Azure Heights Pokémon Laboratory
Main Menu
What's New?
Weird Science

All Attacks








Critical Hit
One-hit KOs
Stat Modifiers

Statistic Modifiers

Many attacks in the game alter the statistics of either the attacker or the victim. Sometimes statistic modification is the sole purpose of the attack causing it (eg, Screech), while sometimes it is merely an occasional side-effect of a damaging attack (eg, Aurora Beam). Statistics can also be altered as a result of the BRN and PAR status changes.

In the Red, Blue, and Yellow games, there are seven statistics of Pokémon that can change during the course of battle: HP, Attack, Defense, Speed, Special, Evade, and Accuracy. Evade and Accuracy are fundamentally different from the other five in several ways, and will not be discussed here. Hit Points are also a special case, in that HP does not change from what we normally think of as statistic modifiers. A few moves (eg, Recover) can increase your Pokémon's HP, while many, many moves can be used to decrease the HP of your opponent.

The purpose of this document is therefore to explain the effect of attacks that alter the remaining four statistics: Attack, Defense, Speed, and Special.

Analysis & Advice

Stat modifiers have two levels of potency: standard and great. The former category includes attacks like Tail Whip and Growth. The latter category includes attacks that the game describes as "greatly" effective, such as Screech and Amnesia.

When a stat-increasing attack is used, and the relevant statistic is starting at its normal value, the effect of the stat mod is easy to describe. Standard mods increase the stat by 50% of its initial value with each usage, while great mods increase the stat by 100% of its initial value with each usage. Things are not so neat and tidy for the use of stat-decreasers, however. Nor is there any simple way to express the effect of stat-increasers on a statistic that has previously been lowered below its initial value.

Fortunately, the following table provides an easy way to visualize the effects of statistic modifiers. On the top row is the full range of possible modifier values — a way to represent the current direction and degree of modification for a given statistic. Ordinarily a stat's mod value is zero. However, standard stat-increasers raise the mod value by 1, while great stat-increasers raise it by 2. Standard stat-decreasers reduce the mod value by 1, while great stat-decreasers reduce the mod value by 2.

The numbers in the bottom row are the statistic multipliers that correspond to each of the possible mod values.

-6 -5 -4 -3 -2 -1 0 +1 +2 +3 +4 +5 +6
0.25 0.28 0.33 0.40 0.50 0.66 1 1.5 2 2.5 3 3.5 4

Let's illustrate these concepts with a few examples in which a Pokémon's Attack statistic is altered. The most common Attack stat mods are Swords Dance (a great increaser, hence a mod value of +2), Meditate, and Sharpen (both standard increasers, hence mod value = +1). Finally, Growl is a standard decreaser, and therefore has a mod value of -1.

(Wait! We just said that statistics have "mod values". Now we're saying that attacks have them too! What gives? Well, technically the numbers ascribed to attacks should probably be called "net effect on mod value", or "mod value deltas", but that's awfully cumbersome. We'll just call these numbers "mod values", on the understanding that we're really referring to how the attack affects the mod value of the stat.)

Example 1. Your Pokémon employs Swords Dance (+2) once, but subsequently gets hit by one Growl (-1). To find the current value of your Pokémon's Attack statistic, sum the mod values of the relevant attacks used to that point. In this case, (+2) + (-1) = +1, so Attack will be at 1.5 times its normal value.

Example 2. Your Pokémon gets hit by 3 Growls (-1, -1, -1) and then uses Meditate (+1) once. The current mod value for its Attack will be (-3) + (+1) = -2. Attack will therefore be at 0.50 times its normal value.

The table specifies two limits of statistic modifiers: they cannot be used to raise a stat to more than 4 times of its initial value (mod value +6), nor lower it to less than 0.25 of its original value (mod value -6). Additionally, no statistic can be raised above 999 or lowered below 1 without the use of a cheating device. Once either of the upper limits (+6 or 999) have been reached, subsequent stat-increasers will fail. Once either of the lower limits (-6 or 1) have been reached, subsequent stat-decreasers will fail, though damaging moves with stat-lowering side effects will still do damage: they will simply fail to produce the side effect.

If neither of the upper limits have been reached, but a stat-increaser would result in a mod value greater than +6 (or cause the stat to exceed 999), the move will raise the stat only up to the relevant limit. For example, suppose a battling Lickitung with an Attack mod value of +5 is instructed to use Swords Dance. The move will not have full function (raising the Attack mod by 2), but will still raise the stat up to the limit (Attack mod = +6). Subsequent Swords Dancing will fail. Similarly, if neither of the lower limits have been reached, but a stat-decreaser would result in a mod value less than -6 (or cause the stat to fall below 1), the move will lower the stat only down to the relevant limit.

The 999 limit is what keeps a trained, L100 Mewtwo from successfully using Amnesia more than twice. His Special is so obnoxiously high that it reaches 999 before the statistic multiplier reaches 4. However, if the Mewtwo is subsequently hit by a Special-lowering Psychic, it will be able to use Amnesia again to restore the stat to 999. Because Amnesia is a great stat mod, only one is needed to fully recover from the effects of two Special-lowering Psychics.

Altered statistic values are temporary: the effect of stat mods is negated if the Pokémon switches away or faints, or if the battle ends. Haze also resets all the stats (other than HP) of both Pokémon to their initial values.

Attack, Defense, and Special
Attack Type Base PP Pow Acc Avg Effects TM HM
Aurora Beam ICE Spec 20 65 99.6 64.7 lower victim Attack (10% chance) - -
Growl NOR NDA 40 - 99.6 - lower victim Attack (100% chance) - -
Meditate NOR NDA 40 - 100 - raise Attack (100% chance) - -
Rage NOR Phys 20 20 99.6 19.9 attacker goes nuts/ raise Attack when hit 20 -
Sharpen NOR NDA 30 - 100 - raise Attack (100% chance) - -
Swords Dance NOR NDA 30 - 100 - greatly raise Attack (100% chance) 03 -
Acid POI Phys 30 40 99.6 39.8 lower victim Defense (10% chance) - -
Acid Armor POI NDA 40 - 100 - greatly raise Defense (100% chance) - -
Barrier PSY NDA 30 - 100 - greatly raise Defense (100% chance) - -
Defense Curl NOR NDA 40 - 100 - raise Defense (100% chance) - -
Harden NOR NDA 30 - 100 - raise Defense (100% chance) - -
Leer NOR NDA 30 - 99.6 - lower victim Defense (100% chance) - -
Screech NOR NDA 40 - 84.4 - greatly lower victim Defense (100% chance) - -
Tail Whip NOR NDA 30 - 99.6 - lower victim Defense (100% chance) - -
Withdraw WAT NDA 40 - 100 - raise Defense (100% chance) - -
Amnesia PSY NDA 20 - 100 - greatly raise Special (100% chance) - -
Growth NOR NDA 40 - 100 - raise Special (100% chance) - -
Psychic PSY Spec 10 90 99.6 89.6 lower victim Special (30% chance) 29 -

All Attack, Defense, and Special Modifiers

You can use the system described in the General section to understand how statistic modifiers will affect your Pokémon's ability to do (and withstand) damage. Simply use the modifier value table to find the multiplier for the relevant statistic, and use the new statistic value at the appropriate place in the Battle Damage Formula. Thus, if the relevant stat (Attack, if using a physical attack; Special, if using a special attack) of the attacker has changed, you should adjust the value of the B variable. If the relevant stat of the defender has changed, adjust the value of the D variable.

An interesting consequence of the way B and D relate to each other in the damage formula is that the interaction between say, Attack-modifiers and Defense-modifiers is multiplicative, not additive. For instance, using either 3 Swords Dances on yourself, or 3 Screeches on your opponent, will each ordinarily result in 4x normal damage. However, using all six of these attacks in a single battle (ie, with a single opponent who does not switch away) will yield not 8x normal damage, but 16x. And Harden combined with Growl is more effective at reducing damage than 2 Hardens. (All of this has been confirmed through testing.)

Tactical Notes

Full descriptions of the physical and special attacks that have statistic-modifying side effects can be found on their individual pages, and advice on their use is available in the Attack Type comparison pages.

None of these moves provide a very reliable means of altering statistics, but Rage suffers an additional limitation: the Attack boost it provides cannot be used to improve the effectiveness of any other attack. The remainder of this section discusses the non-damaging statistic modifiers.

Also worthy of note is Psychic, which, as if Mewtwo needed another advantage, actually gets stronger as it used. It lowers the victim's Special, the statistic on which Psychic defense is based. Acid does a similar trick with the Defense statistic, but it is otherwise a weak attack, and its Defense-lowering side effect does not happen as frequently as Psychic's.

All pure statistic modifiers have plenty of PP, and most are very accurate. Furthermore, any non-damaging stat mod will work on any Pokémon (eg, although it is Normal, Growl can be used to reduce Haunter's Attack) so Type is not an issue. A theoretical consideration for selection is addressed on the Evade/Accuracy page: whether to choose a move that modifies the stats of the Pokémon using it, or one that affects the enemy Pokémon. However, in practice this really comes down to a choice between Growl and Defense Curl for Chansey - not something that should keep you up at night. A potential choice between Screech and Swords Dance for Tentacruel or Lickitung is made easy by Screech's lower accuracy, and there simply aren't any non-damaging Special-lowering attacks to be weighed against Growth or Amnesia.

The major consideration for these moves is therefore whether to include one of them in your Pokémon's moveset. When deciding, keep in mind that offensive stat mods will typically not begin to pay off until the 2nd (for great mods) or 3rd (for standard mods) round following their use. For example, suppose you instruct Hitmonlee to Meditate. On the subsequent round, a Hi Jump Kick will do 1.5x the usual amount of damage - but remember that Lee is already one kick behind (ie, the one he could have done instead of stat boosting). On the next round, the boosted Lee will have delivered two kicks at 1.5x normal power, or the equivalent of three kicks: exactly what he would have done without Meditating! It is only on the 3rd round following Meditation that the serene, contemplative Lee starts to show a numerical advantage in ass-whooping.

The problem with this scenario is that due to his low defensive stats, Hitmonlee may not last long enough to make the stat-boosting worthwhile. Even if he were able to learn Swords Dance, using it would be only occasionally beneficial. A similar problem afflicts defensive stat-boosters: sure it's nice to double or triple your Defense, but not if the enemy has whittled you down to 1/5 of max HP in the meantime. Another thing to remember when boosting a defensive stat is that Critical Hits ignore statistic modifications entirely, and the fastest Pokémon (who are, not coincidentally, among the most popular) tend to CH about 20 to 25% of the time.

Because of these concerns, the best way to make a stat mod worthwhile is to supplement it with an HP-restoring move such as Recover or Rest. Recover is only available to a handful of lucky Pokémon, but nearly anyone can learn Rest, and can use Double Team to reduce the likelihood of rude disturbances during nap time.

Finally, when looking for a way to make statistic modifiers work for you, keep in mind that Growth and Amnesia are particularly attractive, since they simultaneously boost the offensive and defensive capacity of your Pokémon.


In the absence of any other Attack modifiers, the Attack statistic of a burned Pokémon is cut in half. However, things get a lot more complex if other Attack modifiers are also in effect. There be rough waters ahead for the faint of heart, so you can skip past the rest of this section if you aren't interested in the oddities of Attack interactions. If you are interested, but are also impatient, skip down to the end of this section and read the three rules that sum everything up...but it'll probably help you understand things better if you let us walk you through this.

The simplest way to correctly predict the interaction between Attack-decreasing moves and burns is to remember that at a mod value of 0, the Attack of a BRNed Pokémon will be 1/2 of its normal value. For example, suppose a Pokémon with 200 Attack becomes burned. Its Attack will obviously become 100; if it is subsequently hit by Growl, its new Attack will be 0.66 x 100, or 66.

It is therefore possible for a BRNed Pokémon to be at less than 0.25 of its normal Attack. If the victim in the preceding example were hit by 6 Growls, its Attack would be 0.25 x 100, or 25. This is a mere 12.5% of its original Attack!

Attack-increasing moves have an unexpected effect when used by burned Pokémon: they negate the Attack penalty associated with that status change and raise Attack by the expected amount. Thus, suppose our burned friend (presently at 100 Attack) used Swords Dance: its Attack would become 400 (that is, 200 x 2)! The 2nd and 3rd uses of Swords Dance would also have the expected effects: Attack would become 600 and then 800.

Now for the really tricky part. Even if Swords Dance has been used, and the BRN penalty thereby negated, a subsequent use of Growl (or other Attack-reducing attack) in a link battle will affect Attack's mod value in the usual way and reinstate the penalty caused by the burn! So, suppose our friend had used Swords Dance only once, and was therefore at 400 Attack. If it were then hit by Growl, the new mod value would be (+2, for Swords Dance) + (-1, for Growl) = +1. The table tells us that this mod value results in a statistic multiplier of 1.5. However, because Growl was the last Attack modifier used, this multiplier will act upon 100, not the normal value of 200. The Pokémon's Attack will therefore be 150.

In a wild battle, the situation is better for the burned Pokémon: any Attack modifier negates the BRN penalty to that stat. Suppose that a BRNed Pokémon with an Attack of 200 (hence, current Attack = 100) is hit by Growl. This move will change the mod value to -1, which has a multiplier of 0.66. However, the victim's Attack after Growl will be 0.66 x 200, or 132 (rounded down), not 0.66 x 100. In a wild battle, Growl on a BRNed Pokémon will actually increase its Attack. (Yes, this is lunacy. Yes, we have actually tested it.)

It's not known whether this confusing state of affairs was intended by the game's programmers. Regardless, you can determine the current Attack statistic of a BRNed Pokémon under any circumstances by applying the following rules:

  1. Determine the mod value for Attack normally. That is, begin with 0 and subtract 1 for each instance of Growl or Aurora Beam (if it produces its side-effect). Add 1 for each instance of Meditate or Sharpen, and add 2 for each instance of Swords Dance.

  2. When the mod value is determined, use the table to find the appropriate statistic multiplier.

  3. If the last Attack modifier used was an increaser (Mediate, Sharpen, or Swords Dance, apply the stat multiplier to the normal Attack value (that is, what it would be without the BRN). If you're in a link battle, and the last Attack-modifier was Growl or Aurora Beam, apply the stat multiplier to the BRN Attack (1/2 of the normal value). If you're in a wild battle, all stat multipliers apply to the normal (non-BRN) Attack.

Damage Inhibitors
Attack Type Base PP Pow Acc Avg Effects TM HM
Light Screen PSY NDA 30 - 100 - reduces damage from Special attacks by 50% - -
Reflect PSY NDA 20 - 100 - reduces damage from Physical attacks by 50% 33 -

Reflect and Light Screen

Reflect doubles the current value of the D variable in the Battle Damage Formula, if the incoming attack is physical in nature. Light Screen has the same effect, but is only helpful against special attacks. However, this is not to say that these moves double the statistic on which D is based: if this were the case, Light Screen would increase a Pokémon's offensive capacity with Special attacks, but this does not occur.

These attacks can be used in combination with statistic modifiers. No Pokémon that learns Light Screen also learns either Growth or Amnesia, and there are no TMs for any of these moves (though they might still be combined through the use of Mimic). However, a number of Pokémon that are able to learn Reflect can also learn a Defense-boosting attack. As stated above, Reflect doubles the current value of D.

Thus, suppose your Pokémon has a Defense of 100 and uses Harden. This is a standard stat mod, and therefore will boost Defense to 150 in this case. If your Pokémon also uses Reflect, the D variable will become 300 when the incoming attack is physical. Note that the order in which these moves are used has no effect on the final protective value: Reflect simply takes the current value of D and multiplies it by 2. If Reflect had been used first in this example, D would be 200 against physical attacks until Harden was used. Of course, in an actual battle it would be wise to use Reflect first: while building your full defense there will be one round on which you can only have one attack in place, and Reflect provides greater protection with a single use.

Reflect and Light Screen can interact with other defensive moves in surprising ways if the value of D becomes very high. Consider the following examples, in which a cheating device was used to teach the Pokémon moves they could not ordinarily learn:

Example 1. A Mewtwo used Amnesia twice, after which its Special was 999. A second Mewtwo used Surf against the first, and damage was averaged over 10 trials. The Amnesia'd Mewtwo then used Light Screen, and another 10 trials of Surf were done. Average damage was virtually identical!

Example 2. A Golbat with a Special of 181 was taught Amnesia and Light Screen, and used them to defend against a Mew using Surf. Without the benefit of either move, Golbat took an average of 100 HP of damage from Surf (all averages are approximate, as only 4 to 6 trials were done in each case). After one use of either Amnesia or LS, average damage was about 60 points. After both moves were used (in either order), average damage was about 30 points. So far, so good. However, after two uses of Amnesia were combined with LS (again, the order did not matter), average damage was 375 points! Finally, after three uses of Amnesia were combined with LS, average damage was about 52 points.

The explanation for this appears to be that D has a maximum value of 1023 (2^10-1, or 3FF in hexadecimal). When it reaches 1024 (2^10, or 400 in hex), it "rolls over", so that 1024 is equal to 1 for purposes of calculating damage. This theory is consistent with the preceding examples. When the Mewtwo with 999 Special used Light Screen, the D variable against special attacks should have become 1998. If the value resets to 1 after 1023, however, the actual value of D would have been 975 (1998 - 1023): nearly the same as what it was without the use of Light Screen.

Meanwhile, Golbat's problems started when LS was combined with two uses of Amnesia. This would result in a D of 1086 (181 x 3 x 2), but the 375 points of damage is explained if the rollover reduced D to 63 (1086-1023): a fraction of his normal, un-"boosted" Special. A third use of Amnesia (with LS still in place) would result in a D of 425 (1448 - 1023), explaining the damage reduction that was found in the test.

Attack Type Base PP Pow Acc Avg Effects TM HM
Agility PSY NDA 30 - 100 - greatly raise Speed (100% chance) - -
Bubble WAT Spec 30 20 99.6 19.9 lower victim Speed (10% chance) - -
Bubblebeam WAT Spec 20 65 99.6 64.7 lower victim Speed (10% chance) 11 -
Constrict NOR Phys 35 10 99.6 10 lower victim Speed (10% chance) - -
String Shot BUG NDA 40 - 94.5 - lower victim Speed (100% chance) - -

All Speed Modifiers

Speed is modified in exactly the same way as Attack, Defense, and Special, though of course the current value of your Pokémon's Speed is not directly relevant to the Battle Damage Formula. In addition to the attacks listed above, Speed can also be reduced by any move that causes paralysis. In the absence of any other Speed modifiers, paralysis reduces the victim's Speed to 25% of its original value.

The Speed modifiers listed above interact with the PAR penalty in the same confusing way that Attack modifiers interact with the BRN penalty. These interactions are illustrated with specific examples in the section on burns, but in brief, you can determine the currect Speed statistic of a PAR Pokémon under any circumstances by applying the following rules:

  1. Determine the mod value for Speed normally. That is, begin with 0 and subtract 1 for each instance of String Shot, Bubble, Bubblebeam, or Constrict (only if the last 3 produce their side-effect, of course). Add 2 for each instance of Agility.

  2. When the mod value is determined, use the table to find the appropriate statistic multiplier.

  3. If Agility was the last Speed-modifying attack used, apply the stat multiplier to the normal Speed value (that is, what it would be without paralysis). If you're in a link battle, and the last Speed-modifier was String Shot, Bubble, Bubblebeam, or Constrict, apply the stat multiplier to the PAR Speed (1/4 of the normal value). If you're in a wild battle, all stat multipliers apply to the normal (non-PAR) Speed.
Please ignore this subliminal message. Please ignore this subliminal message. Please ignore this subliminal message. DRINK COKE Please ignore this subliminal message. Please ignore this subliminal message. Please ignore this subliminal message.