Elo Worthless
#281
Posted 19 November 2013 - 04:13 AM
With ELO, players will still have roundly W/L ratio around 1. So the factions will also have W/L = 1 .
So... If ELO remain the same. Whatever you do, there will never by any winner - or looser in the CW. So... why CW ?!
#282
Posted 19 November 2013 - 04:27 AM
MischiefSC, on 18 November 2013 - 12:35 PM, said:
I wish you could teach classes. There's a distinct drive for some people that keeps them from enjoying just doing better - they have to be doing better than *someone else*.
Quote
I love me some D-DC.
Edited by Joseph Mallan, 19 November 2013 - 04:27 AM.
#283
Posted 19 November 2013 - 04:34 AM
loupgaroupoilu, on 19 November 2013 - 04:13 AM, said:
With ELO, players will still have roundly W/L ratio around 1. So the factions will also have W/L = 1 .
So... If ELO remain the same. Whatever you do, there will never by any winner - or looser in the CW. So... why CW ?!
How it should be......it shouldnt be a dominant faction in BT.
Think about
Liao and Davion:
God (the BT Matchmaker) dicides to end the war after Liao lost half of theyr systems.
Like before and after intruducing of the Elo System.
Steiner and Kurita:
Steiner Scout lances stormed draconian sytsems. Kurita fought back. W/L about 1.
Clans and Inner Sphere:
well......in fact of cheating Clans (weapon-hacks.) they got time bans till it is fixxed.
The CW is just for those who like fighting in place of the Lore. (and it could used to add more game-modes)
#284
Posted 19 November 2013 - 09:53 AM
Odin, on 19 November 2013 - 02:39 AM, said:
20 games now and i'm at 10 loses and 10 wins. And only 12 deaths. Afking is going faster now that I bought another BJ with my cadet bonus.
What I find interesting is that if I power down at the start the other team totally ignores me even if they're walking right by me. A couple times I've been the last mech alive and forced the other team to cap simply because they couldn't target me and ignored me.
Edited by Sug, 19 November 2013 - 09:57 AM.
#285
Posted 19 November 2013 - 10:47 AM
loupgaroupoilu, on 19 November 2013 - 04:13 AM, said:
With ELO, players will still have roundly W/L ratio around 1. So the factions will also have W/L = 1 .
So... If ELO remain the same. Whatever you do, there will never by any winner - or looser in the CW. So... why CW ?!
What if a large number of competitive units with higher Elo pilots decide to join one faction to work together? I'm thinking that could have an impact, but at the same time, if there are enough other average pilots dropping for both teams at the same time, then maybe the contribution from the higher Elo pilots will be watered down significantly. Hard to say at this point.
#286
Posted 19 November 2013 - 01:57 PM
Sarsaparilla Kid, on 19 November 2013 - 10:47 AM, said:
I don't really know anything about PGI's plans for CW (does anyone?) but Elo should have nothing to do with it.
They've said they're reserving some core worlds for their own uses so we shouldn't be worried about the FC capping Luthien in the first month of CW.
Something as simple as getting a supply bonus or a higher weight limit the closer you fight to your capital or border would help balance things out.
What's that you say? The Capellan's took a bunch of St. Ives planets?? And now there's "civil unrest" on those worlds making them much harder for another faction to hold them?? Mercenaries get a bonus for liberating those worlds?? Wow what a game.
Thing like that opposed to putting nubs against elites to bring the magic 50/50 win/loss ratio to CW will make or break this game. If we can't affect the map there's absolutely no point to CW. PGI might as well give us lobbies and move on to their next game if that's the plan.
Edited by Sug, 19 November 2013 - 01:57 PM.
#287
Posted 19 November 2013 - 03:33 PM
#288
Posted 20 November 2013 - 05:17 PM
Nightfire, on 19 November 2013 - 01:14 AM, said:
I will instead posit this main point for further discussion:
A) It has been stated (paraphrased, don't have the time to search for an exact quote) that the less directly the variable is influenced then the more samples it takes to make a desired outcome observable. In MWO Elo implementation, I think it was stated it takes upwards of 10,000 matches to somewhat accurately determine a players Elo rank. I'm not trying to put words in people's mouths here, working from an imprecise memory is all.
In any random system, given an infinite sample, you will find a sample to match any given criteria. (That part is fact gentlemen, disputing that will not be responded to). The larger the sample size, the more likely you are to match a given criteria.
The method here is to be matched with enough teams, close enough to the desired outcome to place you at your "correct" Elo rank.
My postulation to Elo defenders: At what sample size do you admit that you have as much practical probability of getting that series of matches by pure chance as opposed to by intelligent design?
200-400 matches is enough, at that point you'll be pretty close to your correct Elo - though your Elo should fluctuate as you gain skill or switch to less viable mechs.
There is not a sample size at which random chance would put you at your 'correct' Elo. It would put everyone, I mean absolutely 100% everyone, at exactly average (around 1300 or 1400) after about 400 or 500 matches. It would not create any sort of a curve save for people with only 100 or 200 matches who would be a short distance up or down the scale. There would not be people in the 2,000 range.
if Elo was random there would be no curve. Just a single slightly thick pillar and a sprinkling of sub-200 match people. Also if Elo was random it would disprove the entire field of probability theory and statistics.
Most consumer research takes ideally thousands of samples. At tens of thousands of samples you get the most incredibly detailed mosaics of people and data - for example if I had access to 100% of your telemetry - mechs you own, what you buy and when, how often you change mech types and where, exactly how you fight, etc. I could reasonably accurately predict your age, gender, financial status, which way you vote, marital or familial status, a lot of details about what kind of consumer you are. I'd even have a decent shot at predicting your religious affiliation.
Human behavior isn't random or unlimited. We like to think we're totally unpredictable but that's far, far from the truth. We're pretty consistent little creatures of habit and we fall within a pretty narrow set of behaviors. The difference in the performance of your teammates is actually a pretty tiny slice of behaviors to choose from and leads to a pretty narrow set of results. Picking your impact on that out of the mix isn't nearly as tough as people would like to think.
#289
Posted 20 November 2013 - 05:25 PM
Joseph Mallan, on 19 November 2013 - 04:27 AM, said:
HA! I started gaming in 1981. I still have my badly chipped red D&D dice in my dice box. Out of the countless characters I've created for god knows how many games I even still remember my first character. How goofy is that?
I had a friend sell the entire party into slavery one game. It was actually pretty awesome. I also used to *love* to GM Paranoia. It was great - you could tell you had the right group of players and the right adventure when you never really even had to work out the end of the campaign because there was NO WAY they were even going to survive to the half way mark. I recall one player who gunned another down point blank in the back with a laser rifle and then attempted to set it up to look like a freak washing machine maintenance (that was their job) accident.
Joseph Mallan, on 19 November 2013 - 04:27 AM, said:
I totally agree logically, but for me there's just nothing like running around a corner into that PPC+AC sniper Highlander and watching him slowly backing up and JJing, trying to put a hurt on you and then getting that first ~70 point BOOM in his pie-hole, watching you swing back and forth as you barrel down on him and he realizes "this probably can't be fixed" before you blow him into tiny, tiny pieces. It's when you lead an actual charge of people who actually play 'follow the Atlas' and you just flat out roll over a team until they break, scatter and get run down. I enjoy that a lot.
#290
Posted 21 November 2013 - 11:01 AM
MischiefSC, on 20 November 2013 - 05:25 PM, said:
Ugh...I'm also around that timeframe...where did the years go?
MischiefSC, on 20 November 2013 - 05:25 PM, said:
Surprising how few actually follow the Atlas. Been in matches where the Atlas goes one way and everyone else goes another. Last time that happened, I went with the Atlas by myself, in a Jester no less, and we did a pretty good flank, though the win was not a sure thing, ending up with me and one other friendly left.
#291
Posted 21 November 2013 - 11:47 AM
MischiefSC, on 20 November 2013 - 05:25 PM, said:
HA! I started gaming in 1981. I still have my badly chipped red D&D dice in my dice box. Out of the countless characters I've created for god knows how many games I even still remember my first character. How goofy is that?
#292
Posted 24 November 2013 - 03:05 PM
#294
Posted 24 November 2013 - 03:17 PM
Crazy Billy Joe Bob, on 24 November 2013 - 03:05 PM, said:
Because I am still having a lot of fun, even as a 100% solo player?
How about you? Why are you still here if you hate the game so much? Why are you wasting your time here?
Oh! I see. You don't have a life and have nothing else better to do.
Edited by Mystere, 24 November 2013 - 03:18 PM.
#295
Posted 25 November 2013 - 04:34 AM
First of all, I'm a team player. I try to work in a coordinated fashion with my team and my loadouts are built around team play. This means I do very well when teams act as teams but very poorly when everyone wants to play "hero". Logic suggests that I should find a group to play with but, truth be told, I like staying in PUGS. Unlike many I don't see a need for there to be separate pug/premade queues. This just promotes an atmosphere where "PUG" is synonymous with "disorganized group of freebooters". There is no reason why PUG players can't learn to work as coordinated teams. "Pugger" does not mean "{Dezgra}" and I refuse to accept any intimation that it does.
On with the show:
ELO does seem to need some fine tuning at best or a complete re-working at worst. My observations;
I'll start the day and have a good run with some evenly matched teams and have a good time. A REAL good time with evenly matched battles.
Then depending on the results of those good matches, some odd things happen.
Suddenly I'll lose a couple games in a row and I seem to get placed with better players on average and the roflstomps from my side starts and I get a lot of easy wins.
OR
Suddenly I'll win a couple games and I'll get put in teams where the players can't find their backsides with both hands (not "bad players" just a lot of obviously newbie moves).
The system seems like it's not trying to build me good matches but is putting me on a seesaw trying to maintain the 50% win/loss ratio.
The result is that the game is putting me on an emotional rollercoaster. I'll have a couple good games and win or lose I'll have had some serious fun and feel challenged. Then I get a team that just roflstomps and it's very boring. Then I'll get a couple of good matches and then I'll be roflstomped 2 or 3 times and become absolutely furious with the teams I'm on and the matchmaker.
I don't think this is a good way to try to obtain a 50% win/loss ratio because it sort of means I'm 66% dissatisfied with my gaming experience. I only actually enjoy about 33% of the matches (truth be told the figure is lower).
I am fully aware that for a portion of the playerbase the roflstomps are fun and truth be told I enjoy the occasional roflstomp myself but in the end there's just no consistency on offer.
So technically the ELO is doing it's job.and I have 2027 wins to 1992 losses (1.02 W/L). The way it seems to be doing it, however, doesn't really seem like a workable solution to me.
I would simply much rather be losing 75% of evenly matched games by a narrow margin instead of being handed wins on a silver platter 33% of the time and being almost guaranteed a loss 33% of the time to maintain that 50/50 W/L ratio.
Don't get me wrong, I commend PGI for striving for the 50%. MOST game developers accept that players "are happy with" winning 75% of the time and try to structure the game so their long-term/hardcore players achieve that goal at the expense of casual players. I urge them to keep to the 50% ideal but there really has to be a batter way than this.
Edited by Greyboots, 25 November 2013 - 04:39 AM.
#296
Posted 25 November 2013 - 04:46 AM
This weekend I dropped solo an in groups (4 mans). My solo play was as usual... wildly chaotic! But in Teams dependent on who was in the drop, my performance ranged from useless to god like! As players cycled through my teams the enemy was tougher or weaker.
When I teamed up with Goose, DaZur and his son our Elos must have been fairly close cause we had a bunch of good drops that were sprinkled with some "what just happened?" drops. Diamond Sharks rolled us pretty neatly, I wanna see you guys again! As a tough known opponent can be a great friendly rival
#297
Posted 25 November 2013 - 04:52 PM
Greyboots, on 25 November 2013 - 04:34 AM, said:
ELO does seem to need some fine tuning at best or a complete re-working at worst. My observations;
I'll start the day and have a good run with some evenly matched teams and have a good time. A REAL good time with evenly matched battles.
Then depending on the results of those good matches, some odd things happen.
Suddenly I'll lose a couple games in a row and I seem to get placed with better players on average and the roflstomps from my side starts and I get a lot of easy wins.
OR
Suddenly I'll win a couple games and I'll get put in teams where the players can't find their backsides with both hands (not "bad players" just a lot of obviously newbie moves).
The system seems like it's not trying to build me good matches but is putting me on a seesaw trying to maintain the 50% win/loss ratio.
The result is that the game is putting me on an emotional rollercoaster. I'll have a couple good games and win or lose I'll have had some serious fun and feel challenged. Then I get a team that just roflstomps and it's very boring. Then I'll get a couple of good matches and then I'll be roflstomped 2 or 3 times and become absolutely furious with the teams I'm on and the matchmaker.
It's not necessarily the Elo that needs tuning, but the Matchmaker that then uses Elo to bring pilots together in a team. I have a feeling that you are experiencing this rollercoaster partly for population-related issues. If you're not playing during primetime hours, whenever those might be, when the population on the MWO servers is at its peak, then the Matchmaker doesn't necessarily have enough pilots with a similar Elo rating, or at least within a limited range, to match together within a couple minutes of hitting the Launch button, so it ends up widening the search parameters and pulls in pilots of more disparate Elo rankings, which accounts for the extreme differences in skill you're seeing in your teammates. I think some new game releases, such as BF4, have taken a toll on the MWO player base for now, until something "fresh" comes along to bring them back into MWO, so even primetime hours could be looking on the skimpy side.
#298
Posted 26 November 2013 - 02:28 AM
Edited by SgtKinCaiD, 26 November 2013 - 02:28 AM.
#299
Posted 26 November 2013 - 03:29 AM
Nick Makiaveli, on 11 November 2013 - 04:24 PM, said:
???
No. The idea behind Elo is to pair player of equal skill against each other.
If it would work as you said, you would be forced to play against player FAR over your own ELO if you win a few times in a row. And this is not the idea behind it.
What you mean is the idea that two player with EQUAL Elo have a 50% chance to win. And the whole calculation of the Elo is based on this.
But with Elo it can happen that:
You have a W/L ratio of 2 and still sit on the base ELO
Or that you have a W/L ratio of 0.5 and be one of the top player.
It all depend on your opponents and their Elo. If you are always playing against player with higher Elo, you need way less wins to gain Elo.
And the top Elo player normaly need a higher w/L ratio to keep their Elo. Because its more realistic to play against someone with a lower Elo as his own. And a lose would cost way more in this case.
6 user(s) are reading this topic
0 members, 6 guests, 0 anonymous users