Simulate shield destruction #33
Comments
ghost
assigned 
|
Gathering data like hell to find out what is the logic behind... |
|
May I assist? It feels like a 10% chance with L6s from the edge.
|
|
"It feels like" is unfortunately not really enough. I need to find a way to 2013/6/20 tonyndory [email protected]
|
|
I will try to record some data.
|
|
Any progress on data collection for XMPs and US? |
|
No, I have basically been using my subjective experience as my guide.
|
|
Xosofox, how does one get accurate numbers in order to help you with this? |
|
The only "reliable" way is to intercept the traffic between client and 2015-04-01 19:00 GMT+02:00 Sir Woody Hackswell [email protected]:
|
|
It's really not that hard to do - configure your Android device to talk to a common SSL MITM proxy like Charles, and you can spy on the client/server communication. It'll somewhat increase your server lag, but otherwise have no negative effects. I'm not sure it would violate the ToS - at least reading it, I cannot see any paragraph it would clearly violate, but IANAL. It doesn't seem like the needed information is within the definition of "Content", but if it is then you would run afoul of paragraph 2(d). Paragraph 2(c) is hard to say either way, as it is phrased so broadly one could argue Project Loon is the only compliant way to use Ingress, since all the cell networks involve access through technology not provided by Google. I don't see how using a proxy server you own is any worse than using your own wifi hotspot. |
But reverse-engineering the algorithm that calculates the shield-pop-probability is another story |
|
Hard to know how tough that is until you have the data, and run it through some statistical analysis. Unless someone has tried and failed, we just don't know. If they have, I'd like to see the anonymized data published so more people can try to analyze it. It could be a simple probability calculated in an obvious manner from the "stickiness" property of a shield, or it could be something far more complex like shields accumulating damage over the lifetime the mod is on the portal. If it's the former, a few hundred samples per weapon and shield combination could be enough to figure it out. If it's the latter, it might require controlled experiments with viruses and multiple agents firing to get the necessary data. P.S. I did try collecting the data myself manually for awhile, but I discovered I'm a little too lazy to write it down often enough to be useful, and then they changed the stickiness rules in the middle of my data collection, so I gave up. |
When attacking a portal in game, there is a chance of destroying the shields with each XMP used. Is it possible to incorporate this into the simulator? Testing would need to be done to discover the percentage chance bursters have for destroying shields, but this would be incredibly useful.
The text was updated successfully, but these errors were encountered: