Site scoring system (Feb 2nd 2013)

[gannymede]

On one of these threads recently it was suggested that ranked games not allow players to choose their countries, and it seemed that this idea got overwhelming support, including from some of the mods. Do I remember this right?
Is it too late to consider this?

[sinnybee]

...this is not a suggestion thread actually...

It's probably kind of late to change now anyway?

I'm fine with no country adjustments and all ranked country assignment options--the status quo.
Between the two changes, I'd much rather have add country rating adjustments than take away country assignment choices, because country assignment choices allow a player to try to get a country they may not have played much, a country that they are in the mood to play, a country that they want to practive with, etc.

[super_dipsy]

I can confirm there are no plans to change scoring based on the country you are playing. This would completely unbalance the game dynamics, if you knew at the start that one country would score differently to another.

On the random only country assignment idea, that is exactly what it is.It is a suggestion that has a lot going for it and should be handled through the Suggestions forum. The reason we decided not to go ahead with it at the same time though is we felt the new scoring system was such a huge change that we should not go changing other quite big things at the same time. People have enough to deal with handling the new scoring. Also, as it is quite a big decision we need to give people time to think it through. For example, we need to be clear on whether we want to go just random, or random and preferences, and also do we do this just for ranked games...etc.

This 'churn' issue is exactly why we also decided not to roll in the 2 Orders phase NMRs in a row auto-surrender (changing it from the current 3). With so much going on, we need to give people time to process the current changes before we go changing other things.

[sinnybee]

I can confirm there are no plans to change scoring based on the country you are playing. This would completely unbalance the game dynamics, if you knew at the start that one country would score differently to another.

On the random only country assignment idea, that is exactly what it is.It is a suggestion that has a lot going for it and should be handled through the Suggestions forum. The reason we decided not to go ahead with it at the same time though is we felt the new scoring system was such a huge change that we should not go changing other quite big things at the same time. People have enough to deal with handling the new scoring. Also, as it is quite a big decision we need to give people time to think it through. For example, we need to be clear on whether we want to go just random, or random and preferences, and also do we do this just for ranked games...etc.

This 'churn' issue is exactly why we also decided not to roll in the 2 Orders phase NMRs in a row auto-surrender (changing it from the current 3). With so much going on, we need to give people time to process the current changes before we go changing other things.

Excellent answer.

[LeroyJenkins]

I can never escape the ELO system, can I...

[alvin1912]

I believe the new scoring system is quite good, it encourage people to replace surrendered countries (by not giving score penalty) while give more accurate rating to players. It also encourage players to pick up countries surrendered in early stage as it score more than picking up easily winning or drawing countries.
However it seems the new system don't reward fairly for the effort that players spend on the game (as it directly score according to % of turns you have played), and it give very few incentive for players to join the late game or anti-leader game (games that one players nearly get 18SCs while others remaining players just got 6 or 7 SC each).

As usually people spend more effort for setting up the game (so that's why there are double time for first round), players have to overcome the disadvantages that the surrendered countries faced and have to spend more effort to solve the problem (when compare of maintaining advantages). So usually players have to spend more time when just picking up surrendered countries. In term of effort, the starting usually spent higher percentage of effort in the same period of time in term of a whole game. (effort spent in first two turns in a game / total effort spent > 2 turns / total number of turns of a game). For instance, you have to analysis all order history, read all the past message, knowing other players study the map carefully and build up new relationship. All these need extra effort when compare to maintain a lasting game. So it takes extra effort to join a surrendered game than maintaining a normal game.

In anti-leader games, players spend more effort to build up alliance, rally people to go against leader together. However even if you could prevent leader solo from game, that leader usually could form stalemated line for himself and mean that you would not much possible to get a solo in such game. And then after you have spend lots of effort to rally remaining players and finally success prevent one player solo win, you could probably get a 3-way or even 4-way draw, which mean not much of score you can get. Then as you are joining as late game, probably you could just get 20% of the points and probably mean nothing to a good player. If you want more score, perhaps you would want to defer the draw instead of ending the game at the time it should. (As you could get higher % of score for longer stalling phrase.)

I think it is important to give incentive to players picking up surrendered countries in anti-leader games as it would be much fair to others players of non-leaders which should able to achieve a draw if none are surrendered, though it may not have to play many turns but it still spend many effort.

Would it possible to scale the early % of turns higher?
For example:
you have played % of turns : % of score you got when the game finished
10% 30%
20% 45%
30% 55%
40% 64%
50% 72%
60% 79%
70% 85%
80% 90%
90% 94%
95% 97%
98% 99%
etc.

Would it better shown the effort of players spent while keep encourage players to pick up both surrendered countries in late game (not need players to stall the game just for getting little higher score) and surrendered in early game (which have more opportunity)?
Or it would be difficult to make the % of score depreciate with % of turns taken in programming?

[super_dipsy]

Interesting thought process alvin. What you need to do is get this into the Suggestions thread, where it will get a wider airing. I will move it to Suggestions for you if I can work out how.

[super_dipsy]

Hmm. I think it might be best if you cut and paste it into Suggestions yourself. I will wait to comment on it until you have moved it there, since I think a discussion there would be better than here in the information thread.

[jimbobicus]

Apart from the shield, remember that scores are adjusted if you did not play the country throughout the game. If you join a game on turn 11 and play for ten turns before getting a draw, your result will be calculated at 50%. However, if you surrender then there is no percentage reduction – you score a full loss as if you had played the entire game.

Has anyone considered this potential unexpected consequence...
Suppose you joined a game halfway through and due to your own brilliant play look set to solo. Do you then have an incentive to delay your solo so that you receive a larger portion of the credit? I'm in precisely this position. Its not going to affect how I play, since I've never really been too interested in gaming the system for a few ranking points. However I'm sure there are other people who might.

[super_dipsy]

Do you then have an incentive to delay your solo so that you receive a larger portion of the credit?

There will always be opportunities for people to 'play the system' . In my personal opinion, the opportunities for this are less than they were, but you can never be rid of them. You are correct - the longer you spin the game out for, the greater your percentage result. But think about it - suppose the game finishes after 40 turns. You play 20 of them, and it is going to get you 200points say, halved to 100. if you can spin the game out for another 10 turns, you will get an extra 20 points. But on the down side, you have wasted some of your playing time that you could have been using to get another win. And of course, we also have the balance of Diplomacy - if I was in a game where a player was on 17 and trying not to get to 18, I would quickly go to the others in the game and it would be 100% clear that no-one else is going to get anything unless the player poised for the solo is stopped. Normally you do not get this situation because it all happens quite fast, and suddenly the dam breaks and the winner is there. Allies who had hoped to share a draw gnash their teeth, but the deed is done. Now picture that scenario where you (on 17 say) try to play the game for 10 more turns before you get to 18. In this scenario however, I have 10 turns to talk to the other players in the game and persuade them that we have no choice but to all turn on the leader. I suspect you would never make it to 18 .

In my mind, this makes it not worth the risk although I think many players would take the same view as you and simply say 'I'll have my solo thankyou, and if it only gets me 100 points then so be it, I still WON!'