I Simulated 1,083,021 Battles to Find The Best Gen 1 Pokémon
HTML-код
- Опубликовано: 16 окт 2024
- Using Ai to Scientifically Rank EVERY Gen 1 Pokémon
➤ Here's a more detailed explanation of the ranking process for those interested: docs.google.co...
➤ View the full results: drive.google.c...
➤ View the code: github.com/cRz...
➤ Note: Battle gameplay is not necesserily accurate to our simulations as actual battles were run without a GUI
Join the Discord:
➤ / discord
➤ Twitter: / thesmithplays
➤ Instagram: / thesmithplays
Credits: Craig Livingstone (Production/Code), Weebra (Editing), Aero (Learnset Compositions/General Dataset Building)
Business Inquiries: thesmithplays@ellify.com
➤Music Tracks used:
Via dolce - Arms
Naked Glow 20th anniversay mix - Ridge Racer
On your Way - Ridge Racer
Butterfly Kiss - Persona 5
Tokyo Emergency - Persona 5
Ortiz Farm Tekken - 8
Price - Persona 5
A Fool or Clown Persona 4 - Ultimax
唖然呆然 (disc 2 track 5) - Yakuza 0
Layer Cake - Persona 5
Customer Creed - Yakuza 0
One-eyed Dancer - Yakuza 0
Whirling of Vairambhaka - Genshin Impact
Lee's ending - Tekken 5
Flirt with Bomb - Yakuza Kiwami
Rocket Nuts Groove - Yakuza 0
Just a few notes based on some recurring comments I've seen:
I think a lot of people are misinterpreting how we got movesets, saying things like "x pokemon was missing x move". Every combination of moves with at least 1 damaging move available for each pokemon were tested at each trainer. We were not trying to find the BEST moveset, we were only determining if it could sweep. As mentioned in the video, if a non perfect moveset is tested and it sweeps 3 times in a row, it moves on from that pokemon because obviously a better moveset will still sweep. If a "worse" moveset is tested and it can take out 4/5 pokemon, then another "better" moveset still can only take out 4/5 pokemon, it still keeps the first moveset to get that high score. It generally simply means "better" movesets didn't actually perform any better in that fight. So when you see pokemon with "weird" movesets, this is why. Remember only the top score moveset is kept, EVERY moveset was tested unless a pokemon got a perfect score for some moveset in which case the mon gets a perfect score and the rest are skipped.
I need to acknowledge that the AI is not perfect, we're still finding bugs and there's very likely situations I haven't considered when coding it where it just does not know the correct thing to do. If you've tried writing AI for pokemon yourself, you probably know it isn't an easy task, and requires many many iterations to get right. Currently known bugs include confusion moves being spammed, underestimating double/triple hit moves and dream eater looking at the wrong pokemon for sleep checks. Dragonair was also accidently given access at misty rather than surge and venusaur was accidently given earthquake, slightly boosting their rankings. If you find any more situations like this, please message us in the discord server, as this is where I'm most likely to see it :)
The trainer’s AI is improved from the base game, and is the same as the player AI. This leads to results like charmander being unable to sweep brock. I can see this being a controversial decision but think about it this way, it takes the randomness down dramatically because the AI is not using completely random moves. This makes it feasible to actually run the simulations, because we require significantly less repeat battles. I'd argue without running thousands of battles with a purely random AI you cannot get scientifically accurate results, and this is computationally infeasable. Also generally, if charmander cannot beat brock with decent AI and another pokemon can, of course the one that can is clearly the better pokemon. It is escentially taking into account the worst case scenareo for the original random gen 1 AI, so if they can beat good AI it shows they are consistent. It is even across all pokemon, so the ranking holds.
Our ranking does not take into account grinding time. This is a very fair point, while dragonite can sweep all of blue’s teams, it’s very difficult to get it to a level on par. Our ranking goes by level cap with no stat exp, so in theory dragonite at level cap is more equivalent to a dragonite 10 levels lower with stat exp. This means the ranking is still somewhat accurate, although the most rigorous way to do it would probably be to calculate the total exp you get for each trainer in the game after you catch it and figure out what level you would be. I hope people can understand that this would be extremely time consuming and the process for this video was extremely long already, hence why we did not go down this route. The way we modelled it is as if someone is using infinite rare candies, which I think still has value to a lot of the community.
Blue’s team is split into three and each one is weighted equally. I agree this is an oversight and does weight pokemon slightly to how well they perform in the champion fight. In future we will make sure to normalize fights like this down to one score.
Our ranking does not take into account valuable TMs being used up on the pokemon that maybe could have been used elsewhere. This may be something we can improve on in the future, but I'm not 100% certain on how we'd go about modelling this. If anyone with some math/stats knowledge has any suggestions for how this may be modelled please post them in the discord server!
A bug has also come to my attention where all mons on a trainer's team were set to level cap. Since this is even across all fights it should not effect the results majorly, however it has now been fixed for whenever we do this again.
Also just a note that the code, full results and a more detailed description of the process can be found linked in the description for those interested :)
- Craig/cRz Shadows (Programmer)
Bro, go easy on ditto because he's getting op on the incoming gens with a record of transformations becoming waay too powerful but they will probably slow him/her down probably decreasing stats per transformation.
In case you didn't got it yet, ditto would be able to enter battle with one of his prior opponents pokemons gens allowing him/her to take the advantage since you would be carrying like 10 pokemons instead of 6.
Obviously 4 share one single hp.
Do not take this lightly, everything i predict it happens from any videogames
I love Pokémon. Anyone reading this I could really use some advice. I have a 3ds xl and I have omega ruby on my ds. For a while I was using power-saves to get items and everything for the game. However just last night when I was trying to get ampharos and attached my cartridge into my 3ds it says it got corrupted. Now every time I place my cartridge in the slot remains empty (9) (s) (s) I'm nervous what should I do? Even my powersaves is telling me that my game which I always plugged my game to says it's unsupported. Any advice anything will help I just want to return it to Normal. Is there anything way to fix this?!
One approach to baking in the cost of "this required the use of a scarce TM" would be to think in terms of "if I use this TM on this Pokemon, will that interfere with my using the TM on the fights this Pokemon is bad at?". The TM is only costing you anything if there's some Pokemon you can't beat with the current Pokemon, but can beat if you'd used that TM on something else.
So, for a given TM that helps this Pokemon win at least one fight, limit your search to the fights this Pokemon *always* loses (including with other movesets, since those change over the course of a playthrough and you can always overwrite TMs), and see how many other Pokemon *need* that TM in order to win those other fights. The higher that number is, the more you should penalize the Pokemon for consuming a scarce TM.
@@austinbusumbru123 you have to open it and change the damaged parts.
You're not the first to come to talk about it on RUclips, there are some videos as well
@@Trustee-of-The-Most-High can you tell me how or where I can get it fixed? Because my main issue is that my 3ds won’t recognize my omega ruby cartridge neither does powersaves. It also said corruption last week when it happened
From the charts, it looks like all three Blue variants are scored fully... so this tells us which Pokemon would do best at beating the gyms, the Elite Four, and then Blue THREE TIMES IN A ROW. In other words, it's skewed to pokemon who can beat Pidgeot (x3) / Alakazam (x3) / Rhydon (x3) / Executor (x2) / Gyrados (x2) / Arcanine (x2) at level 61-65. It would be interesting to see how the rankings changed if each Blue encounter was weighted by 1/3.
@smithplays this person knows their stuff please let us know if this is feasible
Also, Starters should only be ranked against the Blue team that they actually face, not all of them.
Or they should have separate rankings for the situation where they are traded in (similar to trade evolutions).
@@Janders797Yeah, and then every other pokemon should be compared against the mean of the three fights.
What about a calculation of "if every pokemon could be your starter pokemon" ?
For that the best metods are the tier lists been made by Jrose and by Scoot Thoughts. Jrose uses game time and Scoot uses real time.
But both of them lowers the points due to necessary grinding, it's not the same if you have infinite candies 😋 slow gyarados and dragonite are not cool, while starters get more love due to the medium slow growth rate
@typhoon008 people forget there are a lot of pokemon with the slow level up growth rate. A gyarados feels no different than arcanine, Exeggcutor, or starmie. It's just that magikarp and dratini are singled out because magikarp has to switch train, doubling the time, and once it gets tackle, it's attack is so bad that each pokemon takes forever to take out. Dratini is stuck with wrap unless you use TM's/HM's. Again, taking much longer to take out each Pokémon, increasing the time it takes to level up. If they can one-shot, it wouldn't be bad. Its why legendaries feel different, despite also being at the same growth rate
I look forward to this sequel
Mewtwo wins
4:51 Fun fact, during my first ever Fire Red Playthrough, I actually fought Surge as the 7th Gym Leader. I wasn't able to read very well at the time, and ended up missing that I needed to teach Cut to a Pokemon to do that. That also meant that I didn't get Fly until after the 7th Gym lol
Incidentally, because I couldn't read very well, I didn't know how to get past the Old Man in Viridian and ended up with a Blastoise before Brock. I thought I had to be stronger before I was allowed past, and didn't realize my mistake until I accidentally walked into the Pokemart one day instead of the Pokecenter.
Jrose11 has entered the chat
Summon him with the forbidden chant.
"RIVAL FIVAL, RIVAL FIVAL, RIVAL FIVAL, BADGE BOOST GLITCH!"
Domain expansion: Badge Boost Beatdown
Scott's thoughts has entered the chat
Jrose needs to educate this man on Farfetchd
@@penpal222I have used fartfechd myself and it is even better than Pidgeot, even though the latter is my favorite.
Learn body slam, fly, double team and with any other attack you can think of they make it a great Pokémon.
"We see a big jump for Dragonite and a stellar performance for Articuno"
Articuno: *went up 50 places, yet the dragon is considered having the 'big jump'*
MY BOY SANDSHREW I AM SO PROUD OF MY BOY
articuno, zapdos, we need a third legendary bird
best i can do is a diet charizard 🤷🏻♂️
You mean better charizard
Moltres in gen 1 had a base special of 125 meaning it's overall rating was 620 instead of 580 after the special split. Charizard special was 85 making it's overall 510 instead of the 534 it is now. Moltres was better by 110 points.@gundispaberzs2281
3:28 Basically, the argument to put Ponyta much earlier, maybe near Bill(25 route).
Makes sense
Ponyta makes more sense in a big open field or something. Maybe East of Cerulean.
@@J05TI Nah, on the east, there is a rocky maze. My other guess would south-east coast, route 15/14. But that could be too late in the game.
@@grinchforest948It's tough. The Safari Zone feels like the perfect habitat for Ponyta, but it's very late. I think as far as the pacing of the game goes, North of Cerulean makes sense. But it's just that the area doesn't seem like Ponyta would roam around in.
How about East of Vermillion, where they added Scyther?
@@J05TI Well, there is also spacious route 1… but that would too early.
Of course a Water and/or Psychic Type would DOMINATE Gen 1 Kanto. 🐐
Starmie solos
This is a very well created video and a monumental effort from your team! Congratulations on producing an absolute banger. That being said, Pokemon is way more complicated than people tend to give it credit for, and even this very methodical approach leaves a lot of aspects about using Pokemon untouched and/or ignored, I'll just list some of them below:
1.The starters should only be able to face off against one of Blue's teams (unless trading is on the cards)
2. There is no anti-grinding factor. For example, Gyarados scores brilliantly partially on account of beating Misty even though it would take 15 levels of grinding to obtain a Gyarados at that point, a cumbersome task. Similarly, its an absolute pain to level up all the Game Corner Pokemon and the Fossils, which gets overlooked here.
3.There is nothing accounting for the time taken in the Pokemon battle. Eg- Arbok probably ranks highly due to Wrap Spam which is almost completely free, but it takes an eternity to win each battle with this strategy.
4. Opportunity Cost is a thing that comes into play in a generation with one-use TM's. Pokemon like Starmie, Dragonite , Jynx , Golduck, Rhydon and Mr.Mime become truly elite with coverage, but may require multiple valuable TM's like EQ, Psychic, TBolt, Ice Beam or Flamethrower to get there. This is especially a pain for the Game Corner TM's. This also means some Pokemon like Dugtrio may be a bit underrated despite having good level-up movepools. I imagine Water types would be even more broken since all of them get HM Surf for free.
5.There is never a system that can account for all the variance in a Pokemon battle. Pokemon like Butterfree or Venusaur that rely HEAVILY on Powder moves' 75% accuracy or non-100% accuracy STAB moves like Fire Blast, Rock Slide and Cross Chop feel intuitively worse to use than 100% accuracy Thunder Wave Electric Types and Flamethrower, Surf for example. This is especially a problem in Gen 1, where crits are based on speed and thus there is a bias towards faster Pokemon.
6.Selecting random movesets with a minimum of 1 attack is a necessity for some of the simulations, but no human ever using those Pokemon would ever run sets like that. Eg- Charizards without Fly, Jynx without dual STAB's. Also, some of the simulations seem kind of impossible for a human, like Beedrill VS Starmie having a 0% success rate even with Twineedle.
That is some of the things that I could think of for now, there's so much to account for! This really is where human judgement comes in and balances out all of the factors with appropriate weightage. I assume the "perfect and scientific" way of ranking Pokemon will be a statistical nightmare with millions of Permutaions and Combinations. Anyways, kudos on making an amazing video, it certainly left me quite intrigued and piqued my interest. The Poketuber community needs more scientific analysers like you and your team. Best of Luck for videos in the future!
Awesome list and I like how deeply thought out the weighting of the list is. Had a blast watching this video. Lovely editing as well. A sub well deserved :)
Make Trasform a priority move.
#justiceforditto lol
To make it work like imposter, how about that it also flinches the opponent for when you transform?
In a recent live he say that they take the sugestion and make this
It would be yet not enough. The issue is about the first pkmn of each trainer not having a good movepool or stats (and typically both), and a restricted number of PP
@@typhoon008 then maybe for gen1 and possible gen2 that Ditto gains the full amount of PP for each attack.
Another additional option is to give Ditto this BST distribution, along with Mewtwo's learning of TMs and some learnable moves from Porygon's and Muk's level up set, strictly the normal moves and acid armor.
HP 95
Atk 95
Def 75
Sp. Atk 95 -> 75 in Gen 1
Sp. Def 75
Speed 95
@@ThanatosZeroLetting Ditto learn a bunch of TMs would defeat the point of it being named "Ditto". It's whole theme is to copy its opponent.
Great video. Methodology all makes sense. I would say there are are a few factors difficult to account for in a similar that make a big difference in how Pokémon play:
Leveling - Slow leveling Pokémon like magikarp/gyarados or Dratini take so long level they play a lot worse. While those in other leveling groups and especially trade Pokémon like Mr.Mime or Farfetch’d will level much faster. Also the level you get then like Eevee being relatively high level when you get them helps.
Utility/Compliments - some Pokémon may only be useful in certain fights but compliment others well. Like Jolteon is really useful if you start with Squirtle since it sweeps all the fights blastoise struggles in.
AI - I suspect some Pokémon hurt from AI not playing them properly. Like with Farfetch’d a player can sweep every fight with setting up swords dance.
Similar to leveling, Ease of Availability also affects how useful a Pokémon is.
I agree with the comment about AI playing poorly as well. I assume the AI does better with moves that don’t require a certain order or strategy. Tauros with hyper beam, nidoking with horn drill and items, anything with a sleep move, anything with the move wrap, stuff like that.
I still don't get how Farfetch'd scored so low given the methodology for ranking. It's available early, has a great learnset, and can beat a ton of matchups throughout the game.
maybe because its a in game trade
@@salamipitza but that's a boon if anything - means it gets boosted experience
To be good you really need to use Swords Dance, makes it great but need that
@@bakuiel1901 yeah but it learns it, so you're not spending the TM. The only thing Farfetch'd costs you is the Body Slam TM
Should be higher. Hm slave jus saiyan
Screw Pokémon master, this is a Pokémon PHD thesis
honestly, professor oak should be big mad
Wait, aren't Masters Degrees higher than PHD's?
@@hyhyd6135 nope
@@hyhyd6135 They're not, but this comment would be pretty funny if they were. :)
@@Lockirby2 Ok
The random mix of whether you use the name of the first form or last form and the pictures not matching is driving me crazy! But I appreciate the effort you put into this
i love the jjk references😭
Someone call Paraspectre, his shroomie boi made it to the top 20!
Sandshrew did well too!
Love this. Im halfway through the vid but just got to your breakdown of how the ranking works. I would love to see a more deep dive into the code/simulation process itself someday!
Fascinating. One flaw with this metric that immediately leapt to my head is, it doesn't account for the practicality of being able to get each tested moveset for a Pokémon in the early game. Some would require an unrealistic amount of grinding, and some would be flat-out impossible to obtain because they require TMs not available until the late game. Still, this seems as good a method for evaluating Pokémon usefulness as I've ever seen someone come up with.
My friggin head tryna deal with the switching names/forms being called out @_@
Great content but that definitely did my head in
I absolutely love this kind of videos! As a huge statistics and simulation nerd I can’t wait to see you do more of this sorts of simulation videos
I get it is an insane amount of work but seeing this run using Yellow Legacy stats and move sets would be fascinating.
Wow!
Thank you to you and your team for taking the time to do all of these calculations. I love your methodology and how you broke it down for us. I'd love to see more videos about the statistics in Pokémon games.
I'm liking and subscribing, and I might join up later down the line!
I remember buying the Magikarp outside Mt Moon in my first playthrough. I was confused as Splash did nothing, but anyway tried to level it up just to see if it learned something useful. I was glad when it learned Tackle, only to be dissapointed again when it did almost no damage either. But I kept trying, and when evolved I was shocked by how much the same Tackle and now Bite was doing. Ahhh the good old times, of course it ended on my very first hall of fame 😁
Excelent analysis, I really loved this video. The charts and information are stunning, but for further, it could be interesting to make every Pokémon "your starter" in order to have a bigger picture, but we understand this is for a Vainilla run. Keep it going 🎉
You are teaching viewers data science. Love your vids. Great, reasonable metrics and data analysis. Amazing data vis.
YOU DID NOT HAVE TO HIT IS WITH THE ORTIZ FARM THEME, good vid pat, I recently found this channel I used to watch your older content and am glad to have found this! Love the Pokémon stuff
Thanks for all your hard work on Legacy!
To be fair, considering that we're talking about standard playthroughs, as in using multiple Pokémon, typically all 6 but anything over 2(1 HM carrier and 1 fighter) means that you don't actually need to be able to sweep every single Pokémon every major encounter has to offer. If you have a glaring weakness that can be covered by another Pokémon and tremendous strengths to pull a lot of weight than you're a perfectly fine incredibly powerful Pokémon. Basically there's a difference between the best Pokémon for soloing the game and for being a part of a cohesive team and these rankings don't showcase the latter at all.
You're right. One thing they could have done is weigh how many pokemon could actually beat such and such trainer's pokemon. Part of the reason Bulbasaur's being placed so low is because it's not taking into account just how limited your options for Brock and even Misty are. (I do agree that Bulbasaur falls off hard late game though. Been saying it for ages)
@@kyklous3657 It's not so much that Bulbasaur falls off hard late game but rather that Grass Pokemon as a whole, and Poison too for that matter, fall off hard due to a lack of good moves. Venusaur is still arguably the best Grass Pokemon in the game(Exeggutor is technically better due to Psychic typing but yeah, Psychic OP...) it's just that his best move is still Razor Leaf which has rather low power for a late game move, even if you consider it crits 100% of the time(cause gen 1 is gen 1). I mean just compare it to Slash that also crits 100% of the time(given the Pokemon in question is fast enough)... Also Grass coverage and weaknesses are far from ideal making it completely unnecessary when you make it to the elite 4.
That's honestly why some ppl love Charmander. Misty is the only real hurdle, while mons like Gyarados and Victreebel round out more powerful teams overall.
@@NeightrixPrime To be fair, most people love Charmander because Charizard is a fire breathing dragon, who isn't even a dragon type for some weird reason. Also Charmander is arguably the cutest of all 3 starters so there's also that. In any case, in gen 1 fire is really an unnecessary type as it has way too many weaknesses and doesn't actually cover too much. Remember that Steel was only introduced in gen 2 and is where Fire begun to shine thanks to being super effective against it. Charizard is still a great Pokemon mind you but not because of his typing but rather his stats.
His typing is actually terrible in gen 1 as he can't actually learn any flying moves, Drill Peck would be amazing on him, and as mentioned Fire moves are just unnecessary. In later generations Fire's super effectiveness against Ice, as well as the above mentioned Steel, make Fire a very good type. However in gen 1 the only Ice type that isn't also a water type is Jynx, who has very low defense and hp so it's really not that hard to take it out with any physical move, not even needing STAB. All the others are Water/Ice Pokemon so Fire's strength against Ice is negated by its weakness to Water.
Both lists say what I've believed for years: Squirtle is the objectively best starter.
Dude regardless if people agree with these lists or not I must applaud you for such a a deep in depth look into the Pokémon like the absolute hard work that has gone into this video is top tier btw first video of yours I have seen instant sub great work appreciate the video I enjoyed it all
I'm a little confused as to why some Pokemon are listed in their pre-evolved forms (Ponyta, Caterpie, Pikachu) while some are listed as evolved (Weezing, Muk, Beedrill)
its really interesting how many times i was just nodding along as these rankings made total sense with my experience doing solo runs.
Absolutely loved this video. Can't wait to see how things change in the next generations, maybe in separate videos to give newer pokemons a bit of room to breathe and appear (like, a video for gen 2 with all available pokemon in the game and another video with only gen 2 mons? idk, but could be nice!).
Also I'd love to see a scientific approach for an in-game team building stimulation. Hear me out: penalizing repeating types (for better type coverage), and rewarding not only early availability, but also "spread-out availability" (meaning if you catch all your Pokemons in one route, you'll have to level them all up instead of keep gradually catching new mons as your adventure continues, which I think everyone prefers). Or maybe a separate ranking rewarding teams that you can build as soon as possible. Edit: Also bonus points for taking into account HMs needed for your adventure!!!
Also if you guys need any help for the programming, I have a degree in physics, i started studying CS this year and I'd love to help with the programming! Keep up the good work 🤠👍🏽
I highly recommend you explain the entire criteria before the analysis starts rather then after
Now combine the charts and determine the best team of six
I hope you do every other generation aswell please, this was fun and interesting, amazing vid
So a few issues.
1: Your simulations just don't seem to be good? Just a cursory look at some of their performances shows matchups pokemon can absolutely win but are showcased as a loss. The simplest example is Charmander vs. Brock but Squirtle also easily beats Misty. You don't seem to go into much detail on how these battles are simulated but the AI for the player seems to be quite bad. Perhaps your simulation isn't taking into account statexp or badgeboost? Are you setting trainer DVs/Statexp correctly. It probably doesn't know how to adapt it's tactics to situations which is why it's always best to give these things a human touch for obvious blunders. Also your simulation numbers are just really really low.
2: I don't really understand what moves you seem to have given them. You state Hitmonlee is bad because it has to use fighting moves (Which are garbage don't get me wrong) but also Hitmonlee learns Bodyslam and like sweeps most trainers with that plus Meditate with the fighting moves reserved for rock/ice types. You later on say you gave pokemon all their different TM moves so I just find this confusing. Like in no world is Lee worse than Hitmonchan, that pokemon is an actual trashfire. A cursory look at Squirtle vs. Misty shows you are using Tackle in that fight instead of Mega Punch. So something tells me your pokemon are not very well optimized.
3: Gyarados being so high is a big sticking point for the real issue that you just don't seem to factor in EXP at all. Based on the footage you seem to level up pokemon to the level of the trainer but this is not how R/B operate at all. Early game pokemon tend to outlevel Surge/Erika and then underlevel the others and vastly underlevel against the elite 4. Gyarados for example doesn't actually evolve until you're ready to go to route 9 in a standard playthrough where you've been switch training him on every fight. He should not be getting a stellar rating for Misty since her EXP is also important to leveling it up. I suppose you could pop some candy early but that has other consequences. Trade pokemon are actually stellar because of this since it cuts their needed EXP for a level by 33% and is a big reason why people like Mime, Jynx, and Farfetch'd but it also helps others like Tangela and Dewgong shine. Trade pokemon actually tend to be on or outlevel the Elite 4 which is why they are so broken. Also how does this list account for overleveled mons like Dugtrio. Level 31 before S.S. Anne is hilariously strong. The logs indicate it is lower which is kinda unfair.
4: The methodology for score is flawed inherently. You seem to only weight the major battles but there are also several route trainers that can make a snag for some pokemon. Grass sucks in Gen 1 because you fight so many Grass/Poisons on routes that they have trouble with clean trainer sweeps and most grass types heavily rely on normal moves because of it; like seriously Route 9 with Bulbasaur sucks. A big flaw in some pokemon is the fact that you have to return to pokecenters multiple times on routes because they waste so much PP clearing them. Also not every major battle is that impressive. Surge for example is easily swept by about any pokemon you get before him outside of Gyarados and holding him in equal weighting to the Elite 4 is very silly. Post Koga pokemon on inherently flawed because there are just so few fights left for them. There are only 2-5 non major trainers the player has to fight after you beat Koga, 1 in Celdon Gym, 2 in Silph Co, and 2 in Viridian Gym. Also some pokemon are just better at fights than others. Squirtle deletes Brock while Bulbarsaur has to painstakingly weaken it down with leech seed giving them the same score is silly. Wrap is theoretically broken cause you can win any fight with it but god it's so slow, and this methodology just doesn't account for that nuance.
5: Availability is always problematic though we've tried your solution in the Fire Emblem community before and it lead to odd unintuitive results. The best way to handle availability we found is to use a log scale related to the number of available competitors. AKA squirtle sweeping brock is weighted lower than Sabrina because every pokemon outside of Moltres can fight Sabrina but only a handful can fight Brock. The idea is that it's more impressive to be good when more competition is available. You seem to inherently drop Rival 1 from the rating and it follows under this logic, only 3 pokemon can fight Rival 1 and they are all good at it so it's not a valuable fight.
As an aside you seem to dislike Sandslash which is weird cause like this is a 25+ year old game and Sandslash has been documented as a top performer for years lol, it always grinds my gears a bit when people do so little research for already documented data. Smogon has several ingame tier threads where almost every mon has been documented against all major fights.
fighting Erica as the 7th gymleader can easily happen by accident tbh. she is not required to advance until then and is tucked away on the bottom of the city behind a wall of trees. You have to use cut to get there.
I taught my Nidoking earthquake, brickbreak, sludgebomb & megahorn...now he is a beast
i remember having a level 50 gyrados against lt. surge because the diglett cave was such a great training spot and would just go back n forth while watching stuff.. good times.
Amazing job with this! Super in depth!
I wonder how the AI was choosing its moves. If the AI doesn't play like a real trainer, that would explain why the Bulbasaur line did ao much worse than it actually plays. The AI would be picking moves that wouldn't make sense like stun spore if the opponent already has a status condition.
Amazing video. Great work. Love that my charizard was so high on the final ranking.
17:00 had me on the edge of my seat.
“And wait for it! Are you ready?!” *mid-roll*
i love the random Jujutsu Kaisen quote when talking about Squirtle
I'd love to see this for other games!
So fun to watch this list!
My favorite team to run was always:
- Slowbro
- Snorlax
- Gengar
- Tauros
- Gyarados
- Raticate
Gyarados, Snorlax and Slowbro are certainly some of my favorites.
Gengar, Tauros and Raticate do have an excellent list of TM moves they can learn.
"Them are Rookie Numbers Man"
Dr. Strange: "Over 14 Million"
I'd love to see a follow-up where you see where they'd place if you could get them as a starter!
Yellow had different encounters and learnsets, will video about this game? Mankey will be higher probably (because you get him before Brock)
"Imagine your 8-year old self leaving on their first Pokemon journey." Actually, I was four. They let a four-year old go on a Pokemon journey unattended aside from my Chikorita.
I would have appreciated the explanation of how the bar graphs are calculated at the beginning of the video. I kept rewinding because I thought I missed something.
I give this ranking a 7.8/10 for having too many water pokemon
Water type is so broken. It's been broken since gen 1 and has never been nerfed. It's insane.
I've always thought it should be weak to Poison; "poisoning the well".
Gives Water another reasonable weakness and Poison a nice buff.
The Water type got a nerf when Gamefreak created the move Freeze-dry, an Ice type move at 70 base power that hit Water types super effectively.
@@homer23422000 But that's not really that big a nerf. Water is still superb, both offensively and, more importantly, defensively.
Just look at the Gen 9 meta. Defensive terra options are usually water or fairy.
@@grrmanOr Normal
@@MadMalMan Normal is more of a tech option from I understand, typically paired with extreme speed for that added oompf
Nice. I imagine it'd take a long time but I'd like to see this in Gen II as well (both the vanilla and your version)
Interesting analysis video! Thanks for uploading!
This new version of the Pokémon rap really slaps!
Someone beat me to it XD
I always told people to not Sleep on the Rattata and Geodude Lines in Gen1. Their ability to hit hard with stuff like Hyperfang or Earthquake is brutal. There is a reason why a single Rocket with a lvl14 Raticate has killed many Pokemon and Golem was able to stand decently (not good but decently) in Gen1s CP Wasteland
Anyone else have the "Gotta Catch Em All" song from the show pop into their heads as we were going through these lists? lol
Missing the other rival fights seems like a huge oversight
Also the bulbasaur slander 😢
I remember watching Gyrados using hyper beam in the anime as a kid, it’s been my favorite pokemon for the past 25 years, and I have used one on every team I’ve ever assembled
Great video! Did I miss any mention to Snorlax or wasn't there any?
scientifically proving fighting types in Gen 1 as doodoowater
Smith plays likes Pokémon?? Glad I found this channel
First like Bulbasaur
If only he knew at the time of writing this
First like Rhydon
I would honestly love to see the AI analyze all possible team combinations and propose the best possible team for each and every game (taking into account version exclusives, and for the games with DLC, HOME transfer exclusive Pokémon).
12:27 YES, FINALLY SOME VINDICATION
Edit: still really good ngl
Edit 2: THE LOBOTOMY IS INESCAPABLE HOLY CRAP
This has inspired me to try some new pokemon, can't wait for these for the other generations.
It feels like this skipped over 10-15. Which are the most surprising parts of the list to me.
Hang on. I'm calling shenanigans. Beedrill getting a 0 for Misty is not correct. We clearly see in your video, that it can 2 shot Starmie, I'm pretty sure it can OHKO if not also two shot Staryu. There's no way it got a zero for the Misty fight. Something is incorrect with the Data. In fact going over the data, unless I'm not understanding how to look at it, there was only one battle against Staryu, where they used a weaker move then what is available to Beedrill. Unless Beedrill got lucky, Furry attack has to hit atleast 4 times to deal more damage then Cut, which you could have at that time, or Twin Needle. Both Cut and Twin Needle essentially power 50 where as Fury Attack needs to hit 4 times in order to out damage either. Also Staryu was given a 3 level boost for some reason, being tested at lvl 21 instead of the lvl 18 displayed in the video. Unless you guys did something to change these Pokemon, Twin Needle will always beat Staryu by turn 3. Your data showing a turn counter going to 5 is the result of the Beedrill player massively misplaying. For that matter, Unless Beedrill has been nerfed in the version of the game you're playing, Starmie would usually not 2 hit Beedrill. In the Data, Starmie got a crit, but that's kinda unfair to hold against Beedrill when if Beedrill had crit, it would also two hit the Starmie with a crit.
Also I just opened up the Zubat fight, and my God, I know Zubat is bad and will always do bad in this, but it's incredibly disingenuous to count battles against it where it only used Confuse Ray. Like come on, I know Zubat is going to lose, but atleast give it a fair shot, and let it use Bite. Gyarados gets to spam Hydro Pump at level 21, but Zubat can't use Bite? Nidoking using Tackle over Horn Attack or double kick also doesn't make sense.
I'm gonna dig a bit more, but some of these battles are completely unfair, especially when compared to what is being given to other Pokemon.
Edit for after more digging on the Misty fights: How come Mega Punch isn't being used? Can't you grab that in Mt. Moon? Unless you get a miss, it's going to be a better attack for Nidoking, Wartortle, Wiggleytuff, or Clefable. For that matter, why are only Geodude and Charmeleon given access to Mega Punch? Also Nidoqueen is given access to the Body Slam, why aren't any of the other Pokemon given access to the Body Slam TM when it's better then most of their attacks? Ekans spams Poison Sting over Bite or once again Body Slam, which is a misplay. Though Ekans probably still loses. Alakzam spams Mega Punch over more Confusion, when Confusion should be doing more damage. Also why was Farfetch'd not tested for Misty? If we are giving Pokemon access to Body Slam, then Farfetch'd would also be available for this fight.
Overall the Data is incredibly Flawed on the Misty fight, with several of the Pokemon either not using their best move, not having access to it despite other Pokemon getting access to them, and takes no unlucky or lucky crits into account. Feeling frisky, so I'm probably going to look into more of the fights, but the testing method looks beyond inaccurate.
Brock fight: Charmander beat Onix and not Geodude, and never got a shot at the sweep, though it'd probably fail so it's not a big deal. Everything else looks fine there.
Surge fight: Golbat is probably going to lose, but why does it have Supersonic over Confuse Ray? Also it keeps using Supersonic for a majority of the fight, even when the opposing Pokemon is already confused. While it wouldn't change the results, Pidgeotto uses sand attack, when it would be better to just attack. Kadabra doesn't use Dig, and uses Flash or Confusion. That one might not matter, but I think it effects the sweep and Kadabra might be able to achieve the sweep if it was using Dig over Confusion. Pikachu uses Quick Attack over Mega Punch or Body Slam. I really wish you guys tried the sweep even if the Pokemon didn't solo the other Pokemon, because I suspect that if Farfetch'd got to set up Swords Dance against Voltorb, who can't stop it from setting up, that it could in fact sweep Surge. Beedrill still uses Fury Attack, despite it usually being worse then both Cut and Twin Needle. Butterfree spams Supersonic, even when opposing Pokemon is already confused. While it probably still loses, I have to wonder why it didn't try to use Sleep Powder on the Raichu, as that would atleast give it a slim chance at victory.
Erika fight: Persian uses Pay Day over Body Slam. Jolteon uses Tackle over Quick Attack or Body Slam. Golem, while it doesn't matter for the results, uses Submission over Body Slam. Marowak might have the luckiest fight here as it somehow get's three freezes. I'm wondering if it really would have beat Victreebel of Vileplume, if it didn't get those Freezes. Hitmochan uses Submission when its resisted, instead of Body Slam. Ivysaur uses Tackle over Body Slam. Mr. Mime uses Confusion over Psychic. Think I'm gonna end there, but there's some weirdness in your program that's making Pokemon under perform where they should, with what moves would be available to them in this test.
Edit2: I lied, one more. None of Agatha's pokemon are attacking Venomoth, so I question wether or not it's position, or Venonats is accurate. Didn't see Venonat's data, so I'm not sure if the same was happening to him.
Ditto = Power Bottom
No power. Just bottom.
Really cool and fun idea for a scientific ranking! I'm curious how this list would be different after the changes from Yellow Legacy!
I’m a bit confused, there’s 150 Pokemon but this list is only in the 80s. I assumed you only did fully evolved Pokemon but why are ponyta, krabby, pikachu, etc. here?
I think a video detailing the reasons for 150 pokes would’ve been a little too long. He probably focused on the obviously better ones, odd choices and fan favorites even if they didn’t get good results in the simulation. And of course, the ones that did the best.
I remember picking up my sister's old yellow version that she never finished. She was at the 8th gym with a fresh caught Articuno, a level 29 Haunter, and everything else in low 20's. I was so confused to how she got that far with such weak Pokemon, but still managed to beat Giovanni. Got Haunter to level 40, caught Zapdos and Moltres and beat the Elite Four with mostly just the Haunter and Articuno. They really are good in Gen One games.
Im curious how Pokemon like Rhydon, famous for being slow and having slow special stats, managed a full sweep against Sabrina or bring good results against Lorelei.
Gen 1 AI will Gen 1 AI. Rhydon is extremely powerful and gets extremely powerful moves, which a player can heavily exploit, meanwhile Sabrina and Lorelei only can deal damage with half their moves and might never actually attack with some of their Pokemon.
@@milsharkie That is true. I can already see Sabrina uselessly spamming recover or her miserable psywave attack while Rhydon oneshots her team. If all her Pokemon only knew Psychic or psybeam or her AI were not dogshit, she would be truly a menace.
@@Alector12 It's also worth mentioning that Rhydon, unlike Onix and the Geodude line, has a stellar HP stat of 105 to somewhat offset its low Special, meaning that none of Lorelei's Ice moves can guarantee an OHKO on Rhydon. And since Gen 1 AI doesn't take double weaknesses into consideration, any Pokemon that has both a water move and an ice move is tossing a coin as to whether they're going to give Rhydon a chance to crush them with Rock Slide.
I'm going to do the calcs and go through Lorelei's entire team; if the numbers seem weird, remember that AI trainers in Gen 1 do not have any DVs.
Dewgong only has Aurora Beam and, specific to Lorelei's Dewgong, is hard-coded to always use Rest on its second turn regardless of the state of the battle. Therefore, Dewgong loses if it doesn't crit with its first Aurora Beam AND get the high roll with it, to boot (71% chance to OHKO with a crit).
Cloyster only wins if he hits Clamp twice and doesn't randomly select Aurora Beam, and Rhydon's Rock Slide is so strong that it is almost guaranteed to 2HKO Cloyster. Note that this isn't "two turns of Clamp damage;" it's "Cloyster has to connect with Clamp twice or roll the 12.5% chance to get enough turns of damage on the first hit, which is less likely than hitting it twice."
Slowbro is the only member of Lorelei's team that isn't weak to Rock Slide, but unfortunately for him, he's only working with Water Gun. The double weakness means that Slowbro still barely ekes out a win, but it's really close.
Jynx is running Ice Punch, not Blizzard; since she gets cleanly OHKO'd by Rock Slide, that means she only wins against Rhydon if she gets a crit (which, considering her speed, isn't that big of an ask).
Lapras is possible for Rhydon to beat, but it's not very likely. Rhydon's Rock Slide will 2HKO Lorelei's Lapras, and Blizzard will have a measly 5% chance to OHKO Rhydon. Well, technically, it has around a 22% chance because of freeze and crits, but y'know. Lapras is faster than Rhydon, and Lorelei will have a 50/50 chance of going for Blizzard, which Rhydon has good odds to survive; assuming he doesn't get critted or frozen, if Lorelei's second move misses (10% chance for Blizzard and 15% chance for Hydro Pump) and neither of Rhydon's Rock Slides miss, then Rhydon wins. Low odds, but definitely not impossible.
Rhydon's got no chance to sweep Lorelei, but he has pretty good odds of winning against any single one of her Pokemon except Slowbro.
Great video !
I do wonder if you did something for most "valuable" sweeps. Like if for example 50 Pokémon can sweep Bruno, is that so much than valuable than sweeping Lorelei ? The rarest sweeps should grant more points in the system.
No matter the rankings I'll still trust Jolteon/Zapdos to have a spot in my team if only to sweep Lorelei.
Thanks for the amazing work !
I'm seeing some bias regarding the comments on Squirtle here... 🤣
Dude, seriously, fantastic job!!
Since you can catch Mew before Misty (that's when I used to catch it at least) it would've been cool to see you guys include it, but I get why you didn't. Great video though!
Oh this is gonna be sweet to see. Can't wait til we get later gens
How did the AI handle fighting Lorelei with pure Poison/pure Fighting mons with only Rage as an attacking move?
Magikarp proved the haters wrong with this one
I still remember my first use of a Master Ball...
I was a wee lad of about 9 or 10 and was lazily just walking through some mid game grass when I came across a new Pokemon I'd never seen before. Curious, I engaged in battle, found it so compelling and unique that I proceeded to whittle it down to low HP and sent every pokeball I had against it until just my Master Ball was left, pondered for a moment and just used it, thinking there'd be another and collected my prize.
That Ditto was mine.
I tought it was articuno that his description
I did the same thing. Ditto seemed so rare and mysterious back then.
4:50 as a kid I routinely fought Erika as the 7th gym leader because I choose squirtle as my starter and Blaine was far easier, and I would rush to Fuscha to get Surf asap.
Worth noting that Venasaur did so bad because the sim didnt know about the toxic/leech seed glitch
Jigglypuff's placement in the top 10 for the first list and in the top 30 for the third list shocked me. I never thought to use him on a team.
Idk how Bulbasaur got that low of a ranking.
I swepted the whole game with a solo Bulbasaur because of the move Growth, Leech Seed, and Mega Drain. I never even evolved the Bulbasaur.
oooh its family not individual pokemon. that makes sense on ditto when you said that my first thoughts were "what about metapod/ kakuna/ magikarp, but the fact its family makes sense.
For anyone who wants to compare, here is the Jrose11 tierlist that I found on reddit.
Mewtwo 50 3:03 1
Alakazam 59 3:06 2
Blastoise 63 3:12 3
Gengar 59 3:18 4
Nidoqueen 62 3:15 5
Nidoking 63 3:19 6
Poliwhirl 64 3:20 7
Hypno 63 3:22 8
Cloyster 56 3:31 9
Mew 54 3:27 10
Zapdos 55 3:26 11
Articuno 54 3:28 12
Starmie 59 3:25 13
Gyarados 55 3:33 14
Snorlax 59 3:39 15
Vaporeon 61 3:27 16
Persian 60 3:28 17
Machamp 63 3:22 18
Mr. Mime 62 3:28 19
Venusaur 64 3:33 20
Dragonite 57 3:51 21
Farfetch'd 63 4:01 22
Tauros 57 4:18 23
Lapras 57 3:59 24
Jolteon 60 4:03 25
Pinsir 59 4:21 26
Gastly 61 4:35 27
Porygon 62 4:22 28
Hitmonlee 62 4:06 29
Raticate 65 4:14 30
Butterfree 64 4:20 31
Electrode 63 4:12 32
Electabuzz 63 4:29 33
Flareon 66 4:08 34
Charizard 69 3:57 35
Poliwag 60 4:45 36
Magmar 72 4:19 37
Tentacool 63 5:03 38
Weezing 69 4:30 39
Arcanine 70 4:41 40
Jynx 65 5:07 41
Bellsprout 65 5:13 42
Staryu 62 5:48 43
Golbat 71 5:11 44
Dugtrio 74 5:01 45
Moltres 66 5:23 46
Aerodactyl 67 5:20 47
Beedrill 66 5:41 48
Hitmonchan 65 5:37 49
Omanyte 69 5:34 50
Chansey 66 5:57 51
Krabby 70 5:39 52
Mankey 67 5:56 53
Slowpoke 74 5:53 54
Exeggcute 66 8:26 55
Bulbasaur 69 6:32 56
Pikachu 69 6:57 57
Scyther 74 6:15 58
Magnemite 68 7:18 59
Dratini 72 7:15 60
Rattata 74 6:29 61
Ponyta 80 6:30 62
Rhyhorn 75 7:10 63
Charmander 81 7:12 64
Onix 76 7:24 65
Ekans 76 7:30 66
Oddish 72 7:14 67
Koffing 73 7:40 68
Geodude 79 7:49 69
Squirtle 80 7:27 70
Jigglypuff 79 7:33 71
Nidoran♂ 78 8:02 72
Diglett 78 8:28 73
Eevee 81 8:30 74
Nidoran♀ 80 8:37 75
Venonat 79 8:49 76
Paras 72 9:22 77
Zubat 82 9:17 78
Pidgey 90 10:48 79
Abra 63 17:18 80
Ditto 100 31:54 81
Weedle N/A DNF 82
Caterpie N/A DNF 83
Maybe for an availability analysis list, you should include all the trainers too. It would be interesting if there was a way to quantity the difficulty of raising a pokemon to an acceptable level as well.
Are we to assume when you say Voltorb, Magnemite, Omanyte, etc that it's that evo line?
It's kinda irony that a Mon that everybody avoid "I can kill you easily but I don't have time to deal with you" actually made it to the top 20 like Raticate lol
I shall not tolerate such electric type disrespect. ZAP GANG RISE UP!
Paras, with its fantastic sleep move, definitely was decent in the original game, where waking up was a turn wasted. It's when that ailment was nerfed is when Paras just completely died.
Glitchless Speed Runners already did all this hard work for you through trial and error.
Fascinating.....now do it for every game and "third" version, too >:D
tbh i would have loved it if you used every pokemon for a whole playtrough, like use mewtwo from the start and see how he goes against all his enemies of the same level. do this with every pokemon and you will get a list which is way more accurate imo
“As for the rest of early Pokémon they couldn’t defeat Brock’s Onyx or Geodude.”
Nidoran: “Am I a joke to you?”
Only in Pokémon Yellow. In Pokémon Red and Blue, they don't learn Double Kick until absurdly high levels.
@@amiciyopre well, at least we still got Mankey in Red.
@@sergioalejandrobricenostap9016 Yes, but you can't catch one before Brock. That's also a Pokémon Yellow thing.
I think gengar at 6 is a bit low, maybe give it amnesia to make it viable
Careful, you'll end up haunting Smith in his dreams with that one.