Quote from: Nexus9090 on Today at 08:58:42 AM

Can you prove it?!?

Prove what exactly? that your local node has an order of magnitude more chances to lose the orphan race compared to a well-set-up pool node? The studies are there for you to read. The burden of proof is on you now; make your own studies that show how you propagate blocks to mining nodes faster than a well-established pool.

I have literally told you that large pools exchange compact block header data that is a few KBs over gigabit connections, while your node needs to propagate an entire block that could be as large as 4 MB. How exactly do the two of you have the same propagation speed?

In fact, even without any relay services, having geographically distributed nodes alone beats any single home node; you need to have nodes that are very close to other mining nodes. Relying on a P2P network alone to send a block from your house in LA to a miner node in Kazakhstan is bizarre; another miner that finds a block in Australia but has well-geographically distributed nodes will likely reach other nodes inside the U.S. faster than you.

There exist many tools and projects on GitHub for this particular topic; thousands of lines of code are written just to test and improve propagation time, and millions in infrastructure and research are spent to achieve the best propagation results. You can deny all of it -- it doesn't mean it's not real.

As for the above test, it's rather useless ; it doesn't matter if your ping results to all peers are 1 ms. Those nodes could very well be just home nodes far away from the nodes that matter (miners' nodes). Block propagation time is what matters, you need to compare the time your node received a block vs. when another well-established node received it. Use UpdateTip on your node to see the timestamp of when you received a block, and compare that to when a large explorer like mempool first saw that block. Run this against a large sample of blocks to get a somewhat accurate result, and mind you, other pools must have seen that block even earlier than mempool.

With that said, it doesn't mean your node will lose every block; it simply means it's at a higher risk, if we were to assume that other pools have a 10-second advantage in propagation time, you can run a simple Poisson process of :

1-exp(-10/600) ~= 1.6653% > chance of a block being found in the next 10 seconds or less, if Foundry finds a block within that timestamp and you both start racing, they will still likely beat you

You may think 1% is low, but it does happen. I haven't even touched on the other potential issues, like your PC or router failing you right about that second. Large corps buy 10k routers not because they want to waste money but because they can't afford the slightest delay or packet loss. While you could have a 100% uptime with your $200 laptop and $30 router, it's just not as guaranteed as the other setup. It's not as bad as some people might explain it, but there indeed exists a higher level of risk when you count on your local node to mine blocks.