Interested in joining the Silph Research group? Learn More »

Silph Study: #002

Published:
10.14.2016

Throw Bonuses in Pokemon GO

Anecdotal reports have long hinted that earning a throw bonus by aiming a ball inside the reticle (circle) improves catch rate, but statistically significant data has been lacking.

The Silph Research group collected a statistically significant dataset (over 4,500 samples) of highly controlled throw data. Researchers tossed a single Poke Ball (repeated if necessary until first contact with the wild Pokemon was made) and recorded their catch success or failure, alongside any catch bonuses. (E.g. Nice, Great, Excellent)

Our analysis has found that earning a throw bonus DOES INCREASE your catch rate.

Variables Controlled For
  1. Only samples collected before the recent Type Catch Bonus mechanic launched were studied
  2. This dataset only includes balls that were thrown straight (See Note 1)
  3. This study was only conducted on Poke Balls
  4. This study excluded Pokemon caught via Incense or Lure Patches
  5. Only Pokemon whose base catch rate is .4 were included (e.g. Rattata, Pidgey, etc)
  6. We used % of Max CP per species as a basic way to estimate Pokemon level (See Note 2)

The Null Hypothesis Vs. Results

The null hypothesis is that all throws have the same catch chance, regardless of getting the throw bonus, or H0: μ1 = μ2 = μ3 = μ4 where μ1 = Normal, μ2 = Nice, μ3 = Great and μ4 = Excellent. With calculated rates of:

Throw Bonus Catch Rate Sample Size Margin of error
Normal* 53.20% 1158 2.88
Nice 57.04% 419 4.79
Great 63.79% 417 4.80
Excellent 69.77% 43 14.94

With those values we are able to reject the null hypothesis in favor of the alternative with a confidence level of 95%. The only ones we cant be fully confident of rejecting is μ3 = μ4 and μ2 = μ4 as the sample size is too small, however it fits the trend and we can reject μ4 = μ1.

* The Normal throw dataset contains data from all ring types, meaning it's average is likely smaller than the maximum ring radius. The disparity between Nice (the largest/full circle) and the random distribution Normal which was undoubtedly a smaller average circle radius, points to the Nice bonus improving catch rates more than the presumed smaller average circle in the Normal sub-dataset. (See Note 3)


Notes

Note 1:

The Silph Research group focused on Straight balls only in this analysis, as different ball types are known to alter capture rates and Curve throw bonuses are anecdotally reported to as well. Curve throw bonuses are the subject of a separate, ongoing study.

Note 2:

A common question regarding this analysis is whether Pokemon Level impacted the throw bonus catch rate percentage.

Pokemon Level does impact catch rate. But fortunately, each throw bonus sample group had a roughly equal distribution of Pokemon levels (roughly approximated by % of Max CP).

The frequency distributions of each throw type bonus can be seen in the following bar chart. Only a few players are able to catch 60-70% and 70-80% mon so that is the reason for their section of each bar being so small. All of the other sections are quite evenly distributed.

Note 3:

It remains possible that there is a separate bonus for timing a throw to when the colored ring is small, regardless of whether the ball hits inside the ring or not. However, since ring size was not recorded for "Non-Bonus/Normal" throws, this dataset's average ring size is most likely even smaller than the Nice dataset's ring size. Despite this, the Nice dataset observed a significant catch rate bonus regardless. Due to this, we believe that the bonus itself is increasing the catch rate, not simply the ring size.


Publication

This finding was shared on our subreddit on Oct. 14, 2016.


Notable Researchers

Contributions to this project were made by many Silph Researchers, but we want to highlight the top 5 contributors to this research for their selfless dedication to furthering our understanding of Throw Bonus mechanics:

  1. CuttyWow
  2. vato915
  3. azindrag0n
  4. iCookiees
  5. charchar_duck