55 replies to this topic

[MustrumRidcully]

HC Harlequin, on 17 January 2013 - 02:07 AM, said:

keep voting


So.. when they fire exactly like SRM's how are they all going to hit? Streaks all hit. Sounds like your way they will miss?

The challenge is implementing it right. The table top rules have it easy here - you roll your to-hit roll, and if you wouldn't hit, you don't fire.

But MW:O is a real time game, so it would basically need to predict whether the missile would hit when you pull the trigger, and if you wouldn't hit, it wouldn't fire. The problem here is that the game cannot predict the enemies evasive actions (intentional or unintentional) - if it stays on its current course, it should be able to predict the hit chance, but if it doesn't... The table top doesn't have this distinction,you can't move out of the way during the attack phase.

But maybe that's not too bad - maybe to compensate for the chance that a false Streak prediction leads to a miss, when the Streaks misfire because of a succesful prediction of a miss, the cooldown is lower than in case of a regular shot?

[Quxudica]

HC Harlequin, on 17 January 2013 - 02:07 AM, said:

keep voting


So.. when they fire exactly like SRM's how are they all going to hit? Streaks all hit. Sounds like your way they will miss?

You must not have read what I wrote, that or didn't understand it. Imagine using regular SRMs on a afk mech standing in front of you, you fire and the missiles hit - just like normal. Now, you move your reticule off the mech and pull the trigger - you'd miss. Wiith streaks, that miss would result in your weapons not firing, no ammo being used and no heat generated, but the cooldown would trigger. Thus the SSRMs would never miss, because they don't fire if they would miss. It's just one idea, and a rough one at that. The point is that streaks currently remove skill from the game when using or facing them, you cannot evade them and you don't need to aim them. Essentially they need to be harder to use to justify being guaranteed damage.

[HC Harlequin]

Quxudica, on 17 January 2013 - 04:52 AM, said:

You must not have read what I wrote, that or didn't understand it. Imagine using regular SRMs on a afk mech standing in front of you, you fire and the missiles hit - just like normal. Now, you move your reticule off the mech and pull the trigger - you'd miss. Wiith streaks, that miss would result in your weapons not firing, no ammo being used and no heat generated, but the cooldown would trigger. Thus the SSRMs would never miss, because they don't fire if they would miss. It's just one idea, and a rough one at that. The point is that streaks currently remove skill from the game when using or facing them, you cannot evade them and you don't need to aim them. Essentially they need to be harder to use to justify being guaranteed damage.

won't work.
see any number of posts regarding ballistics and convergence.
You can't aim at your target to hit. So if the weapon only fires when the reticle gurantees a hit it will always miss because the target will have moved. All the rounds will impact behind the target.
SRM's you have to lead your target by a significant exponent of the hit box both to account for the lag and the ballistic trajectory. If SSRM's are implemented with a client side "guranteed hit" feature then the actual server side trajectory will always be behind the target.

[Helmer]

Excellent Poll. Personally I am avoiding weighing in on ECMs until the "netcode" and ramming are implemented. So avoiding those I see Regional servers, Matchmaking (however they plan to do it, ELO, weight, hybrid) , and more content as being the 3 biggest things that effect the largest cross section of the community.

Matchmaking, and content would improve, almost, everyones experience. Regional servers would only effect those in those regions, however, I feel for those players having to endure high pings. Their experience has been less than desirable.


Improving the New Player experience really should be part of the poll. However that is improved, whether its by a tutorial, a training ground, or matching up with in game volunteers, new player retention effects everyone.


Cheers.

Edited by Helmer, 17 January 2013 - 07:11 PM.

[Merky Merc]

I feel like if the netcode was fixed and lights were consistently shootable, then ECM would also become much less of an issue. It would only make missile boaters grumpy, and I can live with that.

[Quxudica]

HC Harlequin, on 17 January 2013 - 07:03 PM, said:

won't work.
see any number of posts regarding ballistics and convergence.
You can't aim at your target to hit. So if the weapon only fires when the reticle gurantees a hit it will always miss because the target will have moved. All the rounds will impact behind the target.
SRM's you have to lead your target by a significant exponent of the hit box both to account for the lag and the ballistic trajectory. If SSRM's are implemented with a client side "guranteed hit" feature then the actual server side trajectory will always be behind the target.

You took the example far to literally. Regardless, as I said in that post you quoted it was just one rough idea, of which there have been many. The ultimate point is simply that Streaks need to be changed in a significant manner.

[twibs]

This poll is...odd

Why not make multiple choice poll and limit the choices to three. Wold be much MUCH more clearer since I bet ECM is on everyone's list but the place may vary.

trying to read those three polls all together is just confusing.

[HC Harlequin]

twibs, on 17 January 2013 - 08:04 PM, said:

This poll is...odd

Why not make multiple choice poll and limit the choices to three. Wold be much MUCH more clearer since I bet ECM is on everyone's list but the place may vary.

trying to read those three polls all together is just confusing.

Well, it's an extension of an earlier poll located here.
http://mwomercs.com/...-fix-one-thing/
Where it became obvious that the general opinion is netcode first.
Additionally there were comments on that poll and a related poll about having a prioritized list. And you can't limit the choices to 3. it's just multiple choice.

[HC Harlequin]

MustrumRidcully, on 17 January 2013 - 04:41 AM, said:

The challenge is implementing it right. The table top rules have it easy here - you roll your to-hit roll, and if you wouldn't hit, you don't fire.

But MW:O is a real time game, so it would basically need to predict whether the missile would hit when you pull the trigger, and if you wouldn't hit, it wouldn't fire. The problem here is that the game cannot predict the enemies evasive actions (intentional or unintentional) - if it stays on its current course, it should be able to predict the hit chance, but if it doesn't... The table top doesn't have this distinction,you can't move out of the way during the attack phase.

But maybe that's not too bad - maybe to compensate for the chance that a false Streak prediction leads to a miss, when the Streaks misfire because of a succesful prediction of a miss, the cooldown is lower than in case of a regular shot?

Client side hit detection you mean? Either that or a variant of client authentication and forced object orientation. I think that if you introduce a client based automatic damage code it will be too easy to hack. And that's no different than the way streaks are now. If they automatically damage on client side fire then you have magick unseen missiles that you can't dodge or see causing ghost damage on your mech.

[HC Harlequin]

Helmer, on 17 January 2013 - 07:11 PM, said:

Excellent Poll. Personally I am avoiding weighing in on ECMs until the "netcode" and ramming are implemented. So avoiding those I see Regional servers, Matchmaking (however they plan to do it, ELO, weight, hybrid) , and more content as being the 3 biggest things that effect the largest cross section of the community.

Matchmaking, and content would improve, almost, everyones experience. Regional servers would only effect those in those regions, however, I feel for those players having to endure high pings. Their experience has been less than desirable.


Improving the New Player experience really should be part of the poll. However that is improved, whether its by a tutorial, a training ground, or matching up with in game volunteers, new player retention effects everyone.

I don't see a reason to not use all 3 methods of matching. If you first do an ELO comparison then at the end of the match do your ELO adjustments but do it based on a component/tech level percentage then you are essentially using both systems. That is, your end match ELO adjustment is varied based on the comparison between component values. If the winning team has a higher CV than the losing team then the winning team gets a smaller ELO adjustment. If the winning team has a lower CV then it gets a larger ELO adjustment. Higher ELO teams with higher CV will get much reduced adjustments and lower ELO teams with lower CV will get much greater adjustments. And high ELO teams with lower CV will get a higher than normal ELO adjustment and vice versa.
If you add in the ability to choose weight limits on random drops with the obvious repercussion of "Failure to find a match" if the chosen weight limit is unmatchable with available players then you allow teams with light/medium makeups to go up against other light/medium teams which will lend credance to playing the medium mech chassis instead of taking a heavier same task oriented chassis. This, of course, will make the varied light/medium mechs that have the same task orientation as heavy/assault mechs more viable to play.
And I am adding New player support to one of the selections

Edited by HC Harlequin, 17 January 2013 - 08:38 PM.

[anonymous175]

#1-3

[Helmer]

HC Harlequin, on 17 January 2013 - 08:36 PM, said:

I don't see a reason to not use all 3 methods of matching.

That would be the Hybrid option. Incorporating multiple sources of data. Hopefully we see something akin to this in the future.



Cheers.

[King Arthur IV]

is memory leak the same as netcode? like same problem?

Edited by King Arthur IV, 17 January 2013 - 09:25 PM.

[HC Harlequin]

King Arthur IV, on 17 January 2013 - 09:25 PM, said:

is memory leak the same as netcode? like same problem?

No.
The 4 FPS bug/Atari bug is the memory leak generated by changing your mechs in mechlab or being forced to change your mechs in mechlab. The memory leak causes the 4 FPS/Atari bug. This is why you have people that don't change their mechs while pugging that have no problems with the video bug over long periods of play. You can slightly reset your video from the Atari bug by Ctrl/Alt/Del out of the game and on the Windows selection (Start Task Manager, etc etc) hit the Cancel button then reopen your game client. You will still have the 4FPS bug but it will usually reset to the Atari bug in about a minute or so.

[King Arthur IV]

HC Harlequin, on 17 January 2013 - 09:46 PM, said:

No.
The 4 FPS bug/Atari bug is the memory leak generated by changing your mechs in mechlab or being forced to change your mechs in mechlab. The memory leak causes the 4 FPS/Atari bug. This is why you have people that don't change their mechs while pugging that have no problems with the video bug over long periods of play. You can slightly reset your video from the Atari bug by Ctrl/Alt/Del out of the game and on the Windows selection (Start Task Manager, etc etc) hit the Cancel button then reopen your game client. You will still have the 4FPS bug but it will usually reset to the Atari bug in about a minute or so.

that case memory leak is next. i want to play on higher setting but i crash more when i do. currently im playing with low, its alittle sad.

[HC Harlequin]

King Arthur IV, on 17 January 2013 - 09:52 PM, said:

that case memory leak is next. i want to play on higher setting but i crash more when i do. currently im playing with low, its alittle sad.

Don't go into mechlab and don't switch anything out between matches and that will minimize the impact of the memory leak. Additionally, closing and restarting your client periodically (every 4 matches or so) will essentially eradicate the memory leak bug. I personally have never had to clear the cache folder manually. I just restart client, but there are players that will clear their cache folder along with restarting their client.

[Solomon Ward]

twibs, on 17 January 2013 - 08:04 PM, said:

This poll is...odd

Why not make multiple choice poll and limit the choices to three. Wold be much MUCH more clearer since I bet ECM is on everyone's list but the place may vary.

trying to read those three polls all together is just confusing.

ECM is not on my list.

To me Collision is THE game changer and therefore my #1.

After that i want to see pre/post game interface / player lobbies.
I "lost" so many nice and / or good player because i couldn´t remember their name.
Probably the reason i am still pugging 100% of my time.

I want to see regional server but i am fine with them coming further down the road.

[Green Mamba]

Team with the Most Streaks has a huge advantage(I am not talking just about lights) and usually wins.Some rebalancing is needed

[twibs]

HC Harlequin, on 17 January 2013 - 08:21 PM, said:

Well, it's an extension of an earlier poll located here.
http://mwomercs.com/...-fix-one-thing/
Where it became obvious that the general opinion is netcode first.
Additionally there were comments on that poll and a related poll about having a prioritized list. And you can't limit the choices to 3. it's just multiple choice.

well you can always ask nicely for them to pick only three, besides I think the % values would've been indicator enough of what was #1 but maybe that's just me.

[HC Harlequin]

twibs, on 17 January 2013 - 11:16 PM, said:

well you can always ask nicely for them to pick only three, besides I think the % values would've been indicator enough of what was #1 but maybe that's just me.

well. Someone asked that I extrapolate into a more prioritized list. And frankly on the previous list the netcode got to almost 60%