Total Posts:12|Showing Posts:1-12
Jump to topic:

Playing with ELO formulas

Ore_Ele
Posts: 25,980
Add as Friend
Challenge to a Debate
Send a Message
7/29/2011 1:59:09 PM
Posted: 5 years ago
So I'm playing with different formulas of ELO for DDO and was wondering how should it lean?

Obviously we would want something that rewards debating more experienced members and discourages newb sniping (as ELO traditionally does), but we don't want it to be so against debating new members that we give off an elitist vibe and leave new members out.
"Wanting Red Rhino Pill to have gender"
kohai
Posts: 380
Add as Friend
Challenge to a Debate
Send a Message
7/29/2011 3:20:49 PM
Posted: 5 years ago
FIDE has a rating calc

http://en.wikipedia.org...
1) Whatever has contradictory attributes does not exist.
2) The Biblical God has contradictory attributes.
3) Therefore, the Biblical God does not exist
Ore_Ele
Posts: 25,980
Add as Friend
Challenge to a Debate
Send a Message
7/29/2011 3:59:09 PM
Posted: 5 years ago
At 7/29/2011 3:38:47 PM, darkkermit wrote:
Less experienced debaters will debate less experienced debaters. I don't see the controversy.

Personally, I don't want a gap to form of "less experienced vs less experienced" and "more experienced vs more experienced" to the point where it becomes increasingly difficult to bridge that gap.
"Wanting Red Rhino Pill to have gender"
Ore_Ele
Posts: 25,980
Add as Friend
Challenge to a Debate
Send a Message
7/29/2011 4:41:41 PM
Posted: 5 years ago
I currently have a floor of 1,000 with a starting of 2,000 (the numbers themselves don't really matter). Though the floor is pretty much meaningless. At the formula I'm using, someone would have to go 0 - 14 (all against new members at 2,000 range) to hit the floor, and if they went against veterns with higher scores, it would take even longer.

Also, someone would have to go 13 - 0 to reach 3,000 (if facing only new members).

Additionally, for someone to go from 3,000 up to 4,000 would take 17 wins against new members (all at 2,000 each), while would only take 10 wins if facing those of the same level as them. And 20 wins to go from 4,000 up to 5,000 if facing only new members (compared to still 10 if facing current members).
"Wanting Red Rhino Pill to have gender"
Logic_on_rails
Posts: 2,445
Add as Friend
Challenge to a Debate
Send a Message
7/29/2011 10:21:09 PM
Posted: 5 years ago
At 7/29/2011 3:59:09 PM, Ore_Ele wrote:
At 7/29/2011 3:38:47 PM, darkkermit wrote:
Less experienced debaters will debate less experienced debaters. I don't see the controversy.

Personally, I don't want a gap to form of "less experienced vs less experienced" and "more experienced vs more experienced" to the point where it becomes increasingly difficult to bridge that gap.

This isn't normally a problem in chess, depending on how big your difference is. Ie. On chess.com in blitz games I have a rating normally in the 1750-1800 range. I 'create' matches with a minimum of 1800 and easily find matches. My average opponent is about 1840, so on average I'm playing up 65 points.

Of course, there is a point to which you don't take matches. Ie. It's pretty rare for me to see 1950+ matches as they are probably not available for my rating (there's also less players at higher levels) and they don't want to waste their time (ie. a 250+ gap is really, really hard to overcome) . Conversely, I don't accept lower rated challenges unless I know the person because I want to vs. higher rated opponents.

1 thing to note - it's easier for a bad debater to match a good debater than it is in chess. This is because one could merely write a bit better or find better arguments etc. However, a little bit of preparation won't help a beginner defeat an experienced player in chess.
"Tis not in mortals to command success
But we"ll do more, Sempronius, we"ll deserve it
phantom
Posts: 6,774
Add as Friend
Challenge to a Debate
Send a Message
10/8/2011 10:49:12 PM
Posted: 5 years ago
At 10/8/2011 10:38:15 PM, F-16_Fighting_Falcon wrote:
The big question I have with ELO is this: would it take into account forfeits? Because if someone wins by forfeit, it isn't really a win at all.

What if the only reason your opponent forfeits is because you post a damn good argument that he has no chance of refuting?

Also people have forfeited and still won the debate.
"Music is a zen-like ecstatic state where you become the new man of the future, the Nietzschean merger of Apollo and Dionysus." Ray Manzarek (The Doors)
F-16_Fighting_Falcon
Posts: 18,324
Add as Friend
Challenge to a Debate
Send a Message
10/8/2011 10:51:45 PM
Posted: 5 years ago
At 10/8/2011 10:49:12 PM, phantom wrote:
At 10/8/2011 10:38:15 PM, F-16_Fighting_Falcon wrote:
The big question I have with ELO is this: would it take into account forfeits? Because if someone wins by forfeit, it isn't really a win at all.

What if the only reason your opponent forfeits is because you post a damn good argument that he has no chance of refuting?

Also people have forfeited and still won the debate.

I guess it should count then. I was referring to those debates where your opponent leaves after posting none or just one argument, and all the RFDs just read "Forfeit", "FF", "Obvious" etc.
RoyLatham
Posts: 4,488
Add as Friend
Challenge to a Debate
Send a Message
10/10/2011 9:05:18 AM
Posted: 5 years ago
At 7/29/2011 4:41:41 PM, Ore_Ele wrote:
I currently have a floor of 1,000 with a starting of 2,000 (the numbers themselves don't really matter). Though the floor is pretty much meaningless. At the formula I'm using, someone would have to go 0 - 14 (all against new members at 2,000 range) to hit the floor, and if they went against veterns with higher scores, it would take even longer.

Also, someone would have to go 13 - 0 to reach 3,000 (if facing only new members).

Additionally, for someone to go from 3,000 up to 4,000 would take 17 wins against new members (all at 2,000 each), while would only take 10 wins if facing those of the same level as them. And 20 wins to go from 4,000 up to 5,000 if facing only new members (compared to still 10 if facing current members).

That seems reasonable to me. We do get some very sharp new members, so if it's too difficult to advance they're more likely to get discouraged. Any ELO system would be a step up compared to the present system of just counting wins.
phantom
Posts: 6,774
Add as Friend
Challenge to a Debate
Send a Message
10/10/2011 11:55:04 AM
Posted: 5 years ago
At 10/8/2011 10:51:45 PM, F-16_Fighting_Falcon wrote:
At 10/8/2011 10:49:12 PM, phantom wrote:
At 10/8/2011 10:38:15 PM, F-16_Fighting_Falcon wrote:
The big question I have with ELO is this: would it take into account forfeits? Because if someone wins by forfeit, it isn't really a win at all.

What if the only reason your opponent forfeits is because you post a damn good argument that he has no chance of refuting?

Also people have forfeited and still won the debate.

I guess it should count then. I was referring to those debates where your opponent leaves after posting none or just one argument, and all the RFDs just read "Forfeit", "FF", "Obvious" etc.

Yeah but I don't see how a system could differentiate between the two.

Also those rfds would apply to my first example.
"Music is a zen-like ecstatic state where you become the new man of the future, the Nietzschean merger of Apollo and Dionysus." Ray Manzarek (The Doors)
Ore_Ele
Posts: 25,980
Add as Friend
Challenge to a Debate
Send a Message
10/10/2011 12:01:03 PM
Posted: 5 years ago
At 10/8/2011 10:38:15 PM, F-16_Fighting_Falcon wrote:
The big question I have with ELO is this: would it take into account forfeits? Because if someone wins by forfeit, it isn't really a win at all.

Yes, but since nearly all of those are from new members, they will have only a minor boost to one's ELO ranking, rather than a large boost, as they have now.
"Wanting Red Rhino Pill to have gender"