I do a lot of network related work. And stability > latency.
If you have two players of equal skill, equal mechs, meeting on equal terms. Both have an average ping of 80; Player A fluctuates between 70-90 and Player B fluctuates between 40 and 120.
Player A will have the advantage in ping simply because of stability. This is with nothing like HSR, much like the old skool shooters like CS1.6, DoD, and UT99.
However, in the day and age of HSR and it's like (and often ilk, as the old games had lag leading that was compensated for simply by observing the players ping. Whereas poorly implemented HSR often led to hair pulling at unregistered shots.), players have found ways of utilizing it to their advantage.
HSRs are different, some utilize pure client side hit detection like Halo and COD. Where I hit him out in the open on my screen where on his in the 200ms difference he got to cover to only die, to Halo's infamous host-freezing where many players are spiked while the host machine can freely move about and kill opponents. This method is going out of style because of how easily it is abused.
Others utilize server side detection. Some do it well, like BF.... some... not so well. One of the particular reasons why BF does it well is because it can be set to have a floor and ceiling for latency. Players outside those parameters are either booted upon joining or simply not allowed to join. Games such as MWO and the early WOT days (before regional servers), don't implement a hard floor or ceiling, they use is as a "soft floor/ceiling" and optimize the software for those ranges.
Here's where it gets iffy:
What's the range? What's the average that you should seek?
PGI's telemetry says, "The average player has 'X' latency range." And so PGI optimizes their software for THAT range. Falling outside of it doesn't benefit you. Meaning players with 40-ms and 250+ms tend to notice the not working HSR the most.
The closer you reach the "golden ping" the better you should THEORETICALLY be. I've actually spent a fair bit of time looking in to this over the last two years (with the 6 month exception due to service) and I've noticed several things:
- My ping fluctuates between 30-130, but does so predictably. As well as my tinkering with my own ping to troubleshoot.
- During the morning hours of GMT-6, my ping averages 110 +/- 10ms. The Europeans I play with have similar pings ranging from 100-150ms. I have no issues with hit registration whatsoever, and I don't receive ghost damage.
- During the evening hours of GMT-6, my ping averages 63 +/- 30ms. The Canadians/Americans I play with have pings ranging from 10-150ms, averaging 60ms. Matches where my ping was lower than the targets meant some hits not registering or suddenly mechs dying a second or two after getting hit. Matches where my ping was higher, they had unregistered hits or I received ghost damage.
To test my findings, I throttled my evening ping to match my morning ping, and I noticed a decrease in lost hits, however, the ghost damage I received seemed to pick up to match the number of hit I was making (I was hitting more and getting hit more).
TL:DR; There is an optimum ping for server side hit detection and finding it is as much cheating as going into settings and adjusting settings for the maximum eye-candy/performance ratio.
In essence, I find my best performance sitting at around 100-120.
Cheating would be directional throttling. Which is what happened in the "good ole days" of CS1.6, UT99, and Halo2 and is very possible and easily replicated in MWO with varying degrees of success.
Edited by JohanssenJr, 14 April 2014 - 08:52 PM.