for finding things.... the problem is, Media Molecule isn't using any of them. They seem to think this is a great place to reinvent the wheel. They are WAY complicating the entire thing.

Think about this:

Let's say you went into a public library. You want to read about ancient egypt. But instead of everything being organized you find a pile of 20,000 books in the middle of the room - with a few books on top that are the most read and the books on the bottom the least read. And scattered in random places among them are books that someone feels you MIGHT want to read.

What are you going to do? Well.... you'd like to read a book but there's NO WAY your going to find exactly what you're looking for - so instead of just leaving you pick a book off the top. Turns out the books on the top are only popular because a) they are about super heroes or b) 8 year olds find them easy to read. So you start digging through the pile in the hopes that you find a book that you may enjoy.

So how do you make levels visible? You use principles that have always been around for thousands of years and organize them.

How do you accomplish this? You hire the people who designed the Netflix web site.

RangerZero
03-10-2009, 04:24 PM
You know, there has been principles in place for thousands of years for finding things.... the problem is, Media Molecule isn't using any of them. They seem to think this is a great place to reinvent the wheel. They are WAY complicating the entire thing.

Think about this:

Let's say you went into a public library. You want to read about ancient egypt. But instead of everything being organized you find a pile of 20,000 books in the middle of the room - with a few books on top that are the most read and the books on the bottom the least read. And scattered in random places among them are books that someone feels you MIGHT want to read.

What are you going to do? Well.... you'd like to read a book but there's NO WAY your going to find exactly what you're looking for - so instead of just leaving you pick a book off the top. Turns out the books on the top are only popular because a) they are about super heroes or b) 8 year olds find them easy to read. So you start digging through the pile in the hopes that you find a book that you may enjoy.

So how do you make levels visible? You use principles that have always been around for thousands of years and organize them.

How do you accomplish this? You hire the people who designed the Netflix web site.

Man, stop reminding me how this game is ridiculously badly designed...

.