HomeLatest imagesRegisterLog in



Share
 

 Winter 2013 ELO Ratings [FINAL]

View previous topic View next topic Go down 
Go to page : 1, 2, 3, 4, 5, 6, 7  Next
AuthorMessage
Nichigo


Nichigo


Team : Super Friends

Posts : 722

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 94-0-22

Battle Arena Silver
Squire
Winter 2013 ELO Ratings [FINAL] 2ntv8so
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Thu Nov 15, 2012 10:40 pm
Sheet is updated in real time.
If your name is wrongly dormed, PM ChaosSatella and he will correct it.




The sheet where all wars and ranks are stored can be viewed here:
https://docs.google.com/spreadsheet/ccc?key=0ArM63L8AWf7OdHVTTXg1YlVyb2cydGlzc0U3NnBUd2c&pli=1#gid=1

Explanation of the current ELO system (by ChaosSatella) can be found below:

Spoiler:


Last edited by Nichigo on Tue Dec 18, 2012 9:52 pm; edited 23 times in total
Raphael


Raphael


Team : ShadowBright

Posts : 15

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 10-0-3

Winter 2013 ELO Ratings [FINAL] 2ntv8so
Squire
Winter 2013 ELO Ratings [FINAL] 2ntv8so
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Thu Nov 15, 2012 10:44 pm
lol

It's really not like I'm pointing out someone is clearly favoring his own team
Nichigo


Nichigo


Team : Super Friends

Posts : 722

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 94-0-22

Battle Arena Silver
Squire
Winter 2013 ELO Ratings [FINAL] 2ntv8so
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Thu Nov 15, 2012 10:51 pm
There's no favoritism involved. Everything is calculated. I don't have any hand in how the rankings distribute themselves. I simply take the results of the war, plug them into a calculator, and it tells me how to adjust the rankings.

If you want, we can start all over again and you can see how the rankings develop over time...would you like that? Razz
NickV


NickV

Administrator

Team : Toxic Flamingos

Posts : 1411

Duel Points : 100

Duel Season : 1-0-2

Duel Career : 45-0-24

YGOPro Name : NickV

Team Wars Silver
Squire
Weekly Bash Silver
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Thu Nov 15, 2012 11:51 pm
You know you lose elo when you lose a war.. right? There's no way you have the same elo with 4-4 as a team with 4-0.
Shubz


Shubz


Team : Trojan

Posts : 786

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 28-0-15

Weekly Bash Silver
Squire
Winter 2013 ELO Ratings [FINAL] 2ntv8so
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 12:21 am
http://ebattles.net/news.php
try that site..ELO isnt the only thing that should be taken from the results..there are other factors
Zacky


Zacky


Posts : 589

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 34-0-2

YGOPro Name : Zacky

Weekly Bash Gold
Squire
Winter 2013 ELO Ratings [FINAL] 2ntv8so
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 12:34 am
Everyone would love to see these " calculations " mate :3
cause from what it looks ... sigh .... i hope u post them up anyway .
Nichigo


Nichigo


Team : Super Friends

Posts : 722

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 94-0-22

Battle Arena Silver
Squire
Winter 2013 ELO Ratings [FINAL] 2ntv8so
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 2:12 am
: sigh :

This would be much easier if all records were just erased and we started ELO with a fresh slate. Then everyone would see how their rankings progress, and understand how performances are being evaluated. The main problem is that people are seeing these results which have been calculated IN ORDER based on all 18 wars that have been completed since wars began....and wondering how they got there.

I will admit, I was surprised to see that a 4-4 team was tied with 4-0. But ELO doesnt' just make a lose a lose and a win a win. It takes into account the performances of both teams as well as their current rating. If a 1600 rated team plays a 1400 rated team, they're not only expected to win, they're expected to destroy them. If the 1600 rated team only wins by 1 game, they did not meet their expected performance rating, and they will drop.

The sole purpose of this rating system is for other teams to gauge the performance/strength of other teams. It's not a ranking by wins or losses. The reason we can't do a striaght win/loss ranking system is because the way wars are conducted, teams can war AS MANY TIMES agianst ANY TEAM they want. How do you compare a 2-0 winning record to a 15-5 record? Does the 2-0 team get a higher ranking just cuz they have no losses? But they've only played 2 games! You cannot reasonably compare the two records and come to an accurate ranking. The only way you could do that is to have a duel schedule for each season where each team played every other team the same amount of times. That would be even more complicated strcture wise (we'd have to lock teams out of joining mid season...availability would be a factor, etc).

ELO fixes this by rating people based on their performance against people of other ratings. You can't just look at a W/L record and say "it should be THIS". It does'nt work that way. W/L doesnt' tell you much...especially in this setting.

I will attempt to come up with a chart/table/something to show how these ratings came to be. It will take a crap load of time on my part, but i believe in the system, and am willing to put in the work to show you guys that it's much more accurate than win-loss.

As a final note...win-loss right now is an INVALID way of placing people. As stated before, you cannot compare people with different records and also a different amount of total matches. You have to take other things into account. Even if ELO weren't a good way...W/L is still invalid imo.
NickV


NickV

Administrator

Team : Toxic Flamingos

Posts : 1411

Duel Points : 100

Duel Season : 1-0-2

Duel Career : 45-0-24

YGOPro Name : NickV

Team Wars Silver
Squire
Weekly Bash Silver
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 3:17 am
ELO is a very good system, I'm just sure you didn't properly calculate it.

I'm pretty sure every team played for round 1 (let's call it like that). I think there were 6 teams so let's just say 3 go +10 and 3 go -10 (so 3 are now 1510 and 3 are 1490).

Now, if a +10 team would beat a -10, they'd go +5, other goes -5.
If a +10 team would beat a +10 team, they'd go +10.
If a -10 team would beat a -10 team, they'd go +10, other goes -10.
If a -10 team would beat a +10 team, they'd go +15, other goes -15.

That's a rough sketch (made it easy to understand, real elo works with points behind digits aswell) of how ELO works so I still cannot believe Superfriends has the same elo as Shadowbright. You won against LordsofDarkness (low seeded), UltimateBros (low seeded), Blizzard Comets (low seeded) and Genesis (high seeded). Lost against Trojan (low seeded), Shadowbright (high seeded), Dreamstate (at the time both teams were 1500) and one unknown.

As you can see, a lot of low seeds so you simply cannot have that much points as opposed to Shadowbright, who won against a low seed, a high seed, a team where both were 1500 and one unknown, while not dropping any points.
Nichigo


Nichigo


Team : Super Friends

Posts : 722

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 94-0-22

Battle Arena Silver
Squire
Winter 2013 ELO Ratings [FINAL] 2ntv8so
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 3:31 am
^^You're only looking a solid win vs. a solid loss. That would be fine if every war was determine by one duel...but they're not.

As i stated in my main post, I'm calculating ELO to take into consideration margin of victory. Winning isn't the only factor. In order for a +10 team to go up after beating a -10 team, they need to win by a certain margin. The +10 team is already *expected* to win based on their ratings, but the difference in ratings requires them to win by more.

Case in point, I did a rough calculation for the top rated teams (SB/SF) against the lowest ranked team (Lethal Bullets), and, in an 8 heart war, SB/SF would need to win 8-4 in order to go up in ranking. If LB gets in more wins, the higher team's ranking will go down.

This emphasizes ratings differences. You can't just play a team that's severely lower/higher than you, and expect the results to be simply "you won or lost". This system goes much deeper into assessing team's relative strengths.
Sabbaticus


Sabbaticus

Founder

Team : Dreamstate

Posts : 1559

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 32-0-5

YGOPro Name : Sabbaticus

Team Wars Silver
Order of Chaos
Weekly Bash Gold
3rd Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 3:42 am
Nichigo wrote:
Winning isn't the only factor.
As it stands, it seems winning or loosing isn't a factor at all. More like it's completely irrelevant if you win or loose a war.
Nichigo


Nichigo


Team : Super Friends

Posts : 722

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 94-0-22

Battle Arena Silver
Squire
Winter 2013 ELO Ratings [FINAL] 2ntv8so
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 3:47 am
It's not irrelevent. You're expected to win or lose depending on the difference between yours or your opponent's ranking.

What's more relevent is by how much you win or lose by...in other words, every duel counts. It gives more meaning to straight sweeps and close wars. Furthermore, it adjusts ratings based on whether it was expected to be a close game, a sweep, or a moderate win margin.

I'm currently working on a progression sheet for the current ELO ratings. Hopefully up by tonight. Hold onto your panties people.
NickV


NickV

Administrator

Team : Toxic Flamingos

Posts : 1411

Duel Points : 100

Duel Season : 1-0-2

Duel Career : 45-0-24

YGOPro Name : NickV

Team Wars Silver
Squire
Weekly Bash Silver
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 3:57 am
Ah so you're also factoring the duels within the war, didn't notice that.
[KDAp]OneTrackMind


[KDAp]OneTrackMind


Team : [None]

Posts : 136

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 1-0-0

YGOPro Name : OneTrackMind

Winter 2013 ELO Ratings [FINAL] 2ntv8so
Squire
Winter 2013 ELO Ratings [FINAL] 2ntv8so
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 4:14 am
I was actually going to go in-depth of how the ELO system works to clear up the misunderstanding, but the way Nichigo described and explained how he performed the method of calculation tells it all, and so i'm useless. Anyhow, i love the ELO system, it really displays the true dueling strength of an individual team without simply relying on Wins Vs Losses, which is perfect for something like the Warring System. Glad you implemented this system Nichigo, i would love if you produce a table of how you calculate the scores even though i understand how you're gauging the points. Simply because i'm me, and i love it when people put in hard work for my own interests. : P

Keep up with the Good Work Nichigo. Rock On. Headbang
Nichigo


Nichigo


Team : Super Friends

Posts : 722

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 94-0-22

Battle Arena Silver
Squire
Winter 2013 ELO Ratings [FINAL] 2ntv8so
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 12:17 pm
Ok I've completed the new ELO spreadsheet with visible progression in ranks. Please see it here:

https://docs.google.com/spreadsheet/ccc?key=0Ara8lB5sCaqgdGdEZ2RhaDhteXlLU1k4eG93dFUwZ2c&pli=1#gid=3

I made a couple errors in my first run of the ELOs, so these ratings are SLIGHTLY different. They're still around the same values.

Click the "ELO PRogression" tab at the bottom to see how rating changed over wars. The progression list is arranged in chronological order...so the wars happened in that actual order which is very important.

Some things to note:

Expected Margin is the minimum margin a team is expected to win by, given the rating difference. A team that is higher rated is always expected to win. A team that is lower rated is always expected to lose. The difference between the ratings determine how much they're EXPECTED to win or lose by. If you win by a higher margin than expected, you go up because you performed better than expected. The same goes for losses. If you lose by LESS than expected, you also go up because your performance was better.

The case where teams increase in rating after a loss is rare. This typically only happens when a war has a close result when it should have been much more decisive. As you can see on the progression chart, there are only 2 instances where ratings increased after a loss.

Ratings normalize over time. The system smoothens out ratings as more wars are played. You will see large (and sometimes unusual) shifts in ratings while teams are first getting started. Once more games are played, the rating for that team will stabilize at a rating that is more accurately reflective of their average performance (whether that rating is low or high).

This system is meant to make it easy for teams to see how other teams perform relative to themselves by taking actual performance details into account. Win/Loss ratio ranking only hides alot of a team's performance from view because there are an infinite number of ways to win or lose a war. ELO gives you a better look at what a teams actual performance level is.

Pure Win-loss ratio ranking is invalid in a war format where teams have vastly different total number of wars.. There is no way to compare a team that has 2 wars to a team that has 8 wars without considering other factors. ELO allows us to do this. Win-loss ranking would work if every team had the same amount of wars and warred every other team equally (like in sports leagues). Warring on CEA doesn't work like that.


I'm not sure if I can explain it any more clearly. The only other thing I can say is to observe the system over time. It makes more sense as ratings are actually seen going up and down over a season. I will continue to update the progression chart as wars are completed so there is a visible history to how elos are progressing.

PS: I'm adding all this information and the recalculated ELOS to the main post.
Zacky


Zacky


Posts : 589

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 34-0-2

YGOPro Name : Zacky

Weekly Bash Gold
Squire
Winter 2013 ELO Ratings [FINAL] 2ntv8so
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 1:09 pm
Okay i think i finally get this Elo stuff , and maybe theres a little good in it .
A little weird ... but as nichi said , we'll just gv it time ....

But i must also point out the fact that as sabb said this kinda makes winning/losing completely pointless .
EG- if SB wins the next war with lods , we're gonna get negetive points while lods gets positive points ... kinda lame in that aspect .
ShadowServer


ShadowServer


Team : Super Friends

Posts : 520

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 15-0-9

Winter 2013 ELO Ratings [FINAL] 2ntv8so
Squire
Winter 2013 ELO Ratings [FINAL] 2ntv8so
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 1:19 pm
Zacky wrote:
Okay i think i finally get this Elo stuff , and maybe theres a little good in it .
A little weird ... but as nichi said , we'll just gv it time ....

But i must also point out the fact that as sabb said this kinda makes winning/losing completely pointless .
EG- if SB wins the next war with lods , we're gonna get negetive points while lods gets positive points ... kinda lame in that aspect .
If you sweep them as you're expected to for being so many points higher, you will gain points. If you barely manage a win over them you will probably lose some points. You will not automatically get negative points even if you win as you seem to think.
Sabbaticus


Sabbaticus

Founder

Team : Dreamstate

Posts : 1559

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 32-0-5

YGOPro Name : Sabbaticus

Team Wars Silver
Order of Chaos
Weekly Bash Gold
3rd Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 1:29 pm
Well, from the document it seems like Dreamstate actually lost points for winning our war. Then, from what I can gather when looking at the calculations, if we don't sweep UltimateBros in this current war, we will once again loose points for winning our 3rd war in a row. Not even an 8-3 sweep... I think we have to go 8-1 at least to even get any points, or otherwise just go down.

I really think there needs to be an addition point factor based not only on margin, but actually winning or loosing the war. As in +5 for winning, -5 or loosing, or which ever number you care to delegate in accordance with the rest of the scoring.
Nichigo


Nichigo


Team : Super Friends

Posts : 722

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 94-0-22

Battle Arena Silver
Squire
Winter 2013 ELO Ratings [FINAL] 2ntv8so
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 1:58 pm
Sabbaticus wrote:
Well, from the document it seems like Dreamstate actually lost points for winning our war. Then, from what I can gather when looking at the calculations, if we don't sweep UltimateBros in this current war, we will once again loose points for winning our 3rd war in a row. Not even an 8-3 sweep... I think we have to go 8-1 at least to even get any points, or otherwise just go down.

I really think there needs to be an addition point factor based not only on margin, but actually winning or loosing the war. As in +5 for winning, -5 or loosing, or which ever number you care to delegate in accordance with the rest of the scoring.

IDK where you're getting your calculations from, but you will go up with an 8-3 victory. An 8-4 victory will drop you, but only by 2 points.

You guys still aren't getting it. This system gauges your relative strength based on your performance. You don't get "bonus points" for strength because you won a war...this isn't an RPG...this is real life. It's like asking for Extra Credit on a test where you didn't do anything special. The "bonus points" you get come with winning by a larger margin than you were expected to win.

The system isn't biased. It just takes what it's given, and spits out expectations then gives you points based on that. Modifying it isn't going to accomplish anything.

And Sabb, I explained very clearly why you lost points from your last war. If you still cannot understand it after that whole sheet + the explanation under the "NOTES" on the sheet, then I don't know what to say.
Darkness Son


Darkness Son


Team : Super Friends

Posts : 74

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 3-0-1

Winter 2013 ELO Ratings [FINAL] 2ntv8so
Squire
Winter 2013 ELO Ratings [FINAL] 2ntv8so
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 2:06 pm
I really like this system as it takes expected & actual performance into account and is less win-or-lose based. Thank you for putting such a big effort into it Nichi.

I think it would be nice if we could see an "Expected margin of victory" column next to the "Margin of victory" one. This would help understand how points are added/substracted.

Also, in regards to what Sabb said: as I understand it (correct me if I'm wrong), this system evaluates performance only, meaning that it only speaks in terms of "outperforming", "performing as expected" and "underperforming". For example, if a team is expected to win a match by a large margin, and instead they win by a close one, they lose points. And this makes sense because they "underperformed". A way to look at it is to think that they are not in top shape, and an athlete who is not in top shape may get a warning from their sponsors. This is the same. The points you should get by winning, you will get them if you "perform as expected".

All in all this system is much more accurate than the previous one.

P.S.: I think the "Notes" in row 12 are misplaced; they should go in row 13.
Sabbaticus


Sabbaticus

Founder

Team : Dreamstate

Posts : 1559

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 32-0-5

YGOPro Name : Sabbaticus

Team Wars Silver
Order of Chaos
Weekly Bash Gold
3rd Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 2:39 pm
Nichigo wrote:
IDK where you're getting your calculations from, but you will go up with an 8-3 victory. An 8-4 victory will drop you, but only by 2 points.
Ok let me change 1 number in my post. If we win our war 8-4, which is actually by a high margin, we will be doing nothing more than dropping in rankings despite sweeping a team and going 8-4, and never loosing. The fact that we can WIN 2 wars, and LOOSE points for winning both... but then a team is loosing wars (alot of them) and only going up and up. There is clearly something a miss. I understand full well that this is gauged on strength, but there is a VERY high margin for exploitation here. Valkyra told you about this exploitation which you agreed is there, and it's obvious you are one of the teams that even fall under that exploitation. THIS is what I have a problem with. Not the system or method, but there seems to be a missing factor here. Teams who win against weaker teams should get very little points, even if it's just 1 point, but they should not loose points while the loosing team gains points.

Which then brings me to what I said previously... the actual end result of a war is void, winning or loosing actually doesn't matter.
Nichigo


Nichigo


Team : Super Friends

Posts : 722

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 94-0-22

Battle Arena Silver
Squire
Winter 2013 ELO Ratings [FINAL] 2ntv8so
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 3:14 pm
How have we exploited anything when this system just came up...yesterday? I hadn't even figured out how to apply it before then. Watch your implications please.

The current win-loss system can be exploited as well with stronger teams just avoiding other strong teams to stack up wins. Would you say that's a more accurate way to gauge teams' rankings? hm? AT the very least, under ELO, teams will be motivated to duel those of SIMILAR ranking and rank that way. Ranks will still move up and down, and teams will see a large spread of ranks where they will end up dueling other teams.

AS DArkness put much more clearly than I did, this measures performance. Winning and losing are not irrelvent, rather they are integrated into the system as a whole and not the only factor that determines rankings.

If you guys don't like it, i'm open to hearing other solutions. Win-Loss straight up is definitely NOT one of them. If ELO has accomplished anything, it's made at least that much clear.


I would honestly recommend just rolling with this for a while and seeing how it goes. You guys will get a much better feel for it when you see things moving and ranks going up and down. It was a little presumptuous of me to think that everyone would understand it right away with the system applied to all wars up to this point in time =|
Sabbaticus


Sabbaticus

Founder

Team : Dreamstate

Posts : 1559

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 32-0-5

YGOPro Name : Sabbaticus

Team Wars Silver
Order of Chaos
Weekly Bash Gold
3rd Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 4:29 pm
I said you fall within the exploitation, not that you were exploiting. No implication was given.

I know full well that simple win-loss can be exploited. I'm also not saying that this system is bad, only that there is a flaw in it. I do prefer the ELO system but as you have just confirmed, there is a highly exploitable feature in it, 2 actually, but I won't get into. So I will make my suggestions for patching those up.

Winning a war should never result in a negative score. Either 1-2 points, or 0 even, but never negative. Because then you essentially loose when winning, or you are otherwise just playing to loose less. An example would be us having to literally sweep a team just to not loose points, despite winning the war by a large margin. We loose by winning, winning well even.

The other, would be to actually give a defining number to a win and loss. As I said earlier... +5 for a win, -5 for a loss, or which ever number you which to associate with that based on the rest of the scoring. That will truly integrate winning and loosing with performance, because as it stands, you can loose by winning, and win by loosing. I'm all for performance ratings, but the actual winning of a war seems to have been lost, and essentially that is a fundamental part of wars... actually being victorious.
Zacky


Zacky


Posts : 589

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 34-0-2

YGOPro Name : Zacky

Weekly Bash Gold
Squire
Winter 2013 ELO Ratings [FINAL] 2ntv8so
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 4:36 pm
A lot of debate going on about the Elo system , and from looking at sabb's idea and from a little place in my head i think ive got a suggestion , just a suggestion ..

A team which wins a war gets a said amount of points and the team which loses loses the same amount .
the winning team must be given +1 points for each heart / player they win by ,

This way both win-loss ratio plus the margin of victory is taken into consideration . This is pretty much a simpler version of the elo system but it sound a lot more fairer than it .

Again , this is only a suggestion .
NickV


NickV

Administrator

Team : Toxic Flamingos

Posts : 1411

Duel Points : 100

Duel Season : 1-0-2

Duel Career : 45-0-24

YGOPro Name : NickV

Team Wars Silver
Squire
Weekly Bash Silver
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 4:48 pm
What I read here is that it's smarter for a high ranking team to not war a low ranking team. This means less wars will happen and the top team shouldn't even bother. The winning team should gain points, not lose points.
Zacky


Zacky


Posts : 589

Duel Points : 100

Duel Season : 0-0-0

Duel Career : 34-0-2

YGOPro Name : Zacky

Weekly Bash Gold
Squire
Winter 2013 ELO Ratings [FINAL] 2ntv8so
1st Division

Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11Fri Nov 16, 2012 4:54 pm
Exactly nick , Stronger teams shudnt bother to war weaker ones if the elo system is intact .
Cause u can " expect " whatever the hell u want , but a wipe out isnt always what happens :/
Sponsored content






Winter 2013 ELO Ratings [FINAL] Empty
PostSubject: Re: Winter 2013 ELO Ratings [FINAL]   Winter 2013 ELO Ratings [FINAL] Icon_m11
 

Winter 2013 ELO Ratings [FINAL]

View previous topic View next topic Back to top 
Page 1 of 7Go to page : 1, 2, 3, 4, 5, 6, 7  Next

 Similar topics

-
» FINAL Summer 2013 ELO Ratings
» Winter 2014 ELO Ratings
» CEA Winter 2013 Team War Playoffs
» Career ELO Ratings
» 2012 Winter Playoffs Information! Please Read!

Permissions in this forum:You cannot reply to topics in this forum
Chaos Era Academy :: Recreation Room :: Off-Topic Section :: Archives-