442 replies to this topic

[Marcus Tanner]

A few objections to Tolkein's statistics were brought up, and I think they're appropriate enough to deserve some addressing.

OBJECTION: ECM use correlates to player skill and use of premade teams, such that better players use ECM more often. Thus, Tolkein's data reflects player skill and/or the use of teams, confounding the analysis of ECM.

Better players will tend to use the better equipment, all other things equal. So, if ECM use correlates with player skill it means that the best players around will tend to believe that ECM is good to bring. So, even if the data is watered down by this factor it still shows that ECM is better than alternative uses of tonnage/crits.

In short, if the best players around use the thing than that's because it's effective.

---

Tolkein is a good player and he brought ECM in all these tests, so the results will skew toward wins for Tolkein's team and therefore more wins for ECM.

This one actually has merit, as by my count Tolkein's side won most of the matches by a margin of 35:19. That's huge for a set of PUG matches (some players who go out on their own report losing 2/3 of their matches). This effect may be magnified by the occasional 8-player drop where one would expect Tolkein's side (never 8-player) to get stomped.

It may also be relevant that Tolkein appeared to be on a 2-player team, which the matchmaker may often throw in with a 4-player team such that Tolkein may have regularly been playing with 6 people on comms +2 against 4 people on comms +4.

However, despite these problems, Tolkein's side won by a 2:1 margin at best, where ECM-superiority won by a factor of 4:1. The data correlates far more strongly with ECM than it does with Tolkein's skill (or that of Tolkein's wingman).

---

Unfortunately, we don't seem to have any way of getting better data at this point. By all appearances Tolkein has taken every reasonable step to get better data, but unless/until that happens this is what we've got.

[ICEFANG13]

Had some ECM matches today, game is so damn boring now. Dropped in an 8 man, we won one with equal ECM, then lost to 5 D-DCs and 3 3Ls.

Why does this happen? Because ECM is balanced?

[ICEFANG13]

Alright, let me demonstrate what is wrong with ECM.

Ok, everyone, if you could equip ECM on all mechs, what percent of mechs would you not put them on?

I'll say 0%, its such a huge advantage, if I can run it, I would.

[MajorBorris]

Strucker, on 11 January 2013 - 07:16 PM, said:

Not quite sure I'd go that far, to be fair I actually see more issues with missiles incoming these days for the sheer fact that a lot of players are grouping up with either 4 streak raven's / commando's or a combination of the previous 2 mechs with an assortment of streak boating cat's or stalker's.

When you roll those fast build the combination of lagshield and no ability to use 1 or 2 streaks to counter those fast light mechs, it pretty much becomes screwed from the start.

Most of my matches end with me getting swarmed by 4 ECM streak lights and I just DC or run out of bounds, I have no interest in playing in a match when this BS is going on.

Boating streaks was BS, but implementing ECM and removing the most effective means of countering lagshield lights is even more BS.

For those who say just shoot with lasers give me a laser than can counter a light effectively and I will otherwise **** and stop trying to defend your noob ECM builds.

It drives me nuts when noobs play(rage quite) like that, it realy ruins the match.

The reason people loose a match many times is teamates like the above, not the other teams squad build.

Some of you need to spend less time in the forums crying about how someones rock beat there paper and more on the battlefield(fighting), its the only way to improve.

Edited by MajorBorris, 11 January 2013 - 09:09 PM.

[Araara]

Marcus Tanner, on 11 January 2013 - 08:10 PM, said:

A few objections to Tolkein's statistics were brought up, and I think they're appropriate enough to deserve some addressing.

OBJECTION: ECM use correlates to player skill and use of premade teams, such that better players use ECM more often. Thus, Tolkein's data reflects player skill and/or the use of teams, confounding the analysis of ECM.

Better players will tend to use the better equipment, all other things equal. So, if ECM use correlates with player skill it means that the best players around will tend to believe that ECM is good to bring. So, even if the data is watered down by this factor it still shows that ECM is better than alternative uses of tonnage/crits.

In short, if the best players around use the thing than that's because it's effective.

---

Tolkein is a good player and he brought ECM in all these tests, so the results will skew toward wins for Tolkein's team and therefore more wins for ECM.

This one actually has merit, as by my count Tolkein's side won most of the matches by a margin of 35:19. That's huge for a set of PUG matches (some players who go out on their own report losing 2/3 of their matches). This effect may be magnified by the occasional 8-player drop where one would expect Tolkein's side (never 8-player) to get stomped.

It may also be relevant that Tolkein appeared to be on a 2-player team, which the matchmaker may often throw in with a 4-player team such that Tolkein may have regularly been playing with 6 people on comms +2 against 4 people on comms +4.

However, despite these problems, Tolkein's side won by a 2:1 margin at best, where ECM-superiority won by a factor of 4:1. The data correlates far more strongly with ECM than it does with Tolkein's skill (or that of Tolkein's wingman).

---

Unfortunately, we don't seem to have any way of getting better data at this point. By all appearances Tolkein has taken every reasonable step to get better data, but unless/until that happens this is what we've got.

Note that I, too, would like further testing on your hypothesis and I do remember tolkien asking about more data in general for matches with ECM without his implication (aka from random solo players).

However, I'd like to bring a few points to your hypothesis and pre-test conclusion of it. Do feel free to rephrase or clarify the following :

1) You say ECM correlates to player skill, might we have the link to the poll or some kind of logic argument for it? ECM might just be a big fad right now because people realize its the "equip to win" (hyperbole) item and thus, many people are planning on buying a variant of it, skilled or not.

2) You might want to rephrase/clarify what you mean by "...If the best players around use the thing than that's because it's effective". It does kind of mean that it's power is more efficient than any other items. I presume your counter theory would want to go against that (or the phrasing of it, at least)?

3) Nothing to say about tolkien and his friend in the game, it is a factor to be considered and, like I said at the beginning of my post, he DID say he wanted more raw data without his implications.

4) His analysis is based on win/lose PREDICTIONS on either side though, not win/lose rate. Just wanted to clear that out for other viewers. Also, as long as there is no ELO or some kind of rating system, I do not think there is any way to objectively factor player skill in the data.




edit : clarification on point #4

Edited by Araara, 11 January 2013 - 09:20 PM.

[Bagheera]

Strucker, on 11 January 2013 - 07:16 PM, said:

For those who say just shoot with lasers give me a laser than can counter a light effectively and I will otherwise **** and stop trying to defend your noob ECM builds.

Ya know, with a little practice it's actually not that difficult to tag their legs with an AC20. Just saying, especially since well over half of them will come to a complete stop next to one of your teammates for probably a good 40% of the match. Just pay attention, support the teammate they are backstabbing, and take out their leg in 2 shots.

It's not rocket science. Well, it is bullet science, but that's part of why we are all here in the first place.

The point is, instead of just crying about it over and over and griefing matches, maybe you should just voice a well reasoned feedback post in the appropriate consolidated thread and then let it go. After your well reasoned post with suggested fixes, take a break from the game, come back, review your options, and adjust your game play so that they are less of a problem.

My 82kph 4G fears no light mech. And when I get my Trenchbucket HK builds up and running, well, you do the math. 113+kph, a boatload of SRM (not streak) and >304 armor points. Bring it ravens.

[JudgeDeathCZ]

Alex Wolfe, on 10 January 2013 - 03:45 AM, said:

The problem is, with ECM nerfed, guided missiles would also have to be completely rebalanced, or we're all back in the space trenches again. Spamming M1 for 1000 damage is way too good times to come back.

You will have to find cover like us who play mechs w/o ECM....what a shame huh? -.-

Edited by JudgeDeathCZ, 11 January 2013 - 11:00 PM.

[TygerLily]

Ah, gotcha. Well, I think they've taken enough liberties that they could recreate the sprit of the stock's while balancing the game. I'd prefer game balance be the sacred cow than tabletop conventions.

[Ghostbear Gurdel]

As it stand now I cannot use LRMs. One ECM mech on the enemy team prevents all locks, any I die by Autocannon and Gauss fire. I have a founder's Catapult that I have not used in a month, because the ECM makes my LRMs useless.

ECM should only affect hostile mechs within its radius. If a Raven gets near me, sure that can kill missile locks, but if I have LOS on a hostile, and he is in an ECM bubble I should still be able to get missile locks. Right now, using my Catapult is an exercise in frustration, I end up relying on the 4 Med Lasers to actually try to get any kills as opposed to my LRM launchers....

[kalligrapher]

What was that PGI - you're going to deign to answer your players' concerns about ECM?

... nope. My mistake - didn't happen.

[kalligrapher]

Yeah - don't waste time on sorting player issues with ECM - build lots of new maps and mechs.

Hey - wait - you're not doing that either - one new {most likely useless] mech this month and the next new map is next month according to your content declarations. That's a significant addition to the game (yes, for the uninitiated - that's sarcasm) so probably ties up all your resources.

So you're fixing the d/c issues, etc. ?

Er, ... nope - seems as bad as ever.

So .. PGI - can I ask WHAT THE HECK you are doing? Trolling LOLCats & watching random YouTube taking up too much time ?

[Tolkien]

Marcus Tanner, on 11 January 2013 - 08:10 PM, said:

A few objections to Tolkein's statistics were brought up, and I think they're appropriate enough to deserve some addressing.

...

Araara, on 11 January 2013 - 09:16 PM, said:

Note that I, too, would like further testing on your hypothesis and I do remember tolkien asking about more data in general for matches with ECM without his implication (aka from random solo players).

...

I think they only way I will be able to cleanly address these concerns is to either collect data from an eye in the sky observer mode, or to hopefully get match data from the devs.

Anyway, you guys have already pointed out most of the pros and cons of the existing data set and I really can't add much more.

I am still having an intermittent back and forth dialog with Garth and hold a small amount of hope for getting my hands on some raw data.

Edited by Tolkien, 12 January 2013 - 02:04 AM.

[LynxFury]

You've done a great job given the limited amount of in game data that's available. And as you probably already know, no data collect and analysis is perfect--in the real world it's almost always darn dirty and far from ideal conditions (I was a military test officer and analyst at an army test center). None of the objections, or problems with the data or your analysis are enough to really change your conclusions. I do hope someone with the actual in game data actually does something similar.

Edited by LynxFury, 12 January 2013 - 02:21 AM.

[Tolkien]

LynxFury, on 12 January 2013 - 02:20 AM, said:

You've done a great job given the limited amount of in game data that's available. And as you probably already know, no data collect and analysis is perfect--in the real world it's almost always darn dirty and far from ideal conditions (I was a military test officer and analyst at an army test center). None of the objections, or problems with the data or your analysis are enough to really change your conclusions. I do hope someone with the actual in game data actually does something similar.

Thanks for that. It's nice to hear that the professionals have to work with less than ideal data too - and I too hope someone with access to in game data can look at the results as well. Frankly, I hope the >4.1x win advantage I've found is proven to be too high.

Edited by Tolkien, 12 January 2013 - 03:29 AM.

[UrbatOrbi]

ECM is ok!

a ECM-Atlas die how a No-ECM-Atlas!

but the Raven is the Prob - count the shoots hit, add it and wonder......that is by all ravens so.

I can tell you for example, after 300 lrms that xxx raven is only yellow, or after many hits from 2 mls+2 mpls - after 10 hits the leg is light orange (if that hit an atlas leg it is gone!)

Ravens dont die normaly - thx Devs, for this unbalanced mech i ever see - learn to balance a game first - then bring up new mechs

[Marcus Tanner]

Araara, on 11 January 2013 - 09:16 PM, said:

1) You say ECM correlates to player skill, might we have the link to the poll or some kind of logic argument for it? ECM might just be a big fad right now because people realize its the "equip to win" (hyperbole) item and thus, many people are planning on buying a variant of it, skilled or not.

You have a point here. The idea of such a correlation was asserted, but I failed to ask for supporting evidence. Really I just presumed it to be true because effective options are more popular with better players. I can see now that some circular logic sneaked into my head on this one.

Araara, on 11 January 2013 - 09:16 PM, said:

2) You might want to rephrase/clarify what you mean by "...If the best players around use the thing than that's because it's effective". It does kind of mean that it's power is more efficient than any other items. I presume your counter theory would want to go against that (or the phrasing of it, at least)?

Well, the objection was that Tolkein's results measured player skill and not ECM. They correlated the two by saying that skilled players and players in groups use ECM more often than everyone else.

My rebuttal, cited by you, is that skilled players tend to choose good options. This means that if skilled players take ECM more often then we can take that as evidence of ECM being an effective item.

Araara, on 11 January 2013 - 09:16 PM, said:

4) His analysis is based on win/lose PREDICTIONS on either side though, not win/lose rate. Just wanted to clear that out for other viewers. Also, as long as there is no ELO or some kind of rating system, I do not think there is any way to objectively factor player skill in the data.

Well, it was based on win/lose rates... but not for the side Tolkein was on but rather whichever side had more ECM. The test was to see if the "win rate" matched predictions.

Tolkien, on 12 January 2013 - 01:02 AM, said:

I think they only way I will be able to cleanly address these concerns is to either collect data from an eye in the sky observer mode, or to hopefully get match data from the devs.

Agreed. Best of luck to you.

LynxFury, on 12 January 2013 - 02:20 AM, said:

You've done a great job given the limited amount of in game data that's available. And as you probably already know, no data collect and analysis is perfect--in the real world it's almost always darn dirty and far from ideal conditions (I was a military test officer and analyst at an army test center). None of the objections, or problems with the data or your analysis are enough to really change your conclusions.

Agreed.

---

It occurs to me that the effect of Tolkein's skill might theoretically be mitigated with a large-enough sample size by taking a look at the win/loss ratio of the games that were inconclusive due to equal numbers of ECM mechs on either side, with support from taking a look at the difference between the win rate while Tolkein was in ECM-majority and the win rate while in ECM-minority. Taking the inconclusive games as a control group we have a basis for quantifying that variable.

Edited by Marcus Tanner, 12 January 2013 - 06:27 AM.

[StalaggtIKE]

DocBach, on 30 December 2012 - 10:05 AM, said:

We were promised a deep Information Warfare pillar where we'd have several different pieces of equipment that all provided an edge over our opponents and the other equipment, working against each other. What we got was ECM, ECM, and ECM, partially because of the fact that it was implemented with way more features then it ever had in any iteration of Battletech or MechWarrior, partially because implementation of other pieces of Information Warfare is so lackluster; ECM needs to be toned down and other components like Beagle and Narc need to be toned up.

ECM should not provide stealth 'Mech ability to an entire team
(Total Warfare, pg 134): "The ECM does not affect other scanning or targeting devices such as TAG and targeting computers"

ECM should not make missile locks impossible:
(Total Warfare, pg 134): "ECM blocks the effects of Artemis IV fire control systems. Artemis-equipped launchers may still be fired as normal missiles through ECM." and "Missiles equipped to home in on an attached Narc pod lose the Cluster Hits Table bonus for that system if the pods themselves lie within the bubble. The Narc launcher itself is not affected by ECM."

The previous rules state that missile attacks occur as normal, just not augmented with guidance from their advanced systems. Under no rulesets does it say that standard Guardian ECM negates the effect of Streak missile systems.

In disrupt mode ECM should have the following abilities:
-Disrupt Beagle's ability to detect shut down 'Mechs (Total Warfare, pg 134)
"Active probes cannot penetrate the ECM's area of effect. The probing unit would notice it is being jammed, however"

-Negate the tight groupings of Artemis-enhanced missiles (Total Warfare, pg 134)
"ECM blocks the effects of Artemis IV fire control systems. Artemis-equipped launchers may be fired as normal missiles through the ECM, but they lose the Cluster Hits Table Bonus"


-Negate the tight groupings of Narc-enhanced missiles, as well as prevent indirect fire on a Narc'd 'Mech without LOS (more on this later)
(Total Warfare, pg 134): "Missiles equipped to home in on an attached Narc pod lose the Cluster Hits Table bonus for that system if the pods themselves lie within the bubble. The Narc launcher itself is not affected by ECM."

-Prevent spotters in the ECM bubble, or on the other side of, with line of sight passing through a bubble from transmitting target data to team mates outside the bubble
(Total Warfare, Pg 134) "ECM has the effect of cutting off any C3 equipped unit from its network." Even though we don't officially have C3 installed on our 'Mechs, the way that our units communicate target data is what the C3 network does effectively. Being inside the ECM bubble should disrupt your ability to send data back to the rest of your team.

-Make target ID slower to acquire, but not block locks or targeting completely
(Maximum Tech, pg 54) "Though ECM systems can prevent a sensor probe from identifying a unit, they produce powerful distinctive electronic signatures." So sensors know something is out there, it just can't identify it or provide target information like the paper doll.

-Generate ghost targets in a seperate mode from disrupt/counter - bring up the command map with B in ghost target mode and plot a false radar signal that shows up like a target shrouded by ECM - target-able by R, but no target data available
(Tactical Operations, pg 100): "The ECM suite can be tuned to generate "ghost targets" that may affect the ability of enemy units to properly target friendly units. The ECM loses its normal function when used this way."

-Run counter ECM mode
(Tactical Operations, pg 99): "An ECM suite can be tuned to act as electronic counter-countermeasures (ECCM) in order to negate enemy ECM systems."


Beagle should:

-Provide 360 degree scanning and targeting within a 150 meter bubble

(Total Warfare, pg 129): "An active probe can detect any hidden 'Mech if the concealed unit lies within the probe's range." Like ECM, Beagle projects a scanning bubble of 5 hexes around it, equating to 150m of range

-Detect shutdown 'Mechs outside of ECM bubbles
(Total Warfare, pg 129): "An active probe can detect any hidden 'Mech if the concealed unit lies within the probe's range."

-Negate ECM's target acquisition slowdown while outside the bubble

-Identify ECM Ghost targets as being false targets
-Identify users the boundaries of the ECM bubble if they encounter ECM within Beagle's 150m scan bubble
(Total Warfare, pg 134)
"Active probes cannot penetrate the ECM's area of effect. The probing unit would notice it is being jammed, however"

The Narc beacon was rendered completely worthless when PGI failed to include its most important perk (Total Warfare, pg 139) which states:

"Once a Narc pod is attached to a target, all Narc-equipped missiles may be fired indirectly at a target without a spotter"

By this writing, we should be able to maintain locks on targets affected by Narc even if they go behind terrain and cut off line of sight, thus Narc should:

-Tighten LRM/SRM grouping on targets in line of sight
-Allow targets marked by Narc to remain targeted even when LOS is lost so LRM attacks can continue indirectly without spotters
-Last for either an extended duration than current or until the location the Narc hit is destroyed

A couple simple changes would make all of the advanced EW equipment viable, while remaining faithful to the source material without being game breaking or overpowering. ECM would still be useful to take, especially for protection against LRM spotters, as it would prevent them from sending target data to their LRM boats, protect you from Narc which would actually be a useful item if it kept enemies lit up on radar for indirect fire without a spotter like the rules say it does, and give Beagle expanded roles like countering ECM outside of the bubble (inspired by detection rules from MaxTech and Tactical Operations), and serving as a warning system that you are inside the bubble propper, which would be useful for spotters trying to Narc targets or transmit data.

The above changes would make information warfare have counters to each other, beyond just ECM, ECM, ECM. It would make it everything a useful addition to a team, but not an absolute necessity or gamebreaker.

Best idea I've seen.

Edited by StalaggtIKE, 12 January 2013 - 10:32 AM.

[Xandralkus]

ECM is a hard counter against missiles. Hard counters are bad for game design, because it proceeds down the path of metagame paper-rock-scissors and groups exploiting unbeatable permutations, all under the guise of 'tactical' play. It is not 'tactical', it is an abomination unto game design itself.

Redesign ECM so that it becomes impossible to undo or prevent another player's actions with it.

[trebormills]

I like the idea of ecm but the balance seems off- in particular in one battle it seemed that it was no players vs us. They all had ecm we didnt have any. Also those invisible ecm lights shooting rockets up your behind can be annoying. Of course in the match with no ecm we had a load of lrm boats who ***** us from base

If lag screen, ecm nerf- eg heat generation and match maker balance dealt with might be less of an issue. Also my streak srm does not work at all with ecm nerfing it- cant even blind fire the dam thing.

So basically I think ecm good, current version BAD

PS
Just played a load of games without ecm on any side- it was fun
played a few more with mismatched ecm- it was boring lose fest if not ecm majority
to counter the evening rush of ecm- fitted a tag. This was a death sentance,due to as soon as I tagged an atlas, 5 mechs attacked me

Edited by trebormills, 12 January 2013 - 10:42 AM.

[MaxllmuS]

Last time play on 4 mech all with tag, ecm no problem for me )