SGDQ 2015 Game Submissions

2015-04-23 06:29:33 am

station

Here's a google doc with all GDQ games to date.
https://docs.google.com/spreadsheets/d/tlIFz0O8rUzjim2_yIJsIvw/htmlview

Got it from this thread
https://forum.speeddemosarchive.com/post/games_done_quick_marathons_full_game_list_now_with_sgdq2014_final_list.html

TheMG2

2015-04-23 06:31:19 am

The last time I looked at that list, there were a bunch of mistakes.

Cronikeys

2015-04-23 08:14:15 am

I like the ideas presented by Naegleria and pvtcb. I'm willing to create an automated system for accomplishing that as well as collate existing data.

Only interested in doing it if it's in some sort of official capacity, so I'm basically just saying I volunteer to help with such a project if the GDQ staff decides to pursue it.

Antilles58

2015-04-23 09:16:59 am

#FailFish

In general, I think the consensus that seems to be emerging here is that it seems like you guys are shorthanded. From an observer's perspective, it seems like you've gotten great volunteer help during the actual marathons, but if you could use more during the earlier parts of the process too, I would also love to be able to help (I'm a software developer, as well).

I think most people understand and accept the reasons for the delay - even if we weren't exactly thrilled with how it was all handled. Obviously it's too late for this year, but I feel comfortable in speaking for the group in saying that if you need more help, definitely ask us for it! We'll be happy to chip in. Smiley

Cool Matty

Edit history:

Cool Matty: 2015-04-23 11:14:49 am

2015-04-23 11:09:27 am

Crawlathon WR, get down on my level.

There's no algorithm that's going to decide the games list. You forget, romscout and his "advisors" (:P) have to not only judge the game, but who is running it, and what category is submitted. That's not something you can just use basic historical data on. We already keep the database of previously selected/unselected runs, and I believe he does check that information. There's just way more variables to it, and really the only surefire way to judge if a run is any good is to watch some of it.

johannhowitzer

2015-04-23 11:28:56 am

Viking it up

Not to mention, part of what makes a good marathon is tying things together. That's why we have stuff like "megaman block" or "classic NES block" or "WTF Japan" block. The marathon should feel somewhat like a cohesive whole, not just 150 hours of runs tacked end to end.

The computers at the corporate office of the company I work for can't even get ordering right half the time, and that's just basic replenishment based on balance-on-hand. I can't imagine how badly a computerized GDQ selection algorithm would screw up. It might even put The Demon Rush on the schedule...

Cronikeys

2015-04-23 12:45:26 pm

Quote from Cool Matty:

There's no algorithm that's going to decide the games list.

I was talking about automated "run information" that would then be used by romscout and his advisers to speed things up, not automating the actual selection. It would keep track of more than just whether it's been at a marathon before, including things pvtcb mentioned in his post. That coupled with the interactivity of what naegleria suggested (locking / early announcements) seems like a more organized and systematic way of game selection.

This community brainstorming is only happening because of the failure of the existing system (repeatedly missing deadlines).

Crak Atak

2015-04-23 12:45:49 pm

Hax I used all of my strengthz

Correct me if I am wrong. One of the core arguments here was to have data to assist in decision making (which they already do apparently), not to ultimately decide games that get in.

Quote from Cool Matty:

There's just way more variables to it, and really the only surefire way to judge if a run is any good is to watch some of it.

I doubt watching some of a run is the "only" surefire way to judge a run. What if the parts they watch are the only good/bad parts of the whole run?

Cool Matty

Edit history:

Cool Matty: 2015-04-23 01:04:20 pm

2015-04-23 12:58:49 pm

Crawlathon WR, get down on my level.

Quote from Cronikeys:

I was talking about automated "run information" that would then be used by romscout and his advisers to speed things up, not automating the actual selection. It would keep track of more than just whether it's been at a marathon before, including things pvtcb mentioned in his post. That coupled with the interactivity of what naegleria suggested (locking / early announcements) seems like a more organized and systematic way of game selection.

I know what you meant, and it only attempts to solve a small part of a very complex decision-making process. PvtCb basically described romscout's entire job in deciding games, and there's no way to automate something like that. There are patterns but they're not the sort of thing you're going to write a script to manage. There's no nice clean dataset you can pull that information from, and there's definitely no way you could keep it accurate. In the end, you'd swap "deciding the games" for "deciding the data that decides the games" which is only making it worse, not better.

I can't imagine the level of anger people would have if their run was rejected because a script said their run was garbage (particularly if it was assumed on incorrect data/pretenses).

Anyway, the real solution to this is not to set deadlines for things that cannot be accurately measured. Particularly when you have things like a full-time job to juggle on top of it. We're learning, this is just one mistake among many problems we've resolved over the years. It's only going to get better.

[quote="Cronikeys"]
I was talking about automated "run information" that would then be [i]used[/i] by romscout and his advisers to speed things up, not automating the actual selection. It would keep track of more than just whether it's been at a marathon before, including things pvtcb mentioned in his post. That coupled with the interactivity of what naegleria suggested (locking / early announcements) seems like a more organized and systematic way of game selection.[/quote]

I know what you meant, and it only attempts to solve a small part of a very complex decision-making process. PvtCb basically described romscout's entire job in deciding games, and there's no way to automate something like that. There are patterns but they're not the sort of thing you're going to write a script to manage. There's no nice clean dataset you can pull that information from, and there's definitely no way you could keep it accurate. In the end, you'd swap "deciding the games" for "deciding the data that decides the games" which is only making it worse, not better.

I can't imagine the level of anger people would have if their run was rejected because a script said their run was garbage (particularly if it was assumed on incorrect data/pretenses).

Anyway, the real solution to this is not to set deadlines for things that cannot be accurately measured. Particularly when you have things like a full-time job to juggle on top of it. We're learning, this is just one mistake among many problems we've resolved over the years. It's only going to get better.

Cronikeys

2015-04-23 01:03:53 pm

Wouldn't the same data available to Romscout also be available for the proposed database? I don't understand why you would write it off as unmanageable or unclean. It would only supplement and not supplant.

Cool Matty

2015-04-23 01:17:36 pm

Crawlathon WR, get down on my level.

Quote from Cronikeys:

Wouldn't the same data available to Romscout also be available for the proposed database? I don't understand why you would write it off as unmanageable or unclean. It would only supplement and not supplant.

Only if you were to use the info I already mentioned (previous games), which we don't need anything else for. There's no qualitative database of runner suckiness or new skips. Tongue

Cronikeys

2015-04-23 01:25:33 pm

Quote from Cool Matty:

Only if you were to use the info I already mentioned (previous games), which we don't need anything else for. There's no qualitative database of runner suckiness or new skips. Tongue

Well I guess that begs the question of how is it organized now? Does Romscout just have a Google Doc or something to communicate with the other advisers? What's being proposed is a better organizational tool.

Either way, this seems better suited for another thread (perhaps post-sgdq feedback thread) or simply discussed internally

Cool Matty

2015-04-23 01:26:28 pm

Crawlathon WR, get down on my level.

We will reach out to people in the future if we need assistance on this. You can call this resolved for now.

Naegleria

(user is banned)

2015-04-23 04:41:40 pm

The discussion between Cool Matty and Cronikeys was about the Organizers using these tools, but I was more interested in creating them for the people who want to submit.

We get 5 games to submit, and these tools could let us know which games have a better shot, and which ones dont.

For example, if Super Metroid has been in every marathon, then someone who is good at the game can think of that as a really good submission. If someone wants to submit an RPG, then they can look at the history of the submissions for that game and other RPGs that got in over it.

This kind of data/tool would inform mainly the people that want to submit games.

johannhowitzer

2015-04-23 05:25:51 pm

Viking it up

I'm against that, Naegleria, because I would think that would discourage a lot of people from submitting runs they felt would be long shots. We need people submitting things that are long shots! The marathon needs variety to thrive, and much of that variety is going to come from new games or games nobody thought to submit before.

RPGs are a special case because of their length; many RPGs would get in otherwise, there are quite a few with some really interesting tech beyond just bland menu navigation to get around boss loopholes. Other than RPGs, we really need people to be willing to take risks.

Naegleria

(user is banned)

2015-04-23 05:53:23 pm

In that case, then a list of all accepted games, and why they were accepted. This would include the long-shot games that got in.

kirbymastah

Edit history:

kirbymastah: 2015-04-23 10:33:47 pm

kirbymastah: 2015-04-23 10:33:21 pm

2015-04-23 10:29:42 pm

<(^_^)>

I don't think it should really matter (note: not saying it would be useless) since it would only be applicable if you have more than 5 games / 10 categories to submit anyways, and most serious runners shouldn't really be doing that in the first place. It'd be helpful on paper, but not really that useful in practice (more informative than useful)

Naegleria

(user is banned)

2015-04-24 02:14:10 am

Probably true, but I like information, so I'll support the idea on those grounds alone

JTBeezy

2015-04-24 06:21:59 pm

words on words on words

If all goes according to schedule, what time can we expect the game list on Monday?

Shaddex

2015-04-26 07:12:54 pm

Cuts are beginning/in progress right now. https://gamesdonequick.com/submission/all

Antilles58

2015-04-26 08:07:31 pm

#FailFish

Error 502 FTW Wink

TheSoundDefense

2015-04-26 08:52:18 pm

Final game list has been posted!

theruseman

2015-04-26 08:52:44 pm

Rest in peace pac-man world 2 and uncharted 2

Antilles58

2015-04-26 08:57:19 pm

#FailFish

Ok which way to the salt mines entrance?! Wink

Now BRING on those volunteer sign ups!!!! Smiley

CosmykTheDolfyn

2015-04-26 09:06:48 pm

NowOwnsAFreaking Plane

Ok, I hope this serious question does not get lost in a sea of salt, but...
The comment for my accepted Kirby's Avalanche run states "Bid war of Kirby's Avalanche vs. Dr. Robotnik's Mean Bean Machine". However, Dr. Robotnik's Mean Bean Machine itself is rejected as a run.
I'm assuming it's still a bid war between the two titles due to the status comment on Kirby. Am I correct in this assumption, or is Avalanche the only accepted run out of these two identical games?