Yeah rake does seem to make sense the more I hear from users. I'd still make it dynamic with a cap.
I do want to mention one more thing. BTC poker is slim pickings. Recently someone well known in poker came up with a BTC site. It wasn't the best but it was nice enough for a v1.0 and it worked and took BTC and had the backing of a mid-range pro players name, and guess what 6 people showed up. I haven't visited it lately but I think my craigslist ads get more hits.
If I was going to hit the BTC poker market today I would do it in one of two ways. Either host it from home or on a hack job to start so it costs nothing to run and can keep floating for years until it takes off - or option 2 make an interface so cutting edge that everyone wants to play it.
I seen it before I even started coding this. I just wanted something with passive revenue and had spare time. All black-hat SEO, mining, and service models that were the alternatives required too much capital or had markets like this but with higher overhead mostly in time since I can do all the work, but too much time. Like grinding out content and websites for niche marketing, putting S7 miners on free electricity when pools are petahash rates, a secure tumbler on a dedicated host(expensive) etc..
I even looked for some overseen edge with low-level blockchain stuff to exploit in mining..
The server I wrote is on a x86 thin-client behind my NAT barely taking up any KWH. I have a parse.com server too if it gets too heavy and the database is portable. I didn't pay with anything but time so far. I wrote most of it in evenings inside a month.
If it doesn't work out I'll turn it in to a intranet kiosk solution and sale licenses..
Bad business model you have developed.You should considered how to implement the revenue model before deploying the application. Advertisements is certainly not a good idea for poker sites.Most of the websites around earn by rake ,you should be doing that as well.
I wanted a poker client either way. If BTC fails I license a kiosk intranet solution or as a WAN solution for a few grand a license. Whatever under-cuts the other 1-2 people already doing it.
