1017 replies to this topic

[StUffz]

steelblueskies, on 02 January 2013 - 04:27 AM, said:

then yes naive. thermal vision is ir sensing. all mech have ir with a tt hex to meter range of 900 meters. if you can see it with thermal vision so can your mechs' targeting and tracking system. mech standard radar was 720 meters. magscan magnetic anomaly detector slightl less, and seismic only about 3 hexes/90 meters for target acquisition.

Call it IR or Thermal, both are quite correct because you can see the heat status of your opponent/friend from far distance. Main point here is that ECM can not shadow or cloak this and allows manual targeting with ballistic and energy weapon at long range.

steelblueskies, on 02 January 2013 - 04:27 AM, said:

you do not have iron sights.
your mech does the pointing and shooting and all the rest involved for you thanks to its software intelligence, you merely convey to it what to shoot at to the best of your ability. or did they add barrelcams to all the weapons and i missed it? or were you that guy outside his mech mid match laying on it's arm with what looked like a dreamcatcher standing in front of you on the arm?

Before we are going into deep discussion if we have iron sight or not.
With Iron sight I mean the targeting system (your crosshair) which you are still handling manually. And I disagree that your mech does the pointing and shooting. The mech only adjusts to the coordinates where you have pointed and shooting is also done by the pilot.

[ClaymoreReIIik]

Hello ladies and gentlemen,

please excuse the rather long post but in order to be constructive I try to explain the point I am trying to make.

I feel the current approach to electronic warfare is rather unbalanced.

If you take "rock, paper, scissors" as the definition of a balanced game the current implementation of ECM creates a game where ECM equipped Mechs are "Rock" and "Scissors" at the same time and come with a special rule saying "Paper cannot be played".

While I do agree that lock-on weapons like LRMs and Streaks needed a counter-strategy I think the current "you can not lock at all in a radius of 200 meters around any ECM that runs in disrupt"-approach is just plain too strong to stay.

While I do agree that the idea to have a counter-mode against ECM is an absolute necessity I kinda feel that giving that counter mode to the ECM-module itself is just plain stupid. "The only way to beat rock is to choose rock", would be the analogy here.

So here is what I think would be a good fix:
ECM itself:
Keep ECM at 2 modes. Mode 1 would be "cloak" that is similar to how disrupt works now. You can not obtain a hard lock outside of 200 meters on an ECM-cloaked target. Tag from outside of cloak-radius allows for indirect fire. Lock-on times on cloaked targets are trippled (but locking on cloaked targets stay possible at all times, just takes forever and then some).

This will make light mechs still largely protected from Streaks while giving the Missile players at least a theoretical chance to beat them. (After all if you manage to get the lock you did a hell of a dance in a tincan, might as well be rewarded for the effort.)

Mode 2 for the ECM is disrupt. This mode will not allow the target that the ECM-mech has hard-locked to use any lock-on weaponry. The idea that an ECM-Commando should have a payback option to get back at the Streak-a-pults makes a lot of sense and yes I agree its really fun. This option should stay in because it further reinforces the "scout vs LRM-boat" or "Rock beats Scissors" that a balanced game needs. Instead of just making it a "radius around you"-bomb it should be a "one target at a time"-scalpel though, IMO.

To stay in the above mentioned analogy here comes the "Paper" to beat above mentioned "Rock". Beagle Active Probe (BAP) should have 2 modes as well. Mode 1 would be "improve" that works as BAP works now. This mode will be utterly obsolete against ECM-cloaked targets, but will be good against targets that have no ECM-protection. Mode 2 will be "counter" that works like ECM in counter-mode works now. It will cancel the effects of one ECM in range around you. Additionally BAP in counter-mode should provide a "lock-on time reduction" and "target info gathering"-speed bonus against the countered ECM-mech. This additional bonus is not really necessary though.

Those changes will have the following effects. "Rock" (ECM mech) will beat "Scissors" (Mech without Electronic Warfare gear) because it will have a harsh impact on Scissors ability to target and/or do fire support. "Paper"(Mech with BAP") will beat "Rock" because it will be able to take "Rock's" advantages away and makes the "Rock" it's countering an easier target.

Scissors (Mechs without EW-equipment") will not have sacrificed the tonnage and slots for the ECM or BAP and therefore have a slight advantage when it comes to the "bang for the tonnage" comparison.

Right now taking ECM is strictly better then not having ECM. Right now ECM-equipped lights are (for none-US-based players) nearly impossible to kill unless you pack ECM yourself so your streaks still work. The direct fire guns are impaired because of the lagshield and streaks are useless against ECM unless you can counter it.

If you choose my solution you can give all Mechs the ability to pack ECM and BAP and do not have to restrict the usage to certain chassis either. Missile boats will make sacrifices to pack BAP, Brawlers will make sacrifices to fit ECM and some guys might just choose to play without EW-equipment in order to be stronger against either.

[StalaggtIKE]

Simply have ECM to closer resemble TT rules. ECM was never meant to be an all encompassing anti-everything, thus its light weight and critical slots. In BT lore, it was simply there to remove the advantages provided by electronic warfare devices such as BAP, Artemis, NARC and C3. It's weakness being its small range of 180m. It did not provide any stealth; that is for later tech such as Stealth armor and Null Signature System. All of which are heavier and caused heat, while activated. For ECM I propose the following change:

• Remove the stealth bubble; add stealth armor and null signature!

It will continue to do the following:

• Disable enemy bonuses from BAP, Artemis and NARC to all allies within 180m ECM bubble.
  
• Disable enemy's targeting data, lock-on and location of his allies within 180m ECM bubble.
  
• Distort enemy's minimap causing a low signal within 180m ECM bubble

One of the things that MWO's version of ECM has done is eliminate missile spam, more specifically LRM and SSRM. It has worked as a band-aid. These are issues that should also be addressed. I propose the following changes:

• LRM - no lock-ons without LOS or painted target from NARC/TAG (which will work without los). Of course ECM will continue to remove the target sharing benefits of NARC or TAG within bubble.
  
• SSRM - each tubing must require a lock-on in between each shot/volley. So you can no longer continue with a chained barrage of missiles if target escapes your targeting retical.

Source: Guardian ECM Suite, Stealth Armor, Null Signature

Edited by StalaggtIKE, 02 January 2013 - 07:52 AM.

[AlanEsh]

Bleh -- give ECM capability to every mech. My biggest beef with ECM right now is that I see too many of the same mechs, and we've lost a lot of variety in our matches because of this.

[StUffz]

Angelicon, on 02 January 2013 - 07:14 AM, said:

Bleh -- give ECM capability to every mech. My biggest beef with ECM right now is that I see too many of the same mechs, and we've lost a lot of variety in our matches because of this.

The difficulty I see here is that at it ends into a grande stampede and that friend or foes are being shot down. That's why I would not allow ECM to all chassis or to all mechs in one match as well.

[LynxFury]

StalaggtIKE, on 02 January 2013 - 07:03 AM, said:

, more specifically LRM and SSRM. It has worked as a band-aid. These are issues that should also be addressed. I propose the following changes:

Not sure why we call this spam, particular with regard to LRMS--with current maps there are very few places a mech can't make a safe approach to an LRM boat. But we can complain about pew pew and AC spam I guess.

Quote

LRM - no lock-ons without LOS or painted target from NARC/TAG (which will work without los). Of course ECM will continue to remove the effects of NARC or TAG within bubble.

TAG should work just fine inside the bubble--especially for the firing mech.

Quote

SSRM - each tubing must require a lock-on in between each shot/volley. So you can no longer continue with a chained barrage of missiles if target escapes your targeting retical.

An excellent suggestion.

--

Quote

Bleh -- give ECM capability to every mech

Bad idea, given there are no downsides, it would be on every mech. We might as well just get rid of radar if we go this route.

[StalaggtIKE]

LynxFury, on 02 January 2013 - 07:37 AM, said:

Not sure why we call this spam, particular with regard to LRMS--with current maps there are very few places a mech can't make a safe approach to an LRM boat. But we can complain about pew pew and AC spam I guess.

Spam is probably not the right word. I was more so referring to missile boating. Imo, LRM was not much of issue before ECM. However I believe this change would make the game more enjoyable for everyone. As well as add a bit of risk, while rewarding teamwork.

Quote

TAG should work just fine inside the bubble--especially for the firing mech.

I agree. It should continue to work, just not for target sharing.

Quote

Bad idea, given there are no downsides, it would be on every mech. We might as well just get rid of radar if we go this route.

Agreed.

Edited by StalaggtIKE, 02 January 2013 - 08:52 AM.

[Mudhawk]

Tolkien, on 01 January 2013 - 05:32 PM, said:

...
Lots of Numbers
...

This means that even with this small sample set I can state that ECM is a match decider with 97.9% certainty. The other way of stating this outcome is that you are 3.3x more likely to win the match if your side has a larger number of ECMs on your team. I expect this latter number to reduce with a larger data set, but so far ECM seems to be the best way of predicting the outcome of a match - I would gladly bet money using this.

Hello Mr.Tolkien,

thanks for the work you put into this. It's refreshing to say the least to see people base their deductions on facts instead of oppinions.
Even if those facts might only show a small fraction of whats going on:
a.) because they only cover a fraction of matches and
b.) because we (hopefully) deal with humans and not bots.

Before I explain the part about humans contrary to bots I'd like to state that I am aware about ECM being less than perfect and that PGI will have to either introduce better counters or lessen ECM's effect in some way. Personally I'd preffer the later part because I believe that even with C3 down a mech should still be able to target his LRMs if he himself is way out of ECMs reach. But thats a gut-feeling, not a scientific evaluation.

So back to bots and the human factor.
A bot crunches numbers: tonnage, speed, armament and such. After applying a formula there's usually a result along the lines of 1+1=2, 2>1 or the likes.
With humans things are usually not so simple. So 1 plus 1 does not always equal 2 and while 2 may be larger than 1 the smaller number still has a fighting chance. If it dosn't give in from the start because, come on, 2 > 1...right?

There are numerous scientific studies about a psychological effect called self-fulfilling prophecy.
Basically it goes like this:
If you have a strong expectation that something will happen you might, often unconsciously, work towards that end.
A strong believe to fail increases the chance of failure.
A strong believe of succcess increases the chances of success.
Right now what people believe most strongly in game seems to be "Biggest number of ECM equals victory".
Therefore people start the match counting ECMs and basing their expectations on the numbers they see on their side and somewhat later on the opponents side.
That fine and it's basically the same with LRMs, ACs Tonnage and whatsnot.
You want to know what your team can potentially bring to bear and you want to know the same thing about the enemy.
Only with what people have come to expect from ECM, wether it is true or not, people count their own numbers and go:

"Okay, we've got 1 ECM" (Example)

So they run around, meet the enemy and :

"Hurray, they've got no ECM, we win!" (Exibit A.)
or
"They've also got an ECM, no easymode for noone..." (Exibit B.)
or
"Oh my gosh, they got two or more ECM, we are soooo doomed! DOOOMED!!!" (Exibit C.)

Especially in the last case I've witnessed a lot of people only putting up token resistance and simply awaiting the outcome that they deem inescapable.
It is kind of hard winning a battle when faced with that kind of morale.
Now if this was Henry V i'd give a rousing speech and we'd be off "once more into the breach". But I'm sure by now you have noticed I'm anything but Shakespear. So that's not an option.
I let my weapons do the talking instead and more times than not I take some evil fiend with me to Hades. Who probably turns out to be my teammate in the next drop. But thats a different story.
I can, like you Mr. Tolkien, only talk about PuG battles with the occacional sprinkling of Premades. But from what I read in these forums the situation in the 8vs8 can't be much different.

"We've got 8 ECM and THEY got 8 ECM. We're dooomed! DOOOMED!!!" (Joke)

For my part, I think ECM is a powerful defence. Maybe too powerful sometimes. But you can't beat me just with ECM because it causes about as much direct Damage as BAP or TAG (to anything but the eye).

Like Terry Pratchet remarked in his wonderful book Lords and Ladies: From the point of view of an arrow, chainmail is just a loose assortment of holes.
The same goes for my Autocannon and ECM.

Once more thank you for your hard work Mr.Tolkien,
may we meet in the dropzone soon.

Mudhawk Out


Some Links for SFP:
http://en.wikipedia....illing_prophecy
http://www.wisegeek....ng-prophecy.htm
http://www.merriam-w...self-fulfilling

Edited by Mudhawk, 02 January 2013 - 08:47 AM.

[Willie Sauerland]

Tolkien, on 01 January 2013 - 11:18 PM, said:

You're implying that this was 15 trials of Team 1 versus Team 2 - it was not! It was 15 Trials of random teams, where one of the two teams had the pair of me and my friend on it. Since the hypothesis doesn't have anything to do with which team I am on, this too is a non factor.

Over a representative and sizeable sample set of the population factors like Team 1 having all free mechs, or team 2 having a 70 ton advantage will average out. The only outcome that will be correlated with my presence on a team is that it will occasionally happen that we will drop against a 4+4 or a synch dropped 8 man. Since my win loss is ~50% the effect of this superior communication/coordination against my team seems to be <1% on average - again, seems to be a non factor.

Please don't use strawmen just because you don't like the result of the experiment so far.

I will add more data but right now the outcome has a >97% certainty that numerical ECM superiority tends to lead to victory (and anecdotally by a factor of 2.3-3.3!)

Before I can take your objections seriously you will have to point out exactly what does not average away over random trials of random teams over multiple days such as this data set.

Who said I didn't like the experiment? I just think your methodology is flawed and as such does not rule out independent variables. As such, the implication you are making is inherently false. Add to this that the document you cited is clearly laid out for larger populations (it even says so in the document) I would state your statistical analysis is wrong.

However, for the sake of argument, I have over the course of the last week done some data gathering of my own which I would like you to analyze using your methods used for ECM. I had planned on offering these any way since I have been watching the original thread and noticed its flaws.

The methodology I used was quite simple, I had 1 friend who joined in counting the numbers in 15 PUG matches where a minimum of 1 item being counted was present. I will include the match breakdowns in a spoiler, but I would be very interested if your analysis supports my varied hypothesis. This should remove any notion of strawmen you might wish to perceive and will hopefully prove the wrong argument you are making...

Unfortunately, I am a network engineer and my statistical abilities are significantly rusty such that I do not trust them. I think yours are far superior to mine and I am sure you will not fudge the numbers to protect the claim you are making here.

Hypothesis 1: LRMs are overpowered. The team with more LRMs should win.

Spoiler

Team 1 | Team 2 | Result | Expected outcome
0 | 2 | Loss | Yes
1 | 0 | Loss | No
1 | 0 | Win | Yes
2 | 0 | Win | Yes
0 | 3 | Win | No
2 | 1 | Win | Yes
1 | 3 | Loss | Yes
3 | 2 | Loss | No
1 | 1 | Win | Inconclusive
2 | 0 | Win | Yes
2 | 2 | Loss | Inconclusive
3 | 1 | Win | Yes
1 | 2 | Loss | Yes
0 | 2 | Loss | Yes
2 | 1 | Win | Yes

Hypothesis 2: Machine guns are overpowered. The team with more machine guns should win. (This was the first time I have ever seen a cataphract sport 4 machine guns in a PUG match by some poor player - I LOL'd)...

Spoiler

Team 1 | Team 2 | Result | Expected outcome
4 | 0 | Win | Yes
2 | 4 | Loss | Yes
2 | 2 | Win | Inconclusive
2 | 4 | Loss| Yes
0 | 3 | Loss | No
4 | 2 | Win | Yes
1 | 0 | Win | Yes
2 | 2 | Loss | Inconclusive
4 | 4 | Win | Inconclusive
1 | 0 | Win | Yes
2 | 2 | Loss | Inconclusive
0 | 4 | Loss| Yes
2 | 0 | Win | Yes
4 | 2 | Win | Yes
1 | 2 | Loss | Yes

Hypothesis 3: Awesomes are just that. The team fielding the most Awesomes should win.

Spoiler

Team 1 | Team 2 | Result | Expected outcome
1 | 1 | Win | Inconclusive
1 | 0 | Win| Yes
1 | 0 | Loss| No
2 | 0 | Loss| No
1 | 0 | Win| Yes
0 | 1 | Loss | Yes
1 | 0 | Win | Yes
1 | 0 | Loss | No
0 | 2 | Loss | Yes
0 | 2 | Loss | Yes
1 | 0 | Win | Yes
0 | 1 | Loss | Yes
1 | 1 | Win | Inconclusive
1 | 0 | Win | Yes
0 | 1 | Loss | Yes

I will be interested in your outcomes.

Edited by Willie Sauerland, 02 January 2013 - 09:10 AM.

[Tolkien]

Well assuming the numbers as posted are accurate (apparently posted right after you got home for the evening).


For data set 2 - Hypothesis that numerical superiority of machine guns is a predictor of match outcome:
11 matches have conclusive outcomes
10 match hypothesis
1 contradicts hypothesis

This gives a 1 sigma uncertainty of 0.953 matches
This means the measurements could be off by 4.72 sigmas before the hypothesis would be contradicted.

erf(4.72/sqrt(2)) = 0.9999976, or better than 5 nines of certainty that a higher number of machine guns will tend to indicate which team will win.



Data sets 1 and 3 are identical with the data set I posted
13 games
10 matches
3 contradictions
1 sigma uncertainty at 1.52 matches.
This means that the outcomes could be wrong by up to 2.30 sigmas before the hypothesis would be neutralized/wrong.
Assuming a normal distribution etc. this gives erf(2.30/sqrt(2))= 0.979

So assuming you're not making up the data and were able to accurately count number of LRMs, number of machine guns (counting awesomes is about as easy as counting ECMs so no debate on that) then these are the results.

Personally though, given the convenient timing and relative difficulty counting LRMs and machine guns compared to what I was counting in 2 of your 3 examples I have doubts about the provenance of the data.

I think you're just trying to be controvertial, especially since the outcomes for 2 of your 3 data sets match what I already posted just with a switched order.

[StalaggtIKE]

Tolkien, on 02 January 2013 - 11:11 AM, said:

Personally though, given the convenient timing and relative difficulty counting LRMs and machine guns compared to what I was counting in 2 of your 3 examples I have doubts about the provenance of the data.

I think you're just trying to be controvertial, especially since the outcomes for 2 of your 3 data sets match what I already posted just with a switched order.

How does one count all the LRM and MG?
Over 6 people in a match with MG? Where does one find such scrubs?

My initial thought was BS.

Edited by StalaggtIKE, 02 January 2013 - 11:27 AM.

[Caleb Lee]

Noth, on 19 December 2012 - 04:18 AM, said:

Essentially what you are saying is ECM is not OP because ECM counters itself? Typically that is a bad sign. If your ultimate solution for something to to use the exact same thing the enemy is using, that is a bad sign as far as balance goes in gameplay.

What didn't you see about Tag in their build? Tag counters ECM.

I for one do not like that they made TAG 750 meters when it should have been a shorter range counter. They SHOULD have simply made ECM increase lock on time by 3-4x and reduce the range at which locks could be achieved to say 400-500 meters unless BAP is equipped and sensor boosting modules which would extend that range somewhat.

Make any mechs that have ECM require line of sight to lock on and not relay. That makes TAG viable.

They also need to limit how many total ECM mechs can actually drop in a match. Right now it's whoever brings the most ECM wins. I.E. 4 Ravens and 4 DDCs or mix that up with some LRM support like Stalkers but still, seeing 4-6 ECM running around is nuts.

It's supposed to be rare, not the norm.

As for the LRM and SSRM boaters whining, all one can say is L2P. TAG counters things for the most part, but guess what, it requires teamwork. When you pug without at least forming a 4 man with some strategy, you roll the dice and take your chances.

[MadaO]

Caleb Lee, on 02 January 2013 - 11:32 AM, said:

They also need to limit how many total ECM mechs can actually drop in a match. Right now it's whoever brings the most ECM wins. I.E. 4 Ravens and 4 DDCs or mix that up with some LRM support like Stalkers but still, seeing 4-6 ECM running around is nuts.

Instead of QQ'ing about too many ECM mechs you should learn to adapt. We can't all expect the game to impose rules according to each of our individual opinions. If you can't get your team to all bring a 1.5 ton piece of equipment, you should at least L2P and counter their ECMs with TAG. As you claimed, TAG works, especially if you're in a team. Use it then and stop whining.

[steelblueskies]

MadaO, on 02 January 2013 - 12:07 PM, said:

Instead of QQ'ing about too many ECM mechs you should learn to adapt. We can't all expect the game to impose rules according to each of our individual opinions. If you can't get your team to all bring a 1.5 ton piece of equipment, you should at least L2P and counter their ECMs with TAG. As you claimed, TAG works, especially if you're in a team. Use it then and stop whining.

hey another guy making arguments beaten within inches of their life because reading long threads isn't important enough, but being a special snowflake and posting IS.

[Wildstreak]

Tried emailing this in but was told to post it.
-------------------------------------------------------

The change to ECM is way too much, not only different from original BT (not too worried about everything being canon) but overpowering. It is right now NOT ECM, Electronic Counter Measures, a better description would be ECD, Electronic Cloaking Device.

BT ECM is only supposed to counter Active Probes , Artemis , NARC and C3 Computers. Never was it designed to stop missile lock-ons or render someone invisible from detection, maybe you could make an argument for shorter ranges on detection up to 25% but not undetectable.

Better would be not to change detection range but at some long ranges, when hitting R or whatever key to target a unit, instead of getting the exact type, you just see the weight name so at long range you may get ASSAULT but not know which type until closer. Allowing ECM to counter NARC and Artemis affects LRM fire using those systems but it should never make a unit LRM proof at long range. Maybe it also cuts down number of LRM hits when not using NARC or Artemis since countering those systems does exactly that, cut down missle hits.

As for Streak SRMs, since those work either by allowing all missles to hit or not fire at all, just let ECM cut the damage in half. Not canon but I can understand the problem mentioned with Streaks and this would be in line with what ECM already did. Since ECM cuts down number of LRM missles hitting with NARC and Artemis, it makes sense to have ECM allow only half the Streak missles to hit.

What is currently in MWO is not ECM. I hope some type of fix is made sometime because currently it is too good for too little expense.

[Noth]

Caleb Lee, on 02 January 2013 - 11:32 AM, said:

What didn't you see about Tag in their build? Tag counters ECM.

I for one do not like that they made TAG 750 meters when it should have been a shorter range counter. They SHOULD have simply made ECM increase lock on time by 3-4x and reduce the range at which locks could be achieved to say 400-500 meters unless BAP is equipped and sensor boosting modules which would extend that range somewhat.

Make any mechs that have ECM require line of sight to lock on and not relay. That makes TAG viable.

They also need to limit how many total ECM mechs can actually drop in a match. Right now it's whoever brings the most ECM wins. I.E. 4 Ravens and 4 DDCs or mix that up with some LRM support like Stalkers but still, seeing 4-6 ECM running around is nuts.

It's supposed to be rare, not the norm.

As for the LRM and SSRM boaters whining, all one can say is L2P. TAG counters things for the most part, but guess what, it requires teamwork. When you pug without at least forming a 4 man with some strategy, you roll the dice and take your chances.

Did you read the thread at all? TAG still takes 4 seconds to gain a lock with. If you are TAGing for your team and not on TS, it can take up to 10 seconds for them to react to the TAG.Further the nature of TAG makes it difficult to keep faster mechs under the beam, and even small movements can throw off the beam. ECM doesn't have to aim, doesn't need LOS to the mechs it covers to be effective. TAG is not an effective counter. The only effective counter is ECM itself and that is why you see ECM stacking.

The fact that we should limit the number of ECM mechs to drop in a match is a statement to how OP the item is.

[VampireMoose]

When is this going to get fixed?

Spent over 4 minutes in a fight with a raven with ecm who just shut down my lrm boat and sat there constant missles while laughing in general chat.

Sorry having a scout mech that can take apart a mech that is totally dependant on missles is a joke.

When I purchased the Cat mech ECM wasnt in the game, if i had known then what i know now I would not of bothered.


give a method for a Missle boat to be able to lock on.

[Gendo]

VKleita, on 02 January 2013 - 12:45 PM, said:

When is this going to get fixed?
give a method for a Missle boat to be able to lock on.

rofl! you havent got the point...
ecm is just another (intended) lrm nerf that turns em from "rarely usefull" into "completely unplayable".

ecm should only have effect on the mech carrying it - 200m range is too much of a tactical advantage for a 1,5t part.
lets be honest guys: if 1/3 (at least) of the community switches to ecm mechs, there MUST be something wrong (op?) with it...

[Willie Sauerland]

Tolkien, on 02 January 2013 - 11:11 AM, said:

Well assuming the numbers as posted are accurate (apparently posted right after you got home for the evening).


For data set 2 - Hypothesis that numerical superiority of machine guns is a predictor of match outcome:
11 matches have conclusive outcomes
10 match hypothesis
1 contradicts hypothesis

This gives a 1 sigma uncertainty of 0.953 matches
This means the measurements could be off by 4.72 sigmas before the hypothesis would be contradicted.

erf(4.72/sqrt(2)) = 0.9999976, or better than 5 nines of certainty that a higher number of machine guns will tend to indicate which team will win.



Data sets 1 and 3 are identical with the data set I posted
13 games
10 matches
3 contradictions
1 sigma uncertainty at 1.52 matches.
This means that the outcomes could be wrong by up to 2.30 sigmas before the hypothesis would be neutralized/wrong.
Assuming a normal distribution etc. this gives erf(2.30/sqrt(2))= 0.979

So assuming you're not making up the data and were able to accurately count number of LRMs, number of machine guns (counting awesomes is about as easy as counting ECMs so no debate on that) then these are the results.

Personally though, given the convenient timing and relative difficulty counting LRMs and machine guns compared to what I was counting in 2 of your 3 examples I have doubts about the provenance of the data.

I think you're just trying to be controvertial, especially since the outcomes for 2 of your 3 data sets match what I already posted just with a switched order.

Actually, it isn't difficult at all. An ECM Commando can hide very easily and target respective enemy mechs and determine loadouts with the target information gathering module. The other side is easily obtained by spectating and getting killed in a Commando is equally easy. In other words, the numbers are quite accurate for the given number of PUG matches (I have never seen a Machine gun in an 8v8 matchup because they are so inferior). My numbers were accurate and have been collected since my Christmas break. Unfortunately, you posted in another thread your ridiculous numbers which required me to post the flaws at that time.

I'm sure you will look upon my numbers with skepticism especially since it seems to prove Machine Guns are overpowered. Anybody will find this notion ridiculous which is exactly how your ECM numbers appear.

You have proven nothing but have implied a very sensational and dare I say, timely yet flawed "analysis". The fact everybody believes it and you have even provided documentation which clearly shows it is flawed is only adding fuel to an already ridiculous fire.

Your methodology is flawed and the fact you refuse to believe it only shows that you wanted statistics to prove your point of view. There are only three kinds of lies Sir: Lies, Damn Lies, and Statistics.

LRMs are overpowered.
Machine Guns are overpowered.
Awesomes are overpowered.
ECM are overpowered.

Perhaps I should collect data to see if flamers are really overpowered next - this methodology is awesome in that it is so flawed it will prove any hypothesis....

StalaggtIKE, on 02 January 2013 - 11:20 AM, said:

How does one count all the LRM and MG?
Over 6 people in a match with MG? Where does one find such scrubs?

My initial thought was BS.

Incorrect. The number of machine guns were counted - not the number of people carrying them. However, the hypothesis was that the number of machine guns should determine who would win - not the number of people carrying them. Reread the methodology.

Same is true with LRMs.

Awesomes were very easy to count and shouldn't create any discrepancy.

Edited by Willie Sauerland, 02 January 2013 - 01:41 PM.

[Captain Commander]

Won't spend another penny on this game until ECM is adjusted. As it stands there are only 4 mechs worth their weight. Rest of them are questionable at best, feel like I'm nurfing the team bringing anything but an ECM loadout 'mech. Balance bad, you should feel bad.