[Battle] Editing Anticipation to Avoid Super Effective

[Torterra18]

Ok, it isn't exactly asking for help, but more like asking for examples, but anyways, I want to make Anticipation give, like, a 30/20% accuracy nerf for moves that would hit SE (Super-Effective), I know that I'll do it in the file where the evasiveness functionality of Sand Veil/Snow Cloak is done, but, I asked a while ago how to do it, and they told me to define the SE thing, the problem is, Idk how to do it because I, stupidly, didn't ask for examples, which I can learn more easily VIA analysis of other parts of the code, so, what stuff do you recommend me to look into so I can do that task?

 

[Lunos]

I asked a while ago how to do it, and they told me to define the SE thing

What you were told is to declare a local variable that you were trying to use at the top of the function you were editing, before actually making use of it.
Why? Because as you were also taught, you can't just throw in a local variable potatoes in some if statement at GetTotalAccuracy and expect the function to know what value is potatoes supposed to hold or what is it supposed to mean at all.

Anyway, here's a diff for a copy of the latest Pokeemerald-expansion Master branch.

diff --git a/src/battle_script_commands.c b/src/battle_script_commands.c
index 6f2403d4ea..231062afbc 100644
--- a/src/battle_script_commands.c
+++ b/src/battle_script_commands.c
@@ -1856,6 +1856,7 @@ u32 GetTotalAccuracy(u32 battlerAtk, u32 battlerDef, u32 move, u32 atkAbility, u
     u8 defParam = GetBattlerHoldEffectParam(battlerDef);
     u8 atkAlly = BATTLE_PARTNER(battlerAtk);
     u16 atkAllyAbility = GetBattlerAbility(atkAlly);
[color=SeaGreen]+    bool8 isHitSE = GetTypeModifier(gBattleMoves[move].type, gBattleMons[battlerDef].type1) >= UQ_4_12(2.0);[/color]
 
     gPotentialItemEffectBattler = battlerDef;
     accStage = gBattleMons[battlerAtk].statStages[STAT_ACC];
@@ -1919,6 +1920,10 @@ u32 GetTotalAccuracy(u32 battlerAtk, u32 battlerDef, u32 move, u32 atkAbility, u
         if (gBattleMons[battlerDef].status2 & STATUS2_CONFUSION)
             calc = (calc * 50) / 100; // 1.5 tangled feet loss
         break;
[color=SeaGreen]+    case ABILITY_ANTICIPATION:
+        if (isHitSE)
+            calc = (calc * 80) / 100; // 1.2 anticipation loss
+        break;[/color]
     }
 
     // Attacker's ally's ability

What this does is to boost the power of moves that would do Super Effective damage against the first type of the opponent.

You can expand on these changes if you want to take into account secondary and tertiary types.

Here's a freebie, you can call DebugPrintLevel inside the statement for case ANTICIPATION to verify that it's working.
Ex:

diff --git a/include/config.h b/include/config.h
index 2ad84f29cd..f55ec4f4a2 100644
--- a/include/config.h
+++ b/include/config.h
@@ -6,7 +6,7 @@
 // still has them in the ROM. This is because the developers forgot
 // to define NDEBUG before release, however this has been changed as
 // Ruby's actual debug build does not use the AGBPrint features.
[color=Red]-#define NDEBUG[/color]
[color=SeaGreen]+//#define NDEBUG[/color]
 
 // To enable printf debugging, comment out "#define NDEBUG". This allows
 // the various AGBPrint functions to be used. (See include/gba/isagbprint.h).
diff --git a/src/battle_script_commands.c b/src/battle_script_commands.c
index 6f2403d4ea..9cae76a9af 100644
--- a/src/battle_script_commands.c
+++ b/src/battle_script_commands.c
@@ -1856,6 +1856,7 @@ u32 GetTotalAccuracy(u32 battlerAtk, u32 battlerDef, u32 move, u32 atkAbility, u
     u8 defParam = GetBattlerHoldEffectParam(battlerDef);
     u8 atkAlly = BATTLE_PARTNER(battlerAtk);
     u16 atkAllyAbility = GetBattlerAbility(atkAlly);
[color=SeaGreen]+    bool8 isHitSE = GetTypeModifier(gBattleMoves[move].type, gBattleMons[battlerDef].type1) >= UQ_4_12(2.0);[/color]
 
     gPotentialItemEffectBattler = battlerDef;
     accStage = gBattleMons[battlerAtk].statStages[STAT_ACC];
@@ -1919,6 +1920,14 @@ u32 GetTotalAccuracy(u32 battlerAtk, u32 battlerDef, u32 move, u32 atkAbility, u
         if (gBattleMons[battlerDef].status2 & STATUS2_CONFUSION)
             calc = (calc * 50) / 100; // 1.5 tangled feet loss
         break;
[color=SeaGreen]+    case ABILITY_ANTICIPATION:
+        if (isHitSE)
+        {
+            DebugPrintfLevel(MGBA_LOG_DEBUG, "This is being reached! Calc is %d: ", calc);
+            calc = (calc * 80) / 100; // 1.2 anticipation loss
+            DebugPrintfLevel(MGBA_LOG_DEBUG, "And now it's %d:", calc);
+        }
+        break;[/color]
     }
 
     // Attacker's ally's ability

 

[Torterra18]

Thanks for the help Lunos, but, do you rememember that I was working in an older version?, well I had to improvise a bit, but, can I use a code similar to that one if I want to implement AI compatibility?

PS: I've figured out how to make explosion change type, because, while surfing thru the code to figure out how dynamictype works, I've found the code for Revelation Dance, which I completely forgot it existed, and then copypasted it :)

 

[Lunos]

Thanks for the help Lunos, but, do you rememember that I was working in an older version?

Yes, and the code is mostly the same.
You don't have switch statements, but you can check the ability of the attacker with defAbility just fine.

Ex:

    else if (defAbility == ABILITY_TANGLED_FEET && gBattleMons[battlerDef].status2 & STATUS2_CONFUSION)
        calc = (calc * 50) / 100; // 1.5 tangled feet loss
[color=SeaGreen]+   else if (defAbility == ABILITY_ANTICIPATION && isHitSE)
+       calc = (calc * 50) / 100; // 1.5 anticipation loss[/color]

 

[Torterra18]

Ex:

    else if (defAbility == ABILITY_TANGLED_FEET && gBattleMons[battlerDef].status2 & STATUS2_CONFUSION)
        calc = (calc * 50) / 100; // 1.5 tangled feet loss
[color=SeaGreen]+   else if (defAbility == ABILITY_ANTICIPATION && isHitSE)
+       calc = (calc * 50) / 100; // 1.5 anticipation loss[/color]

I did just that Lunos

 

[Torterra18]

Btw, I tried adding multiple types compatibility, I did it like this

bool8 isHitSE = (GetTypeModifier(gBattleMoves[move].type, gBattleMons[battlerDef].type1) >= UQ_4_12(2.0)
                && GetTypeModifier(gBattleMoves[move].type, gBattleMons[battlerDef].type2) >= UQ_4_12(1.0))
                || (GetTypeModifier(gBattleMoves[move].type, gBattleMons[battlerDef].type1) >= UQ_4_12(1.0)
                && GetTypeModifier(gBattleMoves[move].type, gBattleMons[battlerDef].type2) >= UQ_4_12(2.0));

 

[Lunos]

Btw, I tried adding multiple types compatibility, I did it like this

bool8 isHitSE = (GetTypeModifier(gBattleMoves[move].type, gBattleMons[battlerDef].type1) >= UQ_4_12(2.0)
                && GetTypeModifier(gBattleMoves[move].type, gBattleMons[battlerDef].type2) >= UQ_4_12(1.0))
                || (GetTypeModifier(gBattleMoves[move].type, gBattleMons[battlerDef].type1) >= UQ_4_12(1.0)
                && GetTypeModifier(gBattleMoves[move].type, gBattleMons[battlerDef].type2) >= UQ_4_12(2.0));

There's no need to do it so convolutedly when you can just do:

bool8 isHitSE = (GetTypeModifier(gBattleMoves[move].type, gBattleMons[battlerDef].type1) >= UQ_4_12(2.0)
              || GetTypeModifier(gBattleMoves[move].type, gBattleMons[battlerDef].type2) >= UQ_4_12(2.0));

Explanation: If the type of the move stored in the variable move is super effective against the defender's 1st type or 2nd type, isHitSE will return TRUE.

 

[Torterra18]

I did it like that because the pokemon I was testing with avoided a move it's first type was weak to, but it's second type resisted, resulting in a neutral move, but it avoided it anyway

 

[Buffel Saft]

I think it would be better to call GET_MOVE_TYPE(), then calculate the effectiveness using CalcTypeEffectivenessMultiplier() and checking to see if the result is greater than or equal to UQ_4_12(2.0). That way you'll account for edge cases like Freeze-Dry and third types, and your code will keep working if CalcTypeEffectivenessMultiplier()gets updated at some point.

 

[Torterra18]

Ok thx, unfortunately, I don't really understand it and I may leave it like that, however, once I publish it, I may release the source code, and try to learn about C for a next, more ambitious hack

 

[Torterra18]

I think it would be better to call GET_MOVE_TYPE(), then calculate the effectiveness using CalcTypeEffectivenessMultiplier() and checking to see if the result is greater than or equal to UQ_4_12(2.0). That way you'll account for edge cases like Freeze-Dry and third types, and your code will keep working if CalcTypeEffectivenessMultiplier()gets updated at some point.

However, what if I just, y'know, use code snippets to better understand how it works in VSCode?

 

[Lunos]

Ok thx, unfortunately, I don't really understand it and I may leave it like that

diff --git a/src/battle_script_commands.c b/src/battle_script_commands.c
index 83ac411a1c..eacfc5c9d9 100644
--- a/src/battle_script_commands.c
+++ b/src/battle_script_commands.c
@@ -1861,6 +1861,11 @@ u32 GetTotalAccuracy(u32 battlerAtk, u32 battlerDef, u32 move, u32 atkAbility, u
     u8 defParam = GetBattlerHoldEffectParam(battlerDef);
     u8 atkAlly = BATTLE_PARTNER(battlerAtk);
     u16 atkAllyAbility = GetBattlerAbility(atkAlly);
[color=SeaGreen]+    u8 moveType;
+    bool8 isHitSE;
+
+    GET_MOVE_TYPE(move, moveType);
+    isHitSE = (CalcTypeEffectivenessMultiplier(move, moveType, battlerAtk, battlerDef, TRUE) >= UQ_4_12(2.0));[/color]

     gPotentialItemEffectBattler = battlerDef;
     accStage = gBattleMons[battlerAtk].statStages[STAT_ACC];

 

[Torterra18]

diff --git a/src/battle_script_commands.c b/src/battle_script_commands.c
index 83ac411a1c..eacfc5c9d9 100644
--- a/src/battle_script_commands.c
+++ b/src/battle_script_commands.c
@@ -1861,6 +1861,11 @@ u32 GetTotalAccuracy(u32 battlerAtk, u32 battlerDef, u32 move, u32 atkAbility, u
     u8 defParam = GetBattlerHoldEffectParam(battlerDef);
     u8 atkAlly = BATTLE_PARTNER(battlerAtk);
     u16 atkAllyAbility = GetBattlerAbility(atkAlly);
[color=SeaGreen]+    u8 moveType;
+    bool8 isHitSE;
+
+    GET_MOVE_TYPE(move, moveType);
+    isHitSE = (CalcTypeEffectivenessMultiplier(move, moveType, battlerAtk, battlerDef, TRUE) >= UQ_4_12(2.0));[/color]

     gPotentialItemEffectBattler = battlerDef;
     accStage = gBattleMons[battlerAtk].statStages[STAT_ACC];

Ok thx, sorry for being annoying

 

[Torterra18]

So I pasted the code where it originally was and the compiler told me that isHitSuperEffective is used but not declared, changed it a bit and the same error persisted, legacy version problem I guess? sorry.

 

[Lunos]

So I pasted the code where it originally was and the compiler told me that isHitSuperEffective is used but not declared, changed it a bit and the same error persisted, legacy version problem I guess? sorry.

Post the function of code where you're using that variable, or else it's gonna be hard to give you any specific answer.

 

[Torterra18]

u32 GetTotalAccuracy(u32 battlerAtk, u32 battlerDef, u32 move)
{
    u32 calc, moveAcc, atkHoldEffect, atkParam, defHoldEffect, defParam, atkAbility, defAbility;
    s8 buff, accStage, evasionStage;
    bool8 isHitSuperEffective = (CalcTypeEffectivenessMultiplier(move, moveType, battlerAtk, battlerDef, TRUE) >= UQ_4_12(2.0));

 

[Lunos]

u32 GetTotalAccuracy(u32 battlerAtk, u32 battlerDef, u32 move)
{
    u32 calc, moveAcc, atkHoldEffect, atkParam, defHoldEffect, defParam, atkAbility, defAbility;
    s8 buff, accStage, evasionStage;
    bool8 isHitSuperEffective = (CalcTypeEffectivenessMultiplier(move, moveType, battlerAtk, battlerDef, TRUE) >= UQ_4_12(2.0));

It's not isHitSuperEffective that doesn't exist, it's clearly moveType.
moveType is a variable that you have to add, that is meant to hold the type of the specified move via GET_MOVE_TYPE.
Since you're trying to pass it as an argument for CalcTypeEffectivenessMultiplier, you have to declare it and assign it a value before you assign a value to isHitSuperEffective, which is where you're calling CalcTypeEffectivenessMultiplier. After all, and I told you this before, you can't feed the code a variable potato that is not declared anywhere and expect things to just work automagically.

Now, the problem with that is that GET_MOVE_TYPE is a macro that directly fetches the type of a specified move and stores it in the variable that you specify as its 2nd parameter, it's not something that you can assign directly as a value for a local variable, you have to declare the variable as is, and then invoke GET_MOVE_TYPE somewhere below to store the specified move's type in it.

That's the reason why in the example I showed to you above I declared the variables moveType and isHitSuperEffective without assigning an initial value to them, and then proceed to give them one just below their declaration.

 

[Torterra18]

Ok thx!