EDIT 2024-08-26:

This post prompted fellow FFXIV fisher Tyo'to Tayuun to share some data which directly contradicts the conclusion of this post. You can find the writeup on his blog which I have linked below:

Surface Slap

The evidence he has provided is much stronger than what I have shared in this post. For the sake of transparency I will leave this post up, but I would highly suggest disregarding what I wrote here in favour of the stronger evidence. Either way, I'm happy to be proven wrong and to finally get some answers about this subject.

Introduction

Hi, I'm Kiku. In 2019 (late Stormblood) I wrote a blog post on the Lodestone explaining various fishing mechanics in FFXIV. At that time, there were not many resources related to fishing readily available to the public, and my goal was to make knowledge of fishing mechanics more easily accesible for players interested in fishing. When Shadowbringers came out, I gradually updated the guide with information about the newly added skills at that time, up until 2021.

I'm writing this post in 2024 and the availability of quality fishing resources is much better, which is partly why I haven't updated my guide for changes released in Endwalker. However, one thing I'd like to address in this post and for which I caught a lot of flak for when I first wrote it in 2021 was this line from Part 2 of my fishing guide:

Some people weren't happy with me for claiming (without evidence, admittedly) that Surface Slap does not influence bite rates for big fish. The popular sentiment I see in fishing discussions is that, when targeting a big fish, it is advantageous to use Surface Slap because it would raise the bite rate of the big fish by removing the slapped fish from the pool. On the surface, this seems like a reasonable interpretation of the skill's tooltip. However, the reason I have never removed the above quote from my fishing guide is because, contrary to the prevailing opinion, I personally do not believe that Surface Slap increases the bite rate of big fish apart from a singular edge case (The Unconditional). To clarify, I believe that the developers intentionally programmed Surface Slap to not have an effect on modifying big fish bite rates unless a big fish is the only remaining option in the pool after using Surface Slap.

It's think it's fair to criticize me for making unsubstantiated claims, but in the 5 years since I wrote that line I have not seen a single person provide any sort of hard evidence or data to disprove what I originally claimed. Sure, to some degree the burden is on me to provide proof of my claims, but I don't think it is unreasonable to verify what is popularly believed to be true.

Today, I am writing this post to share that I have finally conducted my own tests on Surface Slap and I will be documenting the results of these tests below. Truthfully, I have been wanting to conduct a study on Surface Slap since the end of Shadowbringers, but I have had neither the time nor the motivation to finish this project until now (only a few days before Dawntrail releases, I might add).

Testing methodology

I conducted a study on big fish bite rates by examining bite rates at a single fishing hole under the following conditions.

1. The big fish must be a straight catch and have 100% uptime to make it easier to collect the data.
2. Bites while Surface Slap is active will not be hooked so that Surface Slap need not be constantly reapplied.
3. The big fish must be the only !!! bite in the pool to be able to identify it without catching it and breaking Surface Slap.
4. A minimum of 1000 casts with and without Surface Slap active to provide a decent sample size.
5. A fish will be picked after determining the ratio of each fish in the pool, and Surface Slap will be used on this and only this fish.
6. Chum and Hi-Cordials will be used to speed up data collection.

These conditions allow me to estimate the baseline bite rate of each fish in the fishing hole without Surface Slap. After establishing the baseline, I will check the bite rates while Surface Slap is active and see if the bite rate of the big fish is significantly impacted by Surface Slap using statistical analysis methods.

I chose The Gambler at the fishing hole Shirogane for this study. Apart from fulfilling the first condition above, there isn't anything special about this fishing hole other than the fact that my player estate is located at Shirogane and it is convenient for me to access it.

Results of 1034 casts without Surface Slap ( Raw data here )

Individual breakdown

Some bites went uncaught either to due the fish escaping, or me failing to press hook. These failed casts are included in the dataset to more accurately represent the bite ratios. The reason I did over 1000 casts was because I selected a fish to Surface Slap only after reaching 1000 casts, and it was necessary to catch it a few times to reapply Surface Slap due to hitting amiss.

Results of 1000 casts with Surface Slap used on Glass Tuna ( Raw data here )

Glass Tuna was selected as the fish to Surface Slap. Supposing that Surface Slap does have an effect on big fish bite rate and that the effect is proportional to the Surface Slap target's ratio, I chose the fish which had the highest proportion in the fishing hole. Looking at the data, I suspect that Striped Fugu has an equal bite rate to that of Glass Tuna, but it is not terribly important which fish is chosen as the Surface Slap target as long as it is consistent for all 1000 casts.

Analysis of results

As you can see, a 0.38% difference in the legendary bite ratio is not terribly high, which already suggests that Surface Slap made little to no difference. However, we can use statistics to make a stronger statement.

I apologize in advance for any mathematical errors in this section, as it has been quite a while since I've touched upon statistics in my personal life. Regardless, it is beyond the scope of this post to explain the reasoning behind the formulae used; please verify on your own if you do not trust my calculations. I will do my best to try to briefly explain some concepts for those without any knowledge of statistics.

Simply put, we want to determine whether or not the two samples differ significantly based on whether or not Surface Slap is used. We will start with what is called a "null hypothesis" and see if the data is strong enough to reject it. The null hypothesis for this analysis is:

Using the numbers from the dataset, we can calculate something called a z-score; if the value of the z-score is between a certain range, then we cannot reject the null hypothesis. We will set a 95% confidence level for our test, so the range will be from -1.96 to 1.96. The formula used for calculating the z-score is as follows:

We'll substitute the variables using our data as follows:

${\displaystyle {\hat{p}}_1}$ is the proportion of population 1: 0.0542 (rounded here, actual value used for the final calculation is 56/1034)

${\displaystyle {\hat{p}}_2}$ is the proportion of population 2: 0.058

${\displaystyle {\hat{n}}_1}$ is the sample size of population 1: 1034

${\displaystyle {\hat{p}}_2}$ is the sample size of population 2: 1000

${\displaystyle \hat{p}}$ is the overall sample proportion calculated by this formula:

Where:

${\displaystyle Y_1}$ is the actual number observed in population 1: 56

${\displaystyle Y_2}$ is the actual number observed in population 2: 58

This gives us:

Which equals: 0.056 (rounded here, actual value used in the final calculation is 114/2034)

Plugging all the values into our first equation gives us this:

Which gives us a z-score of -0.3765. Because our z-score falls between the range (-1.96 < -0.3765 < 1.96 ) as determined by our selected confidence level, we cannot reject the null hypothesis.

So what does this mean?

The analysis suggests that no significant difference in bite rate for the big fish was observed between the two populations when Surface Slap was used. This does not necessarily mean that the the null hypothesis is true, only that the evidence did not prove the null hypothesis to be false. At the very least, the test did not demonstrate any connection between Surface Slap and big fish bite rates, which lends some credence to my original claim.

Conclusion

Let's be clear about something: This post is not about attacking anyone who believes that Surface Slap works the way the tooltip advertises. I put together this post out of my own personal interest in the subject and to finally have some data to back up my original claim.

I am also not saying that you should never use Surface Slap when hunting a big fish. If you re-read the original quote from the guide, I suggest a strategy of using Surface Slap to potentially increase the number of attempts during a limited window. If there's anything to take away from this post, it's that there's no sense in just using Surface Slap on any random target for the sake of "thinning out the pool".

I'm sure there will be some who will read this post and disagree with the evidence and my conclusions. I encourage everyone to perform their own trials and try to replicate the results shown here, especially in other fishing holes. I highly doubt it, but there could be a chance that the developers made Shirogane a special fishing hole that interacts with Surface Slap differently. I'm joking, but it's always good to have more data.

A counterargument from the armchair developers that I don't want to hear is something along the lines of "it would be harder to code Surface Slap to work a certain way". First of all, none of us here has or ever will lay eyes on the backend code, so the presumed difficulty of such an implementation is entirely speculative. Secondly, anyone who has spent any meaningful amount of time in professional development would know that if the business logic dictates that something must behave a certain way, whether it would be more difficult to implement it that way is largely irrelevant unless there is some kind of unsurmountable technical challenge. Lastly, my personal assessment is that it would not be terribly difficult to implement some kind of check in the code to maintain a target bite rate for Surface Slap; the solution would likely not be trivial, but I do think it is rather embarrassing for any competent developer to claim that such a check would be too difficult to code.

Preemptively addressing some questions