Chaining Research Project
Greetings, Smogon. My Chaining Guide has now been up for almost 5 months, and thus far seems to have been successful. Now, however, I feel it is time to examine some of the methods in the guide, specifically what I define as an "optimal" and "bad" patch formations. After doing some research consisting of reading other guides, experimenting while chaining, and actually listening what people have to say to me, I've decided that my methods need review. I am certain that some of them are correct and relatively sure of the rest. Although I don't believe any of my methods to be incorrect, I feel as though there are more correct methods out there. Finding those methods, as well as testing the ones I have forced you all to accept, is the purpose of this project.
What you can do:
If I intended to do this all by myself I probably wouldn't have gone through the trouble of telling everyone about it. As you might have guessed, I'm looking for experienced chainers to assist me in this project. I'm not withholding any information in my guide, so there are a few who know nearly as much about chaining as I do. If you're one of these people, I'd certainly like your help. If you're not, that's okay too. The baseline tasks I will be asking of you really aren't very difficult, so anyone who knows what they're doing will inevitably be as valuable an asset as they want to be. I want as many people as possible, so don't hesitate to sign up, no matter how much you think you can contribute. The one thing you need to guarantee me is that you will follow your assignment with care, precision, and efficiency. If you can't do this, please do not sign up.
The One Prerequisite:
I lied, there are actually two things you need to guarantee me. The first is mentioned above, the second is this: You must be able to easily recognize the formations you are looking at when chaining. If you have any trouble doing this, you should not participate in this project. You will soon see why having this skill mastered is of the utmost importance.
What you get out of it:
Not much, to be honest. I might offer some of my pokemon, but I'd rather make this about improving the community's knowledge of chaining than getting prizes. You will all get recognition in the guide of course, but other than that the only thing I can really offer you is an encouraging slap on the ass. Just know that you'll be helping to improve our comprehension of chaining.
What your exact assignment will be:
This really depends a lot on how many people sign up, so I can't tell you right off the bat. However, I figure this will be the first people ask, so I'll give you a general sense...
. You will be chaining... as much as you can handle.
. You can request certain tasks if I offer them or if you come up with good ones.
. You will be keeping very precise data on each formation and how many successes and fails it has. More on that in a bit.
Data Collecting and Reporting:
Extremely important!! Please Read!!
. Only Pokemon with 20% catch rates or below can have data taken on them!! No swarms, Static/Magnet Pull, or Starlys!!
. In your data, specify which pokemon you are chaining, separate the data by pokemon (don't combine data of pokemon with different catch rates)!
. Rather than assigning each person a formation, I want everyone to collect data on all 5 formations, since they will all appear for everybody equally (in theory of course).
. Once again, you must keep exact records. It's not enough to say "Formation 3 broke every once in a while," or "Formation 5 it teh pwnzor!!1!eleven!1!"
. Instead of statements, I want data such as "Formation 3 broke 5/26 times" or "Formation 5 did not break in 37 tries." Pretty simple.
. This is why I named the formations, so it would be easy for you to keep record of. I would recommend keeping track by making tally marks on a piece of paper (that you won't lose) or on a spreadsheet/word document/other electronic source. If you EV train in a similar fashion you could record data that way, since it's something you're used to. Bottom line: whatever gets the job done accurately!
. Report the data to me, and only to me. Do not post it in this topic. You can do this via PM if you must, but I'd prefer it through email or AIM, because I hate PM's. I'm not sure when I will want the data from you by, suggestions anyone? I don't wanna rush anyone, especially because my schedule's quite busy as well.
AIM: Wow its Diesel
Email: derrym4790@yahoo.com
What I know:
I am certain that these formations are among the best in existence, and have a very small chance (<2%) of breaking chains.
4 4 4 4 4 4 4 4 4
4 3 3 3 3 3 3 3 4
4 3 2 2 2 2 2 3 4
4 3 2 1 1 1 2 3 4
4 3 2 1 Y 1 2 3 4
4 3 2 1 1 1 2 3 4
4 3 2 2 2 2 2 3 4
4 3 3 3 3 3 3 3 4
4 4 4 4 4 4 4 4 4
The Jesus Formation, works almost 100% of the time.
4 4 4 4 4 4 4 4 4
4 3 3 3 3 3 3 3 4
4 3 2 2 2 2 2 3 4
4 3 2 1 1 1 2 3 4
4 3 2 1 Y 1 2 3 4
4 3 2 1 1 1 2 3 4
4 3 2 2 2 2 2 3 4
4 3 3 3 3 3 3 3 4
4 4 4 4 4 4 4 4 4
The next best thing, almost never fails.
What I want to test:
My Current Buddy Theory:
. Buddies cannot be on an axis or touching an axis.
. Buddies must be a 2- or 3-away.
. Buddies cannot be directly adjacent or diagonal to a 4-away (no touching).
. Buddies must be in line with the 4-away or...
. Buddies must have a coordinate greater than that of a 4-away. This means if a 4-away has the coordinates of (4,2) then the buddy must have the coordinates of (2,3).
Basically this means that buddies are one of the following 3 squares, and enterable 4-aways are the following 4 squares.
4 4 4 4 4 4 4 4 4
4 3 3 3 3 3 3 3 4
4 3 2 2 2 2 2 3 4
4 3 2 1 1 1 2 3 4
4 3 2 1 Y 1 2 3 4
4 3 2 1 1 1 2 3 4
4 3 2 2 2 2 2 3 4
4 3 3 3 3 3 3 3 4
4 4 4 4 4 4 4 4 4
Formations #1, 2, and 3:
I want to test this theory, or rather its limitations, with the following 3 formations. The first one is a formation I have basically accepted as one of the safest formations possible, and doesn't break any of the rules, but I want to confirm its value. The second fails to meet the final criterion, yet I have entered it more frequently of late, and it hasn't broken any chains. The third formation uses a patch touching an axis as a buddy. This one I certainly have my doubts about, but seems to be the most promising of all formations using this type of buddy.
4 4 4 4 4 4 4 4 4 | 4 4 4 4 4 4 4 4 4 | 4 4 4 4 4 4 4 4 4
4 3 3 3 3 3 3 3 4 | 4 3 3 3 3 3 3 3 4 | 4 3 3 3 3 3 3 3 4
4 3 2 2 2 2 2 3 4 | 4 3 2 2 2 2 2 3 4 | 4 3 2 2 2 2 2 3 4
4 3 2 1 1 1 2 3 4 | 4 3 2 1 1 1 2 3 4 | 4 3 2 1 1 1 2 3 4
4 3 2 1 Y 1 2 3 4 | 4 3 2 1 Y 1 2 3 4 | 4 3 2 1 Y 1 2 3 4
4 3 2 1 1 1 2 3 4 | 4 3 2 1 1 1 2 3 4 | 4 3 2 1 1 1 2 3 4
4 3 2 2 2 2 2 3 4 | 4 3 2 2 2 2 2 3 4 | 4 3 2 2 2 2 2 3 4
4 3 3 3 3 3 3 3 4 | 4 3 3 3 3 3 3 3 4 | 4 3 3 3 3 3 3 3 4
4 4 4 4 4 4 4 4 4 | 4 4 4 4 4 4 4 4 4 | 4 4 4 4 4 4 4 4 4
Formation #4:
Forsworn's formation. It has a nice ring to it, but does it work? This type of formation interests me the most because it doesn't follow any of my rules. If successful, a formation like this seems to leave the door open for so many other formations. This is one reason I think the project is important, to discover more formations like this.
4 4 4 4 4 4 4 4 4
4 3 3 3 3 3 3 3 4
4 3 2 2 2 2 2 3 4
4 3 2 1 1 1 2 3 4
4 3 2 1 Y 1 2 3 4
4 3 2 1 1 1 2 3 4
4 3 2 2 2 2 2 3 4
4 3 3 3 3 3 3 3 4
4 4 4 4 4 4 4 4 4
Formation #5:
1-, 2-, and 3-aways are of the opposite patch type as the chain was begun with, 4-away is the same. I'm very confident this formation is at least decent after trying it out, I don't know how I didn't think of it before.
4 4 4 4 4 4 4 4 4 | 4 4 4 4 4 4 4 4 4
4 3 3 3 3 3 3 3 4 | 4 3 3 3 3 3 3 3 4
4 3 2 2 2 2 2 3 4 | 4 3 2 2 2 2 2 3 4
4 3 2 1 1 1 2 3 4 | 4 3 2 1 1 1 2 3 4
4 3 2 1 Y 1 2 3 4 | 4 3 2 1 Y 1 2 3 4
4 3 2 1 1 1 2 3 4 | 4 3 2 1 1 1 2 3 4
4 3 2 2 2 2 2 3 4 | 4 3 2 2 2 2 2 3 4
4 3 3 3 3 3 3 3 4 | 4 3 3 3 3 3 3 3 4
4 4 4 4 4 4 4 4 4 | 4 4 4 4 4 4 4 4 4
Alternate Hypothesis: On the Serebii guide I read about a "Line Theory," which I will give the gist of. Basically, if any two patches are in line with one another, either vertical or horizontal, the formation is unsafe. I believe this hypothesis to be mostly incorrect at the least. From what I've tried, it does not seem very effective. What success I've had with it seems no more attributable to the theory than to luck. Furthermore, my "Jesus Formation" directly contradicts the Line Theory, which obviously makes me question it.
If anyone would like to defend this theory, test this theory, or even prove to me it works, please do so.
Please post any other suggestions or hypotheses, but make sure you've had success with them so you don't end up wasting everyone's time with them.
How to Apply:
I would imagine most people who are interested in this project are the same people who most in Look At What I Chained, in which case I probably have a decent sense of how good a chainer you are. Nonetheless, just post a little bit about how you are as a chainer (brag if you must, I suppose), and what your experience level is and whatnot. Again, you don't have to be pro, just be able to meet my requirements and fulfill the tasks I ask of you. As you can see this is a pretty light screening process, so I'm placing my trust in you.
Participants:
1. Diesel
2. forbidden
3. Shelcario
4. notorious
5. Pimplup
6.
7.
8.
9.
10.
Questions? Please ask! Here if possible, AIM second, PM if necessary. :)
Thanks for your help!!!
<3 Diesel
