Pokemon's Weirdest Calculation

[@waluigiisthebest2802]

Damage calculation was definitely not made with low level Pokémon in mind. Considering that in the main game, Pokémon only have tackle and growl at those levels.

[@DigitalRag3X]

Egg moves at level mon has been a thing since 4 gens ago

[@TheJpiscool]

@@DigitalRag3X And the damage formula has been a thing since 7 gens ago what's your point

[@polygondeath2361]

What do you mean main game? Plenty of pokemon can have access to far more moves via tutor, egg moves, tm/tr/hm, and some have different moves

[@thedon6079]

@@polygondeath2361 by the time u have access to that you will have far higher level pokemon.

[@polygondeath2361]

@@thedon6079 but we’re talking about LC. You aren’t ever going to encounter a situation in game where you have a corphish with adaptability, aqua jet, and life orb against an onix with eviolite. Not to mention the evs and ivs involved.

[@Regnorash]

Just checked and without sun this is actually a guaranteed ohko, even without the crit. Crazy how sun turns it from ohko to 1 damage

[@UnuUO]

Makes you think that the damage calc might need an overhaul so that weather comes into the equation later

[@Menosh]

@@UnuUO I think this is such a niche situation that it's not worth worrying about. How often are you battling with level 5 pokemon that have weather effects?

[@MOLEMITTS12]

@@Menosh A good amount, in LC, apparently

[@codetaku]

@@UnuUO The fix wouldn't be to change the order, it would just be to store intermediate floating-point numbers properly in your registers. They gain nothing from repeatedly casting to int. There is no optimization to be found in their method--only madness. They can cast to int at the very very end instead.

[@lynneecho4598]

@@codetaku i'm not great with computer sciences, but wouldn't casting to int each time reduce the complexity of each math step? maybe it's an optimization specifically for bigger numbers that might require more memory to storage in between steps

[@FreezaiPokemon]

For some clarity, "rounding down" is an oversimplification of what's going on. What's actually happening is integer division because data is stored in the Int data type. They could just as easily have used floating-point numbers and solved the "issue" (if you believe this to be a problem) but these oddball situations are so rare

[@saltybalot3021]

Or why not just calculate all at once instead of calculating step-by-step?

[@rex_melynas]

There are some float calculations already, they don't do value * 3 / 2 for STAB, for example, they do it * 1.5. so they could probably store it as a float nowdays and do the floor at the end. I don't think doing that Will make it much more expensive to calculate

[@PrismariLaura]

The amount of storage space that would be needed to solve this is entirely insignificant compared to current game sizes. Like, it's on the level of bytes, maaaaaaaybe single kilobytes if they really go ham on floats. Sword & Shield are 10gb in size. That's several orders of magnitude bigger than what we're talking about. These sizes are only relevant for Gen1 and Gen2. Perhaps Gen3 could've justified keeping them as integers, but it's absolutely not justifiable Gen4 onwards.

[@blazernitrox6329]

Storage was nothing to do with it. The GB, GBC, GBA, and (if memory serves) the DS don't have dedicated hardware for floating point arithmetic the way modern systems do. They could use software floating point, but it would be prohibitively expensive in terms of CPU cycles. I suspect that the only reason they still do that nowadays is as a hold over from those old systems.

[@PrismariLaura]

@@blazernitrox6329 that could be the case. I'm not too familiar with the architectures of those systems. In that case, assuming the DS also lacks the hardware capabilities to efficiently use floats, this should have stopped being a problem in gen6. Otherwise, it should've stopped in gen4.

[@sarasaland4709]

Switching Onix in on Corphish went from no brain to galaxy brain.

[@zacharymogel9500]

Alternate title: how damage calculation works

[@ElshazamReformed]

"watch for damage calculations in 0.5x A presses" is a way more accurate title.

[@oodogla]

But that title would attract less views.

[@addisontwede4148]

If that was the title I would not be here

[@bobjoe3492]

@@addisontwede4148 Not all clickbait is bad clickbait

[@addisontwede4148]

@@bobjoe3492 exactly right.

[@nicklewry3854]

Keep this series running, I’ll watch every video in it. This stuff is fascinating

[@miinesweep]

Only crazier damage calc are damage overflows, but this is absurd too

[@porkeman9007]

My Steelix got ohkoed by bullet punch

[@XtheGAMEmaster]

@@porkeman9007 Could you explain a bit how that happened? It sounds really interesting

[@porkeman9007]

@@XtheGAMEmaster max value a stat can be is set to around 655 (not sure) I was playing mix and mega, where the mon gets a stat increase that a certain mega stone gives to the intended mon I gave Steelix aggronite which increased base defense by 50 making it 250 With EVs it made the defense 658 and it crossed the max and started from 0 giving it only 3 It has been fixed and now even if EVs are extra it stops at 655

[@mr.krustykrust]

@@porkeman9007 and I’m pretty sure if you make it so you have the stat overflow to exactly 0, it made you basically unkillable on a specific defensive side. Cause I know eternamax Eternatus can reach that in Pure Hackmons but I don’t play PH so idk how it exactly works

[@airtempest8945]

@@mr.krustykrust The most common time it happens is Gen 2 OU, since Thick Club Marowak with 3 swords dance at max attack actually underflows attack, which is why some people don't run max attack on Marowak, so they can do 3 SD instead of capping out at 2. Those who aren't running SD/only intend to do 1 or 2 SDs can just run max attack.

[@VersacePokemon]

It’s interesting that “level” is involved in the damage calc equation. I always assumed for example a level 10 Pokémon and a level 30 Pokémon with the same attack stats, using the same move, vs the same defense etc. would deal the same damage regardless of level

[@Qualicabyss]

It balances out the fact that 2 defensive stats increase while only one offensive stat, without this moves at low levels would almost always ohko.

[@VersacePokemon]

@@Qualicabyss ahh I see. Because both hp and def/spdef increase with level vs just attack/spatk.

[@12jswilson]

@@VersacePokemon this came up in a discussion for draft league where someone proposed allowing Mega Metagross but only at level 85. The stat reduction makes it more manageable on paper, but I pointed out the level effect on damage would make it just outright terrible.

[@sebastianlucas704]

It is involved. He showed it on screen.

[@TheKitKatThe]

@@sebastianlucas704 yeah lol, they know, they just said they find it interesting

[@thatsmean2929]

So little cup is essentially where the pokemon calculator breaks down.

[@CBuiscuit]

I would be so pissed if I hit an Onix for that little damage with a 4x effective move, even if my Pokémon was 10 levels lower.

[@jimmymackinnon8474]

That’s just the way you are. You have an anger problem and you are too impulsive.

[@something_sorta_serious831]

@@jimmymackinnon8474 shut up, Meg

[@Mr.XY97]

@@jimmymackinnon8474 It's not that deep lol

[@victorlinares4137]

@@jimmymackinnon8474 wierdo

[@imab0t274]

@@jimmymackinnon8474 ikr, god forbid this person experience basic human emotions

[@13kevyaga31]

the only thing crazier is pomeg berry glitched HP Water Spout/Eruption in Emerald. These moves consider current HP/max hp in their equation so if you have 65535 HP, you will deal monstruous damage

[@pigeonskins3887]

isnt it based on hp% remaining?

[@willphillips2522]

Does it? Wouldn’t the same overflow cause damage to round weirdly?

[@Nick-ci8ve]

@@pigeonskins3887 the glitched hp would be a lot more/less than 100%

[@XtheGAMEmaster]

@@pigeonskins3887 because the Pokémon’s current HP would be significantly higher than it’s max hp, your hp would be OVER 100%, which continues to add base power to the move.

[@zenonkuropas543]

@@pigeonskins3887 water spout and eruption are based on current hp

[@davidahia8311]

"I looked forward in time to view alternate futures and see all the possible outcomes." "And in how many of them do we win?" "1 in 16 baby, and here let me do the math for you."

[@invalid3354]

You forgot to include the crit chance And no, i'm not fun at parties

[@Mewseeker]

@@invalid3354 The video already established that the crit chance didn't do anything in this case. :D

[@dylanb2990]

@@invalid3354 wow imagine being lame and wrong

[@Im_Marco]

@@dylanb2990 guy's literally invalid

[@archmagusofevil]

School: always avoid rounding until the very end. If you don't, your calculations will be off. Pokemon: Pfft, like it would really matter that much. Onix: How cute, it crit it's water attack.

[@David08225]

If you run those same damage calculations without intermediate rounding, the correct damage range would have been 24.8~29.2.

[@henriquepacheco7473]

if they used floating-point numbers within the calc and cast to int only at the end, this move would always ohko in this context. Wild shit.

[@psymar]

​@@henriquepacheco7473Or even if they did what NetHack does (an even older game dating to the 1980s!) and simply multiply by 1000, do the divisions, and divide by 1000 (although you do need more bits if you're going to do this without overflow, which may be why they didn't; NetHack is a PC game, which even in the 1980s you could use a Long Integer for 32 bits. The reason NetHack does this is back in the 1980s floating point calculations were SIGNIFICANTLY slower than integers)

[@panzerldr5724]

These short explanation videos are actually excellent. Short, yet thorough explanations of interesting interactions and meta details in competitive Pokemon. Even if already knew a few of these, such as the Nidoking video, they're still pretty satisfying to watch. Kudos to you, keep it up my guy.

[@luukvankrugten7253]

These short videos were you explain super random Pokémon battle facts are honestly amazing to watch. To the point, fun and easy to follow. Every time I try to think how or why is it that they banned this random Pokémon or how does a 4 times super effective move deal so little damage? And I am never able to come up with the full solution. And then when I watch the video it all makes sense and falls into place. Keep it up man 🙂

[@aparnajana8030]

The concept of spread damage is that-Spread moves like EQ/icy wind deal 25% less damage than they would do in singles The only scenerio where spread moves deal their usual damage output (alike singles ) is when ur opponent lacks an Ally (that is, u damaging only a single opp. in a doubles battle) Btw Thnx for the info...those roles have a wide gap

[@blue_elite964]

or if their ally protects

[@aparnajana8030]

@@blue_elite964 I never thought bout this tbh But I don't think that protecting an Ally would grant full damage output on the other The 75% reduction still applies even if u miss on one of ur opposing mons but hit the other one

[@blue_elite964]

@@aparnajana8030 While that may be true, that is because it was *supposed* to hit the target, but the roll made it not. If one target protects, the move becomes a single-target move, and does full damage, because it literally is not targeting the opponent that protects (I think you should probably check though, im going off of prior knowledge so i may be wrong)

[@aparnajana8030]

@@blue_elite964 Since the miss part keeps the roles at 75% only I guess it applies to protect as well Cuz same logic applies to both conditions IMHO

[@DoctorSpacebar]

Does it still spread if there is an ally, but only one enemy?

[@nerdyfalco408]

I feel like corphish vs onix isn't the only time this kind of bs happens in LC

[@gengarzilla1685]

LC is certainly one of the jankiest formats by a good margin.

[@alexread6767]

And this is why you don't round down during calculations :)

[@thebestworst8002]

If you get into computer science, everything integer truncates when the value is a decimal. It rounds down to keep things simple and since the damage calculator probably wasn’t made for these extreme scenarios in mind, rounding down keeps everything simple

[@tetrikitty1686]

@@thebestworst8002 You don't have to truncate at the 1s place, though - doing it at 0.1 or 0.01 would have worked.

[@thebestworst8002]

@@tetrikitty1686 no integers always truncate to the 1s place in computers. There is no truncating to decimal unless you use double variables which are more complex

[@abrahamkutfaz1628]

@@thebestworst8002 The calculator would honestly prolly work better if the calculations were done in doubles, and then casted into an integer after all the calcs.

[@thebestworst8002]

@@abrahamkutfaz1628 it probably would but since this stuff barely matters once your Pokémon’s level is above like 20, it’s just simpler to keep everything as an integer and not deal with any rounding errors double variables tend to have

[@jacobchencarrasco2959]

I’ve always dreaded getting into the nit-n-grit of damage calcs but I enjoy the way you explain each step. Definite a sub from me 👌🏼

[@asterlofts1565]

Maths + RNG game = Pokemon. Nice.

[@davidahia8311]

Let Corphish do 50% against Onix, RNGesus. Don't be a coward. Let it kill Onix. Let it show that snake what it deserves. Freezai do you like Littles? I feel like LC is so cute, but somewhat jank as you can see here.

[@silverkyrie4714]

To be fair, I don't think LC is actually as janky as seen here. This is just a rare and very specific example, not like a regular occurrence you'd see all over the format.

[@xjjay554x]

@@silverkyrie4714 It is by far the jankiest format besides gen 1. There's also the whole bullet seed rock blast problem, berry juice sturdy, and plenty of other wacky things.

[@davidahia8311]

I guess "as" is doing some work here, you're right. This is likely the upper bound of LC jank, you're right, but it's one of a decent sized list of mons and strats that either work counterintuitively or are broken.

[@longbeing]

Videos like this are an interesting look at the math and programming behind Pokémon. I could actually imagine a teacher using one of these as a fun example for a comp sci class.

[@Mewseeker]

I can't wait for when the Pokémon anime will reference this by having a young Corphish dealing near non-existent damage with Aqua Jet to a young Onix. :P

[@sagacious03]

So basically, rounding down at every step with damage this low & most multipliers being less than 2 makes the damage so low that it ends up being 1 regardless, & the damage formula on numbers this low brings it close enough to 0 that rounding at every step MAKES it 0, hence, the bare minimum 1 damage. THEN Type Match-Ups & Abilities. It's only when you get that 1-in-16 chance max roll where it multiplies your 1 damage by 1 instead of by 0.99 or less to make it 0 that your multipliers matter. Do I have that right? It's like a crit, but, in large part because of the order of operations & rounding, instead of that "crit" (The max roll.) doing 1.5 or even 2x damage, it does 10x damage, lol. Great video! Thanks for uploading!

[@TheTwistedKris]

Min 1 dmg is after type match-ups, aside from that you got it. The x10 is actually from the STAB, type, then item (1x2=2, 2×4=8, 8x1.3=10.4=10)

[@Interesting_Failure]

So in short, the actual crit mechanic does nothing for it but it has a 1 in 16 chance to score a "secret crit"

[@ericwolf9664]

More like has a 1 in 16 chance to do appropriate damage

[@cyanideytcuriousseadoggo]

1/16 to do some OK DAMAGE AND NOT BUGGER ALL.

[@taechuK]

I love how you need to crit on your crit do deal damage in this case

[@thefatpikachuchannel7672]

Crit + stab + super effective + adaptability boost + 4X weakness + life orb = 5% damage Strength

[@kin5922]

I love these so much, it's knowledge that I won't use on a day to day basis but it's just nice to see things layed out like this.

[@aidendoig4542]

i cant believe they didnt use floating point numbers and just chop off the decimal at the end. the only reason i can think they wouldnt is the original game boy might not have had enough memory to do precise calculations, and then on future consoles i guess they kept it this way out of tradition. wild.

[@michael_betts]

The original games (rby and gsc) did it in a single byte, and divided by 2 for stats between 256-511, and 4 for stats above 512. Since stats cap at 1k that should be fine. But with light screen and thick club and stuff you can cause overflow. Thick club marowak needs 13 attack IV in gen 2 so that it's post-swprds dance attack doesn't roll over into single digits.

[@n0ame1u1]

Forget the original game boy, even the DS didn't have the hardware to do floating point operations (although they can be emulated in software)

[@user-et3xn2jm1u]

This is my new favorite channel, I know nigh bupkis about pokemon but the way you explain it is interesting, straightforward, and pleasant.

[@skillboom2]

Alternative title "why most tournaments aren't at lv 5"

[@Inhuman1105]

W Darth nihilus pfp

[@krumplin8992]

You'd think that weather would be a modifier with all the axillary effects like life orb ect

[@ronnieketchum1241]

Yup. Damage Calcs kinda fall short in LC. Or you can have a higher processing unit and round down at the very end.

[@thomasmctavish6819]

This is really well explained! Great video!

[@philbb23]

ngl these videos are massively entertaining! would like some longer ones too on more detailed topics similar!

[@metleon]

In FFXII, there's a [Weapon Atk - Defense] term in there, so if your equipment sucks enough, you won't even do a single point of damage. I think most RPGs have similar terms, but put character strength and weapon attack together before doing the subtraction, so high level characters will still do damage with weak equipment.

[@Edujs23]

Fire emblem damage calculation is literally just, strenght + weapon power - enemy defense Some characters are litteraly unusable because of their low strenght cant keep up with enemies defense

[@ianidiot9297]

@@Edujs23 I can't really think of any instances of this within Fire Emblem. Most units are able to do actual damage, even *Sophia,* for crying out loud. The only instance I can really imagine is a strength-screwed Neimi from Fire Emblem 8.

[@matthewkuscienko4616]

@@ianidiot9297 About Sophia, you do realize that while she's considered to be a bad unit, it has nothing to do with her damage output, right? From my understanding, she's actually a pretty decent damage dealer, while having a combination of low speed (resulting in her being double a lot) and being extremely fragile are her main banes

[@ianidiot9297]

@@matthewkuscienko4616 She has the lowest base stats of any player unit in Fire Emblem, not including the old dragon man you know and love. That's why I brought her up.

[@CagnolinoBingolino]

I'm so glad that I found your channel bro. I love your content.

[@DragonaxFilms]

I love stuff like this. I don't even play Pokemon, but seeing the math and code behind game mechanics is super cool to me. It's such a fascinating topic.

[@hppern3971]

Ngl, a lot of these sorts of quirks with Little Cup often make me wish LC was level 100 instead. Would be far more inviting to new players from other tiers.

[@nickspringer5755]

but then rip berry juice sturdy

[@wezen89]

I'm now visiting back the video that got me into freezai. Best Pokémon channel even

[@AmbiambiSinistrous]

You bring out the beautiful in PokeMath - thank you for making this series!!

[@psc6483]

This guy:*explains damage calculation inclusive weather condition and so on Me:“it’s because that’s the special onix from the Pokémon episode

[@asterlofts1565]

Original video of the battle, where I can see complete my friend? :)

[@jared805]

Is your english 2nd language

[@asterlofts1565]

@@jared805 No. :(

[@asterlofts1565]

@@jared805 I bad in english :(

[@neanyanya]

@@jared805 don't be rude.

[@neanyanya]

Pokemon showdown :)

[@MidnightWolf0801]

I was literally looking for a ASMR video in my recommendations.... i found this, and I fell asleep to this instead because you know “MATH”

[@brandonsoto1197]

I Finished my career only to keep math away for me, but still chasing me with this one.

[@Mewseeker]

Yeah, in the world of video games, math in various form is a pretty big deal.

[@mini-_]

I've been playing Pokemon since the first generation and had no idea there was a random multiplier in the damage calculation! Of course now that I know about it, I can definitely think of times where I've seen it happen.

[@ohnoo4468]

have you been playing competitively? I feel like there's no way you wouldn't know, considering that damage calcs always give you a range

[@mini-_]

@@ohnoo4468 No I haven't played competitive, other than a little bit of Showdown over the years. Damage range is something I've always had a hunch about, but never knew for sure until recently.

[@ohnoo4468]

@@mini-_ Well now you know for sure lol that's cool tho. I was just curious.

[@mrgranadas2051]

This weird intersection of math and pokemon is going send me down a rabbit hole

[@fidelluz2942]

my question is why? integer numbers are faster and probably were used in red&blue but nowadays the hardware is so much better so why don't use float numbers all the way then rounding in the very end?

[@JrTroopa]

For the same reason Nidoqueen still can't breed. Gamefreak is allergic to change.

[@blazernitrox6329]

I suspect less "allergic to change" in this case. Moving to floating point would only serve the purpose of fixing very bizarre edge cases such as this, and would come at the cost of changing an algorithm that's literally the foundation of how competitive play works. The change may be relatively minor, but as we see above that change can literally make corphish go from "it has a small chance of actually being a threat to Onyx" to "it is literally always a threat," and a player who doesn't know that change happened (or simply forgets that it happened) could potentially lose the battle because of it.

[@matthewgagnon9426]

@@blazernitrox6329 That will literally only happen *once* to someone before they learn "Oh yeah, the formula changed." Keeping a bad formula around just because it has always been that way is the reason why Pokemon has stagnated as a franchise for so damn long.

[@n0ame1u1]

The original Gameboy didn't support floating point numbers. The Gameboy Color didn't support floating point numbers. The GBA didn't support floating point numbers. THE NINTENDO DS didn't even support floating point numbers (nor did the DSi). The 3ds was the first Nintendo handheld to have proper hardware support for floating point numbers. Of course it's possible to emulate floats in software, but that's very slow, and not really something you want to do too often. It probably would've worked fine since damage calculations don't happen that often on the scale of how fast computers run. They could also have just implemented rounding division using integers. If a//b rounds down, you can do (a + (b//2))//b instead to round to the nearest integer instead (and b//2 can be computed efficiently as b>>1). Basically this works because mathematically, floor((a + b/2)/b) is the same as floor(a/b + 1/2), which is the same as rounding to the nearest. (Although this still wouldn't be perfect for very small numbers in the denominator, like 3 or 5)

[@GrouchyRaccoon]

Huh. This is such an obscurely specific set of circumstances yet it explains so much about the mechanics.

[@dorkmork7349]

I never knew I needed random pokemon facts this much

[@funnycat9962]

This is why you wait until the end of your calculations to start rounding, kids.

[@Dosa-xp1nj]

Great effort and work!

[@blackimbreon9451]

Only deals 1 damage would be better clickbait than only deals 5%

[@CursedXyster]

it's not clickbait either way though

[@gavrielfeigenbaum9274]

Underrated content, please make more

[@bleuj4528]

Someone commented the same hit without sun is a guaranteed OHKO, so I tried to calculate it based on this video and turns out the min damage would be 20 < 21 which is Onix's HP. Turns out NOT EVERYTHING is rounded down. I looked at some of the smogon calculator's source code, and at some steps in the calculation a different rounding function is used instead of the floor function. This functions rounds anything > 0.5 up (instead of the usual >= 0.5!!) and anything

[@manahive6049]

Really good video about Pokemon damage calculation, never really got how it worked until now!

[@manahive6049]

also imagine being that 1 guy who got this exact scenario to happen in a match

[@quetzalthegamer]

If the calculation didn't round down at every step, this would be a guaranteed kill every time. Lol

[@matthewkuscienko4616]

In hindsight, this isn't all that surprising to me, as I remember playing in the battle tower in gen4 one time and a Chansey I was battling against was dealing pitiful damage to my Aggron using Focus Punch, which at the time was kinda wild to me that a base 150 power move that deals x4 super effective damage was barely even scratching my Aggron

[@imab0t274]

Makes sense, you’re taking a mon with 5 base attack and hitting one of the most physically defensive mons in the game. It’s gonna do pitiful damage regardless of base power and type effectiveness.

[@matthewkuscienko4616]

@@imab0t274 That's technically true, bit it's still hilarious to me that it did that little of damage when normally you'd expect a move with that high of a base power that is dealing 4x super effective damage to be a guaranteed one-shot

[@Lordmewtwo151]

@@matthewkuscienko4616 Given the attack/defense difference between Aggron and Chancey, guaranteed one-shot doesn't make sense. That said, I agree that scratch damage doesn't make sense either. Maybe about 20% of your max hp (possibly 10% depending on damage roll)?

[@Lordmewtwo151]

@@imab0t274 True, but you'd probably expect it to do more than scratch damage. I gave a percentage of max hp in my reply to OP's reply to you that I think should be reasonable taking into account the base power and 4x weakness. Yes, 1/5 max hp is bonkers low for an attack of that power and effectiveness, but again, between Chansey's nearly nonexistent attack and Aggron's defense, of course it would be lower than it should.

[@matthewkuscienko4616]

@@Lordmewtwo151 If I recall correctly, I don't think it was even quite 10% (granted it was so long ago I can't exactly remember). Either way, as I have been saying in my previous comments, it's honestly shocking that a move with that huge a base power that deals 4x super effective damage was only doing so little, even taking into account stats. Again, normally you'd expect such a move to be a guaranteed one-shot, but in this case it couldn't have been further from it if you tried

[@tysonfontanez]

This is why you don’t use ints when doing division with small numbers lmao

[@yupa5221]

I was unprepared for the math lesson but it is still a good explanation.

[@guiguipop3658]

The craziest thing is that Onix will only do 5% back to corphish, without any gimmicks

[@JaceGarrdok]

This feels like a massive pkmn shitpost, but in reality this is incredibly interesting and insightful.

[@Cristhian318]

Great video, I learned something new.

[@zad_rasera]

This is what they teach in Trainer's School

[@Nachosaurio167]

Corfish: Why won't you die?! Onix: Damage calculator son!

[@BaneRain]

I love this kind of stuff. Subbed

[@kassygo1375]

So that’s how it works. Incredible video!

[@dreizehn.]

I just started playing Pokemon so this is why the damage seems random

[@der_basic]

Damn i've been playing double battles for so long and didn't even know that spread moves only deal 75% damage o.O

[@puppieslovies]

It does seem like an oversight that the zero damage check is done at the last step and not between every step, but I guess they didn't expect low level pokemon to have weather or items It would be interesting if the leveling system as a whole got a rework in the future now that there are more spinoff titles which have no obligation to follow main series conventions

[@jamesmattila-hine1133]

This made the mathematician in me so happy

[@natsu6156]

This has been the most random thing youtube recommend to me

[@dwcd_5341]

You just made math enjoyable thank you this should be taught in schools

[@zacharyadler4071]

Anyone else confused how 3 is obtained from the initial calculation at 1:18? (2 x 5 / 5) + 2 = 4 4 x 40 x 18 = 2880 2880 / 22 = 130.9... floored is 130 130 / 50 = 2.6 floored is 2 2 + 2 = 4, not 3

[@MegamanStarforce2010]

the extra hilarity that it says 5-50% but like 15/16 rolls are 1 damage so the corphish could literally be spamming aqua jet and just lose lol

[@tysondennis1016]

Onix has a crazy defense bc it was designed more as a boss battle than an playable character.

[@Cass731]

Extremely fascinating… subbed for more

[@Illeea]

Without rounding until the very end, 26-30 damage is done killing onix every time. Add floating point calculations to damage formulas.

[@zeywop]

i miss little cup :( i've tried to play it a couple times since gen 6 but i almost always would have a 2+ minute queue into the same opponent

[@EllTheSylveon]

nitpick: the numbers aren't rounded, but rather truncated

[@BabusGameRoom]

This is awesome. What a fun fact!

[@thepainsight]

and This is why you don't "round up" or "round down" until the final stage of calculations.

[@TheChristmasNinja12]

Did I miss it, or did he fail to mention that a crit always uses the 100% damage calculation?

[@BullyWiiPlaza]

Very cool, they should have just used floating point values. Maybe this is related to their very C old code base or micro optimizations.

[@mondooowada3537]

i love these videos man!

[@ericborczuk135]

why oh why doesn't pokemon use floating-point numbers until the last step

[@ericborczuk135]

My science teacher from high school would cringe at this

[@tiexiaowang7939]

All this could have been avoided if they just rounded once at the end lol

[@gypsysprite4824]

looking at the above comments, and skipping over the technical it's less so that the system is actually rounding down, and moreso that it doesn't read the decimal portion of the number to begin with, and never has throughout the series. after all, why change fix that works perfectly fine EXCEPT in an extreme edge case like LC

[@jgnantes]

Funny how putting the random calc to the end of the process would have literally solved everything

[@tonytang7550]

Well, I can say I learned something new today

[@ElshazamReformed]

Dude, I Wana hear the classic pannnenkoek Mario 64 music just for the reference.

[@stellarknight04]

This is why i keep Hydro Pump on Corphish in LC for physically bulky mons (e.g. Koffing)

[@Kyluh.4]

Alt title: doing math to calc damage

[@KeertikaAndFallenTree]

LC is a wild tier.

[@wezen89]

I forgot the exact numbers but if you make a damage calc of 1000 attack mon hitting a 1 defense mon and the damage range is from 7000% - 200000% lol

[@dudono1744]

1 defense mon doesn't exist

[@wezen89]

@@dudono1744 It's just a damage calc not in a real battle

[@pantherstripes4380]

Very intresting, didn't know that this was possible and I play pokemon a lot

[@vulpony2130]

This is actually insane

[@LRXC1]

Are these just shorter reuploads of videos you’ve already done?

[@StepBaum]

this mans gonna climb the sub count faaast. Good vide tho :)

[@drakenpriest]

This vid poped up randomly but it makes me feel that pokemon damage calc is a broken think and needs fixing.