58 replies to this topic

[DaZur]

giganova, on 19 March 2014 - 10:32 AM, said:

Zing? I'll have to invoke hypothesis non fingo for that one... Nice dodge again, BTW

You are dead-set on dragging me back into this discussion ain't ya?

​Fine. As I stated earlier... Because it's a statistical anomaly the Elo make-up is irrelevant in terms of discussing "why" a stomp occurs. The anomaly occurs as frequently in top-tier games as it does in bottom-dweller games... It's essentially a equal opportunity mechanic of mitigation of force strength. Only difference is low-Elo involved matches will inherently be far more decisive / obvious.

The strength of a players Elo or even the team composite Elo doesn't trump the math behind force attrition.

Edited by DaZur, 19 March 2014 - 11:06 AM.

[giganova]

Well, let's put it this way. For all the situations where PGI had said "working as intended" when referring to the matchmaking system, it makes little sense for them to overhaul or "fix" the MM if nothing were wrong... and there is something wrong...

Perhaps if MWO was an objectively better game, it would have a larger community and the faults of the MM would not be as obvious. ...And the faults of the MM are obvious, at least to me, inexplicable, but still obvious...

[DONTOR]

I see stomps very very rarely, mostly on Skirmish where teams prepare for killing only, via meta 4 mans or otherwise. If you want a better gameplay experience I highly suggest Conquest rarely get stomped, or stomp. Breaks teams up more and there are alot of lights and so on.

[DaZur]

giganova, on 19 March 2014 - 11:03 AM, said:

Well, let's put it this way. For all the situations where PGI had said "working as intended" when referring to the matchmaking system, it makes little sense for them to overhaul or "fix" the MM if nothing were wrong... and there is something wrong...

Perhaps if MWO was an objectively better game, it would have a larger community and the faults of the MM would not be as obvious. ...And the faults of the MM are obvious, at least to me, inexplicable, but still obvious...

Non sequitur...

I don't care if you use the existing MM, tonnage based MM, BV based MM or any other previous community based pairing system offered... "stomps" will still occur.

Again... stomps themselves is not a derivative of the player pairings. They only influence the severity of the stomp when it occurs.

That said and to be fair in this discussion... Obviously the more skilled a player is the faster they recognize the ebb/flow of a drop and have the knowledge and experience to put for the most logical course of response.

[Joseph Mallan]

DaZur, on 19 March 2014 - 10:56 AM, said:

You are dead-set on dragging me back into this discussion ain't ya?

​Fine. As I stated earlier... Because it's a statistical anomaly the Elo make-up is irrelevant in terms of discussing "why" a stomp occurs. The anomaly occurs as frequently in top-tier games as it does in bottom-dweller games... It's essentially a equal opportunity mechanic of mitigation of force strength. Only difference is low-Elo involved matches will inherently be far more decisive / obvious.

The strength of a players Elo or even the team composite Elo doesn't trump the math behind force attrition.

I don't know DaZ... He likes using $5 words an a ten penny novel like you do... Could be a superlative engagement of Cognition mastery!

[DaZur]

Joseph Mallan, on 19 March 2014 - 11:14 AM, said:

I don't know DaZ... He likes using $5 words an a ten penny novel like you do... Could be a superlative engagement of Cognition mastery!

Kind'a like I'm arguing with myself a little bit eh?

Edited by DaZur, 19 March 2014 - 11:17 AM.

[Apnu]

I solo pug most of the time. And I tend to drive mediums and when, its bad, heavies with a high meta build.

I too see things that the OP is talking about but I've aways chalked it up to being stuck with bad players. My TDR-5S stats say I have a .9 W/L but a 2.22 KDR. This means I'm often getting kills but losing the match anyway. Now given that Elo, right now, only considers W/L, I"m doomed to spend time in that mech with newbies shooting at the sky or taking a the Champion Stalker and contributing 10 measly points of damage to the game.

Also, the tonnage imbalance is really bad right now. I can't wait for 3-3-3-3, really. I'm sick of winding up on a team of pugs with two or three locusts, three or four mediums, three or four heavies and one or two assaults against a 4-man in all high meta atlases and stalkers synch dropped with another 4-man of high meta heavies.

Fine. I think the problems I've seen and the OP has complained about say more about the quality of players and the weakness of Elo scoring than anything else. I know PGI is looking at changing Elo and the match maker, and I can hardly contain my frustration over that. But the quality of the players sometimes makes me think the matchmaker hates me. I'm often on the team full of newbs who instantly fly in different directions, totally out in the open and I just sit and watch them die as they go at the enemy blob in groups of 2 and 3. I've often wished PGI gave us a surrender option, it sucks to grind out a loss when your teammates are killed because of their own ignorance.

Another annoying thing: I seem to be on the team that drops short a player or two. I'd say that's taking the short bus to autofail, but some of the best games I've been in was pulling out a win when starting 2 short. That's magical.

[giganova]

Sorry, but ELO threshold tweaks and MM tuning are indicative of a systemic problem. This is the very definition of non-sequitur: https://static.mwome...g/paul/LM02.png

[wintersborn]

I agree OP.
Bad and uneducated players are crying Nerf for no reason other than they now have a weapon that kinda works and don't know how to deal with it.

The MM is the biggest issue for what ever reasons.

[DaZur]

giganova, on 19 March 2014 - 12:07 PM, said:

Sorry, but ELO threshold tweaks and MM tuning are indicative of a systemic problem. This is the very definition of non-sequitur: https://static.mwome...g/paul/LM02.png

So now a red herring?

And I'm the one sidestepping eh?

Clearly your derision goes deeper than just the your premise of MM inequities as thrice you've leaned into a semi-veiled dislike of the developer segue.

[giganova]

Nope, I'm pretty sure the broken MM has been the underpinning of my argument from the start, and yes, it reflects badly on the developers.

[DaZur]

giganova, on 19 March 2014 - 04:09 PM, said:

Nope, I'm pretty sure the broken MM has been the underpinning of my argument from the start, and yes, it reflects badly on the developers.

In farness... You opening post inferred "ELO and matchmaker discrepancies that PGI and its fanboys will never admit to having". It was your interjection to my response to the OP that it's mildly disingenuous to pin stomps wholey and erroneously on the MM, that prompted my issued premise.

Edited by DaZur, 19 March 2014 - 04:36 PM.

[Aresye]

Getting a little heated in here, and I'm seeing a pretty common argument amongst the statistical probabilities of force attrition vs Elo matchmaking algorithms.

Both arguments are right, so I don't really see the need to argue over it.

Let's talk about force attrition. It's a very real thing, and the general rule I've found is once there's a 3-4 kill difference between teams, the match has been decided (obviously with exceptions from time to time). It indeed has a compounding and exponential effect.

Where Elo and MM come in, is who exactly gets the first kill? That is where the skill level between teams becomes apparent, as the more experienced team will often stick together whereas the less experienced team spreads out and players go off wandering into the enemy team solo. Now, before somebody says something about communication, leadership, etc, let's just clear this up right now: It doesn't work.

It doesn't matter if you stick with your team. It doesn't matter if you take command, put tactical points on the battlegrid, and tell your team the exact strategy to win the game. They WILL NOT listen. I can't tell you how many times I have better SA on a certain situation, warn my teammates about it (which they either ignore or don't even notice) with plenty of time to react, and they end up doing exactly 100% what I told them not to do, and they get slaughtered. Now we're 0 - 4. Add in the effects of force attrition, and boom, you've got a stomp.

So yes, while force attrition does have an effect on the outcome of a match, the better team will usually get kills first, and the pre-match determination of which team is better is left to a computer that balances out good players with bad players in an effort to average people's W/L ratios.

[Daggett]

Aresye, on 19 March 2014 - 01:06 AM, said:

I seem to always get a good blend of teammates that decide the best time to test their weapons is into my back armor before I even move. Meanwhile on the other team all I see is an incredibly competent group of players that on multiple occasions felt like I was dropping against a 12-man premade. Coordinated attacks from multiple angles, lights leading teammates away to get slaughtered by the main force, etc.

It's funny that almost every competitive game has such moments, otherwise this old meme would not exist.

Have you checked your Win/Loss ratio? Maybe you just overvalue the negative games like we do with traffic lights which are more often red than green...

My ratio for example is above average, and i consider myself as an above average player. So from my point of view my stats make sense, and i see no serious imbalance in matchmaking.

Edited by Daggett, 19 March 2014 - 06:17 PM.

[DaZur]

Aresye, on 19 March 2014 - 05:47 PM, said:

<snip> Good post. ,snip>

Absolutely... the team fastest to the tipping-point (We both agree around 3-4) far more often than not emerges as the dominant team and nets the stomp.

That said, My point is that Elo itself is not the cause of stomps, force attrition and compounded force strength is... Elo is merely an accelerant and or predictor of the end result.

I just find it a little frustrating when players resign to blaming Elo and the MM for stomps when in reality is a fairly obvious statistical cause / effect mechanic...

Edited by DaZur, 19 March 2014 - 06:14 PM.

[Aresye]

Daggett, on 19 March 2014 - 06:01 PM, said:

Have you checked your Win/Loss ratio? Maybe you just overvalue the negative games like we do with traffic lights which are more often red than green...

My ratio for example is above average, and i consider myself as an above average player. So from my point of view my stats make sense, and i see no serious imbalance in matchmaking.

From the 3hrs or so of playing last night:
Kills/Death: 13/11
Wins/Losses: 4/9

Archived:
Kills/Death: 1918/726
Wins/Losses: 930/578

I really can't make any informed judgement on the stats according to last night, but ever since I returned from deployment last December I've seemed to be in a perpetual state of losing 3-4 matches for each match I win, and I'm pretty sure it had to do with Elo changes while I was on deployment.

My theory is that considering most of my wins/losses and kills were established prior to going on deployment, ever since I've gotten back I've been put into, "Elo Hell," as some call it. IIRC, I was winning ~ 2/3 of my matches prior to deployment, and (likely not coincidentally) now I'm losing 2/3 of my matches. If this system is designed to try and get everyone as close to a 1:1 as possible, then I've got a longgggggg period of losing ahead of me.

Essentially, I'm being continuously punished for the foreseeable future because of the games I played prior to Elo even existing. How exactly, do they expect people to stick around if that's the case?

Edited by Aresye, 19 March 2014 - 06:18 PM.

[DaZur]

Aresye, on 19 March 2014 - 06:17 PM, said:

Essentially, I'm being continuously punished for the foreseeable future because of the games I played prior to Elo even existing. How exactly, do they expect people to stick around if that's the case?

While it's not the answer you want... You're not entirely off base. Clearly your Elo is in the process of establishing your new / correct Elo based on the latest telemetry...

As a frame of reference I myself has a 30+ game skid before the ship righted and and began my slow climb back up. Admittedly I did garage my Assaults and took up Heavies (which better fit my play-style) and that definitely did aid in my recovery.

[Aresye]

And that more or less brings the discussion around full circle, and back to the original question of, "Do you enjoy playing games knowing that your playing experience is mostly determined for you?"

I'd rather drop and let the cards fall where they may. I see nothing enjoyable in winning over a few days, losing for a few straight days, winning for a few days, losing for a few straight days, and repeating this process ad nauseum.

[BLOOD WOLF]

I was just in a game, where my team got rolled, the problem is nobody is interested in what their team is doing, rather they are only worried about what mech they can shoot at. Im not gonna blame other people, because you cant blame a team without blaming yourself(life lessons and skills). It has nothing to do if a player is new, as long as they are doing their best to support the team that is all that matters. I don't care if i win or lose when im going solo but for the love of Kerensky, Some players need to realize that THIS IS A TEAM GAME. IF I AM AT TERRA THERMA AND I SACRIFICE MY MECH TO GIVE YOU GUYS A GOOD OPENING THEN TAKE IT.

I should note, that players are often to blame for a team lacking in coordination. Yes during battle you cannot always text but you should always give good advice to players even if they don't need it. The comms in text chats are dead silent and people only bother to type after they die and are observing someone else. If you are skilled player then take command, because i will tell you this, i have never been in a solo game where we could not have won, if we gave it our all. If somebody would just give their team an extra push(even a new player then its gonna be worth something). What does blaming get us>? it allows a person to vent whats on our chest but what good can come from it/

In conclusion, if you know what your doing, direct the team, take command if you can and help out. maybe those people you call newbies love the game just as much as you and want to be better but how can they when people are to busy blaiming them for what they don't know or what they could have done.

PS. By taking command, i have led countless Solo games to victory, but i only do so when i feel like it( and that is a negativity but sometimes people have to learn without someone telling them first)

Edited by BLOOD WOLF, 19 March 2014 - 07:11 PM.