[Formula] Ping Efficiency

[FuriousFang]

Hey Everyone!

I made something up since I can't code at my laptop atm. Basically, it tells people the efficiency of your server. Traditional Efficiency is always under 30% (energy is released into the environment- or in this case, ping is used to do processes). By multiplying *Ping out/Ping in* by 100, you get the efficiency of your processes. We can't find the actual percent error unless you make a test in the server saying the arrival time - the executed time.

Formula:

(Ping In / Ping Out) * 100

Examples:
-----------------
Let's say your ping from pingtest.net is 12ms (like mine ). Well, you'd log into your server, take a reading, and plug the numbers into the formula.

(12 / 32) * 100 = 37.5%

Use of this thread: You can compare the efficiency of sources to see which one is more efficient than the other. (ex. Elite-CoEmu vs. CoEmu Nano OR Impulse vs. Hybrid) Try it out! =]

Sincerely,
Fang

EDIT:
#request close. Proven to be inaccurate.
That's why I'm not a math major! xP

[Syst3m_W1z4rd]

i don't get it.

[bone-you]

That's not making much sense. Ping in / ping out? What do those mean?

[FuriousFang]

Ok, let's do it together =P

1. Turn on your server (hosted at your computer to reduce network lag and to produce accurate results).

2. Run a Ping Test at ... This is your ping in (ping going into the server- now I know that ping doesn't go into the server but this is where ping is generated. It's the server's response time).

3. Directly after the results are displayed, record the ping from the Conquer client. It's right next to FRPS. That is your ping out (from the server). Now use my formula and u'll get efficiency.

Before you flame btw, you might say that it's the opposite since the real formula is (energy out/energy in) * 100, but since ping is generated inside the environment (the server) instead of outside, the fraction is to the -1. (I'm not a math major btw... I just took Physics. This math isn't really that hard). =]

[Korvacs]

Ok, so according to your calculation my efficiency is 103%

[ChingChong23]

thats not correct.

simply host a server on your local server, login and see your ping in conquer, this can give you a rough idea on the latency of your server (i get 20-30 ping, but thats because my incoming packets get processed every 20ms) if i were to handle them as soon as they arrive i assume i'd be seeing 5ms or so.

[Korvacs]

Originally Posted by ChingChong23

thats not correct.

simply host a server on your local server, login and see your ping in conquer, this can give you a rough idea on the latency of your server (i get 20-30 ping, but thats because my incoming packets get processed every 20ms) if i were to handle them as soon as they arrive i assume i'd be seeing 5ms or so.

5ms? Yeah that wouldnt be anywhere near the actual value. A standard ping contains no data and is handled by the network card thats why you see such low values for pings. But the conquer client isnt actually pinging, its sending a packet to a game server, they are entirely different systems, and therefore will have entirely different results.

You should try handling them as soon as they arrive, i suspect you will be disappointed.

[bone-you]

"Pings" are actually handled by windows but yes, pings in games usually are standard packets that just compare timestamps to show how fast your packets are processed, not exactly latency to the server... at least in the case of these slow C# servers. When pinging it is roundtrip btw because there is no perfectly accurate way to figure out one way time because both computers will have a different time even if by a few milliseconds which will throw off your results which is why a timestamp sent by one computer is just sent back to detect the time it takes and is cut in half = average travel time.

[†he Knight]

Wrong Section?

[FuriousFang]

Idk. I thought it might just be a way to compare sources.

[bone-you]

If you're looking to compare performance then use timers in the server themselves to gauge performance of the individual functions and such. A lot of people on will degrade ping performance on any kind of server (even if not noticeable) so that's not a very accurate display.

[FuriousFang]

Originally Posted by bone-you

If you're looking to compare performance then use timers in the server themselves to gauge performance of the individual functions and such. A lot of people on will degrade ping performance on any kind of server (even if not noticeable) so that's not a very accurate display.

Yah, I guess you're right. This isn't a very accurate test.

[InfamousNoone]

Plug it into the Quadradic Formula and good2go imo.

[KraHen]

Big O notation - Wikipedia, the free encyclopedia

[FuriousFang]

Originally Posted by InfamousNoone

Plug it into the Quadradic Formula and good2go imo.

Ah ha xP
Nice.
#request close. Proven to be inaccurate.