Red hack: Shin Pokemon Red/Blue/Green/JP builds (Bugfix, AI, and QoL patch)

[SPazzzi95]

Thanks for the consistancy fixes as always.👍

 

[LinkNinjaMaster]

Never hurts to go back through the evos_moves file and check for things I missed. I general, I'll set moves to be learned earlier if there is a difference between Red and Yellow.

Ah, I see. Well, I hope Butterfree and maybe others get small favorable tweaks to their movesets in a future update.

It's less involved to press B when asked to switch your pokemon. I'm running out of menu option bits.

Oh ok I understand, yeah I guess it's no big deal to just spam B when defeating an opposing Pokémon.

You are overestimating how much StatEXP affects early-game trainers. It actually does very little to affect enemy stats in the early game because the bonus points it grants is a percentage based on your level. (...)

Consider a level 9 trainer pokemon where I've set the StatEXP to 640 per level.
(...)
Because of rounding, that's a mere +1 it gets to each of its stats.

Ok I did not know how much Stat Exp they were getting, but the early-game trainers and first rival battles felt considerably more difficult with Set mode on. Do trainers in Switch mode retain their average 8 DVs across the board like Bulbapedia says is the case for vanilla Gen 1, or do they get their DVs randomized as in Set mode? because if they stay the same as in vanilla then maybe that's the reason why they felt a bit harder than they should have when in Set mode, since even the Bug Catchers in Viridian Forest can have up to three extra points in their stats (compared to unmodded Gen 1), which at low levels do make a noticeable difference.

For instance, I don't know if this example was confirmation bias, but I was consistently losing the first rival battle with my Bulbasaur in three different attempts (even with the potion in two of them), but when I switched to, well, Switch mode, I beat him without even needing the potion. So I'm guessing the rival got lucky with his attack and defense DVs those few first attempts, even if that gave him just one point in any of those important stats.

 

[jojobear13]

Ah, I see. Well, I hope Butterfree and maybe others get small favorable tweaks to their movesets in a future update.

The next update has Butterfree learning moves earlier per it's Yellow movelist.

Ok I did not know how much Stat Exp they were getting, but the early-game trainers and first rival battles felt considerably more difficult with Set mode on. Do trainers in Switch mode retain their average 8 DVs across the board like Bulbapedia says is the case for vanilla Gen 1, or do they get their DVs randomized as in Set mode? because if they stay the same as in vanilla then maybe that's the reason why they felt a bit harder than they should have when in Set mode, since even the Bug Catchers in Viridian Forest can have up to three extra points in their stats (compared to unmodded Gen 1), which at low levels do make a noticeable difference.

Every trainer normally has DVs of exactly 9,8,8,8 (and a derived 8 for HP). SET mode randomizes the trainer DVs to be at least 9,8,8,8 (letting HP be any number between 0 and 15). Even still, DVs scale as a percentage of level such that the difference between a 0-DV and 15-DV stat is 30 points at level 100. This means that the difference between an 8 and a 15 in a stat is a mere 1 point up until level 10, and the difference is 2 points up until level 20.

It could be that what you are experiencing is the full combination of several things:

• A couple more stat points from DVs
• A couple more stat points from stat exp
• More HP derived from stat exp and DVs
• Because the Speed stat only matters if it is higher than the opponent, many trainer 'mons will move first just by having even a single point of Speed more than you.
• Badge boosts are deactivated in SET trainer battles, so you miss out on a 12.5% DEF boost after beating Brock at a point when everything is still hitting you with physical attacks.

All this syngergizes together with an AI that is not shy about using damaging moves (save for youngsters and cueballs). The 1-in-256 miss bug is fixed. And the 25% failure rate for enemy status moves is also gone. You take slightly more punishment, over slightly more rounds, in a more consistent way. It all starts adding up.

For instance, I don't know if this example was confirmation bias, but I was consistently losing the first rival battle with my Bulbasaur in three different attempts (even with the potion in two of them), but when I switched to, well, Switch mode, I beat him without even needing the potion. So I'm guessing the rival got lucky with his attack and defense DVs those few first attempts, even if that gave him just one point in any of those important stats.

The math supports what you're feeling. Let's look at the Attack, Defense, and Speed comparisons of the three match-up possibilities for the first rival battle:

Player Squirtle vs. Rival Bulbasaur
9-11 ATK vs. 10-11 DEF
11-13 DEF vs. 10-11 ATK
9-10 SPD vs. 10-11 SPD
A fairly even match. Bulbasaur has a higher SPD and its DEF is on par with your ATK. But your DEF will likely outclass its ATK. You may be able to tank and outlast the opponent.

Player Bulbasaur vs. Rival Charmander
9-11 ATK vs. 10 DEF
9-11 DEF vs. 11 ATK
9-11 SPD vs. 12-13 SPD
Charmander has the clear advantage with higher SPD, ATK that likely outmatches your DEF, and DEF on-par with your ATK. It is likely to out-damage you and hit you with a crit.

Player Charmander vs. Rival Squirtle
10-11 ATK vs. 12-13 DEF
9-10 DEF vs. 10-11 ATK
11-13 SPD vs. 10 SPD
Squirtle has a slight advantage with ATK slightly above your DEF as well as a DEF that outmatches your ATK. But you will be going first each round and have a slightly better chance of a critical hit. Squirtle will tank and probably outlast you unless you score a lucky crit.

The differences are not huge, but the odds are stacked against you the longer the battle plays out. And now the rival is going to be picking his damaging move much more often as opposed to the vanilla games. Picking bulbasaur in particular is going to be an uphill battle.

 

[jojobear13]

Thanks for the consistency fixes as always.👍

Thank you for the keen eye.

 

[keojk123]

Could somebody be so kind to post a working patch with the space world back sprites and front sprites. I've literally tried everything in the instructions multiple times and can't for the life of me seem to get it working correctly.

 

[jojobear13]

Could somebody be so kind to post a working patch with the space world back sprites and front sprites. I've literally tried everything in the instructions multiple times and can't for the life of me seem to get it working correctly.

The next version will streamline this process using makefile tags. For now, the master_dev branch already has this streamlining. You can try the newer instructions there for battle sprites if you wish, so long as you don't mind playing a beta.

 

[MStern]

The daycare man will not accept any pokemon with HM moves. I assume this was originally intended to avoid daycare-related softlocks, but I'm pretty sure that's impossible to achieve in RBY. That restriction should probably be removed for quality of life sakes'.

 

[jojobear13]

The daycare man will not accept any pokemon with HM moves. I assume this was originally intended to avoid daycare-related softlocks, but I'm pretty sure that's impossible to achieve in RBY. That restriction should probably be removed for quality of life sakes'.

I'll look into it. The biggest thing going for it is the fact that you can travel between the daycare and the Cerulean pokemon center without any HM moves.

 

[LinkNinjaMaster]

[...] It could be that what you are experiencing is the full combination of several things:

• A couple more stat points from DVs
• A couple more stat points from stat exp
• More HP derived from stat exp and DVs
• Because the Speed stat only matters if it is higher than the opponent, many trainer 'mons will move first just by having even a single point of Speed more than you.
• Badge boosts are deactivated in SET trainer battles, so you miss out on a 12.5% DEF boost after beating Brock at a point when everything is still hitting you with physical attacks.

All this syngergizes together with an AI that is not shy about using damaging moves (save for youngsters and cueballs). The 1-in-256 miss bug is fixed. And the 25% failure rate for enemy status moves is also gone. You take slightly more punishment, over slightly more rounds, in a more consistent way. It all starts adding up. [...]

That makes sense and I had completely forgotten about the badge boost not being there. I guess that is outright "hard mode" now lol. I'll just need to play differently if I decide to give Set mode another go later. (But if after a few updates you still have a menu bit free it would be pretty cool if you could have two Set modes, one for the vanilla mode and one for the harder version :P.)

Anyways thanks for your super-detailed answers, I really appreciate it.

 

[jojobear13]

So for some reason I got on a youtube binge about genetics. This resulted in a shower-thought of "There are 4 DV values used across 2 bytes. What if I were to treat them like genes? They basically already are. What if they could be crossed using a Punnett Square similar to how things actually work in reality?"

This piqued my interest and I ended up designing a new feature for Shin Pokemon. Time to brush up on your high school biology!

**Cloning Genetically Modified Pokemon**
This process clones your 1st spot pokemon and modifies it with DNA from your 2nd spot pokemon.
It does this by treating a pokemon's two bytes of DV values as if they were two genes with two alleles each.
Gene-A has the Attack DV as its hi-side allele and the Defense DV as its lo-side allele.
Gene-B has the Speed DV as its hi-side allele and the Special DV as its lo-side allele.
The A-genes from the two donor pokemon will be mixed to make the clone's A-gene.
The B-genes from the two donor pokemon will be mixed to make the clone's B-gene.
Mixing two genes is done via Punnett Squares, and a random result is chosen based on its ratios.
Within a Punnett quadrant, a hi allele makes the upper digits and a lo allele makes the lower digits.
If two hi allels or two lo alleles fall within a Punnet quadrant, their order is randomly selected.

Example: Mixing gene hexadecimal values of $AB from one pokemon with $CD from another pokemon.
One of the four quadrants will be randomly chosen as a new value containing two DVs.
__________C__________D
_____xxxxxxxxxxxxxxxxxxxxx
A___| AC or CA |____AD____|
_____xxxxxxxxxxxxxxxxxxxxx
B___|____CB____| BD or DB |
_____xxxxxxxxxxxxxxxxxxxxx

Talk to the new scientist in the fossil room of the Cinnabar Lab to get started.
For a small fee, a gene sequence listing the clone's new DV alleles in order is randomly determined.
If the clone's genes are acceptable, you may pay a larger fee to gestate the clone.
Pro-Tip: This makes it possible to selectively breed for shiny pokemon in a more realistic way.

Example of selecting for a shiny pokemon:
- A shiny pokemon in gen 2 has the 2-byte hex value of $XAAA for its DVs.
- Wherein 'X' is the Attack DV composed of the 4 bits yy1y.
- And 'y' bits are "don't care" values that can be either 0 or 1.

- The first goal is to get $AA as the value for the B-gene.
- The minium requirement is to cross two pokemon that both have at least one $A in their B-genes.
- This minimum requirement results in a 25% chance of getting a value of $AA.

- The second goal is to get $A as the lo-side allele value for the A-gene.
- The minium requirement is that one of the donor pokemon have at least one $A in its A-gene.
- This minimum requirement has a 12.5% chance of fulfilling the goal if the donor $A is a hi-side allele.
- Note: Said chance increases to 37.5% if the donor $A is a lo-side allele.

- The third goal is to get a 'X' value (previously described) as the hi-side allele value for the A-gene.
- The minium requirement is that one of the donor pokemon have at least one 'X' in its A-gene.
- This minimum requirement has a 12.5% chance of fulfilling the goal if the donor 'X' is a lo-side allele.
- Note: Said chance increases to 37.5% if the donor 'X' is a hi-side allele.

- Assuming only the most minimal requirements are met, the chance of a shiny is [12.5% * 12.5% * 25%] = 1/256.
- Keep in mind that the default chance of a shiny pokemon is 1/8192.
- You can boost the chances higher by increasing the number of donor $A alleles and optimizing allele placement.
- The chances of getting a shiny by crossing two shiny donors is 50%.

 

[306090]

Hello I'm playing your rom on blue pokemon, I love it.
I just couldn't find the pokedex diploma. Can you tell me how to get it? I got 150/150 pokedex I tought that when I complete it I could have the diploma but idk what to do anymore :(
The "old man in viridian" I can't find too! :c

 

[jojobear13]

Hello I'm playing your rom on blue pokemon, I love it.
I just couldn't find the pokedex diploma. Can you tell me how to get it? I got 150/150 pokedex I tought that when I complete it I could have the diploma but idk what to do anymore :(
The "old man in viridian" I can't find too! :c

Same way as in the original games. The diploma is not an item. It is a screen you have to view. I made it so viewing the diploma sets an event flag. You only ever have to view it once.
Here is how to view the diploma: https://youtu.be/4-hMBNoIhrQ?t=152

The "old man in Viridian" refers to the traditional method of activating the missingno glitch.
Here's how it was done back when I was a kid: https://www.youtube.com/watch?v=Nh3X8zSmTwY

 

[306090]

Same way as in the original games. The diploma is not an item. It is a screen you have to view. I made it so viewing the diploma sets an event flag. You only ever have to view it once.
Here is how to view the diploma: https://youtu.be/4-hMBNoIhrQ?t=152

The "old man in Viridian" refers to the traditional method of activating the missingno glitch.
Here's how it was done back when I was a kid: https://www.youtube.com/watch?v=Nh3X8zSmTwY

Ahhhhhhhhhhhhh OH my god hahaaha
Thanks you so much, love you xD

And again your game is wonderful.

edit: Old man glitch doesn't work to me but it's ok.

 

[jojobear13]

Ahhhhhhhhhhhhh OH my god hahaaha
edit: Old man glitch doesn't work to me but it's ok.

Keep surfing. I made it a rare encounter due to it being a superboss. It can take a while.

 

[306090]

Keep surfing. I made it a rare encounter due to it being a superboss. It can take a while.

Ohh ok hehe

 

[jojobear13]

I have pokemon cloning and gene selection implemented and mostly working. Here's a video showcasing the feature. https://youtu.be/WnVbVSLUEzY

 

[jojobear13]

Here we have an example of how the new gene cloning feature can be used to select for a shiny pokemon.
https://youtu.be/-U12sgSX64o

To refresh, a shiny pokemon in gen 1 & 2 is determined based on four DVs. The attack DV must have the 2nd-least significant bit be set (i.e. values of 2, 3, 6, 7, 10, 11, 14 or 15), and the other three DVs must all be 10.

This means the key "allele" for shininess is DVs of 10. Donor-1 needs to have a 10 for either the speed or special DV. Donor-2 needs to have the same. One of the donors should ideally (but not necessarily) have a shiny-valid value for its attack DV with the other donor having a 10 for either its attack or defense DV.

As can be seen in the video, using the right donor pokemon with the right DVs can produce a shiny pokemon with a high degree of accuracy compared to the original 1-in-8192 odds.

A real-world example of this can be seen with dog breeding. Dogs can carry the allele for the "merle" coat coloring; this allele dilutes melanin in order to express bluish-gray fur in place of brown or black fur. Breeding two merle-carriers together, even if the coloring is not expressed, has a chance of producing a puppy with two merle alleles. These dogs, appropriately called "double merles", have striking white coats caused by an inability to produce melanin. Phenomena like this served as inspiration for gen-2's shiny feature. Though I must add that breeding for double merles is frowned upon due to their high chance of being born blind and/or deaf.

 

[jojobear13]

Announcement: Moving forward, Shin Pokemon will be upgrading its patch format from IPS to the more modern BPS-delta format. There are several reasons for this change.

1. BPS supports checksums, so patchers will throw an error if applying to an incorrect base rom.
2. Rather than encode a list of changes (as with IPS), a BPS patch encodes the steps needed to create a brand new rom file from an old rom file. This results in a 101 KB .bps patch as opposed to a 447 KB .ips patch.
3. The BPS format accounts for shuffling data around memory banks such that .bps files do not run the risk of containing non-novel object code.

BPS patches can be applied with FLIPS (https://www.romhacking.net/utilities/1040/).

 

[MStern]

- The current Lite ips patch release is the same as the full version.
- It also has TMs never get consumed. That part might be an improvement, since shin's placement of purchasable TM's is quite haphazard and hard to remember.
- When adding the PC to the daycare man, the sprite was also added to the fishing guru's house in Vermillion.
- I hate the current implementation of the Old Rod. If you want to catch a specific pokemon like Poliwag when it would be most useful, it takes an unreasonable amount of time and annoyance to do.

 

[jojobear13]

- The current Lite ips patch release is the same as the full version.

Noted. I don't update the dev patches often. The dev branches are for work-in-progress, and sometimes they can get a bit wonky.

- It also has TMs never get consumed.

I found that in one of my playtest streams. Fixed it last week.

- When adding the PC to the daycare man, the sprite was also added to the fishing guru's house in Vermillion

So it looks like Gamefreak set the daycare map file to also be used for the route 12 and vermilion fishing guru houses (to save space). Easy fix.

- I hate the current implementation of the Old Rod. If you want to catch a specific pokemon like Poliwag when it would be most useful, it takes an unreasonable amount of time and annoyance to do.

I've been working on re-doing this so that the good and old rods use the wCurMap value to determine what gets hooked. Old rod will hook magikarp/poliwag or magikarp/goldeen depending on the map constant, and it always generates an encounter. The good rod can hook 3 or 4 pokemon from the GoodRodMons list depending on the map constant.