Besides movesets and teammates stats, I also started working on ways to classify teams into different archetypes such as Rain Stall or Weatherless Heavy Offense, as it should give us an interesting insight on the evolution of the metagame. I'm still working out the best formula, so if you're also in interested in seeing such stats for November and, like me, have a ton of PO team files sitting in your hard drives, I'd be very grateful if you could send them me so I can confront my script to them.
-
Welcome to Smogon! Take a moment to read the Introduction to Smogon for a run-down on everything Smogon, and make sure you take some time to read the global rules.
Official(!) Smogon University PO Stats -- September 2011
- Thread starter Antar
- Start date
- Status
I'd be really curious to see how you plan on working the algorithm. The simplest idea would be to take the ratio of offensive stats on a team (max(atk,spa)+spe, after EV, nature and item modifications) vs. defensive stats (hp x (def+spd)).
Another measure would be the prevalence of stall moves (status moves, entry hazards and phazing).
This is exactly what I was talking to you about, though--division of labor where we each work on metrics best derived from each of our usage stat files.
Politoed had a huge rise in June, which resulted in a centralizing effect in August.
But everybody got their usage reduced, even Politoed...
I bet Politoed cracks the top 3 and Tornadus cracks the top 10 now that Excadrill is uber.
I doubt tornados will make a jump into the top 10 , especially considering how it's hovering between OU and UU at the moment.
Tornadus is a pure Flying-type all things considered, so the weakness to common attacking types like Rock, Ice and Electric do not serve its SR weakness any better. It's Choiced set is powerful, and you can safely rely on Hurricane/Focus Blast/U-turn to get the most out of it offensively. Thundurus was able to abuse Prankster better, learned a solid set up in Nasty Plot and with its move coverage turned the metagame upside down; lastly its typing was a huge benefit to support HP Ice (which usually hit for super effective x4) and Focus Blast almost perfectly. Grass Knot and Hammer Arm for teams that were better prepared for the standard set could not handle its presense without gaping a huge whole in them.
I havn't been on PO for a while... I need to get better at battling.
| 515 | Illumise | 1 | 0.000% |
(assuming you were asking for the BW OU Rated...)
Ctrl+F tends to be very friendly toward the user when searching for a specific name in a list. (If only you could ctrl+f irl...)
prolly a whole bunchload of people couldn't accept that Smogon banned Thundurus and went to WiFi?
That's definitely not it, because Thundurus is banned on Wi-Fi as well (most people 'sav, sadly, so it's not like there's really any excuse).
Gastro *is* a good poke for weather teams. It works. So people probably figure, why mess with it?
It's really weird to see awesome pokemon like Tornadus and Mew be such on low in usage, especially Mew. The stallbreaker set shoulda been top 10 tbh. Either the SpDef based to be more annoying and the Def based to wall shit better, this thing's a beast. The NP set is such a boss as well. Tornadus I cant say too much, Specs set and Taunt Hammer Arm LO sets are there, I really doubt Tornadus will stay UU lol
So, just a little taste of what's coming next month, I've rewritten my log reader script. Previously, all the thing did was parse a battle log and pull out the names of the pokemon that appear in the battle.
Now, here's what it generates:
Okay, so what are you looking at? The first two sections give (1) the name of the pokemon that appears in the battle, (2) the number of KOs that pokemon got in the match* and (3) the number of turns the pokemon was in the battle.
And what about that last section? It tells you what happened with each and every matchup.
With this data, I'll be able to tell you not only who the best sweeper (most kills per battle) is and who the best staller (most turns out per battle), but I'll be able to generate a matrix which will tell you what happens when x goes against y. So, for instance, we might find that Ludicolo is the best counter for Kyogre (random example--not saying it's true), because in almost every matchup between the two, Kyogre is either KOed or is scared out.
So something to look forward to :)
*I'll probably be taking this out, since it can be derived from the last section.
Now, here's what it generates:
Code:
ZcX
Mienshao: 3,6
Espeon: 0,4
Volcarona: 1,6
Abomasnow: 1,1
Scizor: 1,6
Cloyster: 0,1
***
Darkrais shadow
Ferrothorn: 0,3
Gastrodon: 0,3
Heatran: 3,9
Landorus: 0,1
Conkeldurr: 3,5
Gengar: 0,3
***
Ferrothorn vs. Mienshao: Ferrothorn switched out
Gengar vs. Mienshao: Mienshao switched out
Gengar vs. Scizor: Gengar switched out
Ferrothorn vs. Scizor: Ferrothorn was KOed
Conkeldurr vs. Scizor: Scizor was KOed
Conkeldurr vs. Espeon: Conkeldurr switched out
Espeon vs. Gastrodon: Espeon switched out
Gastrodon vs. Volcarona: Gastrodon switched out
Gengar vs. Volcarona: Gengar was KOed
Heatran vs. Volcarona: Volcarona was KOed
Heatran vs. Mienshao: Heatran switched out
Landorus vs. Mienshao: Mienshao switched out
Abomasnow vs. Landorus: Landorus was KOed
Abomasnow vs. Conkeldurr: Abomasnow was KOed
Conkeldurr vs. Espeon: Conkeldurr switched out
Espeon vs. Heatran: Espeon was KOed
Cloyster vs. Heatran: Cloyster was KOed
Heatran vs. Mienshao: Heatran was KOed
Gastrodon vs. Mienshao: Gastrodon was KOed
Mienshao vs. Conkeldurr: double down
---Okay, so what are you looking at? The first two sections give (1) the name of the pokemon that appears in the battle, (2) the number of KOs that pokemon got in the match* and (3) the number of turns the pokemon was in the battle.
And what about that last section? It tells you what happened with each and every matchup.
With this data, I'll be able to tell you not only who the best sweeper (most kills per battle) is and who the best staller (most turns out per battle), but I'll be able to generate a matrix which will tell you what happens when x goes against y. So, for instance, we might find that Ludicolo is the best counter for Kyogre (random example--not saying it's true), because in almost every matchup between the two, Kyogre is either KOed or is scared out.
So something to look forward to :)
*I'll probably be taking this out, since it can be derived from the last section.
....
wow.
wow.
That should be interesting info. Pretty good feature.
Wow, thats amazing, I might make a stall team simply to work with your stats, or not, whatever.
Moves, items, EV spreads and abilities will all be tracked starting with November.
- Status