Total Posts:52|Showing Posts:1-30|Last Page
Jump to topic:

The elo should be scrapped

Garbanza
Posts: 1,997
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 3:09:40 AM
Posted: 2 years ago
It was supposed to give the instigator an advantage in scoring and stop noob sniping. But a failure in coding means the challenger has the advantage, which makes people cautious about starting risky debates which is bad.

Also, debaters with high Elo's are too scared of losing because they lose too much.

We should go back to the system of overall wins.
mishapqueen
Posts: 3,995
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 9:12:53 AM
Posted: 2 years ago
Spam bump
You cannot choose whether or not you will live by rules, but you can choose which rules you will live by. --Me

"I was wrong. Squirrels are objectively superior to bunnies in every conceivable dimension."
--Joey

"Silence is golden, duct tape is silver" --PetersSmith

Nunc aut Numquam
Wylted
Posts: 21,167
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 9:53:08 AM
Posted: 2 years ago
At 10/27/2014 3:09:40 AM, Garbanza wrote:
It was supposed to give the instigator an advantage in scoring and stop noob sniping. But a failure in coding means the challenger has the advantage, which makes people cautious about starting risky debates which is bad.

Also, debaters with high Elo's are too scared of losing because they lose too much.

We should go back to the system of overall wins.

None of that I true. I start risking debates all I sometimes go into debates, expecting to lose.
UchihaMadara
Posts: 1,049
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 5:01:07 PM
Posted: 2 years ago
At 10/27/2014 3:09:40 AM, Garbanza wrote:
It was supposed to give the instigator an advantage in scoring and stop noob sniping. But a failure in coding means the challenger has the advantage, which makes people cautious about starting risky debates which is bad.

Also, debaters with high Elo's are too scared of losing because they lose too much.

We should go back to the system of overall wins.

The elo-based leaderboard is far from perfect, but it is much more accurate than the wins-based one.
Subutai
Posts: 3,187
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 5:26:01 PM
Posted: 2 years ago
At 10/27/2014 3:09:40 AM, Garbanza wrote:
It was supposed to give the instigator an advantage in scoring and stop noob sniping. But a failure in coding means the challenger has the advantage, which makes people cautious about starting risky debates which is bad.

Also, debaters with high Elo's are too scared of losing because they lose too much.

We should go back to the system of overall wins.

You make a good point, but your suggesting essentially causes similar problems. If we go to a system of overall wins, a debater who noob snipes to 250 wins would be superior to a great debater with 25 wins.

ELO is not a perfect system, but nothing is. No system is smart enough to give a truly accurate representation of a debater's skill. I'd say, of the systems on DDO now, ELO is the closest, but yes, it does have problems.
I'm becoming less defined as days go by, fading away, and well you might say, I'm losing focus, kinda drifting into the abstract in terms of how I see myself.
debatability
Posts: 1,160
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 5:30:50 PM
Posted: 2 years ago
At 10/27/2014 3:09:40 AM, Garbanza wrote:
It was supposed to give the instigator an advantage in scoring and stop noob sniping. But a failure in coding means the challenger has the advantage, which makes people cautious about starting risky debates which is bad.

Also, debaters with high Elo's are too scared of losing because they lose too much.

We should go back to the system of overall wins.

a system of overall wins would just make the problems you've mentioned worse
UchihaMadara
Posts: 1,049
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 5:34:46 PM
Posted: 2 years ago
At 10/27/2014 5:26:01 PM, Subutai wrote:
At 10/27/2014 3:09:40 AM, Garbanza wrote:
It was supposed to give the instigator an advantage in scoring and stop noob sniping. But a failure in coding means the challenger has the advantage, which makes people cautious about starting risky debates which is bad.

Also, debaters with high Elo's are too scared of losing because they lose too much.

We should go back to the system of overall wins.

You make a good point, but your suggesting essentially causes similar problems. If we go to a system of overall wins, a debater who noob snipes to 250 wins would be superior to a great debater with 25 wins.

ELO is not a perfect system, but nothing is. No system is smart enough to give a truly accurate representation of a debater's skill. I'd say, of the systems on DDO now, ELO is the closest, but yes, it does have problems.

What do you think of factoring some sort of 'debate quality' metric into calculating how much elo is gained or lost?
Subutai
Posts: 3,187
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 5:36:38 PM
Posted: 2 years ago
At 10/27/2014 5:34:46 PM, UchihaMadara wrote:
At 10/27/2014 5:26:01 PM, Subutai wrote:
At 10/27/2014 3:09:40 AM, Garbanza wrote:
It was supposed to give the instigator an advantage in scoring and stop noob sniping. But a failure in coding means the challenger has the advantage, which makes people cautious about starting risky debates which is bad.

Also, debaters with high Elo's are too scared of losing because they lose too much.

We should go back to the system of overall wins.

You make a good point, but your suggesting essentially causes similar problems. If we go to a system of overall wins, a debater who noob snipes to 250 wins would be superior to a great debater with 25 wins.

ELO is not a perfect system, but nothing is. No system is smart enough to give a truly accurate representation of a debater's skill. I'd say, of the systems on DDO now, ELO is the closest, but yes, it does have problems.

What do you think of factoring some sort of 'debate quality' metric into calculating how much elo is gained or lost?

It would be near impossible for humans to make an objective judgment of debate quality, much less a computer. It's a good idea, but it'd be very head to put into code, and even if it were, there'd be an inevitable element of subjectivity in what constitutes debate quality.
I'm becoming less defined as days go by, fading away, and well you might say, I'm losing focus, kinda drifting into the abstract in terms of how I see myself.
dylancatlow
Posts: 12,244
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 5:38:00 PM
Posted: 2 years ago
At 10/27/2014 3:09:40 AM, Garbanza wrote:
It was supposed to give the instigator an advantage in scoring and stop noob sniping. But a failure in coding means the challenger has the advantage, which makes people cautious about starting risky debates which is bad.


First of all, how do you even know this is true? Secondly, there's no reason to scrap the entire elo system because of a failure in coding. Clearly the next step would be to fix the error. And lastly, how big is the advantage?

Also, debaters with high Elo's are too scared of losing because they lose too much.

They are only particularly scared of losing to someone with low elo.


We should go back to the system of overall wins.
UchihaMadara
Posts: 1,049
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 5:42:34 PM
Posted: 2 years ago
At 10/27/2014 5:36:38 PM, Subutai wrote:
At 10/27/2014 5:34:46 PM, UchihaMadara wrote:
At 10/27/2014 5:26:01 PM, Subutai wrote:
At 10/27/2014 3:09:40 AM, Garbanza wrote:
It was supposed to give the instigator an advantage in scoring and stop noob sniping. But a failure in coding means the challenger has the advantage, which makes people cautious about starting risky debates which is bad.

Also, debaters with high Elo's are too scared of losing because they lose too much.

We should go back to the system of overall wins.

You make a good point, but your suggesting essentially causes similar problems. If we go to a system of overall wins, a debater who noob snipes to 250 wins would be superior to a great debater with 25 wins.

ELO is not a perfect system, but nothing is. No system is smart enough to give a truly accurate representation of a debater's skill. I'd say, of the systems on DDO now, ELO is the closest, but yes, it does have problems.

What do you think of factoring some sort of 'debate quality' metric into calculating how much elo is gained or lost?

It would be near impossible for humans to make an objective judgment of debate quality, much less a computer. It's a good idea, but it'd be very head to put into code, and even if it were, there'd be an inevitable element of subjectivity in what constitutes debate quality.

Fair enough. I mean, people already usually give their opinions on the debate's quality within their RFD's (i.e. stupid, bad, good, great, excellent, fascinating), so there could be some sort of 1-10 slider scale thing to try quantifying that, which would make it much easier to incorporate into a formula. Subjectivity is gonna be there no matter what, but there could be certain benchmarks like 0 for full FF, and 10 for 'front page material'.
Subutai
Posts: 3,187
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 5:44:46 PM
Posted: 2 years ago
At 10/27/2014 5:42:34 PM, UchihaMadara wrote:
At 10/27/2014 5:36:38 PM, Subutai wrote:
At 10/27/2014 5:34:46 PM, UchihaMadara wrote:
At 10/27/2014 5:26:01 PM, Subutai wrote:
At 10/27/2014 3:09:40 AM, Garbanza wrote:
It was supposed to give the instigator an advantage in scoring and stop noob sniping. But a failure in coding means the challenger has the advantage, which makes people cautious about starting risky debates which is bad.

Also, debaters with high Elo's are too scared of losing because they lose too much.

We should go back to the system of overall wins.

You make a good point, but your suggesting essentially causes similar problems. If we go to a system of overall wins, a debater who noob snipes to 250 wins would be superior to a great debater with 25 wins.

ELO is not a perfect system, but nothing is. No system is smart enough to give a truly accurate representation of a debater's skill. I'd say, of the systems on DDO now, ELO is the closest, but yes, it does have problems.

What do you think of factoring some sort of 'debate quality' metric into calculating how much elo is gained or lost?

It would be near impossible for humans to make an objective judgment of debate quality, much less a computer. It's a good idea, but it'd be very head to put into code, and even if it were, there'd be an inevitable element of subjectivity in what constitutes debate quality.

Fair enough. I mean, people already usually give their opinions on the debate's quality within their RFD's (i.e. stupid, bad, good, great, excellent, fascinating), so there could be some sort of 1-10 slider scale thing to try quantifying that, which would make it much easier to incorporate into a formula. Subjectivity is gonna be there no matter what, but there could be certain benchmarks like 0 for full FF, and 10 for 'front page material'.

That's a very good suggestion that'd be easy to implement into the ELO code. Of course, there'd still be subjectivity, but any system will have that. I support this system.
I'm becoming less defined as days go by, fading away, and well you might say, I'm losing focus, kinda drifting into the abstract in terms of how I see myself.
dylancatlow
Posts: 12,244
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 5:51:19 PM
Posted: 2 years ago
At 10/27/2014 5:34:46 PM, UchihaMadara wrote:
At 10/27/2014 5:26:01 PM, Subutai wrote:
At 10/27/2014 3:09:40 AM, Garbanza wrote:
It was supposed to give the instigator an advantage in scoring and stop noob sniping. But a failure in coding means the challenger has the advantage, which makes people cautious about starting risky debates which is bad.

Also, debaters with high Elo's are too scared of losing because they lose too much.

We should go back to the system of overall wins.

You make a good point, but your suggesting essentially causes similar problems. If we go to a system of overall wins, a debater who noob snipes to 250 wins would be superior to a great debater with 25 wins.

ELO is not a perfect system, but nothing is. No system is smart enough to give a truly accurate representation of a debater's skill. I'd say, of the systems on DDO now, ELO is the closest, but yes, it does have problems.

What do you think of factoring some sort of 'debate quality' metric into calculating how much elo is gained or lost?

I like this idea a lot. Any ranking system which places Mikal at #1 is clearly in need of some adjustment.
fazz
Posts: 1,617
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 6:03:13 PM
Posted: 2 years ago
At 10/27/2014 5:51:19 PM, dylancatlow wrote:

I like this idea a lot. Any ranking system which places Mikal at #1 is clearly in need of some adjustment.

For he is the chosen one: https://www.youtube.com...
dylancatlow
Posts: 12,244
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 6:05:12 PM
Posted: 2 years ago
At 10/27/2014 6:03:13 PM, fazz wrote:
At 10/27/2014 5:51:19 PM, dylancatlow wrote:

I like this idea a lot. Any ranking system which places Mikal at #1 is clearly in need of some adjustment.

For he is the chosen one: https://www.youtube.com...

Sorry, his elo is only 9000, not over 9000.
fazz
Posts: 1,617
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 6:09:05 PM
Posted: 2 years ago
At 10/27/2014 6:05:12 PM, dylancatlow wrote:
At 10/27/2014 6:03:13 PM, fazz wrote:
At 10/27/2014 5:51:19 PM, dylancatlow wrote:

I like this idea a lot. Any ranking system which places Mikal at #1 is clearly in need of some adjustment.

For he is the chosen one: https://www.youtube.com...

Sorry, his elo is only 9000, not OVER 9000.!!!

No. There is an ELO cap. It went over 6 months ago.
dylancatlow
Posts: 12,244
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 6:10:22 PM
Posted: 2 years ago
At 10/27/2014 6:09:05 PM, fazz wrote:
At 10/27/2014 6:05:12 PM, dylancatlow wrote:
At 10/27/2014 6:03:13 PM, fazz wrote:
At 10/27/2014 5:51:19 PM, dylancatlow wrote:

I like this idea a lot. Any ranking system which places Mikal at #1 is clearly in need of some adjustment.

For he is the chosen one: https://www.youtube.com...

Sorry, his elo is only 9000, not OVER 9000.!!!

No. There is an ELO cap. It went over 6 months ago.

In any case, ty for fixing my post.
fazz
Posts: 1,617
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 6:13:28 PM
Posted: 2 years ago
At 10/27/2014 6:10:22 PM, dylancatlow wrote:
At 10/27/2014 6:09:05 PM, fazz wrote:
At 10/27/2014 6:05:12 PM, dylancatlow wrote:
At 10/27/2014 6:03:13 PM, fazz wrote:
At 10/27/2014 5:51:19 PM, dylancatlow wrote:

I like this idea a lot. Any ranking system which places Mikal at #1 is clearly in need of some adjustment.

For he is the chosen one: https://www.youtube.com...

Sorry, his elo is only 9000, not OVER 9000.!!!

No. There is an ELO cap. It went over 6 months ago.

In any case, ty for fixing my post.

;P
bluesteel
Posts: 12,301
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 8:21:25 PM
Posted: 2 years ago
At 10/27/2014 5:26:01 PM, Subutai wrote:
At 10/27/2014 3:09:40 AM, Garbanza wrote:
It was supposed to give the instigator an advantage in scoring and stop noob sniping. But a failure in coding means the challenger has the advantage, which makes people cautious about starting risky debates which is bad.

Also, debaters with high Elo's are too scared of losing because they lose too much.

We should go back to the system of overall wins.

You make a good point, but your suggesting essentially causes similar problems. If we go to a system of overall wins, a debater who noob snipes to 250 wins would be superior to a great debater with 25 wins.

ELO is not a perfect system, but nothing is. No system is smart enough to give a truly accurate representation of a debater's skill. I'd say, of the systems on DDO now, ELO is the closest, but yes, it does have problems.

How is it a good point? It's not accurate.... If you debate someone with 2000 ELO, you gain 35 for a win as instigator and 15 for a win as contender. There's no "glitch" that gives an advantage to the contender. The advantage is the same it's always been: contender generally gets to evaluate the Instigator's case, or at least do a bunch of research before deciding whether to accept. That's why the instigator advantage and contender penalty exist in the ELO calculation.

The claims that we should tether ELO to "debate quality" also ignore how ELO is *already* calculated: it takes account of your opponent's ELO. That's the best proxy for debate quality. In fact, it's more important. If you face a tougher opponent, you get more for a win and lose less for a loss. How difficult your opponent is matters more than some other subjective rating of debate quality.
You can't reason someone out of a position they didn't reason themselves into - Jonathan Swift (paraphrase)
xXCryptoXx
Posts: 5,000
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 8:36:16 PM
Posted: 2 years ago
At 10/27/2014 3:09:40 AM, Garbanza wrote:
It was supposed to give the instigator an advantage in scoring and stop noob sniping. But a failure in coding means the challenger has the advantage, which makes people cautious about starting risky debates which is bad.

Also, debaters with high Elo's are too scared of losing because they lose too much.

We should go back to the system of overall wins.

I only lose about 10 ELO from a loss. Consequently, I only gain about 10 ELO also.
Nolite Timere
Subutai
Posts: 3,187
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 9:40:31 PM
Posted: 2 years ago
At 10/27/2014 8:21:25 PM, bluesteel wrote:
At 10/27/2014 5:26:01 PM, Subutai wrote:
At 10/27/2014 3:09:40 AM, Garbanza wrote:
It was supposed to give the instigator an advantage in scoring and stop noob sniping. But a failure in coding means the challenger has the advantage, which makes people cautious about starting risky debates which is bad.

Also, debaters with high Elo's are too scared of losing because they lose too much.

We should go back to the system of overall wins.

You make a good point, but your suggesting essentially causes similar problems. If we go to a system of overall wins, a debater who noob snipes to 250 wins would be superior to a great debater with 25 wins.

ELO is not a perfect system, but nothing is. No system is smart enough to give a truly accurate representation of a debater's skill. I'd say, of the systems on DDO now, ELO is the closest, but yes, it does have problems.

How is it a good point? It's not accurate.... If you debate someone with 2000 ELO, you gain 35 for a win as instigator and 15 for a win as contender. There's no "glitch" that gives an advantage to the contender. The advantage is the same it's always been: contender generally gets to evaluate the Instigator's case, or at least do a bunch of research before deciding whether to accept. That's why the instigator advantage and contender penalty exist in the ELO calculation.

The claims that we should tether ELO to "debate quality" also ignore how ELO is *already* calculated: it takes account of your opponent's ELO. That's the best proxy for debate quality. In fact, it's more important. If you face a tougher opponent, you get more for a win and lose less for a loss. How difficult your opponent is matters more than some other subjective rating of debate quality.

Why are you arguing with me? I agree with you. I agree that ELO is the best system DDO currently has. I was simply saying that ELO is not perfect, and that a truly perfect, objective measure of debate quality doesn't exist.
I'm becoming less defined as days go by, fading away, and well you might say, I'm losing focus, kinda drifting into the abstract in terms of how I see myself.
Imperfiect
Posts: 128
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 10:07:35 PM
Posted: 2 years ago
At 10/27/2014 8:21:25 PM, bluesteel wrote:
At 10/27/2014 5:26:01 PM, Subutai wrote:
At 10/27/2014 3:09:40 AM, Garbanza wrote:
It was supposed to give the instigator an advantage in scoring and stop noob sniping. But a failure in coding means the challenger has the advantage, which makes people cautious about starting risky debates which is bad.

Also, debaters with high Elo's are too scared of losing because they lose too much.

We should go back to the system of overall wins.

You make a good point, but your suggesting essentially causes similar problems. If we go to a system of overall wins, a debater who noob snipes to 250 wins would be superior to a great debater with 25 wins.

ELO is not a perfect system, but nothing is. No system is smart enough to give a truly accurate representation of a debater's skill. I'd say, of the systems on DDO now, ELO is the closest, but yes, it does have problems.

How is it a good point? It's not accurate.... If you debate someone with 2000 ELO, you gain 35 for a win as instigator and 15 for a win as contender. There's no "glitch" that gives an advantage to the contender. The advantage is the same it's always been: contender generally gets to evaluate the Instigator's case, or at least do a bunch of research before deciding whether to accept. That's why the instigator advantage and contender penalty exist in the ELO calculation.

The claims that we should tether ELO to "debate quality" also ignore how ELO is *already* calculated: it takes account of your opponent's ELO. That's the best proxy for debate quality. In fact, it's more important. If you face a tougher opponent, you get more for a win and lose less for a loss. How difficult your opponent is matters more than some other subjective rating of debate quality.

Sorry but I disagree.
bsh1
Posts: 27,503
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 10:19:16 PM
Posted: 2 years ago
At 10/27/2014 8:21:25 PM, bluesteel wrote:

How is it a good point? It's not accurate.... If you debate someone with 2000 ELO, you gain 35 for a win as instigator and 15 for a win as contender. There's no "glitch" that gives an advantage to the contender. The advantage is the same it's always been: contender generally gets to evaluate the Instigator's case, or at least do a bunch of research before deciding whether to accept. That's why the instigator advantage and contender penalty exist in the ELO calculation.

^ This.
Live Long and Prosper

I'm a Bish.


"Twilight isn't just about obtuse metaphors between cannibalism and premarital sex, it also teaches us the futility of hope." - Raisor

"[Bsh1] is the Guinan of DDO." - ButterCatX

Follow the DDOlympics
: http://www.debate.org...

Open Debate Topics Project: http://www.debate.org...
bluesteel
Posts: 12,301
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 10:29:39 PM
Posted: 2 years ago
At 10/27/2014 9:40:31 PM, Subutai wrote:
At 10/27/2014 8:21:25 PM, bluesteel wrote:
At 10/27/2014 5:26:01 PM, Subutai wrote:
At 10/27/2014 3:09:40 AM, Garbanza wrote:
It was supposed to give the instigator an advantage in scoring and stop noob sniping. But a failure in coding means the challenger has the advantage, which makes people cautious about starting risky debates which is bad.

Also, debaters with high Elo's are too scared of losing because they lose too much.

We should go back to the system of overall wins.

You make a good point, but your suggesting essentially causes similar problems. If we go to a system of overall wins, a debater who noob snipes to 250 wins would be superior to a great debater with 25 wins.

ELO is not a perfect system, but nothing is. No system is smart enough to give a truly accurate representation of a debater's skill. I'd say, of the systems on DDO now, ELO is the closest, but yes, it does have problems.

How is it a good point? It's not accurate.... If you debate someone with 2000 ELO, you gain 35 for a win as instigator and 15 for a win as contender. There's no "glitch" that gives an advantage to the contender. The advantage is the same it's always been: contender generally gets to evaluate the Instigator's case, or at least do a bunch of research before deciding whether to accept. That's why the instigator advantage and contender penalty exist in the ELO calculation.

The claims that we should tether ELO to "debate quality" also ignore how ELO is *already* calculated: it takes account of your opponent's ELO. That's the best proxy for debate quality. In fact, it's more important. If you face a tougher opponent, you get more for a win and lose less for a loss. How difficult your opponent is matters more than some other subjective rating of debate quality.

Why are you arguing with me? I agree with you. I agree that ELO is the best system DDO currently has. I was simply saying that ELO is not perfect, and that a truly perfect, objective measure of debate quality doesn't exist.

I responded to you because random people complain about the ELO system all the time, even while not really knowing the formula. But regular users rarely respond to them with "good point." I just wanted to make sure you knew why OP's point was not a good one.
You can't reason someone out of a position they didn't reason themselves into - Jonathan Swift (paraphrase)
Imperfiect
Posts: 128
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 10:30:02 PM
Posted: 2 years ago
At 10/27/2014 10:19:16 PM, bsh1 wrote:
At 10/27/2014 8:21:25 PM, bluesteel wrote:

How is it a good point? It's not accurate.... If you debate someone with 2000 ELO, you gain 35 for a win as instigator and 15 for a win as contender. There's no "glitch" that gives an advantage to the contender. The advantage is the same it's always been: contender generally gets to evaluate the Instigator's case, or at least do a bunch of research before deciding whether to accept. That's why the instigator advantage and contender penalty exist in the ELO calculation.

^ This.

Sorry but I disagree.
YYW
Posts: 36,263
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 10:32:16 PM
Posted: 2 years ago
At 10/27/2014 3:09:40 AM, Garbanza wrote:
It was supposed to give the instigator an advantage in scoring

No.

and stop noob sniping.

It wasn't supposed to do that either.

But a failure in coding means the challenger has the advantage, which makes people cautious about starting risky debates which is bad.

Also incorrect. He or she who has the most skill is who has the advantage.

Also, debaters with high Elo's are too scared of losing because they lose too much.

There may be some truth to that.

We should go back to the system of overall wins.

We still have the old percentage system.
Tsar of DDO
Subutai
Posts: 3,187
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 10:34:11 PM
Posted: 2 years ago
At 10/27/2014 10:29:39 PM, bluesteel wrote:
At 10/27/2014 9:40:31 PM, Subutai wrote:
At 10/27/2014 8:21:25 PM, bluesteel wrote:
At 10/27/2014 5:26:01 PM, Subutai wrote:
At 10/27/2014 3:09:40 AM, Garbanza wrote:
It was supposed to give the instigator an advantage in scoring and stop noob sniping. But a failure in coding means the challenger has the advantage, which makes people cautious about starting risky debates which is bad.

Also, debaters with high Elo's are too scared of losing because they lose too much.

We should go back to the system of overall wins.

You make a good point, but your suggesting essentially causes similar problems. If we go to a system of overall wins, a debater who noob snipes to 250 wins would be superior to a great debater with 25 wins.

ELO is not a perfect system, but nothing is. No system is smart enough to give a truly accurate representation of a debater's skill. I'd say, of the systems on DDO now, ELO is the closest, but yes, it does have problems.

How is it a good point? It's not accurate.... If you debate someone with 2000 ELO, you gain 35 for a win as instigator and 15 for a win as contender. There's no "glitch" that gives an advantage to the contender. The advantage is the same it's always been: contender generally gets to evaluate the Instigator's case, or at least do a bunch of research before deciding whether to accept. That's why the instigator advantage and contender penalty exist in the ELO calculation.

The claims that we should tether ELO to "debate quality" also ignore how ELO is *already* calculated: it takes account of your opponent's ELO. That's the best proxy for debate quality. In fact, it's more important. If you face a tougher opponent, you get more for a win and lose less for a loss. How difficult your opponent is matters more than some other subjective rating of debate quality.

Why are you arguing with me? I agree with you. I agree that ELO is the best system DDO currently has. I was simply saying that ELO is not perfect, and that a truly perfect, objective measure of debate quality doesn't exist.

I responded to you because random people complain about the ELO system all the time, even while not really knowing the formula. But regular users rarely respond to them with "good point." I just wanted to make sure you knew why OP's point was not a good one.

For the record, I do know it. D=100* (W+9L)/10W (where L is losers ELO and W is winners ELO).
I'm becoming less defined as days go by, fading away, and well you might say, I'm losing focus, kinda drifting into the abstract in terms of how I see myself.
bsh1
Posts: 27,503
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 10:36:00 PM
Posted: 2 years ago
At 10/27/2014 10:30:02 PM, Imperfiect wrote:
At 10/27/2014 10:19:16 PM, bsh1 wrote:
At 10/27/2014 8:21:25 PM, bluesteel wrote:

How is it a good point? It's not accurate.... If you debate someone with 2000 ELO, you gain 35 for a win as instigator and 15 for a win as contender. There's no "glitch" that gives an advantage to the contender. The advantage is the same it's always been: contender generally gets to evaluate the Instigator's case, or at least do a bunch of research before deciding whether to accept. That's why the instigator advantage and contender penalty exist in the ELO calculation.

^ This.

Sorry but I disagree.

Why?
Live Long and Prosper

I'm a Bish.


"Twilight isn't just about obtuse metaphors between cannibalism and premarital sex, it also teaches us the futility of hope." - Raisor

"[Bsh1] is the Guinan of DDO." - ButterCatX

Follow the DDOlympics
: http://www.debate.org...

Open Debate Topics Project: http://www.debate.org...
Imperfiect
Posts: 128
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 10:44:21 PM
Posted: 2 years ago
If you debate someone with 2000 ELO, you gain 35 for a win as instigator and 15 for a win as contender. There's no "glitch" that gives an advantage to the contender. The advantage is the same it's always been: contender generally gets to evaluate the Instigator's case, or at least do a bunch of research before deciding whether to accept. That's why the instigator advantage and contender penalty exist in the ELO calculation.

^ This.

Sorry but I disagree.

Why?

Take note of how bluesteel opens the post stating that if you debate someone with 2000 Elo you gain 35 as instigator and 15 as non-instigator. This is discouraging people from accepting any debates then and makes everyone want to be the instigator; that is really so silly, it should be equal.

I mean, from what I've seen, most of the high Elo people are setting up 'impossible to accept' debates and selectively picking their partners anyway so it's not like they are bravely instigating and should be rewarded for taking the risk as to how tough their opponent will be, they are not taking that risk thanks to 'impossible acceptance' being configurable and socially accepted.

Following this, bluesteel just repeats the first point without ever explaining why the instigator's privilege a good thing in the first place.

The notion that the contender gets to do more research than the instigator is so absurd...

What kind of instigator hasn't already pre-researched it for an indeterminable amount of time? The contender has to compete with other contenders to be the one ot either click 'accept this challenge' first or beg in comments to be blessed for his/her majesty considering them a worthy debating opponent.

Also I disagree with the rest of what bluesteel posted but you didn't quote that part so I will assume that you also disagree to it.
dylancatlow
Posts: 12,244
Add as Friend
Challenge to a Debate
Send a Message
10/27/2014 11:03:37 PM
Posted: 2 years ago
At 10/27/2014 8:21:25 PM, bluesteel wrote:
At 10/27/2014 5:26:01 PM, Subutai wrote:
At 10/27/2014 3:09:40 AM, Garbanza wrote:


The claims that we should tether ELO to "debate quality" also ignore how ELO is *already* calculated: it takes account of your opponent's ELO. That's the best proxy for debate quality.

The only thing ELO takes into account is W/L. What is being won and what is being lost are ignored. Debate quality obviously correlates with ELO, but there's no reason to think ELO couldn't better reflect debate skill if we gave more weight to serious debates.
Blade-of-Truth
Posts: 5,020
Add as Friend
Challenge to a Debate
Send a Message
10/28/2014 1:26:53 AM
Posted: 2 years ago
At 10/27/2014 10:44:21 PM, Imperfiect wrote:
I mean, from what I've seen, most of the high Elo people are setting up 'impossible to accept' debates and selectively picking their partners anyway so it's not like they are bravely instigating and should be rewarded for taking the risk as to how tough their opponent will be, they are not taking that risk thanks to 'impossible acceptance' being configurable and socially accepted.

I can see a potential response for this being that it stops random noobs from wasting the time of an experienced debater.

I still don't think this justifies the bonus instigators receive if they choose to do the "impossible to accept" route. I think a proper amendment to the current system would be to take away the instigator bonus if they choose to do the "impossible to accept" setup. I do think, however, that the option to do so should still be available - as long as the bonus is removed.

I'm curious to see if someone can justify keeping the instigator bonus in effect even after the instigator protects themself with the "impossible to accept" format.

The notion that the contender gets to do more research than the instigator is so absurd...

I agree. If anything, the only benefit of the contender is whether or not they wish to accept the debate. The instigator, being instigator, should not need additional time to research a topic they are instigating. It's reasonable to assume they've already done the research since they are instigating it in the first place.

What kind of instigator hasn't already pre-researched it for an indeterminable amount of time?

Actually, there are several reasons why someone might instigate a challenge without knowing the topic fully themselves. I do get the jist of your point though, and ultimately agree.
Debate.org Deputy Vote Moderator
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
DDO Voting Guide: http://www.debate.org...
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Need a judge on your debate? Nominate me! http://www.debate.org...