Magic-League.com Forum Index Magic-League.com
Forums of Magic-League: Free Online tcg playing; casual or tournament play.
 
 FAQFAQ   SearchSearch   MemberlistMemberlist   UsergroupsUsergroups   RegisterRegister 
 ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in 

Idea how to improve Limited Simulation



 
Reply to topic    Magic-League.com Forum Index -> Limited
Author Message
Burton911



Joined: 09 Jun 2007
Posts: 172

PostPosted: Sat Aug 15, 2009 2:40 pm    Post subject: Idea how to improve Limited Simulation Reply with quote

I just had an idea concerning sealed/draft simulation that i want to share.

Almost everyone of us has used a draft/sealed simulation tool, but always there is that little line somewhere that reads something like: “The boosters may be not like the boosters in the RL”. A German Magic Author wrote some Blogentries and even an entire Article about this Topic and after I read the Article I thought how that problem with the random generated Boosters could be solved. Then I had the following idea a few minutes ago.

What if we just collect the Information of like 300 (or maybe more 300 is just a random Number that sounds god in here Smile Boosters and put that whole thing into a program which generates Sealed Pools and Draft Sets out of the Pool of 300 Boosters. This may sounds weird in the First Place, but if you think about it, everyone gets 5 Boosters at the Prerealeses and if at least 60 People would make Photos of their Boosters after cracking them we would have the 300 maybe 1 week after the Prerealese. Apart from that some People may could write down the Boosters when the open Displays, or you could just take a Digital Camera to your local Draft and beg people to make a photo of their Boosters (u probably should give the camera to the shopkeeper till the end of the draft then, cause otherwise people could say that you abuse the information Wink.
In that way we could maybe even break the Printruns only like 3 to 4 weeks after the release.

Then someone has to write a little program which generates the Pool (and hopefully puts it into MWS Format :> ), this shouldn’t be hard, it only requires a database and a Program which works with the information, I think I could do that but only in C as I missed the Java Part in school, and my teacher thought that random C Programms are more important than teaching us the database thing Sad .

If some people like the idea and are willing to demand the information, we should probably post this on mtgsalavation since there are a lot of people who maybe support such a project.

Sorry for my bad English but I suck at English writing, I even suck at german writing which used to be my First language for 19 Years now Very Happy

If there are any Questions fell free to ask them (my passive English isnt that bad at all Wink

B_911
Back to top
Desensitizer



Joined: 27 Jul 2009
Posts: 6

PostPosted: Sat Aug 15, 2009 7:11 pm    Post subject: Reply with quote

We need to have MTGO players save the first pack they open. Like completely. Because that's identical to IRL boosters. Then we would see how it works. You cannot get only 300 boosters that number wouldn't be nearly large enough for it to be random, especially due to mythic rares.
Back to top
alvaro202



Joined: 05 Apr 2009
Posts: 88

PostPosted: Sun Aug 16, 2009 12:51 am    Post subject: Reply with quote

This is really and truly epically stupid. The thought of compiling a database of BOOSTER packs just to randomly select 1 of many to simulate a booster draft... there are so many solutions to this supposed problem that should have been come to before such an idea was announced.

1st, the reason the random number generator is incorrect is because it is a pseudo-random number generator. Nearly all computer programs, unless specifically designed otherwise, use pseudo random number generation because it is faster and easier to code with. There is no such thing as a program language that doesn't have a pseudo-random number generator.

The down fall of this type of generator, is the randomized output is incorrect. If your program is generating information off of a random number generator to a rhythm, your random numbers will reflect that rhythm and give you a result that is not truly random.

To "solve" this problem a generation unite that is connected to a real random number generator must be used. Only when the randomized cards are truly mathematically randomized will you get good results to draft with.

Some information about true random numbers is located at RANDOM.ORG . True randomization inside of a coding language is difficult to achieve, an outside source other than the PC the program is running on must give the program a truly chaotic source to derive its random number from. When the chaotic source is the internal clock, and then a string of generated numbers is used in comparison to the clock, you are bound to have poor randomized effects.

If you are really interested in solving this problem, contact the programmers of MWS, and have them look into implementing C++ libraries from LavaRnd.org, or to read up on truly random numbers from RANDOM.org and explain the importance of true randomization in simulating a draft environment.
Back to top
Burton911



Joined: 09 Jun 2007
Posts: 172

PostPosted: Sun Aug 16, 2009 6:28 am    Post subject: Reply with quote

Im in a Information Technology School, (dont know how to describe it cause this form of school cause this only exists in Austria in that way) so yes i know that building a real Random Generator is nearly impossible and moreover not desireable as u don't want to have unknown elements in a Programm/Machine.

But i dont think u understand the Problem.
The Problem is that there is a Printrun behind Boosterpacks, and u cant generate Boosters that are equal to original Boosterpacks.
And if the Programm picks 5/3 Boosters out of 300~ it should be enought to speak of Random.
Maybe u think that i want to generate the Boosters at random.
But thats not the idea, the idea is to collect the information of enough Boosters to generate Sealed Pools out of Boosters packs that exist in the RL.
And even if that don't works, atlest the collected information could be used to crack the Printruns.

And the rares and mytics could be choosen Random cause i don't think it would really worth it to crack the pattern behind the Rare Distribution (and maybe there isnt even one).
Back to top
alvaro202



Joined: 05 Apr 2009
Posts: 88

PostPosted: Sun Aug 16, 2009 11:33 am    Post subject: Reply with quote

A simulation of something is just that, a simulation. When learning to fly, you can't even use the best of flight simulation technology to get the hours needed to get your license. The reason is because it's a simulation and they need REAL life application. You may believe this is completely off base from the topic, but when you consider what your trying to achieve it really is not. You are simulating a real life booster draft with a game engine. Whether you use random generation of the booster packs, develop a database to pull booster packs from, or create an algorithm to "properly pack" your booster packs you will still come to the same conclusion. That the simulation will still be a simulation.

Maybe you can find a bunch of people from Magic League (that are playing for free generally for a reason) who are willing to spend money on actual booster packs and then properly record the information en-mass and give it to you to process into a database. This is not really viable as incentives for a player is not inline with what would be necessary to get your results.

But hey if you only want 300 booster packs, which isn't really a statistically large enough pool to give you the real life simulation your desiring, it shouldn't be THAT hard. It's like 10 booster boxes of booster packs that are ripped and then recorded, then sent to you. That is possible, maybe you and a group of 10 friends can get a booster box for each set that comes out and proceed to create this database for the booster drafts. It will truly come down to how much effort you wish to put into this more elite form of randomizing booster packs, and whether or not it would just be better to create a new and improved algorithm to develop randomized booster packs.

Also, even though the packs may be different in the game, the same idea's and thought goes into picking your cards whether or not your drafting with the current setup or in RL.
Back to top
DaWorm



Joined: 28 Jul 2009
Posts: 20

PostPosted: Sun Aug 16, 2009 12:10 pm    Post subject: Reply with quote

The priority should be:

1.) ML needs a working drafting system like the sealed system hosted on this site. Other leagues had/have them too.
2.) ML need real draft Minis
3.) ML needs additional to it's single Elimination tournaments an option to implent swiss tournaments with top 8 drafts
4.) Optimize the cards you get.

Simple as that.
Back to top
CrushU
Level 1 Judge


Joined: 31 Aug 2004
Posts: 99

PostPosted: Mon Aug 17, 2009 11:28 am    Post subject: Reply with quote

I have never understood the fascination certain coders have with 'zomg, it's only pseudo-random!!!111'

The current best pseudo-RNG generates numbers that only repeat after you recurse it something like 10^7 times. If this isn't good enough for your program, then yes, find something else.

I somehow don't think we need anything better than that for drafting.

First of all, you'll never get a truly random number generator from a computer program, because all generators, given a specific input, will return a specific output. (that's the point behind 'seeds') Secondly, get a random occurence or generator, and list its output beside the computer RNG. If you can tell the difference, reliably, then I'll listen to you.

As for acquiring images/lists of boosters... It could work. Maybe. It's a brute force method of solving the problem, where the current method of 'cracking' the printruns and using them is the more 'elegant' solution... and prone to hiccups.

I suspect the ideal is a mixture between the two, but I can't think of what it could be at the moment. Regardless, it is not the RNG's fault. (Er, assuming it's coded right. There are some poor RNGs that recurse after 100~ calls.)
Back to top
StringTheory



Joined: 28 Nov 2004
Posts: 6

PostPosted: Wed Aug 19, 2009 10:24 am    Post subject: Reply with quote

Why not just stick to printruns? The hardest bit is assembling the uncommon print runs, and brute force doesn't help there. Anyone could probably assemble the common print runs from 2-3 boxes of M10.
Back to top
Display posts from previous:   
Reply to topic    Magic-League.com Forum Index -> Limited All times are GMT - 7 Hours
Page 1 of 1

 


Powered by phpBB © 2001, 2005 phpBB Group

All content on this page may not be reproduced without consent of Magic-League Directors.
Magic the Gathering is TM and copyright Wizards of the Coast, Inc, a subsidiary of Hasbro, Inc. All rights reserved.


About Us | Contact Us | Privacy Policy