Help with Search System Using Elasticsearch, Redis, and Weighted Round Robin

Hey everyone! I'm creating a search system for an e-commerce platform and could really use some advice. I'm still learning, so please bear with me. Here's what I'm working on:

I want to integrate Elasticsearch for searching products and Redis for caching those results to speed things up. My goal is to implement a Weighted Round Robin (WRR) system that balances exposure between sponsored products (paid) and non-sponsored ones (regular listings). Here's how I plan to set it up:

- I want to show 70 products per page, out of which 15 will be sponsored from different Elasticsearch indices and the rest will be non-sponsored.
- I intend to create separate WRR pools for sponsored and non-sponsored products to manage their display.

For calculating weights to prioritize products, I'm taking into account product reviews, total sales, and seller feedback. My weight calculation formula looks like this:
Weight = 0.5 * (1 + log(productPositiveFeedback)) + 0.3 * (1 + log(totalProductSell)) + 0.2 * (1 + log(sellerFeedback))

To keep things balanced and allow newer sellers to compete, I'm capping the weight. For example, if the weight exceeds 10, it's counted as 11, and weights over 100 get capped too.

I also have several questions:
1. Does my weight calculation and capping approach sound reasonable?
2. Is my Redis and Elasticsearch setup effective? How many products should I cache in Redis to avoid overloading it?
3. Will my weight calculation interfere with how products are displayed within categories?
4. Could a search for "GTA" pull products that shouldn't be prioritized since they come from a "GTA 5" search stored in Redis?
5. Where should I implement the WRR logic, in Elasticsearch or Redis? What do you think?

I appreciate any help you all can give!