99 replies to this topic

[MisterFiveSeven]

Homeless Bill, on 20 September 2013 - 06:00 PM, said:

Be thankful. I wish I had more of those rounds; as it is, I feel like I have to carry or die. Even when I perform poorly, I'm almost always in the top four on my team. Showing up in fluffly Elo land when I bought my first medium was a dream come true for the first 25-or-so rounds.

I like that Elo grinds against you and makes it harder, but some nights I really wish I could just steamroll bads in a non-ranked queue. I'm not good enough to carry, but not bad enough to escape the responsibility. #firstworldproblems

My only problem with ELO is that it is very punishing to my creativity.

Example:

The Orion is terribad awful as far as I can tell. Any build you want to make, either the Victor or the Cat does it better. At this point I'm pretty thoroughly convinced as the only decent Orion I've got is the Protector (and only because UAC5's were god ballistics for a while).

However, I distinctly remember switching from my Stalker (best chassis in the game) to the Atlas just after I mastered them (like a week after they came out). I thought the Atlas was terrible; it was too tall, and you can hit the crest of the skull from anywhere. After more time, I discovered that I was making noob mistake #1: trying to play an Atlas like a Stalker. One hill humps like a demon, the other bashes your face in in a dark alley. While the stalker is incredibly forgiving and can do both well, the Atlas has to be more specialized because it is such a broad target.

Moral of the story: The new matchmaker is kicking *** at giving good matches (with ELO and time target adjustments); this is excellent. However, it means my perspective on the Orion's effectiveness is skewed and I don't know if I'm just being dumb or if it really is subpar (having success with my mini misery ONI-M w/ac20, x2 PPC regardless).

It also means my x6 spl std385 Awesome will never stop gather dust in some forgotten corner of my hanger.

A little off topic but Homeless, I feel your pain.

Edited by MisterFiveSeven, 20 September 2013 - 06:27 PM.

[Homeless Bill]

MisterFiveSeven, on 20 September 2013 - 06:25 PM, said:

My only problem with ELO is that it is very punishing to my creativity.

[getting used to 'mechs]

It also means my x6 spl std385 Awesome will never stop gather dust in some forgotten corner of my hanger.

100% agree. Adapting to new chassis (particularly when you don't have any efficiencies (FFFFUUUUUUUUU)) is always a pain in the ***, and Elo sure doesn't make that any easier.

Besides adapting to new 'mechs, creativity in general is stifled. Want to load up flamers and just **** around? Too bad.

My Awesomes, as ****** as they are, will never gather dust; I just can't get away from my one true love. Though I totally need to try the 375+STD builds...

[Deathlike]

Homeless Bill, on 20 September 2013 - 06:47 PM, said:

My Awesomes, as ****** as they are, will never gather dust; I just can't get away from my one true love. Though I totally need to try the 375+STD builds...

There is no good build that uses an engine beyond the 350-360 STD that even the mighty Atlas uses. I haven't even considered it just because the tonnage gap beyond it is like giving a Centurion something bigger than a 375XL... you can do more with less (tonnage for the engine).

Edited by Deathlike, 20 September 2013 - 09:25 PM.

[Homeless Bill]

Deathlike, on 20 September 2013 - 09:22 PM, said:

There is no good build that uses an engine beyond the 350-360 STD that even the mighty Atlas uses. I haven't even considered it just because the tonnage gap beyond it is like giving a Centurion something bigger than a 375XL... you can do more with less (tonnage for the engine).

Which is exactly why I haven't tried it. Doesn't mean it's not worth taking for a spin. I mean, the thing is a piece of **** either way. Might as well make a total joke of it =P

[Jonathan Paine]

Problem with elo is what happens when the lax matchmaker puts kaffe+7 good pilots on his side with 4 mediocre ones, and my poor 3 man group gets 4 mediocre and 5 trial mechs

Edited by Jonathan Paine, 20 September 2013 - 11:10 PM.

[Mister Blastman]

Jonathan Paine, on 20 September 2013 - 11:08 PM, said:

Problem with elo is what happens when the lax matchmaker puts kaffe+7 good pilots on his side with 4 mediocre ones, and my poor 3 man group gets 4 mediocre and 5 trial mechs

I had something similar happen a day ago... except it was 3 really good premades vs. my 4 man (two rookies in it) and a bunch of random scrubs. Yeah. 9 times out of 15 that's exactly what happened.

So mm is by far not perfect yet but it is getting there.

The ONE thing I would really like to see is having PUG drops just full of PUGs and no 4-mans at all in these. Some of us really like pugging. I do. Sometimes it was refreshing to just drop in randomly and go around blasting at stuff. With ELO tightening you're at extreme risk of getting shafted now when soloing. That is... if I can get a drop. It took 10 - 15 minutes many times in a row to find one a day ago.

Edited by Mister Blastman, 21 September 2013 - 07:15 AM.

[Grits N Gravy]

Homeless Bill, on 20 September 2013 - 11:24 AM, said:

I think that whole first paragraph is a lot of unbacked assumptions. You should only play people of a similar skill, but due to small the player pool and the matchmaker's lax standards, you'll see top-tier players in with complete scrubs. The Elo hell bit is what comes of it, but that's only because the playerbase / matchmaker can't find enough competent opponents. Even though anyone with a high win-loss ratio has an Elo that is "too low," it's still probably one of the higher Elos in the game.

Again, I entirely disagree. Though the way points are distributed based on assumed win/loss messes things up, I see no situation where maintaining a good WLR wouldn't get you an increasing Elo. The more you win, the more "assumed loss" matches the matchmaker will try to give you, the more "assumed loss" matches you win, the higher your score is pushed.

The way you talk about things, it seems like you think there are "tiers" of players that you move between, and those tiers are completely excluded from each other. That's not how it works in this game based on all empirical evidence. The matchmaker just tries to match up average Elo ratings - it doesn't set up a Tier 1 match and a Tier 2 match. We've all seen matches that have PEEFsmash, kaffeangst, and some guy that doesn't know how to set up his weapon groups. The longer it waits, the lazier it gets about finding matches.

Again, what should happen is irrelevant. What does happen is that good people continue to win more because there aren't enough other good players to normalize them downwards. I find it hard to imagine that anyone with a 1600+ Elo rating has below a 1.5 win-loss ratio.

I would also like to thank you for being competent, civil, and enjoyable to debate.

If you have played thousands of games, win to loss ratios will take tens of thousands of games to trend down. The relationship between games played and changes in win loss rates is exponential.

Take a player with a 4.0 or 80% win to loss ratio with 5000 games. If Elo works 100% perfectly and matches him perfectly resulting in a 50/50 win to lose ratio. Even after an additional 10,000 games playing at win to loss ratio of 50%, his net aggregate win to loss ratio is still 60%. It would take that player over 100,000 additional games to lower his win to loss ratio to 50%.

The anecdotal evidence of win to loss rates trending downwards at even 5% supports the claim that there is segregation of Elo scores even amongst the the top .1% (the 3 sigma) of players. Is it perfect segregation? Obliviously not, but it has been said by staff that only 7% of games are considered out side of tolerance.

There are brackets for Elo, it is precisely how the teams are constructed according to the devs

Quote

the match maker starts trying to make a match for an Elo of say 1300 and will pull in players to those teams closest to those values; however, as mentioned earlier within growing thresholds and those curves will be tuned. Currently it may be a bit 'sloppy' about how it's filling those buckets but over time it will be tuned to be much more precise.

source, http://mwomercs.com/...-making-update/

So you can expect that we will trend to evenly matched games over time, which validates everything I said in my previous posts. As long as we trend towards 50/50 matchmaking, everything in my previous post is statistical fact, not anecdotal observations or feelings.

It's entirely possible to maintain a positive win to loss ratio and have your Elo score trend downwards. The magnitude of the loss only has to outweigh the wins. Take a player with a 1600 Elo score Vs someone with a 1300. Your arbitrary elite score Vs the starting score. If the 1600 wins that mach they only get 8 Elo points, if they loss the match they lose 42 Elo points. Any win to loss ratio below 5.25, that is 5.25 wins for every loss, will net a downward trend in Elo points if the 1600 plays only 1300. That is Elo hell.

It's also very possible that there are plenty of people with little experience with high elo scores due to the K factors I described previously. Theoretically you could get to 1600 Elo in 6 games which would put you in the top two thirds of Elo scores. As noted in the Dev logs the max change is 50 Elo points per match, So if a new player faces a team who's Elo points are 800 above his score, for every match, for 6 matchs, he can get to Elo 1600 in 6 matches.

This also is a good way to explain why you see "noobs" with elites in matches. A strong K factor allows people to gain a lot of Elo quickly; so quickly that they may end up high ranked after a dozen matches and still not know how to work controls. This will always be an issue with F2P games as they continually add new players to the pool and you want high K factors to ensue good matchmaking quickly.

Edited by Grits N Gravy, 21 September 2013 - 10:54 AM.

[Satan n stuff]

I'm at 65% wins, but a lot of that is from before ELO, to get a more accurate number I did the math and got 59% for my 4 most recent variants, that's for 177 matches total. Anyway it feels more like 50% wins at the moment, but I guess that's because we tend to give more weight to our losses. I am curious where that places me along the ELO scale, does it mean I'm considered super 1337 like some people like to think of themselves or am I just above average? Probably not worth thinking about as I'll likely never know my real ELO score. At least I can brag about not being awful.
Edit: in response to the above, I hope the matchmaker will be able to adjust it's target ELO to the currently available players so players on either end of the scale won't have to wait ages for a match, and maybe that way the skill gap between players in a team can be made smaller.

Edited by Satan n stuff, 21 September 2013 - 11:24 AM.

[PEEFsmash]

***** n stuff, on 21 September 2013 - 11:21 AM, said:

Probably not worth thinking about as I'll likely never know my real ELO score. At least I can brag about not being awful.

I think that at some point you will know your Elo score, once the developers of this game realize that the only way to maintain a game for many years is with a thriving competitive scene.

[Grits N Gravy]

PEEFsmash, on 21 September 2013 - 11:49 AM, said:

I think that at some point you will know your Elo score, once the developers of this game realize that the only way to maintain a game for many years is with a thriving competitive scene.

There is actually a varied opinion within game design if giving out this info is good for a game and competitive scene. LoL use to give out the players Elo score, but switched away from it to a different ranking system, while DOTA has not given out their ranking system for matchmaking. Both games have great competitive scenes.

Giving out Elo scores can also create negative behavior which does far more damage to a game than keeping them hidden from players.

[PEEFsmash]

Grits N Gravy, on 21 September 2013 - 12:07 PM, said:

There is actually a varied opinion within game design if giving out this info is good for a game and competitive scene. LoL use to give out the players Elo score, but switched away from it to a different ranking system, while DOTA has not given out their ranking system for matchmaking. Both games have great competitive scenes.

Giving out Elo scores can also create negative behavior which does far more damage to a game than keeping them hidden from players.

You don't have to give out the Elo scores raw, you can do something like a matchmaking score/league placement that is largely Elo based, like SC2 does. Dota has very high-quality external leagues and ladders that allow players to understand their standing. In-game leagues/ladders are coming.

Keeping it hidden only keeps the game uncompetitive in a game like this one where there is not thriving competitive support, sponsors, streams, etc. People have no way to confirm self-improvement except at the very top.

[Mawai]

When ELO is working perfectly you would have a W:L ratio of 1. You would win half your games and lose half your games on average. As your ELO settles you will either have more losses or more wins but in a perfect world, when your ELO is reasonably accurately reflected and the matchmaker has the opportunity to create a balanced match then your odds of winning or losing should be 50:50.

If you are winning 55% of your games over the long term then your ELO is likely still going up.

As for competitions ... I expect that they will have ladders you can opt in to as well as other competitive venues and published standings ... but I don't expect them to publish ELO since it isn't really useful.

Edited by Mawai, 21 September 2013 - 12:16 PM.

[Grits N Gravy]

Mawai, on 21 September 2013 - 12:15 PM, said:

When ELO is working perfectly you would have a W:L ratio of 1. You would win half your games and lose half your games on average. As your ELO settles you will either have more losses or more wins but in a perfect world, when your ELO is reasonably accurately reflected and the matchmaker has the opportunity to create a balanced match then your odds of winning or losing should be 50:50.

If you are winning 55% of your games over the long term then your ELO is likely still going up.

That is not true, there are plenty of ways you can have a win rate of 55% and a declining Elo score, as I noted in my previous post.

However, let's say you and 3 buddies are really good. We'll assume you have an Elo rank of 1600. So you enter the queu, but there aren't enough people to match you at your Elo, so you get stuck with 8 rank 1300 players. Your Team average is 1400 points, if you play a team of 1300 or below rank, you will lose Elo points unless you maintain a win rate of 56%, that is winning 2.8 games out of 5. Those 8 1300 rank players can make that a dicey proposition.

Edited by Grits N Gravy, 21 September 2013 - 12:30 PM.

[Inkarnus]

dude elo works just when all ppl have the same base like in chess
a queen is a queen a horse a horse
but here a queen can be a peon because he she it
does run a new mech and not the cheese of the month

there are to many different factors that abide elo to work properly.

On top the ppl that are skilled carry ppl that arent
wich get there elo increased too inflating elo
and giving all even more Dicy feeling since you dont know
if you playwith scrubs or players that know what they are doing

mostly for me scrubs anyway sigh

Edited by Inkarnus, 21 September 2013 - 12:36 PM.

[PEEFsmash]

Inkarnus, on 21 September 2013 - 12:34 PM, said:

dude elo works just when all ppl have the same base like in chess
a queen is a queen a horse a horse
but here a queen can be a peon because he she it
does run a new mech and not the cheese of the month

"dude elo works just when all ppl have the same base like in chess
a queen is a queen a horse a horse
but here a queen can be a peon because he she it
does run a new mech chess strategy and not the cheese of the month strategy that everyone is using"

Just wanted to show you that your disanaogy is not a real disanalogy.

Edited by PEEFsmash, 21 September 2013 - 12:47 PM.

[Nik Reaper]

Is there a specific reason stated why they decided to go with ELO rather than some of the other systems like true skill and such?

This might be just me but I consider a good player one that can get kills , but not the most important point ( he could be kill stealing or be timid and only go if he is sure he can get the kill, leaving him ether going in the fight mid way near untouched or dying after everybody else in the front died ) , the damage, accuracy, component destruction and time one lives in a match should have some pull in a players rating.

I'm I missing some known point as to why such a system would fail to deliver "good" games?

P.S. bringing up the point even though we have elo just so if a good idea comes up we could try to push for a test next year (optimism to the max, expectations 0). and to find out if there is any reason to support keeping this system.

[MadCat02]

Nik Reaper, on 21 September 2013 - 12:55 PM, said:

Is there a specific reason stated why they decided to go with ELO rather than some of the other systems like true skill and such?

This might be just me but I consider a good player one that can get kills , but not the most important point ( he could be kill stealing or be timid and only go if he is sure he can get the kill, leaving him ether going in the fight mid way near untouched or dying after everybody else in the front died ) , the damage, accuracy, component destruction and time one lives in a match should have some pull in a players rating.

I'm I missing some known point as to why such a system would fail to deliver "good" games?

P.S. bringing up the point even though we have elo just so if a good idea comes up we could try to push for a test next year (optimism to the max, expectations 0). and to find out if there is any reason to support keeping this system.

Elo originated in 1 vs 1 sports thats why

[PEEFsmash]

Nik Reaper, on 21 September 2013 - 12:55 PM, said:

Is there a specific reason stated why they decided to go with ELO rather than some of the other systems like true skill and such?

This might be just me but I consider a good player one that can get kills , but not the most important point ( he could be kill stealing or be timid and only go if he is sure he can get the kill, leaving him ether going in the fight mid way near untouched or dying after everybody else in the front died ) , the damage, accuracy, component destruction and time one lives in a match should have some pull in a players rating.

I'm I missing some known point as to why such a system would fail to deliver "good" games?

P.S. bringing up the point even though we have elo just so if a good idea comes up we could try to push for a test next year (optimism to the max, expectations 0). and to find out if there is any reason to support keeping this system.

All matchmaking systems that aren't purely grind-based use something like Elo, where you are rewarded more for wins against good players, and punished more for losses against bad ones. The magic of a matchmaking system comes in those small tweaks between each system, but the idea overall is more or less the same. The best players would be the best players (and the worst would still be the worst) regardless of if the game was using this version of Elo, that version of Elo, TrueSkill, LoL-type Elo, etc.

Edited by PEEFsmash, 21 September 2013 - 01:09 PM.

[Grits N Gravy]

Nik Reaper, on 21 September 2013 - 12:55 PM, said:

Is there a specific reason stated why they decided to go with ELO rather than some of the other systems like true skill and such?

This might be just me but I consider a good player one that can get kills , but not the most important point ( he could be kill stealing or be timid and only go if he is sure he can get the kill, leaving him ether going in the fight mid way near untouched or dying after everybody else in the front died ) , the damage, accuracy, component destruction and time one lives in a match should have some pull in a players rating.

I'm I missing some known point as to why such a system would fail to deliver "good" games?

P.S. bringing up the point even though we have elo just so if a good idea comes up we could try to push for a test next year (optimism to the max, expectations 0). and to find out if there is any reason to support keeping this system.

Elo systems are used because the are considered good enough. In the world of production we have to balance the aspect of time, money and quality. You don't see a lot of companies developing their own ranking systems, like Microsoft's trueskil, because it is a expensive long proposition. You don't see people licensing that system because of A. cost and B, it would take some tweaking to get it optimized for different applications, and most people don't even really understand what's going in that formula. Therefore Elo gives us the most bang for our buck, it does an a good enough job, doesn't cost anything, and is quick to implement.

[Nik Reaper]

True , but the point is kinda not so much to know if you are good or not , but to make sure that you will almost never be crushed by some one out of you'r league , and even though some find it pleasing, not to let you crush people that both you and them know you shouldn't fight each , and not winning doesn't mean you'r bad per say ( or one is ), it's just that you can't carry.

A game is much more balanced if all players in a match have a similar accuracy, damage and live time then when, with who and at what time you play. right? It makes sence to me but do tell if you have a different perspective.

EDIT: goddam the money, always the money!!!

Edited by Nik Reaper, 21 September 2013 - 01:47 PM.