(Un)Official Smogon University PO Stats -- August 2011

Status
Not open for further replies.
Players with < 1001 must be deliberately losing or being majorly inactive to have such ratings. I mean, players start out with rating of 1000. Don't worry, Shuckle Man, the great majority of the players are accounted for.
A few months back, when the server was loopy, I made a team that accidentally lacked a member and had a few moves missing or wrong (like min happiness + return). I tried to update my team, and played with it, but every time I uploaded it it failed, before I realized, my rating was like 870 or something, pretty embarassing.
 

Oglemi

Borf
is a Top Contributoris a Tournament Director Alumnusis a Site Content Manager Alumnusis a Community Contributor Alumnusis a Researcher Alumnusis a Tiering Contributor Alumnusis a Top Smogon Media Contributor Alumnusis an Administrator Alumnusis a Top Dedicated Tournament Host Alumnus
^can that convo please be carried over to the proper thread, thanks or better yet not at all

bluemon said:
is there a possibility that the OU cut off percentage might change?
no, 3.41% is really the only logical choice when backed by the math. Plus, why would we change it just to please some people?
 
no, 3.41% is really the only logical choice when backed by the math. Plus, why would we change it just to please some people?
What?

The math just states that 3.41% is the corresponding percent if we want a pokemon used 1 out of 20 teams.



I don't really see how "1 out of 20" is the "only logical choice" for the OU cutoff, considering it was chosen, iirc, with some disagreement that it should be 1 out of 10, or other convenient, but for the most part, arbitrary, number.
 
Thanks for the stats Antar! I would like to say I second Malvira on this-

Smogon is a competitive community first, and that is why I believe that there should still be a cutoff (1001) so that the stats are more representative of the competitive community.
Most of those users are running stuff like Tropius or Charizard with a random Haxorus. Not only do those teams consist of the most random pokemon, but also sets like Thunder Wave Thunderbolt Taunt Substitute Thundurus.
 

Oglemi

Borf
is a Top Contributoris a Tournament Director Alumnusis a Site Content Manager Alumnusis a Community Contributor Alumnusis a Researcher Alumnusis a Tiering Contributor Alumnusis a Top Smogon Media Contributor Alumnusis an Administrator Alumnusis a Top Dedicated Tournament Host Alumnus
Well more than anything it would be consistent with what we've done in the past, I guess saying it was the best "math wise" was a poor choice of words lol
 

Katakiri

Listen, Brendan...
is a Researcher Alumnus
Call me a data whore, but I do love usage stats. (Even the PO Server's for what they're worth) God help me I do. Great job with this!

Gen 1 OU: 14 Pokemon
Gen 2 OU: 22 Pokemon
Gen 3 OU: 35 Pokemon
Gen 4 OU: 48 Pokemon
Gen 5 OU: 44 Pokemon

♪One of these things is not like the others~ One of these things does not belong~ ♪


Mienshao, Tornadus, Darmanitan, Cloyster, at this rate BL's shaping up to be a fun tier. ...oh wait.

At least Flygon might have Machamp, Metagross & Swampert to keep it company in UU now. If we get 20 more people to stop using Lucario, we can have a little late-Gen 4 OU reunion going on.

Also, Silly Duck & Cotton Ball, what are you doing so deep into UU? Well I guess Cloyster can check Whimsy for the 1 month Cloyster trolls UU...assuming he doesn't mind Stun Spore. And Porygon should go back to laughing at Gliscor & Dragonite. What the duck is he doing this low when those two are that high? I don't care if Mama Heatran crawled back out of her Garchomp-proof Bomb-shelter she hides in until Chomp's banned each Gen, those two are still stupid in OU.

OH! and Ninetales is above Deoxys-S?! I'm like the fast-talking used cars salesman of Drought but even I'm disgusted by that. Sun too good for 2 free layers of anything you want by turn 2?

What a terrible, shrinking metagame OU is. Time to play a few more hours of it before bed. Terrible.
 
yea i love porygon2. he is my answer to anything wallish or sweeperish on my trick room team. he also has the valuable trace to get poison heal (Beast) or multiscale.

lucario is also underrated. if you dont have the proper checks, one can sweep a team with ease.
 
yea i love porygon2. he is my answer to anything wallish or sweeperish on my trick room team. he also has the valuable trace to get poison heal (Beast) or multiscale.

lucario is also underrated. if you dont have the proper checks, one can sweep a team with ease.
Agreed about Lucario. Swords Dance Lucario with Air Balloon is so easy to set up, and it can defeat pokemon such as Excadrill and Landorus, while still hitting very hard. It should definitely see more use; steel typing allows it to set up on the likes of Dragonite locked into Outrage (it'll hurt, but if Dragonite is weakened a little bit, Lucario can Swords Dance and ko with Extremespeed).
 
another thing that makes lucario potent is the versatility. you see a baloon lucario so you assume its swords dance. you switch in your gliscor to realize that its the nasty plot variant. you switch to your special wall and.... its now at +4. gg.

another thing is extremespeed. the fact that it's +2 priority Means that even infernape's mach punch cant hit it before it dies or a speed boost sharpedo's aqua jet.

lucario can be scary no doubt.
 
Antar,
I have sent you a message with better formatted RU usage stats on it.
Hey get off my turf bro.

Naw just kidding, I could use some help, I have done every tier listed except LC, so if you want you can have that. But I only go until it gets ridiculus, which means I only do half of the statistics, so if you want you can finish the rest of them. Antar also mentioned in the future he will get other stats, like doubles and VGC, I I guess we can split those when they come.
 
Antar,
I have sent you a message with better formatted RU usage stats on it.
Hey get off my turf bro.

Naw just kidding, I could use some help, I have done every tier listed except LC, so if you want you can have that. But I only go until it gets ridiculus, which means I only do half of the statistics, so if you want you can finish the rest of them. Antar also mentioned in the future he will get other stats, like doubles and VGC, I I guess we can split those when they come.
Just letting BOTH OF YOU know, when the next round of stats come out, your services will no longer be needed, because I'll have a script to auto-generate the nicely formatted stats automatically. I just didn't have the time to write the script this weekend.

(also, Hunt, the stats you sent me didn't come out correctly. You probably forgot to use the CODE tags--as it was, it all came out on one line).
 
I'd like to focus the discussion a little bit.

Our first priority right now should be to come up with a new OU/UU (and UU/RU, and, preferrably RU/NU) usage-based cutoff.

There are several components to this.
  1. Do we trust that there were no bugs in my scripts?
    If you have PO save your battle logs, those will work just fine. Usage is as follows (the first script I named "LogReader.py" the second one is "StatCounter.py"):
    Code:
    $python LogReader.py "name-of-log.html"
    Make sure you have a subfolder called "Raw" in the folder where you're running the script. Then, once you've run LogReader on a bunch of scripts, run:
    Code:
    $python StatCounter.py "Raw/Tier-file.txt"
    making sure you have another subfolder called "Stats" for the script to write to.
  2. Do we want to make any modifications to the method by which the stats are collected? This is the one where I'd like the most input (because I assume not too many of you know python). My method currently is to count ALL battles, regardless of players' rating, where the battle lasted longer at least six turns. One suggested alternative would be to only count teams of trainers whose current ranking (end-of-month ranking, for future stat cycles) is greater than 1000 (on this note, does anyone know how I can download a list of current rankings? The PO rankings viewer doesn't let you copy/paste).
  3. Finally, how do we mesh this month's stats with those from previous months, seeing as they were generated using radically different methods? This question should very much tie into the previous one.
 
If you do plan on tiering based on ranking, it would probably be easier to count them based on what there ranking was at the time. Also, sure they could get lucky and get over 1000 for one day, but then again if they got over 1000 just because of luck, I doubt it would affect the stats much, and if they are hovering around 1000, they aren't that terrible.
 
If you do plan on tiering based on ranking, it would probably be easier to count them based on what there ranking was at the time.
As I've pointed out in the past, unless I'm missing something (look at some battle logs, yourself), this is not possible, since battle logs do not contain current rankings, only variations.
 
Oh really, ok sorry.

Well if you want a list you can use the list on the server, although it only contains the currently ratings, so you would need to ask around if someone actually kept track of the last stats. If not, then this time around we would need someone to do that.

Sorry I can't be much of a help, I only know a bit of HTML coding, thats the best I can get.
 
If you don't count battles by lower-ranking players, IMHO it helps to undermine the stats, because then you're no longer representing the entire community. You're only representing the more talented battlers, and I dont think that's fair.
 

Meru

ate them up
is a Forum Moderator Alumnusis a Community Contributor Alumnusis a Tiering Contributor Alumnusis a Contributor Alumnusis a Past SPL Championis a Past WCoP Champion
I'd hardly call representing people above 999 the "more talented battlers". I do think there is a baseline level of competency that should be taken into account for statistics that we're using to deduce something about a competitive metagame.
 
If NU is created soon, it's gonna be full of some beasts. Things like Beedrill, Ledian, and Furret are going to have a hard time fitting in anywhere.
 
@Antar:

First of all, thank you so much for your hard work. Statistics are one of the most important tools that smogon provides.

As far as the statistics go, you should try to make the statistics as similar as possible to what DJD/Rising Dusk did. The more different the methods, the less comparable your stats will be to their stats. This is important because people make tiering decisions and judge centralization based on the stats. If the stats are somehow different from before, then people's perceptions of what counts as centralized/broken/whatever will no longer be applicable. There is a value to continuity and meeting user expectations.

I know that you can't perfectly replicate what DJD/Rising Dusk did because their stats included pokemon not used in the battle. However, I don't think this will make a big difference (in my personal, subjective opinion); the minimal difference in the data when you excluded battles <6 turns somewhat backs this up.

Also, this should probably be a discussion in policy review/the backroom.
 
Wow, thank you for this, it must have took some work. Why exactly can't the Smogon Server use the usage stat plugin ? The PO server doesn't seem to have any issue with it.
 
August looks interesting,

Machamp, Cloyster, Darmanitan and Erufuun are all in range of becoming UU and Gastrdon has the usage of an OU Pokemon.
 
Status
Not open for further replies.

Users Who Are Viewing This Thread (Users: 1, Guests: 0)

Top