NationStates Jolt Archive


Official AIDBCA Discussion Thread

Bluth Corporation
22-12-2008, 03:15
Alliance Star and I had a discussion over on WikiStates about this...I'm totally open to others hosting.

The only practical problem is that the scorination system I use is a rather jury-rigged assemblage of PHP scripts and mySQL databases that I run on my local Apache installation, which is not publicly accessible (and won't ever be--I just keep it for local apps, and local testing of websites I plan on putting up on actual servers).

Using this setup as-is would require a rather thorough knowledge of both PHP and mySQL, as the bits that generate the schedules have the data hard-coded into them. I'd be more than happy to send someone a copy of my present system and explain it to them, but you'd have to have the knowledge to make it work.

The other alternative is to develop your own scorination system, which is fine.

Sometime in the future, I plan on cleaning up the scorination system and putting it up on the same server I use to host WikiStates, so that other people can use the same system I use to run their own domestic circuits if they so desire, or to host the World Tour. However, that particular project has a bunch of more pressing projects ahead of it, so it's at best several months away. In the meantime, if someone else wants to host it'd require one of the above two alternatives.

If someone else is willing to take the trouble, I'd be more than happy to dump it off on that person :D

In a nutshell, here's the high-level description of how my scorination system works:
Schedules are generated randomly...I give it a list of all the stadiums that we're going to be using (the basic rule is that on each of the ten nights, we have the minimum number of shows necessary to ensure that no show has more than 8 particpants) and it randomizes that list and spits it back out...and then if we have two shows a night then the first two shows go the first night, the next two the next night, etc.---same thing for one, three, four or however many shows a night we may actually need.

I pick the stadiums we use by first including all the home show stadiums, and then I fill the remainder that are needed by picking from the "extra stadiums" based on seniority, so that right now Zwangzug and I are taken first since we've been around the longest, then Taeshan, then Alliance Star, etc....and once I get all the way through one extra stadium from each country, if I need more I start back at the top.

After the schedule has been made, the initial scores are chosen at random, separately for each caption. The caption that the corps is strongest in is scored from the range 55-65, second-strongest from the range 53-63, then 51-61, 49-59, 47-57. This means that there is a bias in the initial scores towards your strongest caption and against the weakest caption, but in any particular situation there's no guarantee that your strongest caption will have the highest score and the weakest will have the lowest: the weakest could end up with, say, 56.9 and the strongest with 55.3.

After this, each score is multiplied by a number taken from the number of corps that particular corps defeated in the previous corps. Corps that did not participate in the previous tour have no multiplier, and their initial scores are set. This "seeding multiplier" is the same for each caption score, and is equal to 1 + (number of corps defeated in the last tour/100). So, for instance, the Brigadiers' seeding multiplier this time is 1.13, Legion of the Red Shield's is 1.12, Hockey Playin' Nuts's is 1.11, and so on. The numbers that would have been assigned to corps that, for whatever reason, are not participating in this year's tour are skipped; so there was, for example, no corps with a multiplier of 1.08 (would have gone to Lasersound) or 1.10 (The Silver Knights).

Seeding multipliers and caption strengths are only relevant for calculating the scores for the corps's first show of the tour. After that, scores for each caption are chosen at random from a formula that is designed so that a given caption's scores will USUALLY increase each night, but there is a chance that that score will slip. And as the score gets higher and higher, the greater the odds are of the score slipping and the lesser the odds of the score increasing.

Hope that helps!
Alliance Star
22-12-2008, 05:05
Wow! I never knew about the various seeding numbers, or how caption strengths were rated! VERY COOL! ^_^

You really put a lot of work into this. I'd be interested in hosting, in time. I have a few things IRL to iron out, but whenever it's publicly available, I'd be glad to host one or two of the tours!

Thanks for explaining the initial scoring however. Never knew of that.

Thanks so much for putting so much into this!
Bluth Corporation
29-01-2009, 23:38
Well, the latest World Tour is over...congrats to Taeshan.

For the next tour, I'm going to take the first step in making the scorination system publicly-usable by putting them on an Internet server (the same one I use to host WikiStates, incidentally) so that I can run scorination from school as well as home. They'll still be in a rather bare-bones format, so they won't be in a form that's usable by anyone but me since it'll still require a pretty intimate knowledge of the system to use it, but it's progress. I'm hoping to be able to start massaging them into shape over my spring break in March, but that depends in large part on what else I have to do (I'm afraid I'll wind up doing the arranging for this fall's program for the high school marching band I teach) during that time.

That said, if you want to develop your own scorinator and host yourself before I get mine publicly released, you're more than welcome.

Let's try and see if we can't get participation back up to where it was! It'd be nice to get to the point where we could do a top-17 for Semis/top-12 for Finals like DCI does.
Bluth Corporation
01-03-2009, 03:10
First off--any suggestions on changes to make for the next World Tour? The tiebreaker system seems a little iffy, considering the results from this tour (it didn't affect making Quarterfinals this time, but it did affect the placements of the corps that didn't make QF). One thought is to use the score from the last Competition Night as the first tiebreaker when deciding who does and does not qualify for Quarterfinals, and only use cumulative score if the ties aren't broken then. Ideas?

My spring break is the second week of March, and my backpacking trip I had planned got cancelled so I'd like to use that week cleaning up my scorinator to make it usable by others who host, perhaps as early as World Tour IX or X (and you'd also be able to use it if you wanted to run a domestic circuit too). I'll keep everyone posted on how that's coming along. And as always, if you want to host before then and you have your own scorinator, that'd be great too.

Last, I've finally got a way figured out to incorporate RP bonus. I'm going to write that and test it, and hopefully it'll be ready for the next tour.
Taeshan
01-03-2009, 06:44
I find it kinda wierd that after winning a championship my winner in the next season finishes third and is marred from the thing, and this time it was my only one that had been in every one except the first witch i didn't enter
Bluth Corporation
01-03-2009, 17:56
Yeah, I noticed that too. It's kind of like when the Scouts won DCI in '88 then dropped to 7th in '89, or the Anaheim Kingsmen winning in '72 and then dropping to 19th over the next three years.

I really like requiring some movement among the bottom--especially when we start doing RP bonus (which will hopefully require heavier RP), it will make for some good storytelling opportunities as corps fight to remain eligible for the World Tour. Still, I can see where including a "Past Champions' Provisional" or something like that might be a good idea, where if a corps was a champion within the las X number of tours and is in a situation where it would otherwise be bumped, it can stay and the next-lowest corps can go. Or something.

Thoughts?
Taeshan
01-03-2009, 19:15
that would be cool because they just seem to win and then fall the next year or two
Bluth Corporation
01-03-2009, 20:34
OK, what should be the limit on how far back it goes? I don't like the idea of making it indefinite, because then a corps can just absolutely suck mastodon balls but still stay on the World Tour because it won a single championship a hundred tours ago, or whatever.

3 sounds reasonable to me...anyone else have any thoughts?
Zwangzug
01-03-2009, 21:41
Hmm, I'm not convinced it's necessary. Does make for some interesting storytelling...though I guess I haven't been up to much of that. Still, I don't have any strong feelings. Better a short than long thing, 3 tours seems okay. First overall to last among corps from one country in one tour is pretty..."impressive," for sufficiently negative definitions of impressive. :p
Bluth Corporation
07-03-2009, 00:12
Any other thoughts?