This is a technical post again, for the interested reader. I decided to use my program to make a probability estimation in a more complicated situation than what I already did for
n-HKO lead probabilities (factoring in crits and other random effects of moves, such as misses, status inflictions and stat raises/drops, taking into account items such as Sitrus/Lum Berry and Leftovers as well).
Lately I've had many discussions EV-ing the following Pokemon, informally called FatMence:
Salamence @ Leftovers
Ability: Intimidate
EVs: 52 HP / a lot of Atk / a lot of Def / maybe a tiny bit of SpD / 16 Spe
Adamant Nature (+Atk, -SpA)
IVs: either perfect or 30 SpA / 30 SpD / 30 Spe for HP[Flying]
- Dragon Dance
- Substitute
- Aerial Ace / Hidden Power [Flying]
- Earthquake
It is most commonly seen on cripple + setup teams and sometimes on a Trick team (such as Trick + Grudge Ninetales) because it makes Substitutes that are unbreakable by Struggle after Intimidate, and it sweeps
nearly the entire pool of Frontier opponents after setting up to +6/+6.
In fact, the only non-Frontier Brain Pokemon that can survive its HP[Flying] after setup are
Skarmory,
Lunatone, Solrock,
Zapdos, Slaking, Umbreon, Regice, Cloyster,
Suicune,
Aerodactyl,
Lapras, Articuno and maybe Dewgong if you use Aerial Ace and less-than-max Attack. The Pokemon in red can be dangerous if your Substitute got broken before. Or they might break your Substitute and open up Salamence to Quick Claw / Brightpowder / Focus Band hax after that, for example. Or you simply might get two of them in succession during a bad matchup.
The most logical, straightforward thing to do is simply use Hidden Power [Flying] with 252+ Attack, since that indeed maximizes all the rolls and even eliminates Pokemon from the list above:
Pokemon that have a chance to survive 252+ Atk Salamence EQ and HP[Flying]
| Skarmory (Skarmory-Boss [Noland Silver†]) | HP Flying | 52.9 - 62.7% | 0.4% chance to 2HKO after Sitrus Berry recovery |
| Skarmory (Skarmory-Boss [Noland Gold†]) | HP Flying | 52.9 - 62.7% | 0.4% chance to 2HKO after Sitrus Berry recovery |
| Skarmory (Skarmory-3 [579]) | HP Flying | 58.7 - 69.1% | guaranteed 2HKO |
| Skarmory (Skarmory-4 [675]) | HP Flying | 58.7 - 69.1% | guaranteed 2HKO |
| Skarmory (Skarmory-1 [387]) | HP Flying | 62.7 - 73.9% | guaranteed 2HKO |
| Skarmory (Skarmory-2 [483]) | HP Flying | 62.7 - 73.9% | guaranteed 2HKO |
| Lunatone (Lunatone-1 [223]) | HP Flying | 77.4 - 91.5% | guaranteed 2HKO after Sitrus Berry recovery |
| Lunatone (Lunatone-2 [328]) | HP Flying | 77.4 - 91.5% | guaranteed 2HKO |
| Solrock (Solrock-1 [224]) | HP Flying | 86.4 - 102.2% | 18.8% chance to OHKO |
| Solrock (Solrock-2 [329]) | HP Flying | 86.4 - 102.2% | 18.8% chance to OHKO |
| Zapdos (Zapdos-1 [758]) | HP Flying | 92.7 - 109.6% | 56.3% chance to OHKO |
| Zapdos (Zapdos-2 [769]) | HP Flying | 92.7 - 109.6% | 56.3% chance to OHKO |
| Zapdos (Zapdos-3 [780]) | HP Flying | 92.7 - 109.6% | 56.3% chance to OHKO |
| Zapdos (Zapdos-4 [791]) | HP Flying | 92.7 - 109.6% | 56.3% chance to OHKO |
| Zapdos (Zapdos-5 [873]) | HP Flying | 92.7 - 109.6% | 56.3% chance to OHKO |
| Zapdos (Zapdos-6 [874]) | HP Flying | 92.7 - 109.6% | 56.3% chance to OHKO |
| Shuckle (Shuckle-Boss [Lucy Silver]) | HP Flying | 92.9 - 109.4% | 56.3% chance to OHKO |
| Slaking (Slaking-1 [468]) | HP Flying | 93.4 - 110.1% | 62.5% chance to OHKO |
| Slaking (Slaking-2 [564]) | HP Flying | 95.1 - 112% | 68.8% chance to OHKO |
| Slaking (Slaking-3 [660]) | HP Flying | 95.1 - 112% | 68.8% chance to OHKO |
But of course, some extra EVs in Defense are very welcome while setting up, and you might want to try Aerial Ace to prevent Brightpowder hax against many opponents in general. When using HP[Flying], general good marks are 202 Attack for the 50% against Zapdos, or even as low as 199 with 43.8% to OHKO Zapdos and still OHKO Aerodactyl, bulky Regice and Suicune.
But in this post I would like to give
Skarmory-3 [579], the one with Curse, Fly, Toxic and Rest, some attention. If you played this kind of Salamence (or Gyarados for that matter) before, you're probably familiar with the risk of not 2HKOing (because it uses Curse after the first hit) and letting it setup because of Chesto Rest. If this Skarmory pulls through, it will counter you with +6 Fly eventually, as you won't be able to outstall all of its 15 PP by using Substitute usually.
Of course, if you critically hit somewhere, you win automatically, but in general it can be a very lengthy process, and there's always a chance you won't get it before Skarmory starts using attacks of its own. All-in-all, I kept wondering what the
eventual winning probability is for Salamence, so I used my program to simulate this situation 10.000 times for different Attack stats and both for Aerial Ace and Hidden Power [Flying]. I didn't want to simulate a million battles, because I this PC can only do around 60~70 battles per second as some of them take long. With these settings I needed around 3 minutes per case. How little Attack can you afford to still have really good chance of beating it?
In these simulations, I assumed the following, which I hope reflect the actual game's strategy but to my feeling it does:
- Skarmory uses Curse if its HP is greater than or equal to half its HP (86) and Rest otherwise (first time it consumes Chesto Berry ofc);
- Salamence is at +6 Attack and only spams its Flying-type STAB;
- If Skarmory faints, Salamence wins (duh);
- If Skarmory gets to +6/+6/-6, is not sleeping, and has more than 65% HP left, Skarmory 'wins'.
The last assumption is to prevent having to think about Fly mechanics, but reflects the situation where Skarmory will start using Fly, so Salamence won't be able to use many Hidden Powers anymore in fear of Toxic. While not a losing scenario per se, it sure looks bad for Salamence. This should be taken into account when looking at the final results. Note that the 2HKO from the start is present with decent Attack stat, but the probability of not critting diminishes as well.
The results
In the following table, I rounded the four-decimal ratio estimators to two places a bit arbitrarily, but since it's still a Monte Carlo simulation I decided to go for a lazy cutoff to reduce "arbitrary garbage info" as much as possible. Since it's a "Bernoulli trial" estimator of sorts, it's hard to build a good confidence interval, and even rounding to full percentages leaves room for small error, although not too likely.
| Attack stat | EVs required | AA win probability | HP[Flying] win probability |
| 205 | 252 | 0.74 | 1.00 (not guaranteed) |
| 204 | 244 | 0.74 | 1.00 (not guaranteed) |
| 203 | 236 | 0.72 | 0.99 |
| 202 | 228 | 0.72 | 0.98 |
| 201 | 220 | 0.71 | 0.98 |
| 200 | 212 | 0.71 | 0.98 |
| 199 | 204 | 0.70 | 0.96 |
| 194 (Gyarados) | 252 | - | 0.87 |
Note that some Atk stats are redundant, such as 201 Attack, since it doesn't change the range against Skarmory and only changes the range of Aerial Ace against Zapdos, which is a 2HKO anyway.
Conclusion
Use Sharp Beak if you want the guaranteed win, haha. On a more serious note, use this for your own team!
Of course, I could have made these calculations exact simply by analysing the damage range and crits after. But I wanted to illustrate the power of simulations and use the scenario as a practice case.
If you want to see more of these kind of calculations, please let me know.