Dark Hart
Rejected by Azua
There's no need for subjectivity in this process though.


Welcome to Smashboards, the world's largest Super Smash Brothers community! Over 250,000 Smash Bros. fans from around the world have come to discuss these great games in over 19 million posts!
You are currently viewing our boards as a visitor. Click here to sign up right now and start on your path in the Smash community!
Your win vs Sneak was important, but it was your only real win this period. Sneak's win vs. theo was important because of how well he did this period (beating cam and someone else, I think).past pr's matter.
isn't that what made my win vs sneak worthless compared to his win vs theo? even though i beat the guy who beat theo, it didn't matter cause sneak wasn't pr'd at the time. how is that not taking last ranking into account? if sneak proved AT THAT TOURNAMENT that he was better than theo, and i beat him, why shouldn't that matter PR wise? because he wasn't pr'd at the time. that's why.
yea...to follow up the point i made earlier: how often would these points be tallied up to make a "power ranking"? if we have a system close to what cam was saying, after any given tournament (really, after any given match of importance) these rankings can, and will, change. if a players worth can change on the whim of a win or loss, then the rankings at any given moment have to reflect that. so are we gonna start with some sort of point system and update after every given tournament?
edit: so sneaks win vs theo matters because theo beat cam but my win vs sneak who beat theo matters less? see, i can't decide if you guys are actually using inui-esque data or not. i can't decide if theo beating cam is good because cam was pr'd at the time or if its because he's just cam. i mean in all reality, i shouldn't be on cause i lost to EWB in pools, which i guess is enough in itself to knock me out of contention.
but just tell me which matters most, especially with all of these being at the same tournament:
theo beating cam
sneak beating theo
me beating sneak
it would appear that according the the PR's we're looking at, theo beating cam is the most important. why? cause cam was ranked. then it would look like sneak beating theo matters second most. why? cause theo beat cam + theo was ranked. and it looks like me beating sneak matters least. why? cause sneak beat theo.
to me, that is the only real way i see the data being interpreted. i don't really mind this at all. it's how its worked. the rankings exist to let people know who to beat in order to get on the next ones (or at least thats what it looks like to me). but to say that the past rankings don't matter at all seems a little wonky considering the results. please correct me if i'm wrong, however.
ye prollywell then i would have beaten you and we'd be in a fairly similar ****storm.
edit: actually, probably not, cause then i'd be pr'd. =/
edit: probably not cause i still lost to EWB looooooollllll *u**in pools.
Chris: This is irrelevant inui logic that played no part in anything. This statement made me question whether you understand what PR's represent. PR's do not reflect who was better at what tournament; they reflect who performed best (the ratio of good wins to bad losses).if sneak proved AT THAT TOURNAMENT that he was better than theo, and i beat him, why shouldn't that matter PR wise?
Well, our method does actually look at objective results, and the decisions are usually clear once we get the data straight. These 'close calls' are not 'bad data' in the sense that the data doesn't have an answer, they are 'bad data' because it is very difficult to consider all of the variables at the same time for our puny human brains. Any 'mistake' on the PRs (as far as the PRs being a measure of how well said players performed against each other collectively) is a result of US INTERPRETING THE DATA INCORRECTLY, not a problem of the system itself. As I've said before, there should objectively be no tie for tenth place (well I suppose there is a very very very slight possibility that the data regarding those players is truely circular, but I doubt it). The reason we put the tie there is because the data that would be used to break the tie is obscure and convoluted and we couldn't figure it out (again not a problem with the system, but an inherent human weakness). Either way, those players all performed pretty ****ing similarly this ranking period, so I don't see why it's such a big deal. If one of us was decided on it would pretty much be like that time Michael Phelps beat that guy by 1/100th of a second. Sure, it's a victory but it was still pretty goddamn close.it can get hairy but it's not that big of a deal imo
btw I'm willing to bet that dorsey made a post suggesting a point system well before sneak was in NC lol, it was rejected hardcore iirc. I have always been for experimenting with different models.
1. Try to make the best point system you possibly can, incorporating all necessary details/elements
2. Consider someone else's point system
3. Have the panelists make their list
3. See what lists these systems result
4. Compare the lists that the point systems generate to the list created by the panelists
5. How far off is the system generated list? what is the probable cause of this? what can be done to make it better?
I'm F'd up right now but something tells me that a point system can do absolutely no wrong. It can serve as a loose template for panelists to consider while constructing the rankings. Remember Karn saying his memory sucks? rofl alex is my boy and i know he's a good panelist but when i see **** like that i'm thinking WTF get a point system going, a GOOD ONE, and see what list it generates. Like i just said this should provide a loose template for what the rankings should be like. Then, the panelists fine-tune the list with thier extensive knowledge of the NC scene and various players' skills that they have, because they're... well, panelists. The key thing with the point system (a good one) is that it will be based on unbiased facts and unbiased facts alone. It's worth it to see what kind of list a system like that can generate if constructed properly; not yo mention how easy it would make it for panelists to make decisions. We might even have 3 updates in a year in 2012 if we're lucky
Lightsyde-
That is not a point involving a point system. Your point correlates with how ****ty people are seeded at tournaments, apparently. Reason being: If you generate a bracket and assume that the higher seeds WILL win, it is 100% impossible for what you just said to happen. Sorry josh, but SEEDING should decide how good a win is, not a panel afterward. It's like that in basically all tournaments. Just because wack seeding or bracket johns occur at a tournament, it does not mean that a panel should be made to analyze the circumstances of what happened subjectively. The panel should be doing the seeding, not the interpreting.
--------
and LOL, the PR's should measure who places the best at tournaments
computing both attendance and tournament placing is easy btw, no one is forgetting anything lol. I'm a physics major now and have taken math through calculus 3(not that it's difficult at ALL to make a model for this).. would be no sweat for me, just give me the data and I'd be happy to show you guys..
What I don't get, is why all you *****es are so opposed to doing it right, fair, and by the book. When I first started debating about this, I had a feeling that we'd get to "fluke brackets" and johns... and look, here we are.
What are you even talking about? Stuff like that happens at stacked tournaments all the time. Unless you are top 4 on the PR's, there is an almost definite chance that you will have to play a highly ranked PR player early in a 16 man bracket, and have to play another good player, almost immediately after. Have you really never heard people say, "Man, today sucked. I had a Bye, then had to play PP, then had to play Ali in Loser's after he lost to XXXX unexpectedly"?Karn, the only way I can think of your example even happening, is if a good player that hasn't attended in quite some time gets a lower seed than normal(getting dr. peepee early on, getting lozr in late LB assuming an NC-only tourney). But....... that person hasn't attended anything in forever, so they SHOULDN'T get a wonderful seed(something like sneak said, and even if not, wouldn't assigning a proper seed to this person be a good job for the panel?). Then after doing what you just said, he'd have a great seed for the next tournament. So really.... not a good point D:
Responses in bold. The last half of your post (and the rest of it) is honestly absurd enough that I'm just hoping you're trolling. lmaoLightsyde-
That is not a point involving a point system. Your point correlates with how ****ty people are seeded at tournaments, apparently. Reason being: If you generate a bracket and assume that the higher seeds WILL win, it is 100% impossible for what you just said to happen.
This makes absolutely no sense. In what way is it impossible? How about the instance where you seed a bracket and everything happens exactly as the rankings would predict and there are no upsets? The scenario is equally possible as any other.
You do realize that absolutely all tournament brackets made that are not randomly seeded ASSUME based off a player (or a teams) prior record that one team will beat the other, right? We use PR's (based purely off of set counts, no incorporation of johns or bracket, just pure PvP results) to simulate accurate high seeds. Looking at a bracket, the first round will have Bye's for highly ranked players or matches against players they are likely to beat; that is the literal reward of playing well at prior junctures. If we got to round robin and play every other player/team like they do in a professional circuit (NBA, for example), our PR's would have infinitely more data and be that much more accurate at representing who would beat who...because those people have alreayd played in the past and there is data to suggest what will happen.
Sorry josh, but SEEDING should decide how good a win is, not a panel afterward. It's like that in basically all tournaments. Just because wack seeding or bracket johns occur at a tournament, it does not mean that a panel should be made to analyze the circumstances of what happened subjectively.
The panel has not and does not decide the quality of wins based on anything other than prior PR placement. There is no subjectivity, no opinions, just straight up data. Only in extreme cases, such as really sparse data (aka this period) where there would be a call for that and Kevin took the extra (unpopular) step to keep from deciding anything subjectively. That is why there is a 3 way tie. There is no subjectivity used in the process, whatsoever.
The panel should be doing the seeding, not the interpreting. If an "upset" in bracket happens, then so be it, it's a tournament and interpreting the findings of it subjectively defeats the whole idea of the competition. Everything should be on the line at the tournament, not on the panel's decision. (tournament does the skill measuring, not the panel..) Who the **** cares if someone got out-effort'd/tryhard'd while SDing lol. It was in tournament, and they obviously did bad. Peeps need to learn to be more content with that, not the end of the world(but it is what happens in the real world).
oauisdjglijasd The panel interprets the data only in so as we do not have a machine to process the set count results and translate it. Literally, all the panel does is construct the PR's with the data given. That is the extent of their interpreting. Exactly what you think should happen with this data is happening. No excuses are considered, no SD's, no nothing. They are processed, a list is constructed, we use the list for seeding to accurate prior results for the best possible balanced bracket, more data is produced, 3 tourneys worth of data is compiled and then lather rinse, repeat. That's it.
--------
and LOL, the PR's should measure who places the best at tournaments
computing both attendance and tournament placing is easy btw, no one is forgetting anything lol. I'm a physics major now and have taken math through calculus 3(not that it's difficult at ALL to make a model for this).. would be no sweat for me, just give me the data and I'd be happy to show you guys..
What I don't get, is why all you *****es are so opposed to doing it right, fair, and by the book. When I first started debating about this, I had a feeling that we'd get to "fluke brackets" and johns... and look, here we are.