ジムリーダーの城　対戦考察Ｗｉｋｉ

The main requirement for win in Pokemon is to faint all of the opponent's Pokemon. In order to faint an opponent's Pokemon, the opponent's Pokemon's HP must be reduced to 0. This HP and various related elements change continuously throughout the game.
In this section, I show how perceiving games becomes easier by performing cost management which focuses on HP which is directly linked to that requirement and which changes.

In this section, I give some examples, but I don't take into account damage fluctuations due to RNG or critical or etc.
(These luck elements will be mentioned in a later section)

There is a detailed explanation of this writing method in Japanese articles, but this article focuses on understanding the essence with the minimum necessary explanation.

As the simplest case, let's first consider 1 vs 1 match where only Attack Moves are taken into account.
In this case, both players will take damage (unless you are outspeeding and can OHKO).
If you can reduce HP of your opponent's Pokemon to 0 with an Attack Move before HP of your Pokemon reduces 0 with an Attack Move, you can win.
Here, I call the damage your Pokemon takes before defeating your opponent's Pokemon the "Defeating Cost". In this words, if your Pokemon's HP is higher than the Defeating Cost, you will win.

As the concrete example, let's consider the case of Surf Starmie vs Earthquake Marowak.
The specific assumption are as follows.

Starmie HP:100%, Speed:320, Surf's damage: 80%
Marowak HP:100%, Speed:180, Earthquake's damage: 70%

In this example, each Pokemon have 2HKO, but because Starmie is outspeeding, it will only take damage once before defeating Marowak.
Therefore, Defeating Costs for each Pokemon will be as follows.

Starmie's Defeating Cost: 70%
Marowak's Defeating Cost:160%

Starmie can pay the Defeating Cost, but Marowak can't, so Starmie will win this game.

However, if Starmie's HP is below 70%, Marowak can only take one attack before defeating Starmie, so the Defeating Cost becomes as follows.

Starmie HP: 70%, Speed:320, Surf's damage: 80%
Marowak HP:100%, Speed:180, Earthquake's damage: 70%

Starmie's Defeating Cost: 70%
Marowak's Defeating Cost: 80%

In this case, by changing Starmie's initial HP, Marowak is now able to pay the Defeating Cost, and the matchup is reversed.

In more complex n vs n games, Switching will occur.
On the turn you switch, you will be used Moves unilaterally by your opponent's Pokemon, so the Cost you pay will increase accordingly.
Here, I call the Cost incurred during a Switching turn "Switching Cost".

Now, let's consider the concrete case which is Snorlax vs Marowak, Snorlax switches to Starmie, and Marowak uses Earthquake on that turn.

Starmie HP:100%, Speed:320, Surf's damage: 80% // Switching from Snorlax
Marowak HP:100%, Speed:180, Earthquake's damage: 70%

In this case, Starmie's Switching Cost is 70%.
And the Defeating Costs for Starmie vs Marowak in the next turn will be as follows.

Starmie HP: 30%, Speed:320, Surf's damage: 80%
Marowak HP:100%, Speed:180, Earthquake's damage: 70%

Starmie's Defeating Cost: 70%
Marowak's Defeating Cost: 80%

In this way, even if 1 vs 1 matchup is favorable for Starmie, the matchup can be reversed by paying Switching Cost.

It's hard to win if you lose your Pokemon every time a bad matchup comes up.
So it's important to switch to checker and prevent losing your Pokemon.
By checking your opponent's Pokemon, you can either defeat it or induce a check out of your opponent's Pokemon.

To make a check, you need enough HP to pay both Defeating Cost and Switching Cost.
Although Starmie from our previous example couldn't do this, Suicune which is 3HKO by Marowak's Earthquake can.

Suicune HP:100%, Speed:260, Surf's damage: 75% // Switching from Snorlax
Marowak HP:100%, Speed:180, Earthquake's damage: 45%

In this case, Suicune's Switching Cost is 45%.
And the Defeating Costs for Suicune vs Marowak in the next turn will be as follows.

Suicune HP: 55%, Speed:260, Surf's damage: 75%
Marowak HP:100%, Speed:180, Earthquake's damage: 45%

Suicune's Defeating Cost: 45%
Marowak's Defeating Cost:150%

Suicune can still pay Defeating Cost after paying Switching Cost, so it's possible to check Marowak.
This means you can either defeat Marowak or trigger a switch out of Marowak.

Non Attack Moves do not themselves reduce the HP of an opponent's Pokemon, but they affect damage calculations, thereby affecting Defeating Cost and Switching Cost.
Here, I mention Sword Dance which is one of the Setup Moves as the most obvious example.

Let's consider the similar example to the previous Starmie(switching from Snorlax) vs Marowak.
Snorlax switches to Starmie, and Marowak uses Sword Dance on that turn.

In this case, the Defeating Costs for Starmie vs Marowak in the next turn will be as follows.

Starmie HP:100%, Speed:320, Surf's damage: 80%
Marowak HP:100%, Speed:180, Earthquake's damage:140%

Starmie's Defeating Cost:140%
Marowak's Defeating Cost: 80%

In this example, you can see that the matchup is reversed by using Setup Move.

Note that while this alone may seem the same as using Earthquake on the Switching Trun, the effect of setup lasts until it switchs out, so it is important to note that it will also affect subsequent matchups, such as vs Snorlax.

As mentioned above, in many cases, you need to pay a cost to defeat your opponent's Pokemon.
Therefore, if you are trying to defeat many Pokemon with a Pokemon like called "Sweeper", even if you can manage to defeat one or two, it will become difficult to continue defeating any more Pokemon as the Costs accumulate.

However, if you outspeed the opposing Pokemon and it OHKO, Defeating Cost becomes 0. No matter how much you multiply it by 0, it remains 0, so you can keep defeating the opposing Pokemon over and over again.
Therefore, Pokemon with high Speed (and in such situations) can sometimes play this role and progress towards victory.

Playing with a certain Pokemon can sometimes have a positive effect on the Cost calculations of your other Pokemon.
In this section, I call this behavior "Support" and give some examples.

An understandable example of Support would be Non Attack Moves.
Among them, the easiest to understand would be Agility + Baton Pass or moves like Thunder Wave which use to Paralysis Spreading (ParaSpreading) to support Pokemon with lower speeds like Marowak.
When Marowak can outspeed the opponent's Pokemons like Starmie, it reverses the matchup against them.
Furthermore, when Marowak has Rock Slide and Hidden Power Bug/Flying, and setups with Sword Dance, it can OHKO most Pokemon and sweep to the end without taking any damage.
This is why Agility Baton Pass + Marowak is feared in actual OU games.

As we've seen, you can reduce the Defeating Cost by reducing HP of your opponent's Pokemon and making it easier to defeat them.
In other words, Support is possible even with Attack Moves.

There is a timely topic from an actual OU game, so I use that as an example to explain it.
In the current OU game, Tyranitar is sometimes used as lead. The Tyranitar which is as lead is usually a Mixed Attacker, and has favorable matchups against many Pokemon, so it tries to make the opponent lose Pokemon from the early stages of the game to gain an advantage in the game.
However, by supporting with an Attack Move and then switching to another Pokemon once or twice, you can force to switch out Tyranitar without losing a Pokemon.

For example, in the game below, using Cloyster's Surf in Turn 1 and switching to Exeggutor in Turn 2 can reduce the burden on Exeggutor. If Tyranitar had Crunch or Fire Blast, Giga Drain Exeggutor would have been an unfavorable matchup (both are 3HKO), but thanks to Cloyster's support, Exeggutor is able to 2HKO Tyranitar, making it a favorable matchup.
Although this support reduces Cloyster's HP, it is possible to leave it with enough HP to fulfill the role originally planned which use Spikes and Explosion, and in fact Spikes can be used in this game.
https://replay.pokemonshowdown.com/gen2ou-23024281...

Similar plays can be done with other Pokemon, such as Steelix or Zapdos. If you are running into Tyranitar lead with using Zapdos lead, it is effective to use Thunder on Turn 1 and switch out for a Pokemon like Steelix on Turn 2.

A switch occurs when a Pokemon fails, but there is no Switching Cost for this switch.
Therefore, by sacrifice a Pokemon that does not have an important role, it is possible to support the opponent by saving on Switching Cost.
For example, in the previously mentioned example of Snorlax -> Starmie vs Marowak, if the Snorlax can't do anything against Marowak (it's within OHKO range of Earthquake, it's asleep, etc.) and the Snorlax doesn't have an important role, then by sacrifice the Snorlax, you can switch to Starmie without paying Switching Cost, and you can then proceed with the game in a favorable matchup.

As Non Attack Moves that affect Cost calculations, moves that recover HP deserve special attention, especially Rest which most Pokemon have access to.
If your attacking methods are so poor that you have to wait until running out of PP of HP recovery moves to defeat your opponent's Pokemon, your Defeating Cost will effectively become infinite, as you will be constantly being attacked by your opponent's Pokemon which you cannot defeat.
In the GSC mechanics, this will happen frequently unless you know how to attack efficiently.

For example, against Leftovers Suicune, a normal Marowak with Thick Club is 3HKO with Earthquake, but a Marowak without it is 6HKO.
If this Suicune has Rest, this doesn't mean that Defeating Cost is doubled. How many attacks it takes to defeat a Pokemon with Rest varies from case to case, but at least in this case, 6HKO is clearly insufficient, and Defeating Cost becomes infinity.

Being surrounded by unbreakable walls and being unable to pay Defeating Cost is a common way to lose in GSC OU, so as long as you play this format, you need to be aware of how to break those walls.

In the sections up to this point, I have intentionally not thought about luck.
However, the explanations up to this point have laid the groundwork for thinking about luck, so from here on I will think about luck.

The development of the game branches depending on luck.

For example, if it's Starmie vs Marowak and they both use the attack move without switching out, Starmie will most likely win and a switch against Starmie with sacrificing Marowak will occur.
However, if Starmie's Surf isn't critical and Marowak's Earthquake is critical, Marowak will win and a switch against Marowak will occur.
Like this, the development of the game branches depending on luck.

Since you can't escape luck in this game, it's wise to think about how you can increase your odds of achieving the development you want.
Here are some tips on how to do that.

In this game, it is common to play the game hoping that you will get lucky, as if you were rolling dice, but you should think about increasing the impact of your lucky draw.
Types of impact include improving 1 vs 1 matchups, fulfilling sweep requirement, etc, and in particular, preventing Defeating Cost from becoming infinite due to wall break is often important.

For example, Zapdos' STAB is used more frequently Thunder than Thunderbolt.
This is because Thunder has a higher chance of Wall Breaking, especially against Rest Snorlax.
Thunderbolt does have some advantages, such as a better matchup against Substitute Starmie, but the advantages of Thunder are often more important.

It is important to increase the impact of getting lucky on a dice roll, but it is also important to increase the probability of getting lucky on a dice roll.

I mentioned Zapdos' Thunder and Thunderbolt as examples earlier, but strictly speaking, Thunderbolt can also defeat Rest Snorlax. However, to do so (if you don't consider things like Spikes), multiple critical hits are required, and the probability is low, so Thunder which has a higher probability of getting lucky is preferable.

Furthermore, if you think you can't rely on something like Spikes, it's better to give Zapdos Scope Lens. In this case, you need to 3HKO Snorlax (taking into account Thunder's miss), and without Scope Lens the chance is 6.61%, with Scope Lens it is 12.05%.
To make these calculations, there is a handy calculator like the one below, so use of it with feeling free.
https://image02.seesaawiki.jp/p/d/pbs-thread/rsC7u...

Even if the probability of getting lucky in one dice roll is low, if you roll the dice many times, you will have a higher probability of getting lucky in the end.
Sometimes it is important to try to increase the number of dice rolls.

For example, Nidoking has a low chance of defeating Resttalk Snorlax, and you have to rely on like Earthquake's critical hits or Ice Beam's freeze or Lovery Kiss with long turn Sleep / unluckey of Sleeptalk etc, but if your opponent has a Pokemon like Toxic Skarmory which you can switch to Nidoking with a low Cost, you can exploit that to increase the number of dice roll with Nidoking.
One of the reasons Nidoking is considered a good wall breaker despite its modest attack stats is because Nidoking is Toxic immute, and there are cases where you can increase the number of dice roll like this.

In the sections up to this point, I have made fixed assumptions about the plays your opponent will make, but in a actual game, there are multiple possible plays, and your optimal play may change depending on which play your opponent makes.
Here, I mention predictions of your opponent's plays and optimal plays based on those predictions.

Here I mention the predictions which are necessary even if you know all of the opponent's Pokemon's species and moveset.

Prediction is a kind of gamble, and has elements similar to a dice roll, as discussed in the section of Luck.
This means that you should try to get as much propability and impact as possible, and assume that your opponent will likely play in the same way.

For example, in Zapdos vs Skarmory, it is usually unlikely that Skarmory will stay. If Zapdos uses STAB, Skarmory will be outspeeded OHKO. Even if Zapdos' player predicts something and uses a move like Hidden Power or switches out, Skarmory can only inflict chip damage with Drill Peck or fling Toxic, so it doesn't have a big impact.
Therefore, you usually should predict that your opponent is likely to switch out, and use Hidden Power or switch out (Double Switch, described below) accordingly.

Also, if Zapdos' player thinks losing Zapdos to Explosion is not a big deal, it may be appropriate for Zapdos to use Thunder instead of Hidden Power in front of Steelix or Golem.
Steelix and Golem with the typical moveset do not have any Attack Moves that are effective against Zapdos other than Explosion, so if it thinks there is little impact in using Explosion to Zapdos, it is not uncommon to switch out to Snorlax. Roar or Curse (Steelix) or Rapid Spin (Golem) are sometimes used, but in any case, they do not have an immediately fatal effect on Zapdos.
Therefore, you can use Thunder in anticipation of switching out to Snorlax with low risk. You should consider it as one of option and expect to use it with a certain probability.

I can't write anything for sure about this because there is no answer which can be found through logical thinking, but it is certain that prediction is also influenced by the psychology of the player.
In particular, I feel that the fear of certain patterns which lead to losing has a big impact on play.

For example, switching in Cloyster and setting Spikes against Snorlax is common in OU games. Also, it's sometimes to see Snorlax using Lovely Kiss on Cloyster to punish it but miss it.
Teams using Lovely Kiss Snorlax like this are often intentional, as they have structures which can take advantage of inducing Sleep to Cloyster to progress to win, such as incorporating Golem or Marowak.
In this situation, Cloyster's player was lucky enough to detect the opponent's attempt without suffering any fatal consequences, but it is debatable whether it should predict Lovely Kiss again and switched out to a Status Absorber like Resttalk Zapdos in fear of the attempt. Snorlax's player may predict that the opponent will fear Lovely Kiss and use a move like Double-Edge. In that case, it is correct to set Spikes as originally planned.

Whether you should be timid or brave here depends on the player and the situation, and is not very common.
It may not be something I should write in an article like this, but I think you must more likely to get good results if you rely on your intuition in such situations.
(Human intuition is better than argumentative people realize)

Here, I explain the effect of "Double Switch", which means that it predicts the opponent's switch and switch on the same turn.

Double Switch means that you can't force your opponent to pay Switching Cost, but in exchange you can switch without paying the Switching Cost.
You can gain the benefit of Double Switch when the Switching Cost you can inflict on your opponent this turn is small, and the Switching Cost you would incur if you switch next turn is big.

For example, you have Toxic Skarmory (Drill Peck less) on the field, Snorlax which has rolls of Nidoking checker and wall breaker on your back, and Sleeping Umbreon to turn on your Sleep Clause.
And, your opponent has Sleeping Snorlax on the field and healty Nidoking on its back.
You should predict a switch to Nidoking, but Skarmory can't inflict decent Switching Cost on Nidoking, and conversely, Nidoking can make Snorlax pay decent Switching Cost.
So, Double Switch to Snorlax becomes appropriate.

Here I provide additional information on predictions of your opponent's unrevailed species and moveset.

If you know the general team structure or have done enough scouting on your opponent, you can predict their unrevailed options with little information.
For example, if you know about Borat Offense, you can see Snorlax / Zapdos / Cloyster / Steelix / Exeggutor and guess that the last one is likely to be Resttalk Vaporeon.
Also, if your opponent only uses a few teams and you can identify which team is being used in the case of Cloyster lead, you can predict the remaining unrevailed options only the fact that lead is Cloyster.

In games between skilled players, the game's development is determined to a certain extent.
Doing something outside of that can surprise your opponent and sometimes give you an advantage that you wouldn't normally get.

For example, in a Cloyster lead vs Snorlax lead situation, Cloyster will usually set Spikes on Turn 1 to gain an advantage in the game, and Snorlax frequently predict this and stay to use a move which will give it some advantage.
In SPL up to 2023, there were 10 cases of Cloyster lead vs Snorlax lead, and in 7 of those cases (70%) Snorlax stayed on Turn 1.
For reference, below is Snorlax's player's play from Turn 1 (until either switches out or defeats Cloyster) and Snorlax's moveset.

• Raikou (Body Slam/FireBlast/LovelyKiss/Curse)
• Curse->Curse->Double-Edge (Double-Edge/Curse/LovelyKiss/Rest)
• Double-Edge->Thunder (Double-Edge/Thunder/Earthquake/Self-Destruct)
• LovelyKiss->Zapdos (Double-Edge/FireBlast/LovelyKiss/Rest)
• Double-Edge->Flamethrower (Double-Edge/Flamethrower/SleepTalk/Rest)
• LovelyKiss->Earthquake (Double-Edge/Earthquake)
• Curse->Double-Edge (Double-Edge/Curse/SleepTalk/Rest)
• Double-Edge->Thunder (Double-Edge/Thunder/SleepTalk/Rest)
• Zapdos (BodySlam/Earthquake/Curse/Rest)
• Zapdos (Double-Edge/Earthquake/SleepTalk/Rest)

Here, if Cloyster use Polkadot Bow Explosion or Clamp (with Screech to countermeasure against Curse, Explosion, and Miracle Berry to avoid get Sleep from Lovery Kiss and Paralyze from Thunder) on Turn 1, giving you a high chance of removing the Snorlax.
Snorlax is virtually permanently #1 in Viability Rankings in OU, and often has important roles such as being the only checker against like Zapdos, the only wall breaker, and an Exploder which can sacrifice two or more pokemon.
Deleting Snorlax early is sometimes more important than setting Spikes, and if you no matter what want to set Spikes, there is a method to make it coexist with a pokemon like Forretress.