[Competitive] Tournament Usage Stats [Update Post #30]

[kd24]

So this is an idea that hopefully progresses overtime but it's something I have thought about for a while. While the ladder is a solid way to gather information and a general idea of competitive play, there is a distinction between ladder play and tournament play and as of now, we generally ignore it. I think that for a ladder, you build for long-term success and you expect worse competition, while for tournaments you have much tougher competition while building for the short-term. To me, I think I would get a better idea of what the metagame looks like from statistics gathered from these competitive players.

So where does this leave us? Well, as it happens, our first BW Smogon Tour will be tomorrow (no Easter postponement this season). My goal is to collect teams from our players in the top 16 onward [thanks to locopoke for telling me to cut my original goal of round 1 to top 16, there are a ton of reasons this change in plan is good, which i'll discuss later]. I'll post the statistics here and try my best to get stats from each player. If you're confused, this means:

Top 16 of Tour is up. I PM every player, asking which 6 Pokemon are on their team, and I will also try to get moveset statistics as well. After top 16, I repeat the process for top 8, noting teams that were also used last round. This continues every round until the winner's decided. I'm going to do this for all 3 Tour Days if possible (Easter might fuck with me, which is where I may need more volunteers). I'll post the findings here for discussion or just so you can get an idea of what happens to the metagame.

The problem with doing it straight from round 1?
- Bad players can enter and lose Round 1 with a crappy team, skewing statistics
- Easier for me to not get a hold of a player
- Easier for a player to lie to me [This can still happen, let's hope no one is immature enough to do this]
- Most players in top 16 will be on IRC, most players from Round 1 could be anywhere

I'm making this thread now for general predictions about what this information might show, as well as opening it up to you guys for suggestions on how to collect these kinds of stats. If there is a more efficient way, I definitely want to hear it.

My Predictions:
- The relevant statistics we have (Suspect Statistics of February) showed that 40% of teams are weatherless, with less than 20% being Rain. I am going to predict now that only 10% of the teams throughout this week will be weatherless. I'll go one step further and predict a breakdown, from Friday -> Sunday, of:
~ 55% Sand
~ 30% Rain
~ 10% Weatherless
~ 5% Sun
~ 0% Hail [Prove me Wrong?]

-

will be the most used Pokemon,

will be number 2. Tyranitar is going to be acting as its general badass self, while checking Rain and Sun, and adding pressure to teams by Pursuiting the common Latias and Latios (I think these are tournament staples for a lot of players). Keldeo is just really good, I think it's not appreciated enough right now, but you need to remember that it is the best special attacking water-type and fighting-type in the tier. Maybe Rotom-w and Celebi are used more though? Their utility in countering top threats and the fact that they really just fit on any team help their case.

I should note that if I try to do this for the entire tournament, I think Rain would be > 50% and dominate most of it, as its the easiest weather to use efficiently. I think players will be overwhelmingly prepared for it that only 30% of them are used from then on, but I could be wrong.

Use this thread to post predictions or give advice on how to improve this method. Tomorrow night, I'll post the Friday statistics. I'd like to give my appreciation to Ojama, who has offered to ask for these statistics with me. If you're interested in volunteering, PM me on IRC and I'll tell you what I want you to do.

BTW, if anyone reading this decides to play Tour and makes top 16, I'd love for you to post your team, a short rundown on how or why you built it the way you did, as well as any surprises you had while playing. If you do, I'll have an index of "notable users" [hall of fame so to speak], detailing their highlights. I obviously can't promise anything but that would be the kind of thing I look for in a Community Contributor as well - insightful and a good battler.

Let's hope this thread and task isn't a colossal failure :P. Thanks for reading and let's get some stats :]

 

[Deluks917]

I applaud you doing this.

However if this is a big event (say 128-192) I think top32 would be very useful. Though admittedly this is alot more work.

I tihnk you should offer a way to make this more anonymous. I am sure there are people who would be willing to contribute stats so you could tell how many Landorus were in top16. But not willing to publically post their exact teams.

 

[Joel]

I applaud you doing this.

However if this is a big event (say 128-192) I think top32 would be very useful. Though admittedly this is alot more work.

I tihnk you should offer a way to make this more anonymous. I am sure there are people who would be willing to contribute stats so you could tell how many Landorus were in top16. But not willing to publically post their exact teams.

They could maybe PM the teams, on the promise the team users name wasn't revealed, supposing he wasn't a factor in the specific tourney.

 

[Lavos]

i hope i make top 16 so you can all admire my jynx

 

[MMII]

0% Stall
100% Offense

At least for the guys that'll win anyways.

 

[Kidogo]

Im gonna try to prove you wrong MM...also I gotta say, I may use hail one round if i make top 16 just to screw the stats xD

Cool idea kd!

 

[Soul Fly]

0% Stall
100% Offense

At least for the guys that'll win anyways.

 

[ganj4lF]

This is going to be very interesting. As a ladder only player (too lazy / busy / prolly mediocre to join tournaments), it will be very interesting to see how the metagame is different and why compared to the standard ladder. Do you plan to give these statistics an actual impact on the tiering process, out of curiosity?

 

[MMII]

Back in the PO days using such a concentrated sample of usage wouldn't be very helpful when it comes to tiering policies. However, with the introduction of PS where any and many users are there to play casually online, the usual usage stats have become less valuable than they used to be and I think that having some tournament statistics thrown into consideration would help counterbalance this change. I don't think we should ever reach the point of saying tourny stats > ladder stats but it definitely will help in discerning more accurately which usage on the ladder is justified and which is a result of the influx of new casual players.

 

[kd24]

i have all the info, i'll be posting these tonight. thanks to almost everyone in top 16 for supplying the information and teams.

 

[kd24]

I ended up doing all of this by hand, hence why it is taking forever. There are hopefully very few, if any, errors throughout this. The thing is: I'm not even close to done. While I've tallied usage, I'm only posting so we have some content but the issue is by hand, there are so many things to point out and talk about. A list of stuff I'm currently missing:

- Detailed statistics for every Pokemon, except the top 5
- W/L Records for Pokemon when used on a certain team
- Pokemon's most common Partners
- Moveset / Role Statistics [admittedly, these are lacking anyway]
- A ton of other ideas I have if I had the help or time

While incomplete, these are still interesting and gave good insight. I'll give my thoughts about the stats below them:

Smogon Tour Usage Stats as of 3/29/2013
[Rank] Pokemon: Top 16 Usage, Top 8 Usage, Top 4 Usage, Finals Usage - Total Percentage
- Tied usage is Awarded to the Pokemon used most in early rounds

1.] Jirachi: 9/16, 4/8, 0/4, 1/2 --------- 46.7%
2.] Landorus-T: 5/16, 3/8, 2/4, 2/2 ---- 40%
3.] Keldeo: 7/16, 3/8, 1/4, 0/2 --------- 36.7%
4.] Politoed: 7/16, 3/8, 0/4, 0/2 -------- 33.3%
5.] Scizor: 6/16, 2/8, 1/4, 0/2 ---------- 30%
6.] Starmie: 4/16, 2/8, 1/4, 1/2 --------- 26.7%
7.] Skarmory: 3/16, 2/8, 1/4, 2/2 ------- 26.7%
8.] Dragonite: 3/16, 2/8, 0/4, 1/2 ------- 20%
9.] Rotom-W: 3/16, 2/8, 0/4, 1/2 ------- 20%
10.] Latios: 4/16, 1/8, 0/4, 0/2 ---------- 16.7%
11.] Garchomp: 3/16, 1/8, 1/4, 0/2 ------ 16.7%
12.] Breloom: 2/16, 2/8, 1/4, 0/2 -------- 16.7%
13.] Hippowdon: 2/16, 1/8, 1/4, 1/2 ----- 16.7%
14.] Heatran: 1/16, 1/8, 2/4, 0/2 -------- 16.7%
15.] Thundurus-T: 3/16, 1/8, 0/4, 0/2 --- 13.3%
16.] Landorus: 3/16, 1/8, 0/4, 0/2 ------- 13.3%
17.] Donphan: 2/16, 2/8, 0/4, 0/2 ------- 13.3%
18.] Mamoswine: 2/16, 2/8, 0/4, 0/2 ----- 13.3%
19.] Weavile: 2/16, 1/8, 1/4, 0/2 -------- 13.3%
20.] Tyranitar: 2/16, 1/8, 1/4, 0/2 ------- 13.3%
21.] Chansey: 1/16, 1/8, 1/4, 1/2 ------- 13.3%
22.] Amoongus: 1/16, 1/8, 1/4, 1/2 ------ 13.3%
23.] Gyarados: 2/16, 1/8, 0/4, 0/2 ------- 10%
24.] Celebi: 2/16, 0/8, 1/4, 0/2 ---------- 10%
25.] Venusaur: 1/16, 1/8, 1/4, 0/2 ------- 10%
26.] Ninetales: 1/16, 1/8, 1/4, 0/2 ------- 10%
27.] Dugtrio: 1/16, 1/8, 1/4, 0/2 --------- 10%
28.] Xatu: 1/16, 1/8, 1/4, 0/2 ----------- 10%
29.] Ferrothorn: 2/16, 0/8, 0/4, 0/2 ------ 6.7%
30.] Latias: 1/16, 1/8, 0/4, 0/2 ---------- 6.7%
31.] Kyurem-B: 1/16, 1/16, 0/4, 0/2 ------ 6.7%
32.] Jellicent: 1/16, 0/16, 1/4, 0/2 ------- 6.7%
33.] Forretress: 1/16, 0/16, 1/4, 0/2 ----- 6.7%
34.] Sableye: 1/16, 0/8, 0/4, 0/2 -------- 3.3%
35.] Darmanitan: 1/16, 0/8, 0/4, 0/2 ----- 3.3%
36.] Tornadus: 1/16, 0/8, 0/4, 0/2 ------- 3.3%
37.] Terrakion: 1/16, 0/8, 0/4, 0/2 ------- 3.3%
38.] Tentacruel: 1/16, 0/8, 0/4, 0/2 ------ 3.3%
39.] Gliscor: 1/16, 0/8, 0/4, 0/2 ---------- 3.3%

---
Detailed Statistics
---

Most Successful Teams of Tournament:
Hippowdon / Skarmory / Starmie / Amoonguss / Chansey / Landorus-T
Dragonite / Gengar / Rotom-W / Skarmory / Jirachi / Landorus-T

Non-Weather Usage: 26.7%
Rain Usage: 33.3%
Sun Usage: 10%
Sandstorm Usage: 30%

1.] Jirachi
Total Win/Loss Record: 6-8
Non Weather Usage: 28.6%
Rain Usage: 50%
Sun Usage: 0%
Sandstorm Usage: 21.4%
Most Common Partners:

2.] Landorus-T
Total Win/Loss Record: 9-3
Non Weather Usage: 25%
Rain Usage: 16.7%
Sun Usage: 0%
Sandstorm Usage: 58.3%
Most Common Partners:

3.] Keldeo
Total Win/Loss Record: 5-6
Non Weather Usage: 45.5%
Rain Usage: 45.5%
Sun Usage: 0%
Sandstorm Usage: 9.1%
Most Common Partners:

4.] Politoed
Total Win/Loss Record: 4-6
Non Weather Usage: 0%
Rain Usage: 100% [4-6]
Sun Usage: 0%
Sandstorm Usage: 0%
Most Common Partners:

5.] Scizor
Total Win/Loss Record: 3-6
Non Weather Usage: 55.6%
Rain Usage: 11.1%
Sun Usage: 0%
Sandstorm Usage: 33.3%
Most Common Partners:

---

So I was way off on Tyranitar being most common. Instead, it's Jirachi and Landorus-T. I was pretty spot on about Weatherless being way less than the other weathers. I should have expected Rain to be first. I expected everyone to counter it with sand, but if you're trying to counter Rain, honestly you probably should just be the one using Rain, it's that good. In the competitive scene, these usage stats are pretty telling:

Jirachi is way bigger of a thing there, hence Landorus-T being #2. I highly doubt it's a coincidence, it's more likely that most tournament players realize how good Jirachi is and how Landorus-T is a solid answer for it and Scizor. Landorus-T seemingly is the most successful Pokemon as well, winning 75% of the games when its used. Rain is the weather, and Keldeo is the Pokemon. While the stats don't support this, every game I entered pre-top 16 was usually Rain vs Rain and I, as well as top 16 players I talked to, saw Keldeo in almost every match. I think if I had taken the entire tournament, Politoed would be >50% and Keldeo near 60%. OU staple Terrakion wasn't a thing, this was either due to some extreme variance or maybe just because it really is a poor Pokemon compared to the top 5. It loses vs all of them, and struggles vs Pokemon in the top 10 as well. Tyranitar does as well.

I'm going to continue working on these statistics but volunteers are now not only welcomed but sorely needed. This kind of project is extremely tedious, and I give my thanks to everyone who PMed me their teams so I could build these statistics. I'm thrilled I got all 16 player's teams for the remainder of the tournament. But it has such a long way to go :[ I hope at the very least they're interesting to you, especially as I begin to gather more data.

 

[PK Gaming]

You did all of that by hand? That's hella crazy man (and impressive). I think tournament stats are going to get more and more useful as we start collecting them from major tournaments (i'm talking WCOP, OST, SPL, etc). I'm interested in seeing unknown pokemon (not gimmicks) down the road.

 

[Lavos]

noting that i was eliminated in semifinals, meaning i played six matches, and in those matches i saw five keldeo, four jirachi, three garchomp, and three landorus-t, not counting my own pokemon. didn't face a single rain or sun team, though i used both.

 

[chimpact]

this is pretty incredible, but the legitimacy of some of these stats are questionable. Can't you change your team round by round so that some of the statistics are skewed? (like bringing a joke team to one of the later rounds).

w.e i guess i just depends on the person and their incentive for playing. im only saying this because ctc was just using a bad lilligant team for this tournament and got haxed out lol

 

[kd24]

this is pretty incredible, but the legitimacy of some of these stats are questionable. Can't you change your team round by round so that some of the statistics are skewed? (like bringing a joke team to one of the later rounds).

w.e i guess i just depends on the person and their incentive for playing. im only saying this because ctc was just using a bad lilligant team for this tournament and got haxed out lol

it's a legitimate concern, one that worried me more than lying did (i didn't expect anyone to maliciously lie, but i did expect some people to ignore me). however, i think it's inherently adverted and defended against for 3 reasons:
1.) it's a tournament where the stakes are legitimate [at least, as legitimate as they can be for pokemon]
2.) if you're going to bother to get to top 16, are you going to waste it just to fuck up a random set of tournament stats?
3.) if you use mons to fuck up the stats, you're probably going to lose that round, equating to 3% (and it will continue to drop) usage. if people want to waste their time doing this, i have no issue whatsoever, the stats will continue to reflect the most used pokemon in the competitive environment, and the joke mons will simply be dropped to the bottom.

if ctc had made it to the top 16, i would have gladly written down whichever team he was using and added those mons to my usage. even if he had reached finals, equating into 13.3% usage for lilligant, that number would surely drop in upcoming tours unless the meta shifts towards lilligant because of his success. something similar happened to amoongus, who is currently ranked 23. i actually think amoongus is a great pokemon, but it will likely drop back every time these are updated, simply because people wont be using it often. its current ranking is likely inflated, but that shouldn't affect things long-term.

 

[TUO]

Wow. tbh i expected ST 15 to be filled with the KelTar+Lando cores that are going around. Surprised Rain is so big i honestly thought less people would be using it.

 

[Nysyr]

Lando-T just showing he's a boss as usual, and jirachi seems to be a fix all for a lot of teams, having the best mixture of stats, typings, and movepool to work with defensively.

Not surprisingly they form a good core, with both having U-turn as a bonus. Just throw in water or dragon type and presto, instant momentum.

 

[JackknifedJuggernaut]

Donphan: 2/16, 2/8, 0/4, 0/2 ------- 13.3%

37.] Terrakion: 1/16, 0/8, 0/4, 0/2 ------- 3.3%

I spit up my drink when I saw that.

 

[Lork]

it's interesting that hippowdon is gaining momentum as the superior sandstream setter.

 

[HabibsHotDogs]

it's interesting that hippowdon is gaining momentum as the superior sandstream setter.

I wouldn't call it "superior".

Tournaments, due to their nature, tend to encourage conservative or exaggerated aggressive play and with the threat of Dugtrio (which I am sure a lot of the good teambuilders considered) being significant with most people expecting Tar to be common, Hippodown seemed like a safer alternative to a lot of players and indeed you can see that has paid off!

Hippodown has always been THE most sturdy weather pokemon...

I guess some Gothitelle use would have gone a long way...

 

[Lavos]

considering the current tournament metagame trends, i actually would go as far as to say that hippowdon is a superior sand inducer to tyranitar. it's far bulkier, sets rocks easier, has access to instant 50% recovery, phazing, and a powerful stab earthquake. tyranitar's only real advantages are a decent special movepool and access to pursuit. considering that ttar loses to all the top 5 tournament ou pokemon, whereas hippo loses to only 2 of those 5 (and also isn't trapped and killed by dugtrio), i think it's fair to say that, given the amount of data we have on the tournament metagame, hippowdon outclasses tyranitar.

 

[Darknut07]

it's interesting that hippowdon is gaining momentum as the superior sandstream setter.

Whoa, lets take a big step back here. It seems to me that the person who used Hippo/Amoonguss/ Chansey...... team just used the same team all the way throughout. Amoonguss and Chansey just had one use in each round. Hippo had 2 only in the first round, then 1 throughout. So what happened was one battler likely built a successful team around Hippo, rather than this being a sign of its dominance.

 

[Lavos]

Whoa, lets take a big step back here. It seems to me that the person who used Hippo/Amoonguss/ Chansey...... team just used the same team all the way throughout. Amoonguss and Chansey just had one use in each round. Hippo had 2 only in the first round, then 1 throughout. So what happened was one battler likely built a successful team around Hippo, rather than this being a sign of its dominance.

i used hippo for the first 3 rounds before switching off in favor of sun/rain teams based on who my opponents were...and besides, if the guy made a successful team around hippo, doesn't that kinda showcase how good hippo can be? not sure what your point is

 

[Darknut07]

i used hippo for the first 3 rounds before switching off in favor of sun/rain teams based on who my opponents were...and besides, if the guy made a successful team around hippo, doesn't that kinda showcase how good hippo can be? not sure what your point is

One successful team doesn't mean Hippo is now all of a sudden better than ttar like that guy was saying.

 

[Stratos]

One successful team doesn't mean Hippo is now all of a sudden better than ttar like that guy was saying.

but neither does an ad ignorantium prove that the status quo is accurate. As it stands you've done nothing to back up your claims that Tyranitar is the better setter, and Lavos provides some good points.

However, while Hippowdon is the sturdier weather setter and able to survive long, drawn-out matches, those aren't the kinds of matches we see in today's OU meta. in BW's fast-paced meta, I'd really prefer Tyranitar's ability to sponge any special attack—once—than Hippowdon's more consistent, but less impressive, bulk. Besides, the ability to pursuit many top threats is really near impossible to overvalue, especially when paired with a Pokemon that appreciates the removal of the Latis and Celebi, such as the infamous Ttar + Keldeo. Hippowdon at a glance, gets the upper hand, as it seems better for consistently removing rain and sun, but as kd said, if you're trying so damn hard to get rid of those weathers then why not use them yourself? Besides, even though it can Slack Off its injuries, Hippo has only limited switchin opportunities on most standard rain teams, since it's still weak to Water. I'm not saying that Tyranitar is right for every situation and Hippo is always wrong. I'm saying that Ttar is better on offensive sand, and Hippo is better on defensive, at least imo, and that (again imo) Offensive Sand is better for the meta we're in. but then again there's a reason i always get eliminated from tourneys so early lol, so take what i say with much salt.

ps kd i know it would double your workload but it would be really cool if in the future we could see top 32 instead of top 16; most of the bad players are probably removed after two rounds anyway, and this is a pretty tiny sample size, but it's totally understandable if you don't want to do that given how much you're doing by hand and have yet to do.

Edit @ reply: Fair enough.