I really enjoy Expert Mode, and just wanted to share my experience with it.

In the subreddit, we often get questions of "I just unlocked Cyan Beach Expert. Should I go there?" so I thought it might be helpful to see what it might look like.

Despite how long I've played, I'm not exactly the strongest player since I have a slower playstyle. I consider myself a mid-level player since I have a lot of Pokémon above 30, but none are at 60 yet. I thought it might be interesting for other players to see how a "weaker" player might enjoy Cyan Beach Expert Mode.

Based on how much Strength I typically get in a week, I expected to hit Ultra 1, so I wanted to go when there wasn't an event. I made it to Great 4 and then went on vacation, and it turned out I didn't have cell reception the last three days....

So this is actually a story of my 2nd time at Cyan Beach Expert haha.

I got Fairy, Fire, and Flying; Desserts/Drinks; and the 1.25x bonus to Skill triggers.

As far as Skill specialists go, that would really boost my Sylveon for a healer, and Flareon for pot boosting. It's skill level is at 3 (+12), so my pot needs 1 trigger for Zing Zap, 2 for Éclairs or Pancakes, 3 for Waffles. With no triggers, I can make Macarons.

Besides Flareon, fire types don't really help me for Desserts/Drinks, so I rule Entei and Arcanine out. I would love Togekiss, but mine is from the early game and doesn't have good nature/subskills, and have been looking to replace it for a long time. Braviary could be good, but mine is still a Rufflet.

So after looking at skills and dishes, I can tell that these are the Pokémon I'd likely be relying on:

• Sylveon - Healer
• Flareon - Pot Boost (and some Cacao)
• Delibird - Eggs
• Clefable - Berries (and some Honey)
• Comfey - Corn

In order to make strong recipes, I have a bunch of Ingredients left over from the event the previous week, but it's not going to last. I'll need to use some weaker (below level 30) Pokémon and some off-Berry Pokémon:

• Cramorant - Oil
• Cutiefly/Venusaur - Honey
• Luxray/Mr. Mime - Tomatoes
• Absol/Clodsire - Cacao

So how did my week go?

https://preview.redd.it/qad8t4ajdoth1.png?width=1080&format=png&auto=webp&s=8d33b4657feb00e516ea5480c06b2fb07b6ec441

Day	Pokémon	.	.	.	.	Team RP
Monday	Sylveon-51	Clefable-38	Dodrio-23	Flareon-36	Comfey-37	11,358
Tuesday	Sylveon-51	Clefable-39	Luxray-33	Cutiefly-17	Venusaur-50	11,656
Wednesday	Flareon-38	Clefable-40	Delibird-52	Cutiefly-18	Comfey-38	8,587
Thursday	Sylveon-51	Clefable-40	Absol-51	Ribombee-19	Clodsire-45	13,248
Friday	Sylveon-51	Comfey-38	Cramorant-29	Delibird-52	Clodsire-45	11,925
Saturday	Sylveon-51	Delibird-52	Absol-51	Ribombee-20	Clodsire-46	13,067
Sunday	Sylveon-51	Clefable-40	Absol-51	Ribombee-21	Clodsire-46	13,317

This is just a representative example, I would switch my Pokémon as needed to get the Ingredients I needed. I also used Dedenne on Monday, but other days I was too busy trying to get Ingredients and Pot boosting.

Sylveon is currently my best healer, but I was drowning in Milk, and really wishing I had a good Wigglytuff, since I needed Honey a lot more. But I quickly evaluated all my Cutiefly, and chose one to start raising, and it did remarkably well. I also was short on Cacao, so considered using Pawmot, but I only recently got this one and haven't invested Main Skill Seeds in it yet.

Flareon typically gave me enough triggers. The bigger problem was getting enough Ingredients, so a couple days I had to make smaller dishes than I wanted. Dishes I made:

• "Flower Gift" Macarons: 10x
• "Zing Zap" Spiced Cola: 2x
• "Scary Face" Pancakes: 2x
• "Honey Gather" Chocolate Waffles: 4x
• "Teatime" Corn Scones: 2x
• "Leaf Tornado" Smoothie: 1x

I feel like I did fairly well until I think Thursday or Friday when my Sylveon had an off-day and I wasn't able to maintain a stockpile of Ingredients, and I felt like I was playing catch-up.

Like I said, I had expected to make it to Ultra 1--But I actually made it to Ultra 4, far better than I was expecting based on how I normally do. For this chart, I'll divide out my 5% Expert Bonus so it is a fair comparison for a "first time" post. Compared to my first time at Greengrass Expert, I've come a long ways.

Day	Total Strength	Rank	Strength Gain
Monday	231,687	Basic 5	231,687
Tuesday	556,665	Great 3	324,978
Wednesday	793,436	Great 4	236,771
Thursday	1,022,671	Great 5	229,235
Friday	1,234,913	Ultra 1	212,242
Saturday	1,425,913	Ultra 2	190,954
Sunday	1,796,150	Ultra 4	370,283

My thoughts: The 35% penalty is no joke. It was a struggle getting enough Cacao. My actual first time was pretty similar, Flying, Electric, Fire, same bonus, but with Salads. That week, at times Entei's Helper Boost helped me out a lot with those off-Berry ingredients. But this time, I often had to have multiple Ingredient helpers in order to try and get enough.

In hindsight, I suppose I could have went for a weaker dish that would be easier to get with off-Berry Ingredient Specialists, but when I did have the Ingredients, Flareon really pulled through for me, so I'm not sure I could have done better. Plus, I just enjoy taking advantage of the bonuses if I can. And I felt like they were actually pretty huge.

Anyway, I'll look forward to next time. Thanks for a great week of snoozy studies!

Bonus: my spawns, for good measure:

Day	Drowsy Power	Pokémon	.	.	.	.	.	.	.	Shards	Research EXP
Mon	36,490,650	Dragonair-1	Toxicroak-1	Psyduck-2	Grubbin-1	Psyduck-1	Croagunk-1			8,340	4,586
Tue	58,449,800	Psyduck-3	Croagunk-2	Psyduck-2	Psyduck-2	Grubbin-1	Treecko-1	Croagunk-1		11,410	6,427
Wed	69,147,964	Cutiefly-2	Tinkatink-2	Happiny-1	Squirtle-1	Squirtle-1	Totodile-1	Totodile-1		10,589	6,314
Thur	106,306,695	Dratini-3	Igglybuff-2	Tinkatink-1	Eevee-1	Eevee-1	Cleffa-1	Psyduck-1	Croagunk-1	13,996	7,958
Fri	129,665,900	Tinkaton-1	Ditto-1	Wigglytuff-1	Clefairy-2	Jigglypuff-2	Jigglypuff-2	Eevee-1	Jigglypuff-1	16,458	12,530
Sat	149,716,100	Treecko-4	Dragonair-1	Dragonair-1	Grovyle-1	Toxicroak-1	Croagunk-2	Treecko-1	Dratini-1	24,267	12,326
Sun	188,595,800	Toxicroak-3	Heracross-1	Dragonair-1	Dragonair-1	Toxicroak-1	Grubbin-1	Psyduck-1	Psyduck-1	19,440	10,591

This is an article in a series. For background, see the Introduction. The previous article in the series was on Strength skills.

Charge Energy S

For those that don't know, Energy has five tiers, and a Pokémon with more Energy will provide helps more frequently, so having Energy in the higher tiers is better. The Frequency on the Stats page of a helper Pokémon shows its speed assuming 0% Energy, which follows this formula:

Frequency = Floor((Base Frequency for the species)*(501-level)/500) * Nature * (1 - sum of Sub Skills))

• Nature: 1.075 for speed of help down, 0.9 for speed of help up natures (lower frequency = less time between helps = faster)
• The sum of Sub Skills doesn't include Helping Bonus on the Pokémon's stat page, but is included on the Team's stat page.
• The sum of Sub Skills is limited to a maximum of 35%

This table shows the multipliers to frequency. And again, the multiplier being smaller than 1 makes the frequency smaller, which makes the time between helps shorter, meaning it is faster.

Of course, we normally think in terms of multiplication to actual production. So the third column is the reciprocal of the frequency multiplier in order to show that

Energy	Multiplier	1/Multiplier
80+	0.45	2.222
60-79	0.52	1.923
40-59	0.58	1.724
1-39	0.66	1.515
0	1	1

I should add that these values used to be different, but was buffed in 1.9.0, with no changes to RP skill values.

Energy decreases by 1 every 10 minutes. So without healing, a Pokémon beginning at 100% Energy will naturally lose Energy and hit 0% after 16 hours and 40 minutes.

So how does the RP Formula deal with Energy?

It doesn't.

Well, RP acts as representing 5 hours at 0% Energy. But that is the equivalent of 2 hours 15 minutes of 80%+ Energy. What I mean is that there is no Energy component to the RP formula that it could use as a multiplier to speed.

Which is fair--Energy is normally the same for everyone, and Strength resulting from changes to Energy should be attributed to the thing that changed it. Healing should be in the Skills portion of the RP formula, and having it there makes it easier to compare healing skills to other skills, just so long as the estimated value is accurate.

So what is the effect of gaining one point of Energy?

It depends on when you receive it. Let's break it down into something simpler to think about. There are 144 10-minute intervals in a 24-hour period. A normal Pokémon beginning at 100% and without healing will spend 21 of those 10-minute intervals in the top tier, 20 each in the next two tiers, 39 in the tier above 0% and so the remaining 44 in the 0% tier.

When you gain 1 point of energy, it shifts all remaining intervals in the day to be one interval better. So effectively, it converts one 10-minute interval in the 0% tier to one in the current tier (well, the current tier or the one above, if it gets pushed into the one above).

https://preview.redd.it/5c77jf4yxzlh1.png?width=1226&format=png&auto=webp&s=6f228e595f058e0efcc681470636bb1bcf9c50c3

Keep in mind that an overnight trigger will get banked until the morning, giving a lot more chances for ideal healing. Meanwhile, getting a trigger is based on each help, and the blocks with more Energy have more helps, so there's actually a decent chance of hitting those most effective tiers.

So even though the average Energy block would be about 1.56x production, the average Energy block would be healed to 1.95x production, not counting the limit on triggers that can be banked, the realities of upgrading more than one block at a time, or the effect of multiple triggers.

And when you think about it for multiple triggers, the first trigger converts 0% energy to the helper's current tier, but once all 44 of the 0% blocks are converted, then it converts the next lowest tier to the current tier, which is a smaller boost. So healing has a law of diminishing returns.

When you reach a point where you will remain above 80% Energy the rest of the day, healing provides no benefit. (That value would be 130% Energy before going to sleep for 8.5 hours--or 150% Energy at 3 hours 20 minutes before bed.) And actually, since there is a 150% cap on Energy, you may run into that problem earlier if healing would try to put you above that limit.

From / To	2.222	1.923	1.724	1.515	1
2.222	1
1.923	1.156	1
1.724	1.289	1.115	1
1.515	1.467	1.269	1.38	1
1	2.222	1.923	1.724	1.515	N/A

Obviously, healing 0% -> 80+% is the most ideal. But healing 0% to even the 1-39% tier gives a better effect than than a Pokémon that has already received at least 44 healing. And again, as long as we aren't hitting the cap, healing will always upgrade the lowest Energy blocks first, then the next lowest, etc. The tier it upgrades to depends on where it is currently at.

Besides the timing, the strength of healing scales with the strength of the Pokémon itself. Keeping a strong Pokémon above 80% is better than keeping a weak Pokémon above 80%. As one example, if a fast Pokémon gets a full extra help out of healing, a slower one wouldn't. Not to mention how much strength the extra Berries, Ingredients, and Skills that come from those extra helps.

Since the main portion of the RP formula is the strength from Berries, Ingredients, and Skills for 5 hours at 0% Energy, we can use that to our advantage to calculate the amount of Strength +1 Energy gives.

Strength from healing = RP Strength/30 * (1/New Tier - 1/Old Tier) * amount healed

Where RP Strength is the (Berries + Ingredients + Skills) portion of the RP Formula, and the tiers are the multipliers from the first table.

Now, I can use that to see at what level that fraction equals the RP value. I'll just assume neutral nature/subskill, AAA Ingredients, and let's do the ideal 0% -> 80+% Energy conversion. Charge Energy S is currently the most common Main Skill so I won't list them all, but here are a few examples.

Pokémon	1	2	3	4	5	6
Rattata	49	50	51	54	56	56
Dratini	33	35	37	41	44	46
Bellsprout	44	45	46	50	51	50
Torchic	37	37	37	39	40	38
Dragonair	29	29	29	30	30	30
Weepinbell	30	30	30	30	30	30
Raticate	29	29	29	30	30	29
Combusken	28	28	27	27	25	18
Victreebel	26	26	26	26	25	19
Dragonite	11	12	13	16	17	15
Blaziken	14	13	11	10	7	1

Below the level, the RP Formula overvalues the Skill, while over the level, it is undervalued.

This shows the principle that I said: the strength of healing scales with the strength of the Pokémon. So you can see that with weaker Pokémon, the RP Formula overvalues the skill until higher levels, while stronger Pokémon, it begins undervaluing the skill at lower levels.

You can also see that many values seem to cluster around 30--the reason for that is because there is a big jump in Strength when the second Ingredient is unlocked, and the skill lands somewhere in that big gap.

These numbers actually look pretty reasonable, suggesting Charge Energy S is fairly reasonable in the RP formula. It also appears that it typically stays fairly stable with Skill Levels, with the exception that weaker Pokémon can't keep up with the Strength assigned to higher Skill Levels and stronger Pokémon surpass it more easily. Stable Skill Levels suggests that in general, increasing Skill Levels aren't overvalued.

To be fair, I did use the ideal scenario, converting blocks from the lowest tier to the highest tier, meaning in reality the break-even point would be even higher. But I also used average Pokémon, while we are probably using Pokémon with better stats, so the break-even point should be lower.

Maybe here is where you can point out that I was using the RP formula to do this analysis, and we've already seen that Berries ignore favorite Berries, and Ingredients are very undervalued if you are making strong dishes.

That is true, and if I evaluated Charge Energy S on the "real" strength of Berries and Ingredients, it would show it to be undervalued. But I wanted to look at the skill itself in isolation--that is, if a skill were undervalued, I want to know if it is undervalued only as a side-effect of Berries/Ingredients (and Skills for teammate healing) or if it is more inherently undervalued (or overvalued).

Back to the table, I said they were fairly reasonable values, but looking at it, it seemed that Blaziken is a major exception and the break-even level really takes a dive at high Main Skill Levels.

Looking into it, Blaziken (and others I didn't list that also take a dive) have higher skill trigger rates. And that makes sense--with such a high skill value, a significant portion of the Strength is going back into the Skill itself (which is fixed at all Pokémon levels) rather than into Berries and Ingredients. It is true that healing will give more helps, which in turn leads to more skill triggers. However, the reality is that a single trigger at Level 6 already converts all 0% tier blocks, so any additional triggers will always be weaker than the first--and potentially a ton weaker if we are hitting the 150% cap.

I could make a second chart using a different assumption, like a 1.5x boost to production instead of 2.2x, which would be around 0% -> 1-39% tier or 1-39%-> 80+%. However, when I did that everything just clustered around 60 instead, and made for a kind of unsatisfying chart.

I could try assuming all sorts of average tiers, but what might be more reasonable values?

To calculate tiers, I would need the number of helps, which in turn depends on the amount of healing. Since it references itself, it would require an iterative process. It wasn't a problem before--having more helps does give you more chances to roll the Skill itself--but if I don't assume Energy tiers, then I have two variables that depend on each other.

Well, I ended up using a simulation program I wrote to calculate extra helps and triggers per day. I took the average from two million runs.

Despite my desire to keep all considerations isolated, those morning triggers matter too much, so this does include an 8.5 hour sleep with Inventory/banked skill restrictions for morning Energy triggers. Afterwards, I divided by the skill multiplier I described in my Skills overview to hopefully bring it back to the pure or isolated in an attempt to not just seeing the effects of limited Inventory. (Additional methodology in the comments.)

This time, the table includes a production multiplier (rounded off to not imply precision). Keep in mind that lower numbers are not necessarily that it averages out as healing in the middle of the day, but could be due to healing more than 44 blocks and/or from hitting the cap more often.

Pokémon	1	2	3	4	5	6
Rattata	47 / 2.0x	52 / 1.9x	51 / 2.0x	56 / 1.9x	59 / 1.9x	59 / 1.8x
Dragonair	29 / 2.0x	29 / 2.0x	29 / 2.0x	30 / 2.0x	31 / 1.9x	38 / 1.8x
Raticate	29 / 2.1x	30 / 2.0x	31 / 2.0x	38 / 1.9x	44 / 1.8x	51 / 1.7x
Dragonite	15 / 2.1x	18 / 2.0x	21 / 2.0x	27 / 1.9x	29 / 1.8x	30 / 1.7x
Blaziken	19 / 2.0x	24 / 1.9x	29 / 1.8x	34 / 1.7x	45 / 1.6x	56 / 1.4x

Now there's a more interesting chart. Now that I've done this, I'll revise my earlier opinion--it appears as though Charge Energy S is about even except for how it hits the law of diminishing returns at higher Main Skill Levels, and therefore gets more overvalued by the RP formula.

As before, this is with "neutral" Pokémon--the ones we are using are probably better, and so should hit a break-even point at lower levels.

How far off is the Skill Value for regular Pokémon? I have 10 Pokémon that I use or plan on using that have Charge Energy S. Some are strong and fully evolved, while others I haven't really leveled up yet. Running them through the simulator to get the number of extra helps, looks like the Skill Value should be between .5x to 2.2x of what it actually is, assuming Berries and Ingredients to be accurate.

And then I could apply in Strength from Favorite Berries, actual Ingredient strength--that would give my Pokémon Skill values 1.3x to 6.5x greater than what the RP formula gives.

All right, so if we were building our own RP Formula, how could we handle it better?

One problem is that the skill value comes from a fixed table. For other skills we have looked at so far, I've said could be improved by updating the values in the table, or even making the table dynamic based on your gameplay. But healing skills depend on the strength of the Pokémon themselves, so to be accurate, it should be based on its own stats. It should really follow the formula I provided up above to get the Strength from healing:

Strength from healing = RP Strength/30 * (1/New Tier - 1/Old Tier) * amount healed

A truly dynamic RP could do that by updating itself as it lost or gained Energy, basing the Skill part of the formula on the Strength it would gain triggering right then and there (Initially, converting 0% to whatever it is currently at, then the skill being less effective as it heals more). And I suppose at that point, you could add an Energy multiplier to the whole thing, and RP would drop as it drops tiers.

But I want something more practical. I want an RP value that I could use as a quick evaluation tool, and if I want something more exact, I can use a production calculator or a simulator. In this situation, to make the RP formula better, I would use the formula above, but would need to actually provide an estimate of the average tiers.

For my own formula, I can set RP Strength to just Berries and Ingredients, ignoring the Skills. This avoids an iterative formula, but it also means that it won't consider the concept that more triggers will cause more triggers and the extra Strength from those. So, I can try to make up for that by assuming 0->80+% Energy.

Trying it out with my own Pokémon, the range is .5x-1.5x what it should be, a much tighter range than I had with the RP Formula. Which as far as a rough estimate goes, I think I'm pretty happy with.

Next Week: (or month, or whenever I finish) -- Energy skills (team)

This is an article in a series. For background, see the Introduction. The previous article in the series was on pot expansion skills.

Sorry I didn't make this post last week--I was on vacation, and I ended up needing to take some extra time to make sure my math was right.

Tasty Chance S

When you cook a dish, there is a 10% chance of it being Extra Tasty, which gives a x2 multiplier to the Strength it provides. The chance is increased to 30% on Sunday, which gives a x3 multiplier. Tasty Chance S increases that probability.

When it first came out, my naive assumption for evaluating its Strength was to say, "Oh, it gives +x%? Then that means on average it gives x% of the average dishes I am making." However, that is only true if it reset every meal. I forgot to account for the fact that the effect persists until you actually get an extra tasty meal, or you switch islands, which gives a major impact to how to evaluate this skill.

This skill was the first new Main Skill added after the initial release, and I feel like no skill since has gotten as much analysis as this one. Well, here's one more, and I hope to help make it understandable.

What is the expected value of Tasty Chance S? Let's assume our recipes are all the same size. Then it is not only the x% of the current meal, but the sum of all the partial probabilities that it will hit later on. We can do that by calculating the probability of no success and subtracting from 100%. But we have to be careful--we don't want to count the normal 10% chance of success.

https://preview.redd.it/s54g612iatfh1.png?width=1180&format=png&auto=webp&s=fe293a8ac6b9b8ca21660d0d7a3f69731b793230

• t - Total Tasty Chance (which is the number of triggers multiplied by the size of one trigger, capped at 70%). This is just from the Main Skill, and does not include the base 10%
• m - meals remaining in the week (1-21)

Except we also need to account for the last three meals, since Sunday has a base 30% chance and gives x3 instead of x2. So if I separate m into weekday meals and Sunday meals, that expands the formula to:

https://preview.redd.it/g2qj7tskatfh1.png?width=1188&format=png&auto=webp&s=8a669f49eb594a98ff9418749bf95d37af67b492

That will give us the expected value from Tasty Chance S, but that is the total value after potentially several triggers. What we really want to know is the marginal value--that is, the value of the most recent trigger. So we perform the calculation twice, one with the current Tasty Chance, the second with one fewer trigger, and then subtract.

Here is a graph showing the results. This is a .gif--if it doesn't display right in the post, clicking to open in a new window should make it work.

https://i.redd.it/vn1x38bxatfh1.gif

I am loving this graph. I think it shows several things really well:

Tasty Chance S has a huge law of diminishing returns. This is why early advice was to swap Dedenne out after one trigger, since you'd likely have Pokémon that could deliver more direct Strength than a second trigger from Dedenne could be expected to give. As they have added stronger recipes and more recipe levels, you may find it worth it to go for more than one trigger.

And the law of diminishing returns is stronger the more strong the skill is. The more strength the first trigger takes up, then the less value remains to get allocated to a second trigger.

You can also see that triggers are stronger earlier in the week, which makes sense--earlier means more opportunities to find the benefit. But even late in the week, the first trigger still delivers a decent amount of value--and still more than a second trigger earlier in the week.

Those who have an attention to detail may have noticed a couple things. One is that my graph has a 9% line, and Tasty Chance S skips over that one. Well... Maybe I forgot Tasty Chance S only has 6 Skill Levels when making the chart. I'm leaving it in anyway.

The other thing you might have noticed that is interesting is that for Saturday Lunch and Dinner, getting a lot of triggers actually pushes the marginal expected value negative--the reason for that is because it is making it more likely to be extra tasty on Saturday instead of Sunday, and Sundays have that x3 bonus instead of x2.

One thing I noticed was that as we get earlier in the week, the graph doesn't change so much. What's happening is that the Sunday term gets less and less relevant, and as we add more and more chances, it becomes more and more likely that it will make an extra tasty meal at some point. So for sufficiently large number of future meals, the curve is approaching the shape of this formula:

https://preview.redd.it/val0rpzmatfh1.png?width=1128&format=png&auto=webp&s=bcdb80613699e17d5e838fd51f653ece32ebda99

This is a lot nicer looking, and since t is really the number of triggers times the size of the trigger, it simplifies further, using s as the size of the trigger.

https://preview.redd.it/2n8cytwtatfh1.png?width=1135&format=png&auto=webp&s=8d51b3cc1c2fa0ada27d819c9d074c9b0af7743c

Anyway... to evaluate Tasty Chance S to see where it matches the RP Formula, it will depend on how we use it. Let's pretend we are just looking at the first trigger Monday morning to see the most powerful result. What recipe levels would we need to match what the RP Formula suggests? I'll just select one recipe per tier to use for examples.

Recipes	4%	5%	6%	7%	8%	10%
"Honey Gather" Chocolate Waffles
"Cut" Sukiyaki Curry
"Petal Blizzard" Layered Salad
Jigglypuff's Fruity Flan					1	12
"Overheat" Ginger Salad				7	23	34
"Bulk Up" Bean Curry	1	8	22	34	46	55
Warm Moomoo Milk	70

So if you are using the strategy of putting it in for one trigger, then the RP formula is severely underestimating its strength except for beginner players who are making weaker recipes.

But of course, that is the very best of circumstances, so this would only cater to a playstyle where someone wants exactly one giant blast of power, then ignore it until the next week. And even at its best, that only provides +50% of the average dish you are making, since the other half of the time, it would just get Extra Tasty on its own. (Actually, slightly less than +50% since there's a chance you don't get Extra Tasty in a week.)

Probably a more realistic playstyle is to attempt to get more than one Extra Tasty meal during the week, or to increase the chances of Extra Tasty even further. So we could check at other times during the week, or maybe we leave it in longer, and I could take the average of several triggers. After all, like I said earlier, as dishes have gotten more powerful, that has made it more worthwhile to leave a Pokémon on the team longer.

How about we skip to the other extreme and say we just leave it in the entire week? This calculation will be different than simply averaging the triggers before some meal--leaving it in means accumulating some triggers, cooking a meal, accumulating more, cooking another meal (and so on) and potentially getting one or more Extra Tasty meals along the way, resetting everything.

The math doesn't boil down to some nice formulas to display, but how I do it basically boils down to these steps:

1. Given helping speed and meal windows, how many helps per meal window do I get?
2. For each meal type, given helps per meal window and trigger rate, what is the probability that I get 0, 1, 2, 3 (etc.) triggers? (use binomial theorem) and also for breakfast, treat overnight triggers as not checking until morning
3. For each meal of the week, first calculate the accumulated probabilities for each trigger, then alter the accumulated probabilities by calculating the probability of an extra tasty meal
4. Use the probability of an extra tasty meal to accumulate the contribution from Extra Tasty S
5. Divide the total expected reward by the total number of expected triggers in a week for the average value per trigger

Unlike before, now we have included variables of speed, trigger rate, and meal windows. I decided to normalize it for the graph by putting the triggers per week on the x-axis. I just picked a meal window and stuck with it--altering meal windows does change the values slightly, but this graph should be good enough to get an idea of what is going on.

https://preview.redd.it/0rvxrsobftfh1.png?width=1424&format=png&auto=webp&s=f7bccb8e3cac28da6b37cd4f61730ab8bfaeee45

For this one, you can see that perhaps counter-intuitively, the faster or higher skill rate a Pokémon is with this skill, the less Strength each trigger provides. You do indeed get more total Strength, it's just that the extra strength doesn't grow as quickly as the extra triggers do. If we look back at the previous chart, it means we are spending more time in that lesser-value tail-end portion of the earlier chart, or even hitting up against the 70% maximum more often.

Even though this is much lower, this might be more representative for Berry/Ingredient specialists who are on your team for other reasons and the Extra Tasty S is more of a nice bonus, but it's staying on the team if you need their Berries or Ingredients.

So how does this compare with RP? Since there's a wide range of values and I don't want to fill out some endless tables, I'll pick some representatives that can span from the largest values to the lowest. That means maybe some unrealistic setups, but will cover the range.

• Mudkip, level 1, neutral nature, no healer, Main Skill Level 6 (expected triggers per week: 4.6, expected value per trigger: +34.6%)
• Weavile, Level 30, neutral nature, healer, Helping Speed M, Main Skill Level 5 (triggers per week: 9.3, value per trigger: +25.8%)
• Swampert, level 30, neutral nature, healer, Helping Speed M, Main Skill Level 4 (triggers per week: 16.3, value per trigger: +20.3%)
• Cramorant, Level 50, Sassy nature, healer, Helping Speed M, Main Skill Level 3 (triggers per week: 23.5, value per trigger: +16.5%)
• Holiday Spheal, Level 50, Sassy nature, healer, Skill Trigger M/S, Helping Speed M, Main Skill Level 2 (triggers per week: 41.0, value per trigger: +12.3%)
• Dedenne, Level 70, Sassy nature, healer, Skill Trigger M/S, Helping Speed M/S, Main Skill Level 1 (triggers per week: 50.9, value per trigger: +9.9%)

Keep in mind for the chart, it's really the comparison between value per trigger and the value in the RP formula for the Main Skill Level that actually matters--these are just example setups that can give us those values.

Recipes	Dedenne	Holiday Spheal	Creamorant	Swampert	Weavile	Mudkip
"Honey Gather" Chocolate Waffles
"Cut" Sukiyaki Curry
"Petal Blizzard" Layered Salad				1	5	7
Jigglypuff's Fruity Flan	9	19	21	28	32	34
"Overheat" Ginger Salad	33	40	41	46	50	52
"Bulk Up" Bean Curry	54	60	61	68
Warm Moomoo Milk

This looks a lot more reasonable and even more balanced with Skill Level, maybe something like assigning 100 Strength to every +1% of expected value, which creates this band of

But even with a "leave it in all week" strategy, you can see that the RP formula undervalues it if you are making the largest recipes.

As for the question of how much should it be worth--I suppose a truly dynamic RP formula could take your average recipe strength, then multiply it by the expected percent based on when you are looking at it according to the first chart, so the RP would change every time it triggers, or every time you make a meal.

Personally, I'd rather use RP as more of a generic evaluation tool than an instantaneous tool. I'm already tracking my recipes, so I could track when I actually use Tasty Chance S and build an expected average value per trigger based on that. And from the perspective of wanting a simple formula, "pull in the expected value multiplier" is a lot better than "get the expected value multiplier from this complex formula/procedure."

But I haven't been tracking that, and I keep thinking about how the strategy behind Skill specialists should be different than Berry/Ingredient specialists with Tasty Chance S. One idea I had is to build lookup tables based on how I might use them, and so Skill specialists get their own table.

Anyway, just trying out some different numbers, based on the recipes I'm currently making, I'd say my RP values for Tasty Chance S would be more accurate if they were maybe 4-8 times bigger.

Bulk Up (Cooking Assist S)

This is a combination of Ingredient Magnet S and a weaker version of Tasty Chance S. Those two skills use the same values in the RP formula as each other, while Bulk Up is 1.3x bigger. To match what we had looked at before, let's first see the optimal Monday morning trigger.

Recipes	Level 1	Level 2	Level 3	Level 4	Level 5	Level 6	Level 7
"Honey Gather" Chocolate Waffles
"Cut" Sukiyaki Curry
"Petal Blizzard" Layered Salad
Jigglypuff's Fruity Flan							4
"Overheat" Ginger Salad					1	8	15
"Bulk Up" Bean Curry					9	16	24
Warm Moomoo Milk	3	5	7	11	21	28	34

When we looked at Ingredient Magnet S, we saw that low Skill Levels weren't even on the chart. Meanwhile we saw that Tasty Chance S, high tier recipes didn't make it to the chart. The combined effect pushes things all into the low-tier, high-level corner. The larger RP only managed to bring the lowest tier recipe onto the chart, and would have to be a lot bigger to bring more into the chart.

And now let's look at the "leave it in all week" scenario. For this, I used the same setups as before, with a couple differences: Bulk Up has 7 Skill levels, so I gave the Level 1 Heracross a Rash nature, Main Skill Level 7, and added one next to that as Level 20, neutral nature, no healer, Main Skill Level 6. This gives covers a range of 14.9 to 57.6 triggers per week. Column headers is the calculated average expected value per trigger.

Recipes	3.8%	6.4%	7.2%	10.0%	10.4%	13.5%	16.8%
"Honey Gather" Chocolate Waffles
"Cut" Sukiyaki Curry
"Petal Blizzard" Layered Salad							4
Jigglypuff's Fruity Flan					3	9	16
"Overheat" Ginger Salad				2	10	18	24
"Bulk Up" Bean Curry		2	3	7	16	24	30
Warm Moomoo Milk	5	7	9	14	23	29	35

Instead of creating a nice band like above, it seems to have just widened the range of the low-tier, high-level corner. Basically, whatever RP value isn't taken up by Extra Tasty gets taken up by the extra Ingredients.

I think in order to make it fit better, I have to devalue not only the tasty chance, but the the Ingredients themselves from Ingredient Magnet. That's what I did in my earlier Ingredient articles, by saying "what if half the Ingredients aren't part of the recipe?" However, since Tasty Chance requires actual recipes for the calculation, adding in twice as many ingredients would make the recipes more powerful and therefore the Tasty Chance S part of the skill more powerful.

This is where the "real world" kind of clashes with my theoretical "what if" scenarios--realistically, if you are using twice as many Extra Ingredients as the Recipe calls for... then you are probably not making top-tier dishes.

Anyway, I could use a similar strategy as above to give a more accurate RP value. And using the value of dishes and Ingredients I am currently using, and trying out different expected percent values for Bulk-Up, my values would be more accurate if the RP values were 3-5x bigger.

Next week: Charge Strength skills

This is an article in a series. For background, see the Introduction. The previous article in the series was on Ingredient skills.

My discussion about Ingredients is actually misleading, because I left something out.

Let's say we've taken what we've learned and have come up with some method to give a "true" value for Ingredients. Now let's say that I have the perfect Dragonite, all its Ingredients are unlocked and it is bringing in massive amounts of Fiery Herb. Let's say I evaluate it and see that it can bring in more Strength than any of my other Pokémon. That means I should always have it on my team, right?

Well, no.

Even though Fiery Herb is a strong ingredient, it is only useful as part of a Recipe. If I'm bringing in too many, then I'll use them less efficiently as filler ingredients, or worse, throwing them away for Dream Shards.

There is a law of diminishing returns, and that is why you might see strong min-max players talk about swapping out their Ingredient specialists.

Ingredients might be better thought of as "potential Strength" that only gets converted into actual Strength when it is used.

Cooking Power-Up S

So with that background, let's talk about Cooking Power-Up S, which increases your pot size. The first limitation players encounter is how many Ingredients their helper Pokémon can actually find. But let's say we are able to find a ton. After that, you can think of Recipe size as a soft cap for how strong your Ingredients can get, while the pot size is a hard cap for how strong of Recipes you may use.

I like to think of it like this:

https://preview.redd.it/bzdm2a7d48dh1.png?width=1216&format=png&auto=webp&s=5d2fb3411e6872e0df32d79539ae1d80c56045a5

Ingredients part of the recipe get the recipe bonuses, those beyond that only have their base value. But Ingredients beyond the pot size cannot be added, and so are 0. Increasing the Pot size unlocks their potential--granting their base value if just pot filler, granting recipe bonuses if part of a recipe. If the larger recipe is part of a bigger tier, then that extra boost applied to the initial Ingredients should also be attributed to Cooking Power-Up S.

In order to decide if the game overrates it or underrates it in the RP formula, we need to figure out how to actually rate it. How much Strength comes from one trigger?

I think the obvious way is to assign its value equal to the increased Strength from the recipes we are now able to make. But I need to be careful--because didn't I already apply Recipe tier and level multipliers to the Strength of the Ingredients? In order to avoid double-counting, I cannot attribute Recipe Strength towards both pot expansion and the Ingredients themselves. I think it is fair to apply the difference in Recipes to pot expansion, and then reduce the total Strength of those Ingredients by the same amount.

With that, I can write a program to find where the Strength the RP formula assigns to Cooking Power-Up S roughly equals the difference between two recipes at various Skill Levels and initial pot sizes. We will assume without pot expansion, they are making the largest dish they can, and the two recipes are the same level.

Skill Level	Recipe 1	Recipe 2	Recipe Levels	Pot Size	Triggers
1	"Solar Power" Tomato Curry	"Egg Bomb" Curry	5	15	3
1	"Explosion" Popcorn	"Early Bird" Coffee Jelly	10	36	1
1	Greengrass Salad	"Petal Blizzard" Layered Salad	36	63	1
2	Snoozy Tomato Salad	"Water Veil" Tofu Salad	12	15	2
2	"Mold Breaker" Corn Tiramisu	"Leaf Tornado" Smoothie	12	42	1
2	"Role Play" Pumpkaboo Stew	"Hidden Power" Perk-Up Stew	6	69	3
3	"Lucky Chant" Apple Pie	"Sweet Scent" Chocolate Cake	4	18	1
3	"Limber" Corn Stew	Grilled Tail Curry	37	30	3
3	"Cross Chop" Salad	Greengrass Salad	18	57	2
4	Moomoo Caprese Salad	"Cross Chop" Salad	13	24	3
4	"Steadfast" Ginger Cookies	"Leaf Tornado" Smoothie	25	36	2
4	"Role Play" Pumpkaboo Stew	"Cut" Sukiyaki Curry	23	69	3
5	"Dazzling" Apple Cheese Salad	"Overheat" Ginger Salad	16	24	1
5	Ninja Curry	"Inferno" Corn Keema Curry	28	51	2
5	"Flower Gift" Macarons	Clodsire Éclair	22	78	3
6	Hearty Cheeseburger Curry	"Egg Bomb" Curry	34	18	1
6	"Lovely Kiss" Smoothie	"Leaf Tornado" Smoothie	54	36	2
6	"Calm Mind" Fruit Salad	"Petal Blizzard" Layered Salad	5	51	1
7	Big Malasada	"Flower Gift" Macarons	5	24	2
7	Luscious Avocado Salad	Greengrass Salad	21	42	1
7	"Role Play" Pumpkaboo Stew	"Hidden Power" Perk-Up Stew	33	69	1

Okay, so those are kind of all over the place, so maybe not super helpful.

But when I look at these examples, many of these (and others I saw) have lower level recipes and have extra triggers than what I might expect during actual gameplay, which leads me to suspect that Cooking Power-Up S is undervalued in the RP formula.

This is simple to look at in this sort of example-based analysis. But in actual gameplay, things don't fit so neatly like this. Here are some other factors I've thought of:

• Cooking Power-Up S triggers, but you don't manage to gather the Ingredients you need for the larger Recipe. Value = 0
• Likewise, if it triggers on your non-Skill specialist that was on your team for its specialty, and you weren't even aiming for a larger Recipe, so you don't make use of it. Value = 0
• Cooking Power-Up S triggers, but you need it to trigger a second time to make the Recipe you are aiming for. Value = 0
• Cooking Power-Up S triggers enough times, so gets full value. But then it triggers an extra time that you don't need. Value for that trigger = 0
• You make a larger recipe for Breakfast, but now you are out of Ingredients for Lunch, and so have to make a smaller recipe. Value = ?

So in reality, the value of each trigger is going to vary wildly.

Like with other gameplay elements, I want to create my own way to evaluate this skill. One way might be to run a bunch of simulations to create my own value for the Skill. I tried that, but there was too much variation in the randomness for me to get a sense of it.

So instead, I decided to track pot expansions alongside the recipes I am making, and use historical data to predict how much extra Strength on average I earn from the Skill. This worked well, since it also meant that I can reduce my Ingredient Strength by the same amount to avoid double-counting with my Ingredients.

That does mean that for the bulleted list above, I ended up ignoring the last bullet, meaning the "blame" for any weaker later Recipes goes to the Ingredients, rather than as a secondary effect from Cooking Power-Up S.

Which is probably fair. After all, in the first bullet the "blame" will go to the Skill instead of a secondary effect from not finding enough Ingredients. Pot expansion and Ingredient finding are a team that need to work together to be most effective.

So here's the model for how I calculate the value of my Cooking Power-Up S. If I imagine making some recipe, and then being able to make a bigger recipe if my pot were bigger, I would expect both a larger tier bonus and a larger boost just from being able to include more Ingredients. I visualize it like this:

https://preview.redd.it/m3t923g248dh1.png?width=1397&format=png&auto=webp&s=3b7a2f0fa1f0834bf89d888b69d2c23c821b23b2

So if I've made the bigger recipe, I need to work backwards to get what recipe I would have made instead, and then subtract. So what recipe would I have made?

Initially, I thought about tagging each recipe to say how I might have used each Ingredient instead. But that seemed like a lot of work towards something that still wasn't very realistic. Instead, I decided to create a hypothetical smaller recipe that has the same Ingredients, just a percent smaller so that it would fit into my pot, removing extra Ingredients first.

That makes the calculation a lot easier, and plus, like I said, pot expansion and Ingredient finding are a team, and so its okay I attribute strength of a powerful dish to both.

So for those of you who live for formulas, here's what I did. First, the normal recipe formula:

https://preview.redd.it/0ru50mnuu2dh1.png?width=1420&format=png&auto=webp&s=95875d9da3abf06de3f1f1206e02974408f43762

• i - sum of the Strength for each ingredient (base ingredients or extra ingredients)
• t - boost from recipe tier
• l - boost from recipe level

Then for the hypothetical smaller recipe:

https://preview.redd.it/6z530mqwu2dh1.png?width=1426&format=png&auto=webp&s=5e9d6617f8cf4afccced290df154c7afa850ff30

• t' - hypothetical tier should match my normal pot size
• n - number of ingredients
  • base - number of base ingredients
  • extra - number of extra ingredients
  • pot - normal pot size
  • surviving base - smaller of base or pot
  • surviving extra - max(0, pot - base)

In order to compare Skill levels, I want the average value per meal for each +1 to pot size. Like I said before, true value will vary wildly, but I want an average, so I'm averaging by each +1 to pot size.

https://preview.redd.it/gixgp1rzu2dh1.png?width=1428&format=png&auto=webp&s=95ad4b18d1ff77e00b161f923b17e17c11aba99a

• tasty - average boost from Extra Tasty meals, which is 24.6/21
• b - size of pot boost

Then finally, take the overall average across all meals:

https://preview.redd.it/zvlrr7m3v2dh1.png?width=1084&format=png&auto=webp&s=2635122fc08fe691d9baefc57cb2a12521a8bc99

I am currently a mid-level player, meaning I haven't taken a helper Pokémon to 60 yet, which I imagine is when pot expansion really shines. But I do it every now and again if I can. But I calculated my personal average to be 241 per +1 to pot size, which is about 1.3x - 1.9x what the RP formula predicts.

As I was writing my first draft, the Soul Dew event fell into my data range, so I did a bit better with 448 per +1 to pot size, which is around 3x what the RP formula uses. Even then, I wasn't taking full advantage, so I expect that dedicated min-max players should do even better.

Final Thoughts:

So my thought is that the RP formula for Cooking Power-Up S is probably undervalued for mid- or late-game players who are intentionally using it or taking advantage of it, and is probably only accurate or overvalued for beginner/casual players who are making smaller recipes or ignoring the effect.

Bonus Formula:

For completeness, here is my updated formula for how I evaluate my Ingredients:

https://preview.redd.it/tmkyrld9v2dh1.png?width=1078&format=png&auto=webp&s=4188bb0ab29a52c92fc8bbc05973e4767236eef5

• t* - t normally, but t' for dishes making use of pot boosting
• Note also that surviving extra is the same as extra in normal non-boosted dishes

This does mean that in earlier discussions about Ingredient Strength, I slightly overvalued those in the top two tiers (61% and 78%) since they are currently only attainable through pot boosting. Lower tiers I may also have overvalued, depending on how much pot boosting I want to assume is going on.

But even though many effects can compound (another example is Sneaky Snaking) I want to review each in isolation as best I can, and we will need to keep these other effects in the back of our minds.

Next Week: Extra Tasty S

This is an article in a series. For background, see the Introduction. The previous article in the series was on Ingredient finding.

Skills

The third part of the RP formula is the calculation for Skills. Since there are many kinds of Skills, this section will be broken up into many separate articles, but first I want to talk about Skills in general.

One subtle thing to consider is that we saw that Berries don't take into account favorite berry, and it appears that Ingredients don't take into account Extra Tasty (and assumes we are using a LOT of extra ingredients). But many Skills don't have this kind of week-based variation or ignored secondary effects.

Sometimes you'll hear people say that the RP formula overvalues Skills, but maybe it is technically that the other two are just undervalued, and so Skills appears overvalued if it is more accurate. We will look into that--the skills are all different, and so some are still undervalued. But I think this principle is a good one to keep in mind.

On the flip side, if a player is asking why their Pokémon's RP is so high, we shouldn't be so quick to accuse its high Main Skill level-- if it has a low trigger rate, then even if overvalued, Skills isn't actually contributing much to its RP.

Looking at the RP formula, the Skills value is even more susceptible to error than Berries and Ingredients for a couple reasons.

One is the RP formula assumes that we check every help. I already talked about Sneaky Snacking when I talked about Inventory. However, this is more pronounced for Skills, since a helper Pokémon can only bank one skill trigger, or two if it is a Skill specialist.

This means that in order to be more accurate, the formula I had developed for Sneaky Snaking would need to be further modified in order to account for this. I'm still assuming that I check every help during the day, and so this only applies at night. Since these formulas use the number of helps before full several times, let's put that into its own term in the formula:

https://preview.redd.it/1y7f8nxynfah1.png?width=1047&format=png&auto=webp&s=a3c114995a1005ce9b981a2140839b48aa08d673

• n is the % of helps that happen overnight.
• T is the 24 hours in a day multiplied by average Energy
• h is helps per hour. (so that n*T*h = number of helps at night)
• I is the Inventory
• D is the average number of items per help, calculated by: number of Berries * (1- Ingredient Rate) + average number of Ingredients * Ingredient Rate

With this factor, that makes our previous formula look like this:

https://preview.redd.it/kbuscl42ofah1.png?width=1077&format=png&auto=webp&s=265e075e1f07a6831ec308777c7556c29b912ac1

And now, applying it to Skills, we get two formulas with one additional variable not yet defined

• r is the Skill trigger rate

This is the formula for Berry or Ingredient specialists:

https://preview.redd.it/lxqg5wn5ofah1.png?width=1060&format=png&auto=webp&s=c6e9f774355166caea386a1a8f64742f6ff001c5

This is the formula for Skills specialists:

https://preview.redd.it/5a45fk09ofah1.png?width=1078&format=png&auto=webp&s=0160d05afb0c7ed548f0d5c1e2095a2d5543b20f

Taken together, that would make my updated RP formula look something like this:

(Berries * (1 + Sneaky Snaking) + Ingredients * (1 - Sneaky Snaking) + Skills * (Skills multiplier)) * Miscellaneous

As I mentioned in my earlier article, I'm assuming 8.5 hours of sleep, and naturally decreasing Energy throughout the day, so I'm setting n to about 25% and T to about 37.

Sneaky Snaking often doesn't have an effect when Inventory is large enough. When it does have an effect, it increases Berries and decreases Ingredients and Skills. Changing it to have this skills multiplier means there is always a reduction--even compared with the Sneaky Snaking multiplier. So that is one way that Skills in general can be overvalued.

How strong is the effect? Looking at my Pokémon, applying Sneaky Snaking (not counting removing Inventory from the Miscellaneous factor) made the RP of my Pokémon change +5% to -8%. Applying the effect of only banking 1 or 2 skill triggers overnight can make RP go down up to 3% for Pokémon with a high trigger rate, so it would seem that perhaps most of the effect was applied thanks to Sneaky Snaking, but if I also assign them their max Main Skill Level, they can lose up to 6% of their RP compared with just Sneaky Snaking.

Anyway, I said that there were a couple ways that the Skills value in the RP formula is even more susceptible to error than Berries and Ingredients.

The other way is that even if we maintain the assumption that we check every help, not every Main Skill contributes directly to Strength. How the game deals with it is that it assigns a value to each Main Skill at each level to use in the formula. Having fixed values means that the strength of a Main Skill scales only with Main Skill Level. That may be true for Charge Strength skills, but many skills actually do scale with level, Berry Burst being a prime example.

So how reasonable are these values that they assign? To answer that, we will look at each Main Skill in turn and see under what conditions the assigned value is accurate to decide how inaccurate the RP formula is for skills.

Next week: Ingredient Magnet and Ingredient Draw

This is an article in a series. For background, see the Introduction. The previous article in the series was on both Berries and Inventory.

Ingredients

The second part of the RP formula is the calculation for Ingredients. Unlike Berries which scale with level, Ingredients depend on what recipe it is used in. The formula for Recipes is:

Floor((Round(Sum of Base Strengths of the Ingredients for the recipe * Recipe tier bonus * Recipe Level bonus) + Sum of extra ingredients) * Extra Tasty bonus * Area bonus)

• The game shows you the recipe level bonus as you are making the recipe, for example, level 65 gives a 234% boost.
• The recipe tier bonus is somewhat hidden, but is mostly correlated with the number of Ingredients the recipe calls for. For example, recipes with 103-115 ingredients give a 78% boost.
• Sometimes meals are Extra Tasty, which gives a x2 multiplier (x3 on Sundays).
• Extra ingredients added to a meal only get the extra tasty bonus, they do not get the recipe tier or level bonuses.

So in order for the RP formula to be accurate, it would need to look into the future and see what recipes you will put them in. Raenonx gets accurate results in its Production Calculator by letting you enter in what recipes you plan on making. Neroli's lab runs simulations and gives probabilities of what recipes you might be able to cook with the team you have selected.

But the RP formula is supposed to be a simple calculation, not a time machine or a full production calculator. It is designed to be self-contained and not depend on other gameplay elements. So how does it account for boosts to Ingredients? It takes the base strength of the Ingredients the helper Pokémon finds (average strength per help) and then multiplies it by a fudge factor that is based on the Pokémon's level. I think that's a fine enough method, but does it give good results?

Since we are looking into how far off the RP formula is, I think the best way to explore this question is asking: under what conditions is it accurate?

Last time we talked about Sneaky Snaking, so we won't revisit that--let's continue with the assumption we collect every help. This table answers the question: Given a Pokémon Level, what level would a recipe need to be (at the different tiers) for the RP formula to be accurate? If your recipe is a higher level, then the RP formula is undervaluing your Ingredients, and if your recipe is a lower level, then the RP formula is overvaluing your Ingredients.

Level	18.75%	20.5%	25%	35%	48%	61%	78%
20
25	0	0
30	4	3
35	7	7	4	0
40	11	10	8	4	0
45	16	15	12	7	2
50	21	20	17	12	6	0
55	25	24	22	17	10	5	0
60	29	28	26	22	16	9	4
65	37	36	34	30	23	20	13

I ignored Extra Tasty, which would make levels even lower. But I also ignored extra ingredients as pot filler, which would make levels higher. Those numbers were looking pretty low... what if I instead assumed 50% of Ingredients were filler? Here's a new table:

Level	18.75%	20.5%	25%	35%	48%	61%	78%
10
15	0	0	0
20	4	4	2	0
25	9	8	7	2	0
30	16	15	12	7	2	0
35	22	22	19	14	8	3	0
40	28	27	25	21	14	8	3
45	33	33	31	26	21	15	8
50	38	38	36	32	27	22	15
55	43	43	41	37	32	28	22
60	48	47	46	42	38	33	28
65	57	56	55	51	47	43	38

I suppose there exist players that are just using Auto Cook and making Mixed recipes, but from my perspective, this shows that the RP formula is undervaluing Ingredients.

I imagine that the RP formula was designed to ignore Extra Tasty meals, much like it ignores favorite Berry. If I were re-designing the fudge factors, I'd put in a minimum increase of 24.6/21, which is the average multiplier for Extra Tasty meals (factoring in the 10% chance of Extra Tasty meals Monday-Saturday and 30% chance on Sunday). I would also assume a relationship between pot size and level, so that I could make the fudge factors scale with Recipe level and tier.

Another thing to consider is that not all ingredients are used in all tiers, so we could make fudge factors different by ingredient, but then you would have to change them as they add new recipes. And to be fair to Select Button, they not only have added stronger recipes since the game first released, but they also increased the value of the lower tier recipes since then, too.

Last time, I talked about making a "Dynamic RP" formula that would allow the value to change on external changes, such as favorite Berry. For Ingredients, you could keep track of recipes made and what extra ingredients were used in them, and then make the assumption that Ingredients will be used the same way in the future to create a multiplier for each Ingredient. Obviously it would take away from the simplicity of the RP formula, but it could produce somewhat accurate results without having to go into a full production calculator.

Even with this plan, there's different ways you could make an average. For now, I went with using the last two weeks I've used each recipe type. Based on my current gameplay, my Ingredients are getting between 3.0x to 5.7x their base value, which is far greater than the 1.2x-1.6x that the RP formula is giving me, meaning my Ingredients are even more undervalued than my Snorlax's favorite Berries.

I should also add a note that this discussion has only been on the accuracy of the RP formula. One limitation of even a Dynamic RP is that it is just a flat score. How strong it is will always be based on how you actually use it. So for example in relation to Ingredient specialists, Ingredient spread is important to how strong it will actually be. Same Ingredients (AAA) are more reliable and therefore valuable in general, while mixed-Ingredient spreads are less reliable and therefore more situational. Raw strength values won't capture this value. It is more of a starting place, and then then a production calculator can be used to evaluate it further.

Next week I'll have a brief overview on Skills in general, followed by an article on Ingredient-related Skills.

This is an article in a series. For background, see the Introduction

Berries

The first part of the RP Formula is the Berries calculation. Berries is the most accurate because the Strength of a Berry depends on the Type of Berry and the Level of the Pokémon. I understand the game uses a lookup table for all values, but in the early days someone figured out that it follows a formula:

Strength per Berry = bigger of (Base Berry strength + Level-1) or (Base Berry strength * 1.025^(Level-1)****)

However, before rounding off, it also multiplies by the favorite berry and area bonus, which aren't things that the RP formula considers. Instead, the RP formula simply considers the rate at which it finds Berries, since each help can either find Berries or Ingredients.

So that's the first way the RP formula is inaccurate. Of course, we don't want RP changing every week, so it makes sense that it doesn't consider these factors. Since Area Bonus affects all sources of strength, I think it is fine that the RP formula doesn't consider that. But for favorite Berry, it does mean that it will undervalue any Pokémon who matches the favorite Berry.

One interesting thing though is that in the "Help Me Pick" feature when you build teams, "Berry finding" does account for favorite Berries.

A more subtle way that the RP formula is inaccurate is another assumption: that you collect every help. A Pokémon has a limited Inventory, so if you don't collect for a while, then eventually the Inventory will fill up, and the Pokémon will no longer provide normal helps, but will switch to what is called Sneaky Snaking, and every help will be Berries that are fed directly to Snorlax, without rolling for Ingredients or Skills.

So the Berry portion of the formula is exact when non-favorite Berry, and when not Sneaky Snacking.

It is probably a fine assumption that you collect every help, but then that means that Inventory is meaningless. So before moving on to Ingredients/Skills, lets look at Inventory.

Inventory

The RP formula is a bit inconsistent. It does not include anything for base Inventory, but it does for the Inventory Up subskills, which add to the multiplier in the Miscellaneous portion of the formula.

However, an increased inventory doesn't multiply the strength from Berries, Ingredients, and Skills--it just prevents you from going into Sneaky Snacking.

What would it take to be accurate? The formula could have a inventory factor that changes some percent of Ingredient helps into Berry helps, and reduces the number of Skill helps by the same percent. I would use a formula that looks like this:

Sneaky Snaking percent formula

• p is the percent of the day spent Sneaky Snaking.
• I is the Inventory
• n is a value I would provide for the percent of consecutive helps I am not checking (e.g. overnight)
• h is the helps per hour
• T is a value I would provide that equals 24 hours * average boost from Energy
• Berries is the number of berries per Berry help
• IngRate is the Ingredient finding rate
• AvgIng is the average number of Ingredients per Ingredient help
• This inventory factor would need to be based on how fast it can fill its inventory. The trade-off would be that it would take away from the simplicity of the RP formula.

If I ignore Skills that heal a Pokémon (which I would want kept in the Skills portion of the RP formula anyway) and if I ignore healing from meals and so treat Energy as 100% in the morning after 8.5 hours of sleep and naturally falling throughout the day-- then I would set n to about 25% and T to about 37.

After making this calculation, we could apply it to the main part of the RP Formula this way: (Berries * (1+p) + Ingredients * (1-p) + Skills * (1-p))

Under what situations would the RP formula be the same, making just this change to move Inventory out of the Miscellaneous factor and instead use Inventory to include this Sneaky Snacking factor instead?

I have a spreadsheet where I have the stats of all the Pokémon I have ever caught. Applying the formula, I think it is most interesting if I set them all to level 60 and put them at max evolution. Here's what I see:

https://preview.redd.it/u186eh68g48h1.png?width=1213&format=png&auto=webp&s=d680257ca0cd2c3f98df4b7c156abc0fcd316438

22% had no change. These are helper Pokémon that didn't have an Inventory Up subskill and that won't fill up their Inventory overnight.

Of the 10% that were positively affected by the change, none have Ingredient subskills, and 86% of them are Berry Specialists, the rest are Skill specialists with Berry Finding S.

On the flip side, the remaining 68% affected negatively, most can be attributed to having an Ingredient Up subskill, which no longer provides a boost in the Miscellaneous section. Looking at the others, they are mostly Ingredients Specialists. The few Berry specialists that were negatively impacted all either had Ingredient Finding M or was Slaking, which naturally has a high Ingredient finding rate.

So I think it is safe to say that the RP formula strongly overvalues Inventory subskills. Inventory Up often has no effect on actual gameplay, and when it does, it is provides a boost primarily to Ingredient specialists, though not as large as the RP formula gives.

It's probably fair that they don't include Sneaky Snaking--it takes away from the simplicity of the RP formula. But more importantly, in situations where Berry helps are stronger than Ingredients, and Inventory fills up fast enough, then unlocking an Ingredient Up subskill would make the RP formula go down, which wouldn't be very exciting for a random player to see.

But what I will say is that it might be interesting to make my own personal RP formula that is just a bit more complicated, which still preserves the Berry/Ingredient/Skills split and includes favorite berries/requested dishes/team composition that affect those three things, and so is allowed to change based on how I play the game--a "Dynamic RP"--but would be something would be more useful for a sort of situational comparison without going into a full production calculator.

Next week: Ingredients