Is it always possible to get every instrument to max multiplier?

[Epsilon82]

Rock Band 3 had something like 5 patches, the last of which was more than a year after release.

Sure, but that doesn't mean that a 6th patch or 7th patch wouldn't still have been an extra expense, especially at a time that they weren't exactly swimming in cash. They probably just had a priority list and weren't able to get around to the solo bonus issue. It's regrettable, certainly.

[Meat-Popsicle]

Well, the reason they didn't require a patch to tweak the coin economy was that everything related to it is handled server side

That sounds plausible. So maybe an actual patch would still be complicated and expensive.

5 patches were done on RB3, really? Wow, I guess I don't notice them when they are small files that download quickly. I remember at least 2 that happened early on, which is completely typical these days. There are always tweaks that have been figured out in the time since the disc got printed, and then unexpected issues that come up in the first few hours after release. I never wind up remembering what was fixed or changed from old versions.

When was the 2nd-to-last patch, surely it came out in the first month, when MTV still owned the title? And then that last patch came out much later, and it was totally unexpected, a real bonus. It gave us the "hide songs rated 1-lighter" option, and allowed 2 more songs to be imported from a previous game.

[SirDavidTLynch]

But they didn't do anything about the star cutoffs despite numerous opportunities. Personally, I think that songs without solos are too easy to gold star.

[thatmarkguy]

But they didn't do anything about the star cutoffs despite numerous opportunities. Personally, I think that songs without solos are too easy to gold star.

I agree with this, but I think there was a happy medium to be struck between "ignore the solo bonus" and "add the FC solo bonus to the GS threshold". Something like "add the 95% solo bonus to the GS threshold".

[Epsilon82]

I agree with this, but I think there was a happy medium to be struck between "ignore the solo bonus" and "add the FC solo bonus to the GS threshold". Something like "add the 95% solo bonus to the GS threshold".

Yeah, that's exactly what I was thinking. Or even 90%. The primary issue was that the difference between 100% bonus and missing one note could be tens of thousands of points. Maybe they just didn't think it through very well and assumed that the benefit of getting extra overdrive from solo unison bonuses would offset that. In some cases, it probably does, but in songs with longer solos, it doesn't even come close.

[Epsilon82]

But back on the topic at hand, I was thinking about this some more this morning, and I was thinking that one way Harmonix could potentially neutralize the complexity of the disproportionately punitive lower max multiplier caps at earlier stages of the song would be to simply handle those cases differently.

For instance, we already know that if the game detects that there aren't enough notes to raise the cap by one level, that it simply doesn't count it as part of the equation for determining the highest multiplier for a given checkpoint. So if you are at the highest multiplier on 3 instruments but vocals only has 3 notes, for instance, you'll still get a +3 bump at the checkpoint even though vocals wasn't raised at all. EDIT: the exception seems to be in rare cases when there are no notes at all before a checkpoint, in which case it seems to give +0 instead of +3.

What if they simply extended that logic to situations where the maximum multiplier on a given lane is only 1 or even 2? So instead of simply taking the lowest number of multipliers as long as at least one is available for every lane, they could just rather easily determine what the highest possible multiplier is for a given lane and go ahead and give the +3 bonus as long as all lanes are maxed out to the highest possible level?

It just seems needlessly punitive to hit a player with a low cap boost simply because the structure of the song itself is such that even "perfect play" isn't good enough.

[thatmarkguy]

I think a more community-driven method of determining GS threshold, rather than game-engine-driven, might be the best solution (now that we have a game that, by rule, reads data from the online servers at song start and end). Perhaps, once any song has more than 100 logged scores, if it has less than 2% of logged scores being GSed, then re-strike the GS threshold to be the lowest score out of the top 2% of scores. When the song begins the game would fetch the 'new GS threshold' from the server and use that as the "Duke" score-to-beat.

So no song would be unattainable-for-GS as a result of engine shortcomings - at absolute worst, GS would be attainable by the top 2% of players.

(adjust '2%' to taste.)

[hodayathink]

I think a more community-driven method of determining GS threshold, rather than game-engine-driven, might be the best solution (now that we have a game that, by rule, reads data from the online servers at song start and end). Perhaps, once any song has more than 100 logged scores, if it has less than 2% of logged scores being GSed, then re-strike the GS threshold to be the lowest score out of the top 2% of scores. When the song begins the game would fetch the 'new GS threshold' from the server and use that as the "Duke" score-to-beat.

So no song would be unattainable-for-GS as a result of engine shortcomings - at absolute worst, GS would be attainable by the top 2% of players.

(adjust '2%' to taste.)

A system like that would be too easy to game, as all you would need is a bunch of people playing the song as soon as it comes out and intentionally scoring low to drive down the top whatever percentage.

[Epsilon82]

A system like that would be too easy to game, as all you would need is a bunch of people playing the song as soon as it comes out and intentionally scoring low to drive down the top whatever percentage.

There are other issues with it, although it certainly is an interesting idea.

First of all, I suspect that a change like this would still require a software patch, unless the offline mode doesn't track star scores at all. I highly doubt that the game is calculating star score cutoffs server side, so if they were to do that, it would likely create even more load on their servers.

Speaking of that, I'm very inclined against supporting anything that makes this game rely any more on the server side than it already does, given how poorly the game handles temporary disruptions or timeouts. On top of that, I'm thinking long term here, and I hope that they end up going in the opposite direction, which is to make the game more fully functional off-line with server syncing capabilities. It kind of unnerves me that with things as they currently are if something happens to Harmonix (like going out of business, God forbid) or they just decide to move on and stop supporting Blitz, the game will be crippled beyond belief.

Also, there would be the issue of whether scores would be dynamically adjusted as scores progress or not. If the scores did dynamically adjust (meaning that your Gold-Star run could become a Five-Star run over time), then that would be a significant departure from the way we've all become accustomed to the "accomplishment model", meaning that once you've done something, you've done it and it can't be taken away. But if they didn't dynamically adjust, then you'd have all kinds of weird situations where someone may have a Gold Star credit for a score that wouldn't even be close a month later, or vice versa, which would also be bizarre.

And on top of that, they'd still have to have some kind of base case scenario because you can't necessarily assume that there's a reasonable sample, especially for less trafficked songs.

It's definitely an interesting idea, but for all these reasons I don't think it's the most advisable course.

[thatmarkguy]

A system like that would be too easy to game, as all you would need is a bunch of people playing the song as soon as it comes out and intentionally scoring low to drive down the top whatever percentage.

I understand the theory, but I just don't see it in practice. You're suggesting one hundred people will co-ordinate deliberately bad scores in a song before two people post legitimately good scores.

First of all, I suspect that a change like this would still require a software patch, unless the offline mode doesn't track star scores at all.

I am curious about that myself. I had assumed the powerup costs were baked into the software and that was proven wrong. It would not surprise me at all if the GS value is (or can be overridden by) a value in the start-of-song 'handshake' that goes on between the game and the servers.

I highly doubt that the game is calculating star score cutoffs server side, so if they were to do that, it would likely create even more load on their servers.

It would be one long integer fetched from a database and encapsulated into the data that's already being exchanged between game and server at song start as it is. The extra 'load' would be immeasurably small. Oh, and probably a once-a-day service that scans the song list and does a restrike for songs that meet the restrike criteria (one division-and-compare per row of a ~5000 record database, in off hours, once per day - again, miniscule server-load impact)

If the scores did dynamically adjust (meaning that your Gold-Star run could become a Five-Star run over time), then that would be a significant departure from the way we've all become accustomed to the "accomplishment model", meaning that once you've done something, you've done it and it can't be taken away.

My suggestion was a one-way street. Songs that are seen to have an unusually low rate of GS would have their GS threshold tuned down. Nothing would ever cause a GS threshold to go up. Yes, it's possible your old then-recorded-as-5* score would now be good enough to be a GS score. But the reverse would not be true - you could never have a saved GS score that no longer would qualify as a GS score.