Check out the Rock Band Dashboard app on Facebook!

[Bront20]

My feedback is still simply that adding points to people with the current DLC on their list is a positive feedback loop, and those are typically bad (yes in this case positive is bad and negative is good - a good example of a positive feedback loop is shoving a microphone in front of a speaker it's driving. Bad.).

I still like my "points for bands I hate" idea earlier in this thread to create a negative feedback loop. Plus then just for fun I can put all those points into Muse because I've listened to the rest of their Absolution album and Hysteria is the only song on the whole thing that would be worth playing in Rock Band. So there.

People don't buy songs they hate, so that feedback would be useless to HMX.

[Bront20]

Oh, question- do RBN authors get access to all of the data? I imagine that'd be useful for them.

Define "All the data".

I'm very sure RBN authors can see how well their songs sell. They can probably even see PS3 vs 360 sales of their own songs. The first part is required in knowing how much they are owed when it becomes time to pay out. The second part is still useful info for them.

If you mean the facebook app, it's pretty easy to see how high a band is ranked, just search. More than that, probably not.

[cwilbur]

Right, but again, since Harmonix can't verify what you do and don't like,

Let me say this again. The verification of the motives of ANY SINGLE INDIVIDUAL USER DOES NOT MATTER IN DISCUSSION OF MY POINT. My point has nothing to do with what single users can do to game the system if they try. Let them go, ignore them, they can't be stopped until you create a perfect system, which you can't. So no matter what you build some user somewhere will game the system and it won't be 100% fair. Fine. But the single user gaming the system for some motive, who you bring up again and again, has NOTHING TO DO WITH THE POINT I'M MAKING.

I'm talking about the social engineering (intentional or not) of the behavior of the mass of users, as a group, and how THAT behavior is influenced by the choices HMX makes with how they build the tool and how they add points to it. If HMX makes the system easy to game, lots of users will game it. If they make it hard to game, lots of users won't. If they distribute extra points in a manner that the mass of users will tend to spend those points in a particular way, those points will get spent in a particular way. And those effects will in the long run be FAR stronger than anything any individual user trying to game the system can pull off.

it's literally the same list as you'd have now if you REALLY wanted more points.

And if you're a casual user who doesn't sink a ton of thought into how the system works at all you're going to fill the hate list with, well, bands you hate, and you probably outnumber the "game it to gimme points" user 100 to 1 or so, so from HMX's data gathering point of view the gaming the system user doesn't matter. Heck, that "hate" data is possibly as valuable to HMX than "like" data, because if they release some band that's way up the "like" list only to discover that a large number of users would have had that band on a hate list if one was available*, and the DLC sells far less than the wildly high "like" number would have indicated, well, that's probably something they'd want to know before making the investment in licensing and development.

Fine, so users who think it's important to have lots of points will fill their hate lists with every band that has been confirmed, hinted or rumored. What will they put in their want lists then? Probably bands they actually want, which means HMX gets more accurate "like" data than in the current system where the "like" lists of those people will be filled with rumored bands in the hopes that they'll score more points. Sounds like a win for HMX to me.

* I'm thinking of A7X as a very possible for real candidate here, because I know lots of other old fart RBers with lots of disposable income who buy tons of DLC. Many wouldn't touch more A7X DLC with a ten foot pole and would probably put them on such a "hate" list.

Edit:

People don't buy songs they hate, so that feedback would be useless to HMX.

This appeared while I was typing. I think I already covered how that data actually could be useful, and having the feature could even get more useful data out of GNF's hypothetical "filling out my list just to get more points" people.

If I'm a business**, knowing what stuff I could find myself sinking effort and money into that my more loyal users (because it takes some measure of devotion to knowing what's going on to even know there IS a facebook app) will totally reject is damn well information I want. And if I can offer those users a cookie for releasing a band that they told me they can't stand, which I realize really bothers some people, then I think I might seriously consider giving them that cookie to reward their loyalty and make sure they're still around when I release something they did tell me they wanted.

**VP of a small company IRL, so, holy crap, I kind of am....

[Meat-Popsicle]

This might be a bad system, if HMX was somehow forced to release DLC for the bands with the top number of votes. I'm sure they find this data useful while recognizing the limits built into the model. Step one to scrubbing the data: re-weight all votes as if each person only had 10 votes.

I haven't seen any post saying this wouldn't work, so I'm putting Bush and Huey Lewis on my list right away. And then I'm adding the artist(s) for next week as soon as they are announced on Friday morning. I'm normally a perversely anti-cheating person (I've NEVER played on no-fail mode! I always restart a song if I accidentally hit pause when trying for the overdrive button on my MIDI adapter!) but these points mean nothing, nothing at all, EXCEPT I would be able to list more than 10 different bands. Which would be fun. Or I can give all my points to Juliana Hatfield one week, just for laughs. I mean, even if I somehow accumulated 500 points HMX could look at the data and say, "hmmm, every single vote is from one member, we'll go ahead and ignore that".

[cwilbur]

Step one to scrubbing the data: re-weight all votes as if each person only had 10 votes.

This is so brilliant and obvious I hope HMX is already doing it. I'm pretty sure it would defuse the positive feedback of rewarding likes that I've been worried about. Yeah, that would do it. If HMX is re-weighting the votes like this before assigning ranks then I think they can do whatever they want with extra points without creating bias in the system.

I haven't seen any post saying this wouldn't work, so I'm putting Bush and Huey Lewis on my list right away. And then I'm adding the artist(s) for next week as soon as they are announced on Friday morning.

I think someone from HMX already said in this thread that they can see when votes are entered to prevent this kind of thing, but I think you should do this anyway and report back with your results the next time they give out points for liking because it would be interesting.

[SirDavidTLynch]

The "hate list" idea is actually serious? I don't think labels would appreciate Harmonix revealing the most hated bands on Facebook or encouraging people to bash their artists. Besides, any artist to get high on a list of most hated artists must be doing something right for people to have even heard of them (see: Nickelback).

[cwilbur]

The "hate list" idea is actually serious? I don't think labels would appreciate Harmonix revealing the most hated bands on Facebook

Good point, but there's no reason the hate list has to function exactly like the "like" list. Keep its rankings private if you want, the point is for HMX to get a little more info and have a way to throw a cookie to people who hate something they release (which has really been known to get some people on this forum bent out of shape in the past). They can accomplish both those goals without "revealing the most hated bands on Facebook" (as much fun as that reveal would be, IMHO )

[SuperTigerMario]

cwilbur writes the longest responses in the history of the forum. Seriously every response is at least a paragraph if not a 5 paragraph essay complete with intro, body, and conclusion.

[Bront20]

And if you're a casual user who doesn't sink a ton of thought into how the system works at all you're going to fill the hate list with, well, bands you hate, and you probably outnumber the "game it to gimme points" user 100 to 1 or so, so from HMX's data gathering point of view the gaming the system user doesn't matter. Heck, that "hate" data is possibly as valuable to HMX than "like" data, because if they release some band that's way up the "like" list only to discover that a large number of users would have had that band on a hate list if one was available*, and the DLC sells far less than the wildly high "like" number would have indicated, well, that's probably something they'd want to know before making the investment in licensing and development.

Great selling DLC artists that were on "Hate Lists":
- Nickleback
- Lady Gaga
- Linkin Park

LP and Gaga are some of the top selling DLC of 2011 and 2010 respectively, as is Nickleback.

Hate lists are useless, because the issue isn't to find out "who won't buy our product", it's to find out "Who will buy our product". It's like giving a survay about what you like about cell phones to someone who will never buy one.

You don't ask a vegan which hamburger he prefers, because as a marketer, you don't care, because he's not a potential client.

Let's just use Nickleback as an example. There are websites dedicated to folks who hate them, and you see jokes thrown around here regularly about them being the worst band on the platform. Yet, they've sold very well as DLC for both packs. Had HMX said "nope, folks hate them, we can't release that", well, they'd have likely walked away from a ton of money.

Heck, Apple is a company that has a huge hater following, and continues to do so, yet that doesn't stop them from releasing huge selling products. Many of their ideas have been rejected and dumb and hated upon (iPod, iPhone, iPad all got some criticism initially and were products that many other companies passed over the idea to make already). How's that working out for them?

Hate data can be fun for the users, but it's not useful marketing info when dealing with something artistic like music where tastes can vary greatly.

[GNFfhqwhgads]

cwilbur writes the longest responses in the history of the forum. Seriously every response is at least a paragraph if not a 5 paragraph essay complete with intro, body, and conclusion.

And it's never anything besides 'reward people for things they don't like'.