What? This thread is for all tiers. That's why the stats for more than just OU are in the OP.
-
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.
(Un)Official Smogon University PO Stats -- August 2011
- Thread starter Antar
- Start date
- Status
What? This thread is for all tiers. That's why the stats for more than just OU are in the OP.
As for your 2 questions posted on Policy Review:
1 - Go with A - as Oglemi stated, May and June are rather outdated, thus your stats is more informative; don't discard it. Secondly, it's been 3+ months since the last tier change; UU and RU is ready for an update. We cannot wait for an additional 2 months for the tier change.
2 - Personally your "Full Teams Only" seems more "reasonable" than the original method, but seeing how it represents only 55% of the battles the data may be even more skewed.
I am liking Maestro's method, since it does take into account May and June's stats, but if the data collection method is drastically different between Rising Dusk's method vs your method, then combining the three results may be inappropriate.
If Maestro's method is sound, I'd go with Maestro's Method. If it isn't, I'll go with your original method, since it represents a greater bulk of battles than the full-team method.
1 - Go with A - as Oglemi stated, May and June are rather outdated, thus your stats is more informative; don't discard it. Secondly, it's been 3+ months since the last tier change; UU and RU is ready for an update. We cannot wait for an additional 2 months for the tier change.
2 - Personally your "Full Teams Only" seems more "reasonable" than the original method, but seeing how it represents only 55% of the battles the data may be even more skewed.
I am liking Maestro's method, since it does take into account May and June's stats, but if the data collection method is drastically different between Rising Dusk's method vs your method, then combining the three results may be inappropriate.
If Maestro's method is sound, I'd go with Maestro's Method. If it isn't, I'll go with your original method, since it represents a greater bulk of battles than the full-team method.
Nng... You're either cutting out some or all of the Pokemon of a team that doesn't send out all their Pokemon. In a battle that runs to completion, that can only be the winner's team... Either pivots and leads are artificially inflated, Pokemon that aren't leads or pivots are artificially deflated, or winners, especially landslide winners, aren't always counted. They're all poor options, and I'd say it's for that reason that in particular, ignoring teams that don't send out all their Pokemon, while looking good on paper, is an option best tossed out now.
Out of what remains, there's one point I've seen made several times that makes a lot of sense. Pokemon not used in a battle, even if it's because they're on the winning team, aren't affecting battles in that tier as much as ones that are. For this reason, the first version of the stats may well be the best. However, it changes the expected percentages; it's not really compatible with previous months' stats, and it may require a change in the OU cutoff, especially if it's kept for future months. If the intention is to adopt one of these methods permanently, I feel that the best option is to go with the first, which ignores Pokemon that don't take part in battle, while discarding the previous months' stats due to the time difference and perhaps lowering the OU cutoff in consideration of the change. Could be done by changing the numbers involved, could be done with the crude adjustment to make the totals add up to 600%, or it could just be left as is with the knowledge of the changed expectations of OU.
Out of what remains, there's one point I've seen made several times that makes a lot of sense. Pokemon not used in a battle, even if it's because they're on the winning team, aren't affecting battles in that tier as much as ones that are. For this reason, the first version of the stats may well be the best. However, it changes the expected percentages; it's not really compatible with previous months' stats, and it may require a change in the OU cutoff, especially if it's kept for future months. If the intention is to adopt one of these methods permanently, I feel that the best option is to go with the first, which ignores Pokemon that don't take part in battle, while discarding the previous months' stats due to the time difference and perhaps lowering the OU cutoff in consideration of the change. Could be done by changing the numbers involved, could be done with the crude adjustment to make the totals add up to 600%, or it could just be left as is with the knowledge of the changed expectations of OU.
Dont want to waste your time, but have you tried a light version of your "full-team" approach? Maybe account teams that have at least 5 or 4 Pokemon active rather than 6? This may provide the best of both worlds - like your original method, it would represent a larger bulk of the teams played, while retaining the "reasonableness" of the "full-team" results. Full-team was a bit on the extreme end anyways.
while that's not a bad suggestion, that's BASICALLY what the stats are now. I've posted this graph before, but I'll post it again:
This is a graph of pokemon used per battle (after taking out battles forfeited before the sixth turn). As you can see, it's heavily skewed towards the 11-12 end, with almost 60% of battles containing at least 11 pokemon (meaning ALL teams are guaranteed to have at least five pokemon out). Battles with fewer than eight pokemon account for less than 20% of all battles (or so I'm guesstimating). Their impact on the stats is even LESS significant, since each POKEMON is weighted evenly. So the "original method" is already basically what you're saying. Still, I'd be more than happy to run it and see what changes. Still, if you're curious...
Weighted by Maestro's method (divide by total pokes, multiply by 6)
Code:
Total battles: 236958
Total teams: 413002
Total pokemon: 2265511.0
+ ---- + --------------- + ------ + ------- +
| Rank | Pokemon | Usage | Percent |
+ ---- + --------------- + ------ + ------- +
| 1 | Tyranitar | 89349 | 23.663% |
| 2 | Rotom-W | 79164 | 20.966% |
| 3 | Scizor | 77710 | 20.581% |
| 4 | Gliscor | 77105 | 20.421% |
| 5 | Ferrothorn | 75258 | 19.931% |
| 6 | Politoed | 68318 | 18.093% |
| 7 | Dragonite | 66734 | 17.674% |
| 8 | Heatran | 60301 | 15.970% |
| 9 | Jirachi | 56717 | 15.021% |
| 10 | Latios | 55349 | 14.659% |
| 11 | Skarmory | 48725 | 12.904% |
| 12 | Excadrill | 47724 | 12.639% |
| 13 | Thundurus | 39758 | 10.530% |
| 14 | Reuniclus | 34295 | 9.083% |
| 15 | Haxorus | 33127 | 8.773% |
| 16 | Jellicent | 32243 | 8.539% |
| 17 | Gengar | 32047 | 8.487% |
| 18 | Conkeldurr | 31834 | 8.431% |
| 19 | Celebi | 31372 | 8.309% |
| 20 | Starmie | 31277 | 8.283% |
| 21 | Infernape | 31004 | 8.211% |
| 22 | Forretress | 29684 | 7.862% |
| 23 | Ninetales | 28883 | 7.649% |
| 24 | Gyarados | 28597 | 7.574% |
| 25 | Magnezone | 27667 | 7.327% |
| 26 | Landorus | 27233 | 7.212% |
| 27 | Deoxys-S | 26858 | 7.113% |
| 28 | Terrakion | 26260 | 6.955% |
| 29 | Blissey | 25607 | 6.782% |
| 30 | Volcarona | 25561 | 6.770% |
| 31 | Salamence | 24640 | 6.526% |
| 32 | Latias | 24204 | 6.410% |
| 33 | Gastrodon | 23374 | 6.190% |
| 34 | Tentacruel | 22736 | 6.021% |
| 35 | Espeon | 21372 | 5.660% |
| 36 | Toxicroak | 20686 | 5.478% |
| 37 | Hydreigon | 20492 | 5.427% |
| 38 | Bronzong | 20424 | 5.409% |
| 39 | Vaporeon | 20078 | 5.317% |
| 40 | Breloom | 19400 | 5.138% |
| 41 | Scrafty | 19051 | 5.045% |
| 42 | Mamoswine | 16738 | 4.433% |
| 43 | Lucario | 16662 | 4.413% |
| 44 | Virizion | 15602 | 4.132% |
| 45 | Venusaur | 14441 | 3.825% |
| 46 | Metagross | 14079 | 3.729% |
| 47 | Mienshao | 13358 | 3.538% |
| 48 | Swampert | 13309 | 3.525% |
| 49 | Chandelure | 13109 | 3.472% |
| 50 | Tornadus | 12831 | 3.398% |
| 51 | Hippowdon | 12333 | 3.266% |
| 52 | Donphan | 12103 | 3.205% |
| 53 | Zapdos | 11826 | 3.132% |
| 54 | Darmanitan | 11733 | 3.107% |
| 55 | Dugtrio | 11263 | 2.983% |
| 56 | Quagsire | 11229 | 2.974% |
| 57 | Slowbro | 11219 | 2.971% |
| 58 | Mew | 11040 | 2.924% |
| 59 | Cloyster | 10990 | 2.911% |
| 60 | Chansey | 10971 | 2.906% |
| 61 | Kingdra | 10607 | 2.809% |
| 62 | Jolteon | 10274 | 2.721% |
| 63 | Machamp | 9914 | 2.626% |
| 64 | Abomasnow | 9675 | 2.562% |
| 65 | Porygon2 | 9299 | 2.463% |
| 66 | Roserade | 8533 | 2.260% |
| 67 | Whimsicott | 7971 | 2.111% |
| 68 | Togekiss | 7535 | 1.996% |
| 69 | Azelf | 7262 | 1.923% |
| 70 | Azumarill | 6800 | 1.801% |
| 71 | Tangrowth | 6707 | 1.776% |
| 72 | Victini | 6685 | 1.770% |
| 73 | Weavile | 6632 | 1.756% |
| 74 | Snorlax | 6459 | 1.711% |
| 75 | Arcanine | 6450 | 1.708% |
| 76 | Hitmontop | 6437 | 1.705% |
| 77 | Zoroark | 6348 | 1.681% |
| 78 | Cresselia | 6199 | 1.642% |
| 79 | Sharpedo | 6185 | 1.638% |
| 80 | Electivire | 6014 | 1.593% |
| 81 | Nidoking | 5988 | 1.586% |
| 82 | Sawsbuck | 5961 | 1.579% |
| 83 | Milotic | 5532 | 1.465% |
| 84 | Suicune | 5504 | 1.458% |
| 85 | Dusclops | 5052 | 1.338% |
| 86 | Gorebyss | 4796 | 1.270% |
| 87 | Sigilyph | 4676 | 1.238% |
| 88 | Ninjask | 4654 | 1.233% |
| 89 | Flygon | 4539 | 1.202% |
| 90 | Kyurem | 4358 | 1.154% |
| 91 | Aerodactyl | 4164 | 1.103% |
| 92 | Smeargle | 4139 | 1.096% |
| 93 | Empoleon | 4113 | 1.089% |
| 94 | Blastoise | 3927 | 1.040% |
| 95 | Staraptor | 3927 | 1.040% |
| 96 | Heracross | 3892 | 1.031% |
| 97 | Gallade | 3872 | 1.025% |
| 98 | Ludicolo | 3861 | 1.023% |
| 99 | Charizard | 3814 | 1.010% |
| 100 | Alakazam | 3752 | 0.994% |Edit: by inspection, I've found a few transpositions (Infernape and Starmie; Forretress and Ninetales), but by and large, everything is where it should be, and within fractions of percents from each other. The OU/UU cutoff is in the exact same place.
Edit 2: oh, ha! It looks like I never published the August OU stats for Maestro's method, all teams. You can, of course, derive it yourself from the "original method" stats, but here it is, for your convenience:
All teams where the battle lasted at least six turns are represented. Percent = 100*usage/total pokemon*6.
Code:
Total battles: 236958
Total teams: 473916
Total pokemon: 2421865.0
+ ---- + --------------- + ------ + ------- +
| Rank | Pokemon | Usage | Percent |
+ ---- + --------------- + ------ + ------- +
| 1 | Tyranitar | 95686 | 23.706% |
| 2 | Rotom-W | 84284 | 20.881% |
| 3 | Scizor | 81761 | 20.256% |
| 4 | Gliscor | 81449 | 20.178% |
| 5 | Ferrothorn | 80902 | 20.043% |
| 6 | Politoed | 75124 | 18.611% |
| 7 | Dragonite | 71521 | 17.719% |
| 8 | Heatran | 64319 | 15.935% |
| 9 | Jirachi | 60943 | 15.098% |
| 10 | Latios | 58072 | 14.387% |
| 11 | Skarmory | 51586 | 12.780% |
| 12 | Excadrill | 49093 | 12.162% |
| 13 | Thundurus | 42488 | 10.526% |
| 14 | Reuniclus | 36344 | 9.004% |
| 15 | Haxorus | 35503 | 8.796% |
| 16 | Jellicent | 34192 | 8.471% |
| 17 | Conkeldurr | 34130 | 8.455% |
| 18 | Gengar | 34040 | 8.433% |
| 19 | Celebi | 33407 | 8.276% |
| 20 | Infernape | 32714 | 8.105% |
| 21 | Starmie | 32701 | 8.101% |
| 22 | Ninetales | 32008 | 7.930% |
| 23 | Forretress | 31519 | 7.809% |
| 24 | Gyarados | 31012 | 7.683% |
| 25 | Deoxys-S | 30247 | 7.493% |
| 26 | Magnezone | 29488 | 7.305% |
| 27 | Landorus | 28417 | 7.040% |
| 28 | Volcarona | 28245 | 6.997% |
| 29 | Terrakion | 27355 | 6.777% |
| 30 | Blissey | 26982 | 6.685% |
| 31 | Salamence | 26015 | 6.445% |
| 32 | Latias | 25748 | 6.379% |
| 33 | Gastrodon | 24750 | 6.132% |
| 34 | Espeon | 24039 | 5.955% |
| 35 | Tentacruel | 23934 | 5.929% |
| 36 | Toxicroak | 22186 | 5.496% |
| 37 | Vaporeon | 22044 | 5.461% |
| 38 | Bronzong | 21951 | 5.438% |
| 39 | Hydreigon | 21562 | 5.342% |
| 40 | Breloom | 21085 | 5.224% |
| 41 | Scrafty | 20959 | 5.192% |
| 42 | Mamoswine | 17639 | 4.370% |
| 43 | Lucario | 17498 | 4.335% |
| 44 | Virizion | 16658 | 4.127% |
| 45 | Metagross | 15163 | 3.757% |
| 46 | Venusaur | 15139 | 3.751% |
| 47 | Mienshao | 14377 | 3.562% |
| 48 | Swampert | 14349 | 3.555% |
| 49 | Chandelure | 13982 | 3.464% |
| 50 | Hippowdon | 13597 | 3.369% |
| 51 | Tornadus | 13422 | 3.325% |
| 52 | Donphan | 12884 | 3.192% |
| 53 | Zapdos | 12611 | 3.124% |
| 54 | Darmanitan | 12394 | 3.071% |
| 55 | Mew | 12250 | 3.035% |
| 56 | Cloyster | 12086 | 2.994% |
| 57 | Quagsire | 11882 | 2.944% |
| 58 | Dugtrio | 11779 | 2.918% |
| 59 | Slowbro | 11774 | 2.917% |
| 60 | Chansey | 11585 | 2.870% |
| 61 | Kingdra | 11389 | 2.822% |
| 62 | Jolteon | 10885 | 2.697% |
| 63 | Machamp | 10854 | 2.689% |
| 64 | Abomasnow | 10408 | 2.579% |
| 65 | Porygon2 | 10030 | 2.485% |
| 66 | Roserade | 9174 | 2.273% |
| 67 | Whimsicott | 8725 | 2.162% |
| 68 | Togekiss | 8156 | 2.021% |
| 69 | Azelf | 8092 | 2.005% |
| 70 | Victini | 7063 | 1.750% |
| 71 | Weavile | 7060 | 1.749% |
| 72 | Azumarill | 7053 | 1.747% |
| 73 | Tangrowth | 7008 | 1.736% |
| 74 | Snorlax | 6884 | 1.705% |
| 75 | Zoroark | 6784 | 1.681% |
| 76 | Arcanine | 6770 | 1.677% |
| 77 | Hitmontop | 6711 | 1.663% |
| 78 | Cresselia | 6571 | 1.628% |
| 79 | Sharpedo | 6546 | 1.622% |
| 80 | Nidoking | 6430 | 1.593% |
| 81 | Electivire | 6356 | 1.575% |
| 82 | Ninjask | 6224 | 1.542% |
| 83 | Sawsbuck | 6170 | 1.529% |
| 84 | Suicune | 5905 | 1.463% |
| 85 | Milotic | 5831 | 1.445% |
| 86 | Gorebyss | 5705 | 1.413% |
| 87 | Dusclops | 5356 | 1.327% |
| 88 | Sigilyph | 5148 | 1.275% |
| 89 | Smeargle | 5003 | 1.239% |
| 90 | Flygon | 4750 | 1.177% |
| 91 | Aerodactyl | 4620 | 1.145% |
| 92 | Kyurem | 4540 | 1.125% |
| 93 | Empoleon | 4393 | 1.088% |
| 94 | Gallade | 4214 | 1.044% |
| 95 | Staraptor | 4182 | 1.036% |
| 96 | Blastoise | 4131 | 1.023% |
| 97 | Ludicolo | 4125 | 1.022% |
| 98 | Heracross | 4082 | 1.011% |
| 99 | Charizard | 4022 | 0.996% |
| 100 | Alakazam | 3973 | 0.984% |
So then we should be using either one of these methods yes? IDK what exactly we're waiting for then lol. If possible just pick one of these stats and then combine them with May and June and see what happens.
That would be my choice as well (Option A, Maestro's Method, in terms of the PR thread). In that case, the hippo stays OU. As do Vensaur, Swampert, Chandelure and Mienshao.
Code:
| 46 | Aggron | 988 |3.24509% |EDIT: It should be way higher.
Not Scizor, unless he gets a burn that is....
Well all that UU Swampert stuff was nonsense it seems.
(fuck yeah go swampert you boss)
(fuck yeah go swampert you boss)
Gastrodon switches in:
Rotom-W used hydro pump / volt switch, Gastrodon absorbs water / is immune. Total stop, as now it has to switch and your opponent loses his momentum.
I'm probably missing something here, but the latest PO client shows both teams at the start of a log, so wouldn't that remove the issue of only recording the pokemon that were sent out?
Thanks for your effort in bringing us stats btw!
Thanks for your effort in bringing us stats btw!
The client does, the server does not (currently). Innocent Criminal is working on it. Plus, the Smogon server is a few versions out of date.
Gastrodon takes around 70% from a CB Scizor U-Turn, IIRC.
Antar, are there any other pokemon like gastrodon that, no matter which method of counting you choose, whether it be 'my method', or counting full teams, or normalized counting the teams who reveal 4 or more pokes, will undoubtedly be moving tiers? Perhaps you should mention in the Policy Review thread the option to freeze the pokemon on the borderline of their tier (Hippo & co.) and only move the ones that will definitely be moving. One could argue that the stats aren't precise or accurate enough to be able to declare a close call in either direction, so it's safer to not change.
However, of course there's the argument that the tiers are getting stale (so movement might be for the better), and that they are contracting (so any borderline movement would allow pokes to be used in more tiers and not less), but it might be worth bringing up.
However, of course there's the argument that the tiers are getting stale (so movement might be for the better), and that they are contracting (so any borderline movement would allow pokes to be used in more tiers and not less), but it might be worth bringing up.
I just posted the new potential tier lists in my thread in PR. Haven't done a line-by-line comparison yet, but Tangrowth popped out at me...
[Edit: let's see... Sigilyph... Xatu...]
Celebi and Mamoswine to OU. Tornadus, Darmanitan, Machamp and Whimsicott moving down to UU and Rhyperior moving up to UU. Not sure about anything else.
You meant UU, right? That freaked me out at first, because if Rhyperior is OU, something is seriously wrong.
Yeah, I meant UU.
Maestro, Aerodactyl, Charizard, Sawsbuck, Sigilyph, Spiritomb, Victreebel, and Xatu are headed to RU.
How can E-Vire, Feraligatr, Sceptile, Bouffalant and others be NU... It doesn't make sense.
They're not.
Electivire: RU
Sceptile: UU
Feraligatr hat a lot of stiff competition these days, and Bouffalant is slow as molasses.
why? Electivire and Bouffalant suck
This
The UU itself feels like gen 4 OU, meaning OU AND UU meta both evolved into stronger, faster, etccc
Yeah, 4G OU shares a lot more in common with 5G UU than OU, IMO. Interestingly, I don't really feel like RU is very much like 4G UU...
- Status