I will preface this by saying I am not a NOAA data scientist, my data science background is far more applied and grounded in the real world than they ever hope to be.
So the NOAA original numbers:

FL Reef Survey:

So after considerable outcry, NOAA crunched the numbers again:

So I became curious about their model and data since it has so many people riled up.
The MRIP data is available to the public with the warning that it is raw data blah blah blah.
So the data for 2023 we have 734 samples for Gag's, of which only 235 samples were taken within grouper season. Of those 235, 56 were actually offshore, IE greater than 3 miles so roughly 8% of the dataset is actually relevant to the real world. Not looking real good to begin with.
Moving on, I'm going to ignore the fact the majority of the data collected was out of season etc.
Out of that dataset and looking at unadjusted Catch A which was:

We have a grand total of 194 fish caught in 81 sample events which averages a whopping 2.4 fish per sample.
Moving on to the second part of their statistical equation B1, which is:

Which comes up to 0
Which is funny because even when its 0, the model still assigns weighted values for statistical calculation, so I wondered where that those numbers are calculated from, it comes from releases:

So even when we are good sportsmen and release our out of season fish, statistically we are still counting towards a catch limit.
Speaking of a catch limit, 106000 lbs caught by shore anglers according to the model, right?
That was extrapolated out because one guy got lucky and caught a 3kg keeper in Region 3, doesnt matter that the average catch rate of a keeper from shore was 0.022.
So going back to the original math, 1.7 million lbs total caught:

Based on data that amounts to an average size grouper caught was 159.767mm or about 6.5ins, whcih is really odd since the average weight is 1.07 kgs which is almost 2.5lbs.
Which seems a bit odd, dont you think?
So breaking it down per month....

So Jan-April claimed harvest is adjusted to 0 but yet there is claimed harvest in the unadjusted counts than amounts to 0 length and weight which is strange enough, but we have catch data for May, June, July, and August when the season is closed. In fact the highest average weight was in May, so where is this coming from in the harvest data? The season was closed, are the people doing MRIP so dumb they dont even know what species they are looking at and just filling in blanks? How is the average size grouper caught in August 44mm? Thats 1.7in on AVERAGE and weighs 0.397kg (0.875lbs) on Average.
So the conclusion that I come to is this, the model they use in SAS may be well enough done to calculate harvests, I don't know, I never even looked at the model since the input data is GARBAGE. Frankly it is ridiculous to even try to use this data for anything worthwhile. I would be ashamed to even have my name attached to it, I'm sure the NOAA will say we used the best available dataset which is junk.
I will say this one of my takeaways from this boils down to this realization, if you are being surveyed for MRIP, do not mention any short grouper you catch out of season, in fact dont mention the shorts at all even during season. If we are being penalized for catching shorts that are being released, the best thing you can do is not mention it at all.
Sent from my Pixel 8 Pro using Tapatalk