What Have We Learned From MaGMML1 and 2?
Author Message
Crazy_Hand Offline

Posts: 37
Threads: 3
Joined: Apr 2018
Reputation: 0
RE: What Have We Learned From MaGMML1 and 2?
(13-06-2018, 06:55 PM)Flashman85 Wrote: I get where you're coming from, but I'd argue that the rarity of collectibles is up for debate (eg, look at how much there is to collect in the Battle Network and Zero/ZX games), and that Noble Nickel placement fits right in with item and enemy placement as a component of the level design abilities being tested. If letting amateur level makers place Noble Nickels was a terrible idea, then letting them make a level in the first place was a terrible idea. I don't think Noble Nickels are any more inherently problematic than giving contestants access to Apache Joes and Quick Man lasers—yes, these things can utterly ruin a level if used improperly, but that can be said of practically any asset.

To your point, we do have a guide in the form of the MaGMML wiki, and Crystal Man gives in-game hints about where to find the Nickels you haven't collected yet. I'm fairly certain MaGMML3 will handle level skipping a bit differently, so hopefully this will be less of an issue next time.

Good point about Crystal Man and the MaGMML wiki. I think Crystal Man MIGHT be better replaced with an in-game guide that costs 9999 Bolts, but is more of a direct help though, some Nickels were a pain even WITH the map.

(13-06-2018, 06:55 PM)Flashman85 Wrote: The rating rubric had a lot to do with the wonky tier placement, and the tier designations are an arbitrary way of breaking up the contest into digestible chunks. For MaGMML3, we're re-examining the rubric to ensure everything is weighted appropriately. I've actually proposed devoting a point or two to the level name properly setting expectations for the level.

Custom graphics and music were never part of the rubric, so it's a bit unfair to give an advantage to level designers for them—and if we award that kind of novelty in aesthetics, we need to reward people for EVERYTHING new (eg, custom enemies and gimmicks)...at which point, why would anyone bother using what's in the comprehensive devkit if it won't earn them maximum points?

At the end of the day, the contest is a test of one's ability to organize different elements into a fun, cohesive level. The specific ingredients shouldn't matter; it's what the level does with them that counts.

Besides, custom assets vary in quality; does Taco Man (sorry to throw you under the bus, Taco Man) really deserve the same amount of bonus points as Quarantine Woman and Boil Man for its custom music? Beyond that, it's impossible to consistently identify which assets are truly custom assets, and which are ripped from some obscure game the judges have never heard of. Consider how unbalanced the scores for Mount Sabre would have been.

That's very good to hear. To your point about Taco Man, I do agree. Maybe have custom graphics and sound each give 0-2. 0 for things like sound that breaks your speakers and plain colors/gradients, 1 for bad stuff clearly made in good faith, and 2 for stuff that is at half-decent.

It's just a thought of course, but it would avert the problem.

(13-06-2018, 06:55 PM)Flashman85 Wrote: Again, this all goes back to the rubric. Beneath Sand and Rock is arguably most memorable for its music and bike gimmick, but those only account for a fraction of the total score. In fact, there were complaints about both the music and the bike section in the judge comments, so copying a DKC track and slapping on a bike gimmick weren't enough to reach Tier 10. The level got points for things like the boss fight (which is solid), all the challenges prior to the bike section (which are sufficiently creative and generally well assembled), and the graphics (which are polished enough to belong in a high tier). The judges gave the level several perfect or near-perfect scores across the board, not just where the music and bike section were most relevant.

To some degree, it sounds like you've got different criteria for what qualifies as a good Mega Man level, and that's fine. I think it's the rubric, more than the tier breakdown, that's the main issue here. I agree with you that there are several levels that don't feel like "Tier [whatever tier they're in] quality" to me, but when I look at the score breakdown, the overall placement makes more sense. Judge comments go a long way in justifying scores I disagree with, too. The MaGMML3 rubric currently under consideration has more granularity than the one used for MaGMML2, which should help smooth out some of the weirder score variations seen here.

That's fair enough, considering your points about the Rubric above. At any rate, thanks for reading and considering my feedback.
(This post was last modified: 13-06-2018, 07:37 PM by Crazy_Hand.)
13-06-2018, 07:36 PM
Find Reply

Messages In This Thread
RE: What Have We Learned From MaGMML1 and 2? - by Crazy_Hand - 13-06-2018, 07:36 PM

Forum Jump:

Users browsing this thread:
2 Guest(s)

Return to Top