Interested in joining the Silph Research group? Learn More »

Silph Study: #058


The Lucky Trade Rate Caps at 20%

With the sixth year of Pokémon GO coming to a close this week, we thought it would be a great time to revisit the only game mechanic with an age-related component: Lucky Trades.

Lucky Pokémon require half the stardust to power up compared to normal Pokémon, making them highly desirable for many travelers. Travelers trading two freshly caught Pokémon can expect one Lucky Trade in every 20 trades (5%) over the long run. We’ve previously shown that the only action that can increase the probability of having a Lucky Trade (other than earning guaranteed Lucky Trades by becoming Lucky Friends) is storing a Pokémon for more than one year before trading it. For each year that you keep the Pokémon, the probability of a Lucky Trade increases by 5%. Trading two Pokémon that are more than one year old can increase the probability even more.

Before this study, it was unknown whether the Lucky Rate could increase all the way to 100% or if it levels off for very old Pokémon. If there is a cap, how old do the Pokémon need to be? Researchers recorded data for more than 9,000 trades where at least one Pokémon was more than one year old in order to crack this mystery. Let’s see what they found!

Key Finding

The Lucky Trade Rate caps at 20% for two Pokémon with a total age of 3 years.

Lucky Trade Rate Cap

Researchers traded Pokémon of various ages in order to compare Lucky Rates for different age combinations. As we found previously, the Lucky Rate appeared to be the same for each group of trades with the same total age. We were therefore able to combine data for trades between new Pokémon and 2-year-old Pokémon and between 1-year-old Pokémon and 1-year-old Pokémon, and so on. Researchers completed 2,893 trades where the two Pokémon had a total age of 1 year, 2,162 trades with a total age of 2 years, 2,411 trades with a total age of 3 years, 954 trades with a total age of 4 years, and 728 trades with a total age of 5 years. The plot below shows the Lucky Rate we found for each of these age combinations and their associated 95% confidence intervals.

Our data were consistent with a Lucky Rate cap at 20% when the combined age of the trade Pokémon was 3 or more years (chi-squared goodness-of-fit test, χ²(5)=12.9, p = 0.024). We also considered various other models, but all were inconsistent with the data.¹ (Note that the 20% cap does not apply to the 10 guaranteed trades involving Pokémon caught in July or August 2016.) Therefore, a Pokémon reaches its maximum lucky chance once it turns three years old, and there is no more benefit to be gained by keeping it longer, as this won’t increase your chance of a Lucky Trade.

Bonus: Do two half-year Pokémon have a higher lucky rate?

Thus far, our total age models for Lucky Trades have used a Pokémon’s age rounded down to the nearest year. Following our previous publication, we received a question concerning how the game might handle fractional ages. Namely, if the age in days of two Pokémon total to more than a year, would that contribute to a higher Lucky Rate? Or does the game disregard time spans less than a year altogether?

To test this question, we looked at trades where the two Pokémon were both less than a year old, but whose combined ages added to at least 366 days. For example, we traded one Pokémon aged 100 days with another aged 300 days for a total age of 400 days. If the total age included days, then we would expect the Lucky Rate of these trades to be 10%. If the total age disregards days and only considers full years, then we would expect the rate to be 5%.

The overall Lucky Rate for these trades was 4.84% (N=723, 95% CI [4.74 – 6.59%]), which is around what we would expect from trading freshly caught Pokémon. This indicates that only full years matter when adding the age of Pokémon.² Make sure to keep Pokémon for the full 365 days if you want to reap benefits of a well-aged trade!

Parting Words

These results show that keeping old Pokémon for more than three years does not further increase the chances for a Lucky Trade. In order to maximize Lucky Trades for rare species, a three-year-old Pokémon should be traded for a newly-caught Pokémon instead of another old Pokémon. Similarly, the most efficient trade for a two-year-old Pokémon is a one-year-old Pokémon.

We hope that these results can help travelers make informed trading and Pokémon storage decisions in regards to Lucky Trades. May the odds be ever in your favor, travelers!


Authors: Scientists Titleist, NSM, and CaroKann
Analysis: Scientists CaroKann and Titleist
Graphics and data presentation: Scientists Titleist and WoodWoseWulf
Editing: Scientist Cham1nade and Lead Researcher JinianD
Project Leaders: NSM, Cham1nade, and RDC-DCIfan
Data collection: 250 Researchers traded Pokémon with friends and enemies for this study. These 12 were especially social:

  • amyrobynne
  • aniwol
  • AquaTaq
  • bs4u
  • constituent
  • JinianD
  • JokingJohn
  • LT Deadpool
  • marinaj
  • NSM
  • Rachies
  • RDC-DCIfan

¹ Using a Chi-squared goodness-of-fit test we evaluated various total age predictive models on Lucky Trade outcome. Of the models we considered, the stepped model of 5% Lucky Rate growth per year capped at 20% provided the best fit to our data (χ²(5)=12.9, p = 0.024). At our corrected significance threshold of 0.0083, we were able to eliminate other models at various caps, namely a 15% cap (χ²(5) = 136.7, p < 0.001), 25% cap (χ²(5) = 26.1, p < 0.001), no cap (χ²(5) = 53.4, p < 0.001). In addition, we ruled out a logarithmic growth model using total age plus one (a = 0.0505, b = 0.1029, χ²(5) = 19.0, p = 0.002) and an increasing exponential decay model using total age plus one (C = 0.9419, k = -0.0558, χ²(5) = 41.3, p < 0.001).

² We also considered a modified case where a Pokémon’s age was rounded to a half-year (6 months) before factoring into the combined age equation. To test this, we looked at trades where both Pokémon ages were between 183 days and 365 days (exclusive). The overall Lucky Rate for these trades was 4.82% (N=602, 95% CI [3.32 – 6.75%]). This likely indicates that ages between 6 months to a year are always rounded down to 0 years before calculating the combined age.


Below is a complete table of results. Due to the limited observations in the 6+ year trades, the models discussed in this article only considered trades where the total age was between 0 and 5 years.

Total Age Total Trades Luckies Rate 95% Confidence Interval
0 Years 133,230 6,735 5.06% 4.94 – 5.17%
1 Years 2,893 319 11.03% 9.89 – 12.17%
2 Years 2,162 346 16.00% 14.46 – 17.55%
3 Years 2,411 519 21.53% 19.89 – 23.17%
4 Years 954 213 22.33% 19.68 – 24.97%
5 Years 728 140 19.23% 16.37 – 22.09%
6 Years 56 10 17.86% 7.83 – 27.89%
7 Years 17 6 35.29% 12.58 – 58.01%
8 Years 7 0 0.00% N/A