"Surviving in a draw, yes. But surviving in a solo, no."

Agreed 100%, Mapu. The only survival in a solo is the victor.

"or multiply all final outcomes by 100 to get rid of that decimal and give us whole numbers that are reflective of the real final score"

Good suggestion, Kaner. Whole numbers are much cleaner.

Multiplying the result by 100 has the same effect as starting with each game worth 15000. I do not think the scoring for a solo was ever specifically stated in the discussions before the tournament started, but I think the consensus was WTA, so it should be 15000 for the solo, 0 for everyone else.

"Mapu - you don't think the method of distributing points should be the same across all games? That hardly seems fair."

Some of the ultra veteran players can correct me and add to this, but my understanding is that surviving a solo is actually worse than being defeated (and thus worse than surviving in a draw) because you failed to stop the solo. I think some scoring systems even take this into account.

I don't think any veteran player would say surviving to a solo is worse than being eliminated - the eliminated powers 'lost' in the same way and their actions would also have had an impact in setting the solo up. I know some people who say that to make themselves look better after being eliminated, but I don't think anyone seriously believes it.

Someone throwing the game is often *less* likely to survive than other players.

That said, I don't know many veteran players who would say surviving to a solo is better than an elimination, either. Both cases are losses, both are worth the same.

In the post on pg5 of this thread where Dr Recommended officially launched the torunament, and the first game, he posted "We will use SoS scoring to distribute points for the tournament."

Nowhere do I see that being limited to draws. I would disagree heavily with any push to change the rules, part-way through the tourney, to have solos split WTA rather than SoS.

The official scoring system Sum of Squares (SoS) is a WTA system, SoS is a way of scoring the draws.

mouse:

Although I agree the post you referenced does not state that SoS does not apply to solos, the post on page 1 by kaner, which suggests the SoS scoring, reference a webDip link:

http://webdiplomacy.net/points.php

He also copied part of the information in the link, and it states:

"This is a centre-based scoring system. In a draw, each player gets a proportion of the pot scaled by their centre count:

(SC_count^2) / (sum of all players (SC_count^2))"

Additional information at the webDip link that is not copied to the post states:

"WebDiplomacy has three scoring systems enabled: Draw-Size Scoring, Sum of Squares and Unranked.
In the first two systems, if you win this game by capturing the required supply centres (18 on the Classic map) then you win the entire pot for the game, and receive all those points."

It was my interpretation that when Dr. Recommended used the term SoS scoring, he is referring to this definition, which I think clearly states that solos are scored WTA.

Why is it that Diplomacy players of all people cannot comprehend that "solo" means one?

I'm still waiting an answer to my question: can I play this tournament, please?

Maucat, the tournament is well underway and the games are not open. I will add your name to the list of potential replacements.

Good discussion on scoring, I'll have to catch up and add my thoughts later today.

Bump. Games waiting on two players.

Who are the two donks not signing up? We have eager players begging to play. If you cannot continue, please speak up.

I'm a donk. In less than 4 days I'm away with zero internet access for a week. I flagged this with Doc and instead of having to start a game and put everyone through a pause a day later, thought it made far more sense to just wait on starting the next game when I got back on 20 Dec.

*cough* game sitter feature *cough

Surely a silver level developer knows about this? :)

They hand out silver medallions for just turning up. You should know that RUFFHAUS ;-)

...

Amby is back and will join the games shortly. That leaves one player still not joined. Who is it? I'm in.

I've PM'd the remaining holdout (thanks Cap!)

Are the countries already shuffled?

On it now - please do not enter orders in the meantime.

Done! Have fun all :)

I've asked Captainmeme to PM the player that NMRd. I'll give a day or so more and then pull a player from the substitute list. Is this agreeable?

Game 2 scoring:

Dr. Rec: (30^2)0.0414 = 37.26
bozo: (27^2)0.0414 = 30.1806
Desert Fox: (27^2)0.0414 = 30.1806
Eki83: (25^2)0.0414 = 25.875


Scoring after 2 games:

Hirnsaege: 80.15
Eki83: 76.595
Dr. Rec: 37.26
bozo: 34.3206
Desert Fox: 30.1806
RUFF 8: 14.95
mouse: 0.04

I'm still not convinced we've got the right scoring system, but I don't have a better idea. For example, it seems that bozo reaching the draw in both games, with 10 and 25 centers respectively, should be above me (1 draw, 30 centers) in the scoring. I guess it just shows the value of grabbing any extra centers you can as the game winds down.

*bozo 10 and 27, not 25

Two possible tweaks. First, I calculated the scores the same way, and simply added the number of centers to them:

Eki83: 136.595
Hirnsaege: 124.15
bozo: 71.3206
Dr. Rec: 67.26
Desert Fox: 57.1806
RUFFHAUS 8: 33.95
mouse: 1.4

This gives more weight to bozo and Eki reaching the draw in both games.

The next one is admittedly arbitrary. I awarded 10 across the board for reaching the draw. This again adds weight to bozo & Eki drawing both games, and smooths out the huge range generated by squaring. I don't know if it will work as we get deeper into the tournament, but I think this one gives the fairest representation:

Eki83: 96.595
Hirnsaege: 90.15
bozo: 54.3206
Dr. Rec: 47.26
Desert Fox: 40.1806
RUFFHAUS 8: 24.95
mouse: 10.04


I rather doubt we'll find complete consensus on this, and I hope I'm not starting any arguments. Just want to look at our options before we get further along.

voice from the off, not really knowing what was discussed so far:

does it have to be ^2 ?
any other exponent between 1 and 2 would tone down the differences and still scale to some extent.

something like ^1.2 would still favour the larger size in a draw, but far less drastically.
i.E. if we have someone having 10 centers, some other 11 centers, another one 20 centers, we'd get:
10^1.2 = 15.85
11^1.2 = 17.77
20^1.2 = 36.41
bases difference (for 10,11,20) is 110% and 200%
with exponent 1.2 applied we get a difference of 120% and 230%

So it scales and favours the larger winner – but not so much as a ^2 which favours those who once manage to have a huge victory, even if they fail in all the other games.

1.2 is still an arbitrary factor of course ... but ^2 actually is too.
with expontent 2.0 applied to (10,11,20), we get a difference of 121% and 400%
--> having double centers is worth four times in scoring. having 3 times centers would be 9 times already ...

Players ready for next game:

1. Amby

1. Amby
2. Desert Fox