Our trilogy reaches its quintessential phase with the discovery of another XL mechanic! In a Niantic Helpshift article, it was casually mentioned that Candy XL can be earned by evolving Pokémon. In case you were unaware of this Candy XL source, don’t worry — we forgot too! That is, until one sharp-eyed Silph Researcher and another member of the subreddit both spotted this mechanic in action. So here we go again: our team of researchers enthusiastically donned their mad scientist lab coats to evolve Pokémon by the thousands. Let’s take a look at what they learned!

Key Findings

1. Higher-level Pokémon are more likely to reward a Candy XL when evolved.
2. The relationship between Candy XL and Pokémon level likely follows the distinct stairstepped pattern observed when catching, transferring, and walking buddy Pokémon.
3. Candy cost and evolution items play no significant role in Candy XL drop rates.

Background & Methodology

In order to control for any potential confounding variables, Pokémon were evolved in their obtained state without powering up, purifying, or trading them. The only exception to this rule was the evolution of species that do not have a candy cost to evolve post-trade. Researchers recorded Candy XL numbers before and after each evolution, as well as each Pokémon’s pre-evolution information (species, CP, and IVs, which were used to determine Pokémon level). In addition, they tracked any use of evolution items and noted any free post-trade evolutions.

The Effect of Level on Candy XL

During the course of our investigation, our researchers evolved an incredible 16,170 Pokémon for the cause. Every evolution is guaranteed to produce one regular Candy for the Pokémon family involved in the evolution, thus generating a probability to produce a Candy XL alongside it. But Candy XL from evolution has proven to be a rare commodity. Researchers never observed more than one Candy XL drop per evolution, and more often observed none at all.

We set out to clarify the relationship between Pokémon level and Candy XL received and compare it to the results of our previous Candy XL studies.

Plot of the average number of Candy XL received as a function of evolved Pokémon level with a 95% confidence interval (black). The blue dashed line represents the predicted rate for the stairstep model.

As with catching, transferring, and walking buddy Pokémon, the Candy XL rates from evolution have a tendency to break into distinct level ranges (modeled by the blue dashed line). The range up to level 14 produces negligible amounts of Candy XL, only doing so about 1% of the time. For each stairstep above the first, the rate increases. As we found previously, the probability of receiving Candy XL when catching a Pokémon corresponds to three flips of a weighted coin, which becomes more heavily weighted toward Candy XL drops as Pokémon level increases. For evolution, the probability relates to just one coin flip with the same weighting and level multiplier. The rates for each step range, their 95% confidence intervals, and the modelled estimates are shown in the table below. Please note: Level 23 is an outlier from the model, but not outside our expectations.

Evolving continues to parallel the previously proposed models for catching, transferring, or walking buddy Pokémon, with weighted chances for a Candy XL to be generated alongside each regular candy. However, the rates observed here are much lower than those observed for other methods of obtaining Candy XL, thereby making evolution a less-than-plentiful source of Candy XL.

Factors That Did Not Influence Evolution Candy XL

Other factors were examined for a possible effect on Candy XL drop rates, but nothing other than Pokémon level exhibited any significant effect.

• Evolution Candy cost did not exhibit a significant relationship to Candy XL drop rates after controlling for level range (tested using a logistic regression model, z = 1.37, p-value = 0.17).¹ This includes evolutions done for free after a trade for eligible species.
• Evolutions requiring items showed no significant difference in Candy XL received compared to itemless evolutions (z = -0.63, p-value = 0.53).¹
• First- and second-stage evolutions did not significantly differ in Candy XL rates (z = -0.59, p-value=0.55).¹
• No significant relationship was observed between Candy XL rates and Pokémon species or Pokédex number, but more data is needed to draw a stronger conclusion.

Viability of Evolution as a Means to Accrue Candy XL

The existence of this mechanic raises the question: is evolution a viable way for travellers to gather Candy XL, or should it be treated as no more than an occasional bonus? Under what circumstances is evolution more efficient than converting candy to Candy XL at a rate of 100:1?

This question can be answered by dividing a species’ evolution candy cost by the corresponding average evolution Candy XL obtained for that Pokémon’s level from this study ([Candy/Evo]/[XL/Evo]=[Candy/XL]). Excluding evolutions that cost 100 or more candy, which will always be less efficient than conversion by default, this gives us the following table of average regular candy spent per Candy XL obtained upon evolving:

This result reveals that only a few conditions have better efficiency than direct conversion: 12-Candy evolutions at level 23 and up, or 25-Candy evolutions at level 31 and up.² Obtaining Candy XL through evolution may be beneficial in certain specific circumstances, but is otherwise ineffective.

Parting Words

Candy XL remains a vital and often rare resource in Pokémon GO. Receiving it from Pokémon evolution is a nice bonus reward for an otherwise routine task. Unfortunately, the low drop rate means evolution is less beneficial than other methods of obtaining Candy XL. We’ll explore that and more in our upcoming review on all things Candy XL. Until next time, travelers!

Acknowledgements

Article authors: Scientists DeeDillyDawn and Gustavobc, and Lead Researcher PancakeIdentity
Analysis: Scientists CaroKann and Gustavobc
Graphics: Scientists WoodWoseWulf and CaroKann
Project Leaders: Scientist Gustavobc and Lead Researchers AlertedFancy, archer, and PancakeIdentity
Editing: Scientist Cham1nade and Senior Researcher JinianD
Data collection: Between August 23rd and November 14th, 2021, 130 evolution researchers collected data for this article. These 10 would have made Professor Rowan particularly proud:

• aarghaarghaargh
• alyssa (they/them/theirs)
• CelestialElaine
• Cloudstarcarrie
• Daemare
• Emma
• FatalisticFeline
• HallDaCounselor
• LoudShrugging
• SpelingChanp

Footnotes

¹ For the multiple logistic regression model, we estimated the log likelihood of receiving Candy XL as predicted by Pokémon level, evolution cost, usage of evolution items, and the evolution stage. Using a Wald test on the estimated coefficients for each variable in the regression, we found evolution cost (z = 1.37, p-value = 0.17), item usage (z = -0.63, p-value = 0.53) and evolution stage (z = -0.59, p-value = 0.55) to be poor predictors, whereas Pokémon level was an extremely good predictor for the likelihood of receiving Candy XL (z = 34.13, p-value < 0.001).

² The interaction of this feature with other Candy XL generation mechanisms produces different probabilities which will be investigated in a later article. Trade evolutions will always generate a chance at a Candy XL at no cost for two-stage families (e.g. Shelmet), and share the same standard level breakpoints for three-stage families (e.g. Machop).