The Sea Pony, based on anecdotal evidence, seems to have much greater variation in its drop rate than most drops seen up to this time. Furthermore, there are an unusual number of anecdotes from players on this and many other sites citing extremely protracted drop rates on mains and surprisingly quick drops on alts - often serially.
Based on these anecdotes, my hypothesis is: the drop rate for the Sea Pony is scaled negatively against certain character statistics. Possible candidates that are relevant, objective, and consistent with this general pattern include:
- /played time
- Fishing skill
- Total fish caught
- Player level
Heading off the inevitable "random loot is random," I would point out it is not always that simple, and the WoW Dev Team has demonstrated an increased preference for curving over the years. Quest items, for example, used to have drop rates based on simple probability; today, the chance that most quest items will drop off applicable mobs scales directly with the number of applicable mobs previously looted. Although it is provably untrue that past results influence future performance, it is human nature to so believe, and it may well be good game design to corroborate this human attitude.
Rather than be content with what is a simple assumption, I think it most intelligent to collect data to corroborate or debunk my hypothesis.
To my knowledge, no automated tool exists to simultaneously collect and collate said data regarding character statistics and drop rates. Therefore, I can think of no better approach than for interested parties to simply post their data in this thread. From this, perhaps we can notice patterns (or the absence thereof).