• Check out the relaunch of our general collection, with classic designs and new ones by our very own Pissog!

np: OU Suspect Testing Round 1 - ...wait, I'm not Jumpman16!

Status
Not open for further replies.
It seems more logical to base voter pool size on a qualitative basis, not a quantitative one. I mean, simple logic would show that there is not one unique point in the rating system where you magically change from a common player that knows little about the metagame into a wise guru who deserves to vote in the Suspect Test, so why should there be an objective cutoff?
It would be more sensible to maybe have one "magic number" that gets you voting rights automatically, but people who miss that number should still be able to apply to vote.

Think about it: What is the difference between a 1500 player and a 1499 player? Mathematically insignificant. What is the difference between a 1400 and a 1399 player? Mathematically insignificant. A mathematically insignificant difference should not disqualify a person from voting.

Same logic applies to predetermining voter pool size. Allowing a predetermined of voters to vote and nobody else is ludicrous. Why should nine people get to vote while the #10 position does not (I'm looking at you, Smogon Council Method)? Why should 40 people get to vote while #41 is left out? Once you hit 40 on the ladder, God probably doesn't send you a PM telling you how a metagame should be balanced and what should be accepted into OU, so why should we create a system that advances that notion?

I understand that this process should be expeditious, and we don't have time to delay the Suspect Test to analyze the resume of every potential voter. We tried to fix the time problem and got the Smogon Council (arguably one of the worst systems to date). There is one solution, and that is to use reason. To speed up the process, we can auto-allow a certain number of voters, or everyone past a certain rank. That's fine, as long as you set that bar high. However, we need to subjectively determine people who deserve voting rights, and that can still take time. Maybe we will have to analyze a few people's resumes and make a relatively unbiased decision. We want to optimize the amount of people to analyze, not reduce it to 0.

I don't know if this problem is already resolved, or if it's still being debated, but I'm just putting in my $0.02.
 
Maybe I did get a PM from God o_O

Lolcat I agree with you 100%. The only problem is that at some point their has to be a cutoff. If we put the cutoff at say 1350 for example the person with 1345 could go "Hey I'm just a little lower than that guy why can't I vote. Then when the guy with 1340 see's that he does the same thing. It doesn't matter where the cutoff is, their will always be people left out.

I was thinking that we could put the cutoff point at something like the front page and let people that don't reach the point to submit applications to vote. That way people who didn't reach the required ranking could still vote and it wouldn't leave qualified people out.

PS: 3rd win on Sunday, Superbowl next year
 
Specs Latios Thunderbolt vs +1 252/0 CM Manaphy: 51.49% - 60.89%
Specs Latios Thunderbolt vs +2 252/0 CM Manaphy: 38.61% - 45.54%
Click to expand...
Well, of course with CM Manaphy it's different.

It would be more sensible to maybe have one "magic number" that gets you voting rights automatically, but people who miss that number should still be able to apply to vote.
Click to expand...

If this is anything like the UU process, that's exactly what's going to happen.
 
Just an observation, but using a rating system cutoff might not work so well when you take into account how much to lose people have at the top. Everyone is hitting 1400 and then stopping. They now don't have experience with the last week of the metagame unless they're using alts. Doing something like the Top 25 / Top 5% or something may encourage people to keep playing.
 
What about lowering the requirement to 1300? Looking at the rankings now, this is the top 75-80, and the first two pages of the list. 1300 seems like an appropriate cutoff, as all the good players hover around 1100-1200.

While 1500 was the initial/tentative benchmark, it seems excessive now.

You can apply to vote? What are the prerequisites?
 
Is there any a desired amount of posters? I mean, not more than 50, or at least 20, for example? Because that could help setting a benchmark.

I still think using a Top 50 or something would be better, then it's just tough shit for #51, but #50 just ranked higher (exception if two players score the same rating).
 
Firestorm is absolutely right (again). As far as I know, the rediculously high cutoff of 1500 was to encourage the top players to keep playing and not to just stop when they get to a cutoff point. Else, there should be a change in the way the rating score is calculated, such as your score decreasing over time when you don't ladder. But I think that is up to the folks at PO, not smogon.
 
The rating cutoff system is fine. I can't speak for everybody, but since I hit 1400 I've been consistently laddering on alts. I find it hard to believe that top players would just stop playing pokemon once they reached 1400. The other proposed systems would not work at all due to the number of alts in the top x spots; we would end up with significantly fewer voters than we want. 1350 seems like a reasonable cutoff point. With most players stopping once they reach 1400, it's really hard to consistently gain points once you approach that rating, as you get nothing but +9 -22 games against people abusing darkrai/skymin/deo-a (at least, in my experience).

A quick word on Dory: he needs to go right now. His sweeping abilities are great, but manageable. His real strength comes from rapid spin; not a single useful ghost can safely block spin without risking a KO from a +2 EQ / Rock Slide. Granted, nothing could safely block spin against LO Starmie in gen 4, but the difference is that a mispredict against Starmie didn't mean facing a +2/+2 pokemon with a base 135 attack. Rapid Spin Doryuuzu practically beats stall by himself.
 
I've been playing these days and i also have to agree that 1500 is too much. People at the top, like Firestorm said, really have a lot to lose when every battle is +5 -30 or something like that. Even worse is having to play in a metagame with skymin, inconsistent octilery and stuff like that, since you can lose just to sheer luck.
 
I have to agree with the criticisms of the rating system. To put it bluntly, it's terrible. It makes no attempt to actually predict one's skill level, and puts emphasis only on the present for the most part. There's almost no reward for being a consistent player. And to top it off, it still doesn't address the main problem with Shoddy's rating system, namely the encouragement of alts.

But now I'm guessing it's too late for anyone with a sound statistical argument to have that changed. Too bad willing programmers with a working knowledge of statistics seem to be so rare...
 
We don't need volatility to tell us who's still playing and knows what's going on. The guy who knows the game in and out is the one who got up to the 15-1600s while everyone was still flailing around in the 1300s and can probably type a paragraph about the metagame better than you guys type essays. He doesn't need to be punished for taking a break or even retiring in his prime.

That's like taking away Muhammed Ali or Michael Jordan's trophies because they don't know the way their sport is played now, or young blood is stepping up to take their place. Hello, they're still beastly athletes and what they have to say about the sport still matters even after 30 or so years later. It's not like it's an entirely different game. Let me know when a Gen 3 expert is trying to tell us what Gen 5 is about then we'll talk.

Besides, I don't think anyone plays Pokemon just to get a huge rating, then quit to gloat about it or get voting rights. Those same players are probably on alts still averaging in the top 20s and you don't even know it. Fact is, people who know the game don't need to have it proven with a rating system forcing them to play every day, lest they lose the number signifying their achievement. They'll either be over here telling us what they think or they'll keep playing on other accounts to prove it.

And plus, 2 1/2 weeks is a really short period of time in the grand scheme of things. If I play very successfully even if for only one week before taking a break, you can't take away from the fact that I knew what I was doing at the time and 1 week is not enough time for the metagame to warp itself to the point where I wouldn't know what's going on if I started playing again later.

tl;dr: Volatility is bullshit and I'm glad PO doesn't have it implemented.
 
Well on the rating system, this is the first time we are using it so we are kind of in unknown terriotory with it. We'll have a better system in place for the second round of testing i'm sure.
 
I totally agree with how crappy this rating system is, but that can be discussed another time.

I know that some of you guys worry about the top players not playing, but I can at least vouch for myself. Since I got an account above 1400 I didn't stop playing, but instead put a second above 1400 as well. I am currently on a quest for my third 1400 account so I don't think that it's a matter of the top players stopping once they reach a certain mark. Instead I believe that many of them have continued to ladder and play on alts, to keep up with the metagame and to develp new and effective strategies. I really feel that we should set the cutoff at around 1350, as a number closer to 1400 would result in only about 15 people voting due to the prevalance of alts on the ladder.
 
I agree with idiotfrommars about the top players. Even after I reached 1400 I was constantly playing under alts and experimenting with new teams. Then, I lost a few matches under my main account and dropped 100 points. This system doesn't encourage consistency in players and instead encourages the use of alts. I propose having the cut off rank be around 1330 because I know that there are at least 3-4 people on the first page that have alts. I know a lot of good players that aren't on the first page right now that know what they're doing. Just my opinion though.

Edit: With the cutoff at 1330 there would be around 40-50 people voting excluding the number of alts. I think that this would encompass the player skill pool from good to best.
 
Yeah, I have to agree that the ranking system is annoying. A win and a loss moves you up and down like 10 places at a time. If you're unlucky for a bit and start tilting, you can easily lose up to 30-40 places in the blink of an eye.
 
I'm completely agreeing with firestrom here..Im not really sure what to say about the rating system as i never really tried to seriusly ladder too much before PO because quite frankly..Shoddy sucks and it hates me ;_;....

Well anyway the current system dsn't really encourage consistency. I really just quit laddering after hitting top 5 because its really hard to work with when you drop like 20 ranks for even one loss as after a certain benchmark its really really hard to find players above you so you have to do like "+4 -30".Idk about other players but i wouldn't blame them either if they stopped playing after hitting the rank.

Making the requirement Top X encourages that people keep playing as they have to constantly keep fighting for their rank.If they dont play for a week or so they will most probably fall below top 50 and would be forced to ladder to get it back.But i suppose its too late for that..

Oh and lets not lie to ourselves..Nobody is even trying to get 1500...
 
Yeah there is no point lying to ourselves about ranking here. Everyone seems to stop at 1400, but most likely the system will be readjusted for the next suspect test and then people can truly fight over ranks instead of stopping at a arbitrary cutoff point. I do feel that we should wait for the next test to attempt something like that though.
 
want a prefect example of how Fed up the ranking system is i slump for 2 days go 5-6 for 30 approximately i start in the top 50 (this was on the PO server not when i was not quite ready 2 jump into suspect just then) i come out outside of the top 600 yea if that isn't overkill deranking i don't know what is.
 
The sports example is pretty bad. Sports are a lot slower in evolving, mainly because the rules are generally pretty static over many, many years. That's why Michael Jordan could still reasonably talk about basketball and know what he's talking about. That's why old people who can't actually compete can readily coach teams. On the other hand, this metagame has been much, much more volatile. It was decidedly different two weeks ago than it is now.

Even despite all this, the cold, hard fact is that Shoddy's rating system actually makes an effort to prove its claims through the way in which it's calculated, and PO's rating system makes no such effort and is ultimately pretty arbitrary. Shoddy's rating system actually contains quite a bit of information that PO's utterly lacks, and Pokémon Lab's planned usage of Glixare was/is to be an improvement on that. In addition to this, I would consider the encouragement of alts a bad thing just because it makes general accountability so much harder to enforce.
 
Like I said before, the example is not the point. The point is that volatility and deviation were pointless - all they did was make ladder climbing harder for experienced players. Taking a break from Pokemon is punishable by a lower rating - why? Winning more than you lose is consistent enough to signify that you're pretty good at this game and deserve voting rights. If you get lucky and win vs. a top player with gamebreaking crits, you get a lot of rating, same way you did in Shoddy, but if you go ahead and lose to low ranking player after that, you lose rating in the same respect. If you win consistently, you'll be closer to the top and eventually able to vote.

Shoddy's rating system, for all it's worth, still encouraged players to make as many alts as humanly possible. I know guys who've said one loss in a power laddering session meant it was time for a new account. People keep doing this until they find just the right formula and head to top ten in like, one day. There's simply no way to guarantee a player's consistency. Beating the average player base should be enough to signify that you're good enough to weigh in your opinion and vote.
 
dude look @ my previous post real instance may have exaggerated the record a little but a 2 day slump like that + no more than 5 loses should not impact 1es record that badly.
 
dragonmaster951753 said:
Not just rain but all weather, and overcentralizing is a problem. Who wants to run nattorei on every team just so they don't get ran over by rain teams. Garchomp was overcentralizing not overpowered in gen 4 and that got him banned as an example
Click to expand...

God, I hate these threads so much.
"Who wants to do X to respond to Y" when we can just "ban Y".
It reeks of laziness and an unwillingness to evolve with the game but rather, to engineer it.
Blech...Puke..Vomit..
 
fidgety said:
dude look @ my previous post real instance may have exaggerated the record a little but a 2 day slump like that + no more than 5 loses should not impact 1es record that badly.
Click to expand...
That's a little different than the point Cape is arguing. I do remember Shoddy's system being slightly more lenient with the occasional loss, but I also remember constantly getting pissed off that some random idiot finds a little luck and sends my rating down ~ 20 points.
 
Status
Not open for further replies.
Back
Top