About Oyster Analytics
A Hub for MLB Prospect Projections, Research, and Rankings
Oyster Basics
Oyster Analytics is the passion project of two former college ballplayers who have an unhealthy devotion to baseball. Founded in 2024, Oyster was initially borne out of an idea for a different approach to prospect modeling. Rather than focusing on rankings above all else or making laughably precise projections about individual future stats (let’s be honest, no one can actually say to the point what Luis Peña’s OPS will be in 2031), we wanted to assess players in terms of the distribution of their potential career outcomes. We were also tired of the flaws of regression models, and wanted to create a tool that utilized a more robust approach (we use a system of 32 random forest models).
Since then, Oyster has expanded into a wider array of research. We partner with Down on the Farm Substack to write weekly research articles on all things baseball. We have also worked with our former college program to optimize in-game strategy. We have a few more projects in the works that we’ll be launching over the offseason, when we actually have some time to take a step back and grow our offerings. Stay tuned for many more exciting pearls!
The Oyster Model
We’re biased of course, but we really do think these are the strongest MLB prospect projections out there. But don’t take our word for it, check them out for yourself.
So what makes our projections and rankings so great and different? Well, they’re based on a robust multi-level assemblage of 50 random forest models trained on 15 years of data from all affiliated baseball. These models can capture the complexity of prospect profiles in a way that regression-based models simply cannot. Furthermore, our projection of different levels of MLB productivity is a unique output that allows for an easier understanding of the distributions of the expected future value of a prospect. We are reflexive in our prospect projection, highlighting and learning from our misses, rather than hiding and burying the evidence.
Our Research at Down on the Farm
Since January 2025, we’ve been writing (at least) weekly articles for the Down on the Farm Substack. This is where we go deeper into our prospect evaluations and publish our wider ranging research projects. Some of our favorite research deep dives have been our analyses into how caution is costing teams on the base paths, ABS-related strategy, and where money is best spent in amateur player acquisition. You can also check out our pre-season top 50 hitting prospects, self-audits of our previous hot takes, and Beatles-themed prospect evaluations.
Subscribe to the Substack here, and give Down on the Farm a follow on Twitter and Blue Sky!