The North Carolina Melee Power Rankings! Updated 8/14/14!

[Redd]

PP you should stop by VT on your way back and play a set with me!!

 

[bossa nova ♪]

lol kevin ur such a doodoohead

 

[Dr Peepee]

PP you should stop by VT on your way back and play a set with me!!

Haha I'd like to man but I'll prrrrobably have to be hurrying back that day. You never know though. =p

 

[Moophobia]

What would you guys do if you were half way through a tournament set with someone to find out you didn't have to actually play.

 

[Dark Hart]

Lol at Josh and Kyle

For lack of a better phrase, "circular argument" seems to fit pretty well, Dorsey got it, but off course y'all don't get Dorsey so you would go off on that aribitrary subject to try and explain to me the definition of two words I decided to put together because it sounded right to me just because I apparently didn't use the term correctly so airgo my point's mute

Sure guys

LOOOOOOOOOOLL

 

[Bl@ckChris]

hmm...well, if seeding was done right, then i guess direct results would show who was the best in tournament that series, without any dependence on lack of data...it would say "we seeded right, and the best people advanced. here are the best people" regardless of who they played.

interesting.

but pr's are not thought to say "we seeded right, and the best people advanced. here are the best people". pr's seem to try to get at "out of everyone who played everyone, here seems to be the people with the best wins and least bad losses"

the question is whether the assumption of "we seeded right" can be made. and if that answer is yes, then the results themselves should speak for the skill of the player themself.

if that answer is no, then we need to change how we seed lol.

if a player can get far in bracket without playing anyone notable, then either the definition of who is notable is wrong, or the seeding is wrong...i guess.

so the question is whether we trust our seeding enough to base a PR off of pure results, or whether we have to rely on who beat who to build a structure of how our scene is.

don't really care either way. i think both sides have merit.

edit: moo, i'd quit out and find my real match *shrug*

 

[Dr Peepee]

What would you guys do if you were half way through a tournament set with someone to find out you didn't have to actually play.

Malk did that to me after we played our "tournament set" at my first tourney LMAOOO it was soooo confusing. XD

It still sucks though and kind of lets the air out of my tires or deflates me in a way.

Only other thing I can say about that is check the bracket when you're told to play. I do this at locals even if I'm the one who made the bracket and the guy who walks up tells me we have to play lol.

 

[Twin_A]

PeyhPeyh!!! How long did it take you to get better after Genesis?

 

[Dr Peepee]

2 more weeks LOOOOOOOL

sup twin <3

 

[Moophobia]

I'm curious as to whether people would be annoyed or appreciate the tournament setting. I would probably be pissed because I use a lot of energy to play tournament sets.

 

[NearZzz]

Hope everyone likes the Adventure Time Theme.

1. Dr Peepee - Finn
2. Twitch - Jake
3. Darksyde - Cosmic Owl
4. Everlasting Yayuhz - Ice King
5. I've Jihad It/Ali - BMO
6. Dop - Marceline
7. LoZR - Billy
8. Ron - Eberhardt
9. Crystalnite - Fionna
10. THO - Flambo
10.Lord Karn - Ancient Psychic Tandem War Elephant
10.Sneak - Blastronaut

Me and my friend worked on this together. c:

 

[Dr Peepee]

Updated OP with the picture!

Thanks dude =)

 

[Moophobia]

billy is the party god?

 

[NearZzz]

oops, that's the previous version.

Lozr was going to be Party God originally when I drafted it up since Lozr is the Party God, but then I couldn't get a good enough image.

 

[GofG]

psy-inspired thoughts:

i have commonly held that my bad tournament placings has been directly caused by bad seeding: i have been placed against a peach in either my first or second match in every tournament i have entered since i started maining ice climbers.

this most recent tournament was the first tourney where i did not switch off of iceys upon seeing peach picked and just ****ing played the character i know how to play, weaknesses and all.

oh **** it's a new day, relevantly i should probably make my **** Aasem claim so that blitz isn't a part of the shotgun metagame.

 

[Juno McGrath]

**** Aasem son.

 

[DJRome]

lol, dorsey has taken calculus 3 and therefore can model this.

ive spent way too much time thinking about prs, probably as much as anyone here. and it's quite clear that either you:

1) do what i suggested before
OR
2) spent unnecessary time making a useless model that assigns weighting of variables based on your arbitrary bias

statistically, dorsey, even if you make a model that can predict past prs, current prs, and tournament results, it won't be 100% correct. because the ranking is not strictly linear (ie 7>8 100% of the time, etc), you have no way of making a model that gives the same result each time as smash is a game of probability. and since you have no way of confirming this probability, your model is at best a good guess (especially given the data), which is what the prs already are i covered the random nature of fighting games in a past post, obviously not the first to come to that realization, but it is there if you want to dig for it.

i hope that you don't get to thinking that your elementary math skills (i say this because calculus 3 is at best the most fundamental tool needed for such a model to be anything of value) can magically sprout a model worth using. i feel that my level of math is not very sufficient to make such a model, and i can assure you i know a lot more about math than you do, though it is still quite terrible. nothing against you, but i had to make sure that you understand how complex such a problem is

 

[GofG]

lol, dorsey has taken calculus 3 and therefore can model this.

ive spent way too much time thinking about prs, probably as much as anyone here. and it's quite clear that either you:

1) do what i suggested before
OR
2) spent unnecessary time making a useless model that assigns weighting of variables based on your arbitrary bias

statistically, dorsey, even if you make a model that can predict past prs, current prs, and tournament results, it won't be 100% correct. because the ranking is not strictly linear (ie 7>8 100% of the time, etc), you have no way of making a model that gives the same result each time as smash is a game of probability. and since you have no way of confirming this probability, your model is at best a good guess (especially given the data), which is what the prs already are i covered the random nature of fighting games in a past post, obviously not the first to come to that realization, but it is there if you want to dig for it.

i hope that you don't get to thinking that your elementary math skills (i say this because calculus 3 is at best the most fundamental tool needed for such a model to be anything of value) can magically sprout a model worth using. i feel that my level of math is not very sufficient to make such a model, and i can assure you i know a lot more about math than you do, though it is still quite terrible. nothing against you, but i had to make sure that you understand how complex such a problem is

shut the **** up you stupid fat asian.

e: even the ELO rating system doesn't require calc 3, and it is a much more complicated model than what we're talking about; in fact I can't see how any sort of predictive model which assigned points based on the probability of a player winning a match or on the probability of *anything* useful to our circumstance requiring anything more than simple logistics functions.

e2: i'll admit my expertise isn't in calc (noncommutative ring theorist here); maybe there are outer echelons of calc i'm not aware of which could be used to make such a model, but i doubt it, and i have yet to hear any argument that a simple program which auto-updated a bayesian probability table wouldn't suffice.

e: on the merits of a point system: we already use a point system! we have a theoretical value the panelists ascribe to each of the players, evidenced by match history and tournament placements and other things, which exists in the minds of the panelists and is used to rank the players. if we could simply find a better, more concrete, more deterministic way of modeling that theoretical value, we could do away with the panelists alltogether and just as we have decided on what objective game we are playing, we could decide on what objective metagame we are playing as well.

 

[DJRome]

lol u mad?

come back when you can show how what you're describing would work in practice

 

[GofG]

lol

come look at the source


git://98.158.26.222/home/john/bin/pr.git

obviously not working yet but all the math is there, just working on input/output. hope you can read c

i'm just wondering why you took the attitude that a relatively simple predictive model would require calculus 3. my guess is that you have an elitist duke attitude just like every other engineer/math/sci person who ever came out of duke.

 

[DJRome]

couldn't connect, pass?

 

[GofG]

uhm no password, it should Just Work

e:

john@debian ~$ git clone git://98.158.26.222/home/john/bin/pr.git
Cloning into pr...
remote: Counting objects: 12, done.
remote: Compressing objects: 100% (8/8), done.
remote: Total 12 (delta 9), reused 12 (delta 9)
Receiving objects: 100% (12/12), 3.63 KiB, done.
Resolving deltas: 100% (9/9), done.

e2: ****, when i clone from a machine that isn't the machine located at 98.158.26.222 it fails. yay i get more work to do tonight and it's already 6 am

 

[DJRome]

well, if you want to describe the math here, that's fine. im actually interested in what kind of algorithm you used to compute it

 

[AlcyoNite]

idk wth any of u are saying

but if u have worked out some sort of model, id like to see what it spews out

edit: and wtf why am i THE ONLY WOMAN

 

[DJRome]

All i heard was

I AM THE ONLY WOMAN

 

[GofG]

each player is an instantiation of the object "player" which has two attributes:

name (string)
trueskillvalue (float)

there's a method "update" which takes as parameters two "player" objects (the first being the winner) and then a boolean "perfect" which is true if it was 2-0 and false otherwise.

the algorithm ran essentially does a bayesian update on the probability of a player winning an *entire tournament* of every player object based on their new true skill value after the win (+/- a certain number based on the two tsv's of the players).

i have input some test data, giving kevin a tsv of 2000, lozr with 1500, dop with 1500, and karn with 1000, making the odds 80, 8, 8, and 4 (it's logarithmic and based on an extremely basic formula which i intend to tweak)

eventually the idea is, i'll plug in data from a couple tournaments 3 years ago, then see if I can use that data to predict who won the tournament 2.9 years ago, with the person second-highest most likely to win coming in second, etc, tweaking how TSV affects the odds until it can actually make this prediction accurately. then i'll enter in the (predicted accurately or actual, same) data and use this to predict the next tournament, with the updated TSV.

at this point it will either be obvious that the system will work or that, since people improve at varying rates, the system will require a more consistent stream of data to make it work. either of this is still math-based and objective and, honestly, pretty feasible, and sure as hell didn't require calc3.

 

[DJRome]

so if it gives a probability for a player to win, how will this produce consistent results? correct me if I'm misinterpreting, but doesn't that make it very open to interpretation whether a win in any result is in the higher or lower portion of probability? i would like to hear how the determination is made. the main problem i ran into in using this type of estimation to test values is that it has no way to accurately gauge improvement while also being consistent with past results

on a side note, my reference to calc 3 was not that it was necessary to make such a model (though i would say that it does), but rather that dorsey pointing out that he had taken it meant in no way that he understood how to model it. i can see how you could be offended by such a remark, however.

anyways, this is pretty much what i was aiming to do generally last time i tried to make such a model. unfortunately, the model proved very unreliable and had the added annoyance of factoring in new players and oos players. i would like to see this model work if you can get it going, but it still needs to address the core problems that exist

 

[NearZzz]

Theo, you're Fionna, the gender swapped, female version of Finn.
You're a girl.

 

[Bl@ckChris]

so...gofg is trying to make an algorith that could have accurately predicted placings before, and the top ten would represent "if nc had a tournament, these are the predicted top ten".

sounds legit.

 

[Dorsey]

ROFL, no wonder DJ thinks it's bad. Because HIS way of doing it is bad(the completely wrong way). Pretty hilarious.

No surprise coming from duke.

but rather that dorsey pointing out that he had taken it meant in no way that he understood how to model it.
t

My comment meant nothing of the sort. It's a lot easier than that... rack your brains math genius, I know you can figure it out.

 

[Bl@ckChris]

so dorsey, all you want is a panel/point system that can seed people right and then let the placings from proper seeding speak for themself right?

to you it doesn't matter who you played to get a decent placing because if it was seeded right in the first place, the better players advanced, and that proves it all, correct?

 

[Dorsey]

I'd think it depends on the margin between people in points. Discretion could be used, I assume the panel would want to look in to two people that are close in points. And as far as seeding is concerned.......... Half will be pretty easy to do(dr peepee, lozr, twitch, dop, etc. + lower skilled players), then there's a big chunk that are harder to do, the players who are anywhere from a a little above avg to a little more than above avg(the way I even had to describe that in itself shows that it's kinda hard to seed, but not as hard as some put it on to be though). I'd just take into consideration how these guys have been playing each other when they meet in the last few tourneys(a panel-esque thing to do) and do it as best as possible. An idea would be to let MM's count for seeding instead of help/worsen your chances on the PR(instead, it would indirectly do this with your seeding), and have the PR based on tourney results / attendance. I would scale the players from 0-10 for a reference to determine the seeds, so each person would have an actual "value" to order those who attended to the proper seed. This would be how I'd factor in attendance, by simply adding up the total values of all participants in the tournament giving the tourney a 'value'. So if there were 4 tournaments being considered, then i'd divide each 'tournament value' by the sum of all 4 of them getting a fraction / percentage for each. Then, taking a 0-10 pt system like cam/sneak more recently suggested, i'd multiply the fraction made from the tournament value to the points earned from their placing. Then, when you do this for all 4 tourney's and add up everyone's, everyone will have a value from 0-10 depicting how they did in regard to their placing and general attendance. You don't need some complex model, you just need to do the damn seeding right, and precise in regard to the 10pt scale. I think that's a more fair process, competitively speaking, than a panel deciding what was fair or not afterward.

Yes, this can go wrong in certain situations concerning who played who. Concentrating efforts on seeding will limit this from happening, and with this system it should be easier for the panel to see, and everyone else to see, why changes based on who played who were made, when these changes are needed.

 

[lord karn]

each player is an instantiation of the object "player" which has two attributes:

name (string)
trueskillvalue (float)

there's a method "update" which takes as parameters two "player" objects (the first being the winner) and then a boolean "perfect" which is true if it was 2-0 and false otherwise.

the algorithm ran essentially does a bayesian update on the probability of a player winning an *entire tournament* of every player object based on their new true skill value after the win (+/- a certain number based on the two tsv's of the players).

i have input some test data, giving kevin a tsv of 2000, lozr with 1500, dop with 1500, and karn with 1000, making the odds 80, 8, 8, and 4 (it's logarithmic and based on an extremely basic formula which i intend to tweak)

eventually the idea is, i'll plug in data from a couple tournaments 3 years ago, then see if I can use that data to predict who won the tournament 2.9 years ago, with the person second-highest most likely to win coming in second, etc, tweaking how TSV affects the odds until it can actually make this prediction accurately. then i'll enter in the (predicted accurately or actual, same) data and use this to predict the next tournament, with the updated TSV.

at this point it will either be obvious that the system will work or that, since people improve at varying rates, the system will require a more consistent stream of data to make it work. either of this is still math-based and objective and, honestly, pretty feasible, and sure as hell didn't require calc3.

Is there anyway you can make the TSV be calculated from the tournament data itself?

 

[Dark Hart]

come look at the source

THE MOTHER****ING SOURCE

OF

....

Shhi......

 

[Dark Hart]

We could just stop seeding an let it be random

Sure as he'll would tons of interesting and more pvp match-ups

 

[vZakat]

Karn should have been a whywolf...

 

[Lightsyde]

Jesse:

I had not responded yet because I

a) did not get on SWF after work
b) read this yesterday and realized there is absolutely no way you are going to listen to me.

I don't think our PR's are accurate for numerous other decisions, but your reasons and explainations have been illogical enough that I was convinced you were trolling. Responding to you has had diminishing returns, especially being I'm only explaining the current system which you don't seem to understand and explaining why things like placing are irrelevent (which Karn made a big post about) and you proceeded to further demonstrate your misunderstanding in your responses. You've been taking personal shots at me during your posts or just being condescending and honestly, I've been just trying to not say anything that would lead to more negative feelings because you are my friend.

This is my last response.

Sneak quoting it and saying it doesn't mean anything. It's very easy to incorporate attendance and tournament placing so that they ARE INDEED using the same point system(proportioned to each other). I have already said how to do this. The less good people that are there, the less total point value for the tournament. Attendance is covered. My math is right. And sneak can PM me for more information about it specifically if he pleases, it's painfully obvious that my efforts no matter how right they may be, would be labeled as arbitrary..

Why are you talking in this paragraph as if you proposed some novel concept about an attendance variable? I mentioned it in the first place. You haven't posted an actual drop of your math so you seem like you're just blowing hot air in this discussion, whether you are or not. It's painfully obvious that you haven't contributed a single thing to this discussion that has not already been addressed, discussed, disregarded, reconsidered, rewritten and discarded again. I don't know why you think otherwise.

Why is there even a panel? Is it normal to have a panel to interpret stuff like this?? No, I have never heard of such a thing, therefore I can only conclude that the use a panel/system was done on a whim by the NC smash scene, therefore it is arbitrary. rofl.. just some weird logic that's applicable to just about everything(the whole arbitrary thing).

Unless I'm mistaken, every state has a panel. I think Sneak even posted right after you to let you know that this is purely an incorrect assumption on your part and you are just looking to point fingers. If you would absorb any of the explanations (look at how Chris' posts changed with his understand), you would realize the panel is "interpretting" data only in the sense that the PR's are do not literally create themselves.

This periods exception to that is in the case of Adam/Aasem/David etc being included on the PR's with only 1 tournament. That was seemingly done in order to not completely **** seeding the next time they show up to a tournament and would be "Unranked players".

And josh, when you would get a hot head on these forums and rage about smash on occasion(I know you've gotten a lot better in regard to this), I certainly didn't laugh and consider your words beneath me like you're doing so now. I actually turned a blind eye to it just as most of everyone else, because they either like you or didn't want to deal with the reaction. And if you have trouble reading/comprehending my posts(apparently, you do) then I suggest that you go back and read them again instead of laughing.

This is irrelevant. I didn't respond because I thought you were trolling. That's it. I literally thought your responses were so nonsensical and baseless to the discussion that had happened, that I assumed you were either trolling or completely clueless and talking out of your ***. I was trying not to have to make that judgment but you seem to want to force this point. Again, all your "ideas" have been stated more concisely/organized/clearly by other people in this thread and you can't understand that you are talking big and saying little. You undoubtedly feel the same way about what I am saying; the only difference is...the people who understand PR's (By the way, how many have you been on? how many tournaments have you been to in the last year? You are disconnected enough from the actual happenings that you don't even remember how wonky placings (after 4th) are because they show a lot more about seeding than you'd think. Our attendance is sparse so the data is sparse. The real data of sets (not probabilities, the stuff that actually happened) is reflected in the PR's which are used to seed our tournaments. You forget that this is a fighting game where people have bad match ups. John Wittle, for example, has had the misfortune of playing Peach players early A LOT. If he got PR'd and then lost to an unranked Peach that would throw off the whole bracket and no function is going to be able to compensate for variations like match ups. You're addressing this problem as if it's basketball (which has seasons worth of data for seeding) and where there isn't really a concept like counter picks (not to the extreme we are talking in Melee). Placings are ridiculous to use because stuff like Plank getting 9th at Apex happens (he plays Sheik and his first 2 bracket matches were vs Marth). Not to denounce Plank's skill, but it's pretty obvious that Plank was likely not on anyone's top 10 list, nor would he have gotten his spot if seeding (which he made btw) did not land him against IB and HBK, Sheik's best high tier match up.

Thing is you're flat out wrong. Sorry if you can't handle that, but it seems that you my friend are incapable of grasping the bigger picture, which is:

Thing is saying "you're flat out wrong" means absolutely nothing. This is essentially an ad hominem to make it sound as if you've proved anything I've said wrong. You haven't. Again, you're just resaying irrelavent things that have already been said better.

The panel says there is no way to be perfect about this, being that seeding is flawed for PR usage, therefore the bracket is flawed for PR usage, therefore placement means nothing whatsoever. So, they use their own discretion between opponents and such to avoid these flaws, or at least that is the goal of it all.

What? The panel said there was no way to be perfect about this ranking period due to the data produced by attendance. You are misquoting the reasons the panel gave in order to build a strawman to support your RIDICULOUS case for placings. Honestly, if you ever got to play in more than the first few rounds of a Melee tournament you would see first hand how wildly innaccurate placings can be just because of stuff like upsets, DQ's, counterpicks etc. Again, read Alex's post right after yours where you explains for the 20th time why using placing is horrible.

I say that there is no way to be perfect about this, so just do the seeding as best as possible(and perhaps come up with new methods of seeding better, like I suggested) then just go off of placing/attendance, two variables EASILY(yes EASILY) computed into a model. Being that there's no perfect way to go about this, I feel that something absolute, like a point system, is a much more proper/fair thing to do opposed to relying on an intermediary's interpretation(the panel). You could publish all the data right there with the rankings, and then the panel could do their part in making a few changes where they see fit. Due to the data published, it would be obvious for everyone to see why the panel made their decisions in which discretion is involved.

New methods like you suggested? Again, everything you've said has been discussed over and over again in the last few years. NONE OF YOUR IDEAS HAVE BEEN THE SLIGHTEST BIT ORIGINAL. Just because they are new to you does not make them revolutionary. I've been following this thread for almost 4 years. The fact that you are using placing as a variable completely invalidates any credibility your proposal might have. There's no perfect way to go about this because of the fact that every tournament is going to have a really different mix of players with few constants. Like John pointed out, we already essentially use a masked point system with the way that we weight PR'd wins. The reasons why the panel did what they did are obvious to anyone who understand what the PR's actually represent.

It just seems like a more attractive idea, and apparently several agree.

No, it doesn't. The people who have agreed with you are the ones are also not making logical cases for their arguments and, from my reading, only seem to be named Lucas.

Responses in bold. This is a waste of time.

I'd think it depends on the margin between people in points. Discretion could be used, I assume the panel would want to look in to two people that are close in points. And as far as seeding is concerned.......... Half will be pretty easy to do(dr peepee, lozr, twitch, dop, etc. + lower skilled players), then there's a big chunk that are harder to do, the players who are anywhere from a a little above avg to a little more than above avg(the way I even had to describe that in itself shows that it's kinda hard to seed, but not as hard as some put it on to be though). I'd just take into consideration how these guys have been playing each other when they meet in the last few tourneys(a panel-esque thing to do) and do it as best as possible. An idea would be to let MM's count for seeding instead of help/worsen your chances on the PR(instead, it would indirectly do this with your seeding), and have the PR based on tourney results / attendance. I would scale the players from 0-10 for a reference to determine the seeds, so each person would have an actual "value" to order those who attended to the proper seed. This would be how I'd factor in attendance, by simply adding up the total values of all participants in the tournament giving the tourney a 'value'. So if there were 4 tournaments being considered, then i'd divide each 'tournament value' by the sum of all 4 of them getting a fraction / percentage for each. Then, taking a 0-10 pt system like cam/sneak more recently suggested, i'd multiply the fraction made from the tournament value to the points earned from their placing. Then, when you do this for all 4 tourney's and add up everyone's, everyone will have a value from 0-10 depicting how they did in regard to their placing and general attendance. You don't need some complex model, you just need to do the damn seeding right, and precise in regard to the 10pt scale. I think that's a more fair process, competitively speaking, than a panel deciding what was fair or not afterward.

Yes, this can go wrong in certain situations concerning who played who. Concentrating efforts on seeding will limit this from happening, and with this system it should be easier for the panel to see, and everyone else to see, why changes based on who played who were made, when these changes are needed.

Yes, this is how a point system would more or less work, congratulations, you now understand what Sneak was proposing and we talked about several pages ago.

Hilarious that even in your restating there is point in your system where the panel would have to use more of their own discretion than in our current system; which is apparently what you've been arguing against.

All of this has been said. You are still talking about things that have already been brought and proposing them as if they are your ideas and wondering why no one is paying attention. The only new thing is factoring tourney attendance for weighting the tournament, which sounds cool, but wouldn't actually change much at all. If you had a tournament of 3 really good players and 50 scrubs, that tournament would be worth more than a stacked 8 man bracket...which is ********.

 

[Dorsey]

First bold part: Yes I did show my math.

Second bold part: I know that other states have panels, I clarified that in the post directly after

Third bold part: It is relevant, because you thought my posts were absurd so you laughed at them and considered them beneath you, so I brought up a similar situation in which I have thought your words were absurd, yet chose to not say anything.

Fourth bold part: You called my posts ridiculous again, that proves a lot. also "because of stuff like upsets, DQ's, counterpicks etc".. From the biggest johner in NC smash, that doesn't mean much bro. And I am quite aware of those factors, I said I know how random wins can be between similarly skilled players due to whatever sundry reasons.

Fifth bold part: Name whomever suggested a point system before me. I did over 2 years ago, so please back up what you're saying with some actual proof instead of saying something is ridiculous in all caps, lol.

2nd Response: I've never seen sneak propose this, link the post and I will check it out. I proposed that method a long time ago before sneak even moved here anyway. I have been telling dave that exact method of years.

Don't get mad because I'm trying to get you to be fair in this discussion. Your current tactics don't seem to be working right now, I'd try something else(best bet would be to leave this discussion to the panel / leave that as your last post like said, and not the guy who's happy that he finally got a good ranking).

Also:

Has there ever been an NC tourney stacked with 50 scrubs??? No. Horrid point comparing it to a stacked 8 man bracket as if this analogy would ever happen in a ranking period in our state. And if insane circumstances alike do happen then it's a job for the panel, as I suggested.

 

[lord karn]

I'm just going to restate for the millionth and last time that using placing in our deliberation IN ANY WAY AT ALL weights the previous season's PRs and makes it difficult for new players that deserve it to get on the PRs. One of the FUNDAMENTAL REASONS for why we are using the system we are using is because we don't want previous season's data to affect the current system, we instead want a fair system where a player who demonstrates improvement can make it on the PRs like he deserves. If you even want to think about arguing that placement is important, the very first thing you need to do is explain why you think previous season's PRs should give them an advantage in the new season.

And the reason we have a panel at all is because we are a small community and no one has cared enough to make a program that can calculate it for us. As I've said before, our system is purely quantifiable a program could do all our work for us if we just had one that considers the right things. I'm sure pro sports leagues have plenty of money to throw around to get something similar made for them. As someone suggested earlier, most pro sports leagues have everyone play everyone else at some point or another. In a round robin like that placing is actually an accurate measurement. In fact, our method is pretty much us making a giant round robin and filling in all the sets that actually happened in tournament and seeing who comes out on top (and 'bad data' is when there are a lot of wholes in this theoretical round robin).

Anyways, I agree with Josh that your posts also came off as incredibly condescending and arrogant at first (whether you meant it or not, or were just trolling), so I understand why Josh responded in an angry or hotheaded way. In fact, it annoyed me too, and that's the only reason I came back into this thread to explain myself for the millionth time.

 

[Lightsyde]

First bold part: Yes I did show my math.

Your newest post shows the extent of what your equation would be like and I already addressed why it wouldn't work. It's as if you read the first words of what I say and then respond without reading the rest.

Second bold part: I know that other states have panels, I clarified that in the post directly after

Cool. It was just a good example of you leaping to an assumption to cast doubt on our system despite it being a baseless accusation.

Third bold part: It is relevant, because you thought my posts were absurd so you laughed at them and considered them beneath you, so I brought up a similar situation in which I have thought your words were absurd, yet chose to not say anything.

Again, you misunderstand. I literally thought you were trolling me and did not want to be trolled. It had nothing to do with me judging you to be beneath. It had absolutely nothing to do with you. You were saying things that were either trolling or wrong. I've explained 3 times and you still think this is some personal attack. /facepalm

Fourth bold part: You called my posts ridiculous again, that proves a lot. also "because of stuff like upsets, DQ's, counterpicks etc".. From the biggest johner in NC smash, that doesn't mean much bro. And I am quite aware of those factors, I said I know how random wins can be between similarly skilled players due to whatever sundry reasons.

I pointed out why I thought your post was ridiculous because I was forced to accept that you weren't trolling and really do believe the nonsense you keep talking. I feel like you are looking for things that aren't there. Anyone could have posted this and I would have felt the same way. I am explaining with examples and referencing other posts which have elaborated further.

And stop making this about me wtf. It's like you're trying to make me mad. You are completely outside of the real time happenings of this community and haven't been to a tourney or fest or in A LONG TIME. I could have not even mumbled a john in the last year and you woudn't know. You even contradict yourself in the next sentence and reinforce that what I said about upsets etc is happens and is correct. Which means you took the time to insult me and call me the biggest johner in NC for no purpose; you agreed with me!

Ask Dave about how much I john'd the last time he and I played if you really want context. I will not stoop to your level again and insult your tournament performance or whatever other easy targets there might be. If you look at the OP, my name has number 3 beside it; it's a good thing I stopped johning long enough to get that placing.

Fifth bold part: Name whomever suggested a point system before me. I did over 2 years ago, so please back up what you're saying with some actual proof instead of saying something is ridiculous in all caps, lol.

2nd Response: I've never seen sneak propose this, link the post and I will check it out. I proposed that method a long time ago before sneak even moved here anyway. I have been telling dave that exact method of years.

The first time I opened this thread, James/Pac! was arguing with people about the merits of a point system and the current PR's. You are a child when it comes to your overall length of exposure and membership to NC discussion. I'm fairly certain that Corey, Adam, Kevin, Ph00t, and various other old schoolers could testify to this.

Don't get mad because I'm trying to get you to be fair in this discussion. Your current tactics don't seem to be working right now, I'd try something else(best bet would be to leave this discussion to the panel / leave that as your last post like said, and not the guy who's happy that he finally got a good ranking).

You are the one slinging insults at people and accusing the panel of being biased. I have been making arguments and doing my best to avoid insulting you, which is getting progessively more difficult. Like Alex just said, he took your posts as pretty damn offensive and condescending too. I've been calm and typing these posts in between doing wires at work. My current tactics are not working because your reading comprehension does not seem to register arguments regarding the irrelevance of placing and why your system is bad with the way the data actually comes in vs the ideal way you dream about.

Also:

Has there ever been an NC tourney stacked with 50 scrubs??? No. Horrid point comparing it to a stacked 8 man bracket as if this analogy would ever happen in a ranking period in our state. And if insane circumstances alike do happen then it's a job for the panel, as I suggested.

This is called a counter example. I provided an instance in which your system would produce erroneous results, it does not need to happen. In our current system, there would be no need for the panel to decide anything other than looking at how the relevant players of that tournament did against each other. Your system requires the panel to make judgement calls. You argue that the current system is diluted and I provided an absurd example to show that it's actually yours that would be likely to require arbitration. I could have easy said 10 or 15 or 20 numbers of scrubs and 3 or 4 PR'd players. Don't think those happen? They are called "Asheboro".

For larger numbers of scrubs, see Melee tournaments with PR'd players at gaming conventions.

aosifjaosidjfoiajsf