Steeling into Cut – a Belated Top 8 US (and Canadian) Nationals Report

Hello Hat Lovers!
Hayden McTavish here, new recruit for the VGC With Hats crew. I recently (well, recently when I started writing this report) made top 8 at US nationals, losing to fellow VGC With Hats member Raphael Bagara in top cut. What’s more interesting than how I did, though, was the team itself. It had some fascinating quirks in it, ones I think (because the team did do well) may warrant further examination. That said, I don’t want to present this team as a finished solution – I felt it had some dubious choices and there are definite areas for improvement. The Thundurus matchup in particular is less than ideal, and I had issues playing against Japan sand sometimes, if only because I need to make effective hard reads against salamence/tyranitar leads and it’s difficult to maneuver around bulky water + Aegislash. (actually I got some laughs when I told people I brought a Japan-sand weak team to nats, and while it likely was a poor meta call I don’t believe the matchup was altogether terrible, and the comfort factor of running this team far trumped that issue as I was pretty comfortable in the large majority of my other matchups)

But enough about the team’s flaws! here it is, with ev spreads bolded since they’re the only information players couldn’t have already accessed on pokemon.com:

Salamence @ Salamencite
Salamence-MegaAbility: Intimidate
Level: 50
EVs: 196 HP / 212 SpD / 100 Spe
Jolly Nature
– Draco Meteor
– Earthquake
– Return
– Protect

For Salamence, I used a relatively standard moveset, finding that I preferred Earthquake’s ability to deal with steels (barring Ferrothorn, Escavalier, Scizor, and some more obscure Pokemon like Skarmory) to the recovery that something like roost afforded or any other coverage.

What probably stands out the most is how ridiculously oriented towards special bulk this Salamence is. This is in part because Salamence was among the last of the Pokemon added to my team. As such, I didn’t really have a team built around resolving Salamence’s weak points, and had to instead make salamence less vulnerable to its usual weaknesses, such as Suicune, Zapdos, and Thundurus. This meant I’d need to invest in special bulk to take ice beams and hidden powers.

However, were I to invest in special bulk, it would be difficult to maintain enough offense to reliably KO Sylveon and Gardevoir, and were salamence to be too weak to those pokemon I didn’t have enough support in the rest of my team to really deal with them (as much as it may appear that two steel types is enough to ruin fairy pokemon, it’s really not, especially with both of them being primarily special attackers). I also knew from experience playing mence earlier in the season that being able to KO a hyper voice user, but still being easily KO’d by it, really isn’t a very stable situation and can easily turn against you if you mispredict, become paralyzed, are intimidated, or are in a situation where two Pokemon threaten you.

If I couldn’t reliably OHKO the Pokemon, then, I’d just have to survive their attacks! I therefore used the survivalcalc (http://survivalcalc.appspot.com/ , credit to Nuggetbridge user Stats for creating it), to determine whether it was possible to survive a modest max special attack gardevoir’s hyper voice. Through testing, I quickly determined that I really wanted to be able to outspeed Kangaskhan after mega evolving (which also meant I outsped most charizard before mega evolving), so I used the EVs I would need for that as a parameter. It turns out the highest probability I could get was a 7/8 chance to survive the attack, with a spread of something like 52HP/236SpDef (228 was all that was necessary but 236 optimized the nature). However, there was an alternate Jolly spread of 196HP/212SPDef, the one shown above, that had a 13/16 chance to survive a Hyper voice, disregarding critical hits. I ended up selecting that one because it gave me far more HP than the other spread, and I believe affected significantly my ability to take Adamant Landorus-T Rock Slides.

Looking at the calculations again, it is actually possible to have special bulk while still KOing Gardevoir, but the earlier mentioned problems of paralysis and intimidate do remain, and it’s difficult to still outspeed Jolly Kangaskhan  while having special bulk and KOing Gardevoir.

The flaw with this Salamence is that it fails to consistently KO Amoonguss, since maxHP/Def with a defense boosting nature allows amoonguss to survive the return 15/16 times. This means even in cases where I can KO amoonguss (ie if it’s sassy) I’m not confident that I will, because there’s a chance the amoonguss is bold/relaxed. Thus while this Salamence was built to adapt around gardevoir to some extent, it didn’t adapt to one of Gardevoir’s most frequent partners. I think this problem can be remedied to some extent in a Bo3 environment by memorizing some damage calculations from my other Pokemon on amoonguss so I could determine the spread (and thereby whether or not Salamence has a reasonable chance of KOing Amoonguss), though that is an imprecise solution.  The larger problem, however, is that Rocky Helmet puts me in KO range from Gardevoir’s Hyper Voice! What I should have done, or would look at doing were I to run this again, would be to have Hyper Voice on the set, since that allows me to attack an amoonguss (and severely damage it even when I’m jolly if it’s physically bulky) without activating rocky helmet, put it in range for return, potentially tell from damage output if return will KO normally/ without chip damage, and do some damage to Amoonguss.

Rotom-W @ Choice Scarf
Rotom-WashAbility: Levitate
Level: 50
EVs: 20 HP / 4 Def / 228 SpAtk / 4 SpDef / 252 Spe
Timid Nature
– Hydro Pump
– Volt Switch
– Thunderbolt
– Will-o-Wisp

My most fun pokemon. I used scarf Rotom-W way back in 2012, though that had some notable differences, namely Trick (which is harder to use with mega stones being non-removable). What this Rotom did was answer double genies, mostly the Landorus-T part by outspeeding and OHKOing scarf Landorus-T with Hydro Pump. It has some added power beyond the special attack mark needed for 4HP Landorus so that it can KO bulkier variants, while helping it to deal with charizard since it can get a fast volt switch off that will do a lot of damage while dodging SolarBeam. Will-o-Wisp was fairly situational, but made sense to dodge sucker punch and was helpful in my set against picklesword (Matthew Greaves), where Will-o-Wisp burned his Tyranitar and let me tick away at Cresselia without needing to answer his Cresselia with Aegislash. However, it had mainly been present to help Gardevoir in dealing with Salamence, (burning Salamence so that Gardevoir can take the double edge), where I now have Cresselia and need to just KO mence anyways, so HP ice would likely have been better . It doesn’t KO Landorus, but with helping hand it should, so if Helping Hand were to remain on Cresselia it would make sense as a more reliable Landorus answer (if Cresselia can’t afford Hydro Pump missing and allowing a landorus to U-turn). Will-o-Wisp also was revealed to be flawed in my final round of Day 2 nationals (vs Alex Underhill), where I was unable to effectively lock into Wisp despite it being important to do so against a Tyranitar because it left me in a really bad situation in the case of a Rotom-H switch in (which happened in game 2, where I was left locked into Wisp against a burned salamence and a Rotom-H, and wasn’t able to execute a safe switch out). I don’t know that it’s really terribly flawed, it just won’t always be a reliable response. Dropping Thunderbolt is definitely also an option, or Volt Switch.

Conkeldurr @ Life Orb
Ability: Iron Fist
Level: 50
EVs: 68 HP / 252 Atk / 188 SpD
Adamant Nature
– Mach Punch
– Knock Off
– Ice Punch
– Protect

This is one of the two pokemon I began with in constructing this team, the other being Aegislash. The premise Ben Hickey proposed to me involved Conkeldurr/Aegislash as two Pokemon that could both work well inside or outside of trick room, while having perfect coverage together. Ben suggested Life Orb, and that’s what I started working with, especially with the ability to KO 4HP Salamence from -1atk with Ice Punch, and the ability to KO 4HP Kangaskhan with Helping Hand Boosted LO Mach Punch. Life Orb served me well, especially in increasing the damage done to Thundurus and Amoonguss, although the main Salamence KO I was able to achieve was against Collin Heier, where I caught his salamence as it switched in, where I didn’t need LO to get the KO. In light of some issues with taking Hyper Voice, I’d like to look at an Assault Vest set, but Protect proved invaluable against Kangaskhan players that would assume I didn’t have it and double target the Conkeldurr, buying time for my partner (often my Rotom, which could Volt Switch and possibly put Kangaskhan in KO range from Mach Punch).

The main quirk of this Conkeldurr is that its only fighting type attack is Mach Punch. I elected to run Knock off over Drain Punch/Hammer arm mainly due to a bit of a problem with Aegislash. I found in practice that I really wasn’t using Hammer Arm that much, perhaps because I wasn’t always able to set up trick room or didn’t always pursue that approach. Usually I would find myself just using Mach Punch. The other issue is that my main potential trick room combination, cresselia and conkeldurr, would be unbearably weak to aegislash if I didn’t have Knock off.

The main issue with this set is that I don’t have a way to hit water types effectively. The most I can do is Knock Off, where Drain Punch, especially boosted by Helping Hand, could have helped to resolve my weaknesses to Milotic, Suicune and Rotom-W (though not Azumarill). I’ve also considered Breloom and/or Virizion in this slot, though neither of them deal with Thundurus as well (I don’t find Virizion’s Stone Edge reliable enough, and Quick Guard, while useful, won’t eliminate Thundurus).

Heatran @ Shuca Berry
HeatranAbility: Flash Fire
Level: 50
EVs: 252 HP / 252 SpA / 4 SpD
Modest Nature
– Heat Wave
– Earth Power
– Protect
– Overheat

Heatran is probably the least intriguing of my Pokemon, other than its allowing me to run two steel types, which is nice because of the plethora of resistances it affords me. However, I’m fairly convinced that, aside from normal, steel’s most important resistance is fairy, so I probably could have used another fire type in this slot without losing too much, especially if said fire type worked well against Kangaskhan. I did like having Heatran, though, mainly for its overheat, which allowed me to deal with Aegislash, and its being a useful answer to non-HP ground Charizard. I ran Shuca Berry to allow me to tangle with other Heatran or survive a Landorus attack, although a fundamental issue with this set is that Banded Landorus can still OHKO with Superpower, and of course Kangaskhan Low Kick is difficult to combat. I’ve been considering running a bulkier set as a result, but the issue is that lowering special attack loses the KO on Aegislash with overheat, which at least at nationals I had felt was very important. I’m starting to toy with a bulkier set that foregoes the guaranteed KO, though, as I begin to reassess how threatening Aegislash is (or isn’t) to this team if I make some other changes to it.

I’d wanted to run a 29 speed IV on Heatran, because it would let me underspeed all non-Trick Room Heatran and outspeed most Trick Room Heatran, but I didn’t have the resources to reset for a 29 speed IV, and it wasn’t significant enough for me to try to get that at the time.

Cresselia @ Sitrus Berry
CresseliaAbility: Levitate
Level: 50
EVs: 252 HP / 44 Def / 212 SpA
Calm Nature
– Psychic
– Light Screen
– Trick Room
– Helping Hand

I’m pretty sure that my Cresselia moveset was flawed for what I wanted from it. I know that this movest is considered solid by players like Enosh and Crow (or was after Spring Regionals) but I personally would have preffered psychic and ice beam, and regretted the lack of recovery (though I might have been scared enough by Tyranitar and its sand stream to run chesto rest rather than Moonlight). Helping Hand is a really useful and cool move but I find myself making heavy commitments that may not work out, and I found it not always useful. I liked Trick Room not just for its ability to make Conkeldurr and Heatran “faster” (especially since Conkeldurr often used Mach Punch anyways) but also for its anti-speed control potential – any time the opponent altered speed for their benefit, I could just invert the attack order if I was able to have a spare turn with Cresselia. However, I wasn’t able to use it very well…

I think the bigger issue is that I used an Enosh/Crow moveset but didn’t have mental herb, which I think is important for this kind of Cresselia, where I don’t have Ice Beam to punish Thundurus taunting me.

Thundurus actually became a bit of an issue for the team in light of this, and my plan to use Trick Room as anti speed control, responding to paralysis and the like, fell a bit short with the ease with which Thundurus taunted it. It might work better with the influx of Swagger over taunt on Thundurus now.

Overall, I think the perfect coverage priority afforded by Agislash and Conkeldurr probably negated need for speed control a bit and I could have used another set. Light Screen was crucial, though, and Trick Room was still helpful, even if I often found myself using it just to invert other Trick Rooms.

Aegislash @ Weakness Policy
Aegislash-ShieldAbility: Stance Change
Level: 50
EVs: 252 HP / 92 SpA / 164 SpD
Sassy Nature
– Shadow Ball
– Flash cannon
– Shadow Sneak
– King’s Shield

Against solid players in best-of-three (where there’s added risk of your opponent observing your playstyle) a prediction heavy Pokemon like Aegislash can be painfully difficult to use, at least for me – I often found myself winning game 1 off of intense predictions, then finding myself outpredicted in games 2 and 3 when practicing with friends. I found that my best way to avoid my problems with Aegislash and prediction was shadow sneak and a bulkier Aegislash. Knowing that Wide Guard would be heavily anticipated at nationals, and having talked with a player in the BC invitational who discussed shadow sneak as useful even without Life Orb Aegislash (where Shadow Ball + Shadow Sneak can OHKO 4HP Landorus), I decided to try it, and found that it worked very, very well.

The set let me do a lot of things. First of all, it gave me a safe way to do damage with Aegislash. Where there is often a prediction game involving whether or not Aegislash will King’s Shield, doing no damage and effectively wasting a turn if it wasn’t targeted, or Shadow Ball, doing heavy damage but possibly just getting KO’d first and still doing no damage, I was able to deal a safe, smaller amount of damage. I didn’t often use Aegislash this way unless Weakness Policy was activated and Shadow Sneak would be doing substantial damage (or my aegislash was too low on HP to attack any other way), but it’s something I wish I’d done more often, because there are sometimes situations where Shadow Sneaking can save a match where there is a prediction game, especially if there’s a gardevoir on the field (Shadow Ball barely does more than Shadow Sneak in that situation).

It also let me slant the prediction game heavily in my favor. If opponents expected to be able to leave in a low Hit Point pokemon or anticipated that Aegislash’s only safe move was to King’s Shield, I could Shadow Sneak and KO the low HP pokemon or just get an attack off before fainting. One of the problems with standard Weakness Policy Aegislash at the time was that at lower HP it was pretty difficult to get an attack off, and Shadow Sneak helped with that.

The Special Defense was particularly based on an observation that the main way to deal with Aegislash tended to be special attacks, often strong fire moves. Though I didn’t observe exactly the trend towards strong single-target special attacks that Pokemon.com highlighted in a pre-worlds article, the fact that I both dropped Wide Guard and boosted special defense especially helped me with this. Being able to survive attacks such as Heatran Overheat, then fire back with a boosted Shadow ball, was really helpful. The fact that I would be at low HP afterwards was also mitigated by shadow sneak; Aegislash, when boosted, could still be useful when it didn’t have enough HP to survive an attack .

The special defense also helped with the prediction slant, though. Because I had so much special defense, I was able to take hits in Blade form that other players anticipated to KO. This meant that where players assessed my staying in Blade form as especially risky, it actually was less so, and thus opponents could predict a King’s Shield where I didn’t have reason to King’s Shield. It also was a really helpful backup to have if I mispredicted with leaving Aegislash in blade form and took a heavy hit, or if Aegislash was taunted.

That’s the team! I may edit in some information on what the team’s issues were later, but hopefully this provides enough food for thought. There were a lot of somewhat unconventional decisions on this team, and I’m glad that those changes I made based on observations of what was working in practice ended up working out, for the most part. (though I suppose I’d have to have won nationals to claim the decisions weren’t something that could have held me back)

World Championships

Since this report is so late, I have the opportunity to discuss also my worlds run (My fall regionals team had little to do with this one so I won’t discuss that). I ended up with a 3-4 record, but in reality that was a 4-3 record where I made the mistake of accidentally cancelling participation too soon and receiving a match loss in my final round, so it was a disappointing but not altogether terrible run. I would’ve liked to have done better (bet you weren’t expecting that!), especially with a painful set against Blake Hopper Round 6 where I became overconfident in game 1 and made some mispredictions game 2, both of which lost me a (probably) winning position, but it was still an experience I came out of with a lot to talk about, and some wounded pride mixed with a renewed aspiration to get better at this game and catch up to players like Yosuke Isagi, and, of course, Shoma Honami.

I think my notes in the teambuilding department were similar to the ones I mentioned earlier in the report, of having obsessed over the team more than my playing. I also ended up really second guessing myself, which I think I regret. I ended up swapping out Cresselia (a cool pokemon but my least effectively used at nationals) for a Sitrus Berry Thundurus-Therian before worlds, and while it was incredibly cool it didn’t really help me against the teams of Conan and Demitri, who I battled way too much on the way to worlds, and that – in combination with feeling like I couldn’t resolve all of the team’s issues with Thundurus-Therian as well as a vague unease about my not having speed control – caused me to switch away from Thundurus-T at the last second despite it working well in testing. I ended up replacing Cresselia with Porygon2, a change that was beautiful in theory (thanks to Conan for listening to me discuss my team’s problems for a couple of hours to suggest that, and to have thought it through even when he mostly just didn’t want to listen to me list obscure pokemon that might solve problems in my team, like magnet rise Manectric). Porygon2, for reference, was decided upon because it made the type synergy of my team a bit more effective – despite hurting salamence’s ability to earthquake, its ghost immunity where cress had a ghost weakness helped immensely with aegislash (at least in theory – in reality I think Cresselia’s Light Screen was enough to cover this), and its doubled fighting weakness with heatran was less of an issue with porygon taking little from low kick, and superpower from Landorus-Therian Forme being a three hit-KO most of the time given the attack drop (though this calculation was for Scarf Landorus, not the Jolly Life Orb variation). It also allowed me to set up trick room more comfortably in the face of tyranitar and to KO mence, thundurus, and landorus more easily (remember, my original cresselia didn’t have ice beam), and porygon especially with download is ridiculously powerful, 2HKOing Thundurus as bulky as Rapha’s if it got the special attack boost. It thirdly enabled me to run an actual fighting move on conkeldurr over knock off (I chose hammer arm, though drain punch probably would have been better to keep conkeldurr around) more comfortably, since conkeldurr + Trick Room setter was now no longer a combination that was weak to aegislash. This allowed me to better address my issues with bulky water types, at least in theory. In retrospect I’ve since learned that knock off is for more than KOing aegislash when I have conk cress out – it also fits with resolving pokemons’ weak points so the rest of the team won’t have to compensate and so that conkeldurr can be brought more often, since it allows conkeldurr to be less afraid of cresselia and aegislash when brought. It also gave me a much better Cresselia/Heatran/Amoonguss/Landorus/Kangaskhan matchup to have Knock Off there, so I do regret the choice a bit, especially because I don’t think I ever used hammer arm and there was one situation, in Round 1, where I could perhaps have beaten Hayato if I’d used knock off on his cress on a turn where I whiffed an attack into his Kang’s protect (though the protect was a surprise, so it’s not certain  I would have made that play even had I had knock off).

The Porygon2 set was:

Porygon2 @ Eviolite
Ability: Download
Level: 50
EVs: 244 HP / 188 Def / 60 SpA / 4SpD / 4 Spe
IVs: 1 Speed
Quiet Nature
Moves:
– Shadow Ball
– Ice Beam
– Trick Room
– Recover

I could only get a 1 speed IV on Porygon2, so I elected to use the 4 evs that would otherwise be wasted on speed to boost it by 1 point (just to avoid speed ties with 0 speed aegislash and thereby get information on the speed of the opposing aegislash while reducing uncertainty about move orders)

It wasn’t a bad choice by any means to look at this pokemon, and Demitri actually adopted it as well, with it being able to deal effective damage and wall pokemon with recover (It actually even survived a LO Hammer Arm from my Conkeldurr in practice) to stay around the entirety of a game until nothing can eliminate it. If it doesn’t get hit by Knock Off it can be really difficult to handle. However, immediately after Worlds I switched back to Cresselia in this slot, because I realized something that Jeudy Azzarelli had been trying to tell me a month earlier – that in making changes to improve my theoretical matchups, I always had the potential to fundamentally weaken my team. Though Cresselia was imperfect at nationals, it was still a really good and important pokemon for this team, and it supported salamence well with light screen as well as the Helping Hand salamence needed to really be an effective threat (more important for such a low attack investment salamence). It was also just generally effective and important, I felt, and I‘m more comfortable using the team again now that I put cresselia back. I’ve since been toying with making smaller changes, such as switching Psychic to ice beam (which helps cress be far less deadweight vs thundurus/landorus) and thunder wave over trick room, which prevents issues like the ones I had vs Alex Underhill, where I would end up unwittingly hurting myself by setting up trick room at the wrong time. Although it does hurt a bit to lose my ability to reverse trick room, and I wouldn’t mind a Cresselia that also knew psychic/psyshock so I could eliminate bulky waters, the changes have had a net positive effect, and it would have been interesting to see what would have happened had I been able to figure out smaller changes to cress in the course of working on my playstyle and general play, especially were I noticing more about my team and gathering more information, or had I trusted myself and stuck with Thundurus-Therian (which also would have helped me a lot vs Blake, Hayato, and Shun, my three losses in swiss).

The other change I made was as a result of a half-hearted effort at day 1 scouting. After some cajoling, Caleb Ryor was nice enough to give me some of his notes on day 1 teams, but I hadn’t been doing my own scouting very effectively, having been concerned about what I should have run and being a little distracted by the Pokken Machamp (so cool!). I also didn’t have much time to review the notes with having to go to sleep early enough to be well rested for the tournament, and thus I mostly just noticed my dad’s observations that there were a lot of charizard teams. However, what I didn’t notice, because it wasn’t necessarily surprising enough for my dad to point out, was that Kangaskhan was still bigger (I also didn’t see that Blake pointed it out on twitter until after the tournament). We also didn’t realize that Kang had made day 2 far more often than Charizard (most Charizard teams missed the day 2 invite). There were also limitations on the data, so it wasn’t super obvious (a lot of Japanese Kangaskhan players’ teams weren’t on the team scouting list). I don’t think I’m a player that should necessarily make meta calls at the last second, since I tend to pursue absolute solutions more than ones that just help a matchup, but here what I did was make a meta call on the wrong part of the data, and freak out about making my charizard matchup, already good, ridiculously solid. I swithed to stone edge over draco meteor on salamence. Now, there was more reasoning to that change, because it helped a lot vs Thundurus, but in the end I would have preferred draco meteor, because as Conan had pointed out to me earlier I didn’t have many ways to KO Kangaskhan via special attack, and in general it made Salamence a lot less useful if burned or intimidated. I also switched Salamence to Adamant with a bit of attack so I could KO amoonguss a bit more often, but it made mence slower than some charizard and Kangaskhan, so I really didn’t like that. I’ve since switched back to Jolly/Draco Meteor.

Lastly, I’m not sure I liked my mentality going into this tournament. Although it probably isn’t obvious from how late this report is, I ordinarily abhor procrastination. It’s something that I’m almost always very frustrated with myself for doing. I mention this because I felt like with the large gap between nationals and worlds I wasn’t using my time as effectively as I should have. Although to some extent it’s useful to not be working too much on pokemon, and I definitely had to recognize that I wouldn’t be effective if I only did pokemon, I kind of avoided teambuilding and practicing extensively for this tournament. I probably did put in more thought than I felt I did, but the important thing wasn’t necessarily how much preparation I put in but how prepared I felt, and I definitely felt underprepared this tournament. This shook my confidence a lot, and caused me to make more last second shifts than I probably should have in looking for the perfect team, and to seek the opinions of others on team ideas I really should have been resolving myself through deeper analysis. It’s always tempting at big tournaments, especially where I was travelling with friends, to rely on conversation with others for team ideas, but that’s really bad for me at a certain point, because I usually function through independent thought supplemented by conversation. I also found out pretty close to the competition that I vary a lot more than I thought from some of my friends in how I prepare for events mentally, and how I approach the game, which probably made it more difficult for all of us to prepare.

All right, excuses over! Thank you for taking the time to read through all of this. Feel free to leave a comment below if you have questions.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s