I agree with the intentions of this post although I disagree that ECM should negate seismic even though TT says so. As far as how to implement the way seismic sensors work all that really needs to be determined is just how accurate or sensitive they really are. For this we could look at the abilities of real seismic sensors to get an idea. We can also use the ways these sensors are affected or degraded to similar conditions to get a more accurate idea of what might be a reasonable starting point for the ones represented in game. The reason I disagree with the ECM thing is also based on reality. Why exactly they gave ECM this property in TT I don't know. Perhaps it was best tool they had to balance seismic sensors, maybe they felt ECM needed a buff, or perhaps they just didn't quite understand how seismic sensors work. They work by using a sensor that detects vibrations in the ground. These vibrations are created by the ground being impacted by some object. The greater the mass of the object and it's velocity the stronger the vibration. How does ECM prevent vibrations from traveling thru the earth? How does ECM affect the ability of a sensor to detect these vibrations when the sensor is either physically touching the earth or the mech in such a way as to register residual vibration thru the mechs frame? If neither of those are properties given ECM and we know that ECM doesn't disrupt a mechs internal ability to communicate information within itself it is not possible for ECM to nullify the effects of seismic sensors. If we were to go that route we might as well pretty much just remove them from the game altogether given the popularity of ECM. I could see limiting their effectiveness in certain environmental conditions as well, such as the volcano map lot of seismic activity in areas like that. The large pieces of falling ice in the cave on frozen city would probably send off frequent phantom signals that would make them less useful within a certain area perhaps next to useless within the cave itself. They do need looked at but with a bit of thought there are plenty of counters for them. We just gotta make wise adjustments.
Balancing Seismic Sensors That Make Scouting Viable
Started by James The Fox Dixon, Jul 25 2013 07:38 AM
26 replies to this topic
#21
Posted 27 July 2013 - 09:21 AM
#22
Posted 27 July 2013 - 09:55 AM
The reason why ECM disrupts seismic is that the sensors are deployed above ground and transmits a signal to a base station that receives it through a wireless connection. The base station transmits the information it receives to the mech that is equipped with a seismic module through a wireless connection. ECM disrupts all communications, so this is why it should counter seismic modules.
PGI implemented seismic wrong since according to the tabletop rules it requires special equipment to gather the signal.. The equipment that can pick up the signal, according to the rules, is the following: Active Probe, C3 Master Computer, Improved C3 Computer, Command Console, or Communications Equipment.
PGI implemented seismic wrong since according to the tabletop rules it requires special equipment to gather the signal.. The equipment that can pick up the signal, according to the rules, is the following: Active Probe, C3 Master Computer, Improved C3 Computer, Command Console, or Communications Equipment.
#23
Posted 29 July 2013 - 06:27 AM
Slight bump.
#24
Posted 29 July 2013 - 08:28 AM
All good ideas. I think seismic should not show "pings" however, but should show the direction from which the vibrations are emanating. The more intense the vibration(s), the more intense the signal. That way you can tell if something is moving in an area, but NOT be able to ascertain the number of enemy 'mechs nor their EXACT location.
#25
Posted 02 August 2013 - 06:58 AM
Hotthedd, on 29 July 2013 - 08:28 AM, said:
All good ideas. I think seismic should not show "pings" however, but should show the direction from which the vibrations are emanating. The more intense the vibration(s), the more intense the signal. That way you can tell if something is moving in an area, but NOT be able to ascertain the number of enemy 'mechs nor their EXACT location.
I agree with this.
#26
Posted 02 August 2013 - 11:05 AM
This is a balanced, rational, and reasonable solution to the seismic sensor issue. Well done.
I would also like to see it affected by speed in some fashion; a 35 tonner making anchor turns and jumping makes a lot more seismic 'thump' than a 55 tonner moving slow.
I would also like to see it affected by speed in some fashion; a 35 tonner making anchor turns and jumping makes a lot more seismic 'thump' than a 55 tonner moving slow.
#27
Posted 02 August 2013 - 11:35 AM
Malleus011, on 02 August 2013 - 11:05 AM, said:
This is a balanced, rational, and reasonable solution to the seismic sensor issue. Well done.
I would also like to see it affected by speed in some fashion; a 35 tonner making anchor turns and jumping makes a lot more seismic 'thump' than a 55 tonner moving slow.
I would also like to see it affected by speed in some fashion; a 35 tonner making anchor turns and jumping makes a lot more seismic 'thump' than a 55 tonner moving slow.
Thank ya, I'm glad that you like it.
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users















