I submit a lot of 2 and 3 leg option orders. Sometimes the bid/ask on my submitted combo is wide and if I can find part of my order sitting on the book with a much tighter bid/ask spread, I might be willing to take some execution risk in exchange for getting a better fill. Ideally, I would do this with software and not manually. If there is a benefit to doing that, then second idea would be to scan multiple combo order books and look for low risk, close to arbitrage opportunities. Again, probably only makes sense to do this if the approach can be fully automated as there are bots with exotic networking infrastructure for ultra low-latency connections to multiple exchanges. So I expect all the pure arbitrage opportunities are gone, but there might still be some near arbitrage opportunities left to exploit. Either way, I would need to be able to access the data through an API otherwise, probably not worth the effort.