PBEM - Results thread

Originally posted by ERIKK
ProPain lets agree that this is a "free to use" ladder and that people should decide the game specifics for themselfs (generated map, restarts, location), and that they only have to report who played who and who won.

According to their position on the ladder they will get their points.
Exactly what I meant in my post but you phrase it much simpler and thus better. One restriction I do want rules in place for people who decline challenges with the sole goal to protect their ranking. (Finally I'm no 1 now I won't play anymore!)

Originally posted by ERIKK
The scoring mechanism. You can use mine (very very simple or Mel's (very very difficult)). Or you should have someone get us better one!

Can you make an example? You have 2 PBEM results and some people that agreed to join.
For me the scoring subject hasn't been decided yet. The first system hasn't been discussed very much and I would like people to think that over too. It's simple and easy to use. Also it automatically punishes a loss to a much weaker player (and rewards a win against a much better player) because of the larger ladder drop.( but you may refuse >5 position difference) It doesn't matter if you play a lot or a little and allows 'fast learners' to rise quickly instead of building up score over ( a lot of ) time
 
OK consider the first type of ladder, swapping positions

You are a very good player and have won all your games.
You are first

Another player has lost all his games and is last.

You play together and due to bad food/lighting/luck you lose.
You are now last and the other player is first...
Is this a true representation?
No...

Do you ever play anybody much lower than you?
No...
Why would you? Nothing to gain, everything to lose.

Top players stay top
Bottom players stay bottom.

Right thats that covered then.

:p

Melifluous
 
The Chess Table link is interesting Mel.

Essentially, if the higher ranking player wins they get less points than they stand to lose if they lose the game.

Conversely, if the lower ranking player wins they get more points than they stand to lose if they lose the game.

I like the concept, obviously draws are null and void, and the points will want to be adjusted, but it is a nice idea.

Regarding your last post, I do _not_ agree that a straight ladder system is completely unfair, in fact, those at the top can only go down, and those at the bottom can only go up, which is in fact a better system than the points one if you want to polarise it in to 'top stays top' and 'bottom stays bottom'... :p
 
Just saying at least with the points table even if you are way up there you can still earn points by playing a game...

And that you gain something, swapping means you play a lower player and win you gain nothing.

And I am putting myself forward to handle scores and maps and stuff.

Is that a staff role? Maybe it is...
But I am available.

Melifluous
 
People should comment on the link mel provided here, as it seems like a workable solution to me.
 
Here is the scoring concept that was linked by Mel.

I changed some values, to make it nicer. I also recommend (if we were to use this), that we all start on 1000 points. It will take a few games for people to adjust to a decent place inthe ladder, but it will work well over time.

Rd = Rating difference
H = amount higher ranked player gets if they win, and also the amount the lower ranked player gets deducted.
L = amount lower ranked player gets if they win, and also the amount the higher ranked player gets deducted.

Code:
Rd	H	 L 
0-100	50	50 
<200	45	55 
<300	40	60
<400	35	65
<500	30	70
<600	25	75
<700	20	80
<800	15	85
<900	10	90
>900	05	95

So, if I have 1100 points and ERIKK has 350 (hehe), I am 750 ahead, so if I win I get 15 points and ERIKK loses 15 points. If ERIKK wins he gets 85 points and I lose 85 points. This eliminates all the problems of the simple ladder where you can jump many places.

Comments?
 
Only thing I would point out.

The difference between the differences (if you know what I mean :rolleyes:) in your table is arithmetic, the table online is geometric, this means lower differences cause more change in points. Higher differences less so.

My 2 cents.

Melifluous
 
The differences between the differences in my table are actually zero, since each difference is 100.

If you look at the other table they are not a geometric progression. I get your point, but I don't think the differences are as big as you think.

i.e. my diffs are: 100, 100, 100, 100, ...
so my 'diffs of diffs' are 0, 0, 0, 0

the original table has diffs:
22, 22, 23, 23, 24, 25, 27, 29, 32, 36, 41, 50, 64, 95, 196

so it's 'diffs of diffs' is:
0, 1, 0, 1, 1, 2, 2, 3, 4, 5, 9, 14, 31, 101

This is more of an arithmetric progression at first, with some exponential numbers tacked on the end. The first few values are small enough to be there because of rounding.

EDIT: Removed some anaylsis that seemed a bit dodgy on reading again...
 
Anarres' his table seems right. Only the differences are 100 per step. But I dont think such large differences will emerge. We wont even play so many games.

And what about the initial ranking? We will probably start all together on 1600. In the first months there wont be much movenment in the Ladder I guess...
 
Possibly start positions could be based on perceived skill?

Maybe people say whether they are a deity or a cheiftain player?
Then start at say 1000 for Regent players, 900 for warlord, 1100 for monarch etc.

Melifluous
 
Originally posted by Melifluous
OK consider the first type of ladder, swapping positions

You are a very good player and have won all your games.
You are first

Another player has lost all his games and is last.

You play together and due to bad food/lighting/luck you lose.
You are now last and the other player is first...
Is this a true representation?
No...

Do you ever play anybody much lower than you?
No...
Why would you? Nothing to gain, everything to lose.

Top players stay top
Bottom players stay bottom.

Right thats that covered then.

:p

Melifluous

If this is the way this discussion is going to be held I'm gonna have no part of this anymore.

* Your post shows you haven't read the proposal carefully
* Your post focusses on an extremely unlikely option. Such an unrealistic example can be made up for any system.

and last but not least
* Your post is disrespectfull. Or at least that's how I experience it.
 
Originally posted by ProPain


If this is the way this discussion is going to be held I'm gonna have no part of this anymore.

* Your post shows you haven't read the proposal carefully
* Your post focusses on an extremely unlikely option. Such an unrealistic example can be made up for any system.

and last but not least
* Your post is disrespectfull. Or at least that's how I experience it.
Ah, common ProPain, he is just fooling around. The swapping proposition was already from the table... and speaking of tables, what do you think of what we have come up with (from anarres)?

And Mel, quit fooling around please!

ERIKK
 
Originally posted by ERIKK
Ah, common ProPain, he is just fooling around. The swapping proposition was already from the table... and speaking of tables, what do you think of what we have come up with (from anarres)
ERIKK

What do I think:
1. I came up with two proposals to seriously discuss the pros and cons of both. This is not happening because people dismiss the first proposal without carefully thinking about it. I think that's a shame.

2. I think the scoring system can be quite ok but I don't like the fact you can keep adding points to your score. This is what the 'chess system' does and this means

- It favors smaller maps (faster games)
- It favors people with lots of freetime
- It favors people who have a long history of playing in the system. When a new really good player comes along he has to build up score to overtake a lesser player who has played a lot
- You can become no 1 without ever playing the current no 1. although I don't think this last one won't really be a problem I still think that's weird for a ladder system.

To balance the scoring you should have an 'expire date' on your result like the one proposed in the gotm.

The way I see it:
The first 'swapping' system is more like boxing - You defeat the champ and you're the new champ
The second 'scoring' system is more like tennis - You play enough and get good results you'll be the champ

I prefer boxing :D
 
All I know is that in chess all ranking/scores stay between the initial 1600 (1500 - 2100 (?))

Isn't the "much playing gets you more points" -thing not already taken in that formula?
 
Like Propain I feel a simple method of keeping the scores will be most effective. However, a big con for me for using the first system is that undoubtedly players will reach the top positions that are not interested in playing (PBEM) games against a wide variety of players under all sorts of conditions, showing off their real skills, but trying to maintain their high rankings, by declining everyone who's more than 5 places away as much as possible, and play other top flight members so there will be little rotation between the high and low end of the ladder.

I think we all agree that the reason this much attention is given on how scores and results should be processed is that we all want to see a list of how good current PBEM players are (and want to top that list of course ;) ).

But since there are many ways to play the game, people's opinions on how the game is played best will differ, as each person favours other playing styles than the other. Personally, I already have some ideas on who I think are good players, based on comments posted on CFC in general, but mainly from reading the comments about progress of current PBEMs, and I do not need a score system to tell me.

What I would find most important is a clean and simple table showing which PBEMs are active, who are participating in them, and who is the person that won it, or maybe even a database showing for each player which games he has played and what the results were. Based on that, and comments in the related PBEM game's threads, people can make their own ranking of players. With so many play styles and different game factors involving the results, I think people really should make their own conclusions on who's a good player, and this would make any fancy scoring system only of secondary importance.

Just my two cents..
 
ProPain,

I humbly apologise if I have caused any offence, my tone in most of my posting is always slightly humorous. This, on occasion, causes offence. Never intended, but sometimes taken.

I was merely trying to break down what I had against your first system. I assure you that I considered both of your suggestions with care and your assumption that I have not is in itself a comment likely to cause offence. I dont like being told what I have done, I assure you that I know quite well what I have done.

This was an extreme example that I posted, but is true of any player playing anyone lower than them. Why would they? They have nothing to gain and their position to lose.

Also in your later post, you said that lesser players keep adding to their score by playing a lot.

If YOU had read my proposals properly then you would surely have realised that if you lose a game, you lose points.

Anyway, I have apologised, and I really dont wanna drag this thread into some kind of personal argument.

I think your Idea of a PBEM ladder is absolutely fantastic and would hate it if you were to give up just because of me.

Melifluous

PS. BTW ProPain... Wanna 1v1? :ninja:
 
Originally posted by ProPain
1. I came up with two proposals to seriously discuss the pros and cons of both. This is not happening because people dismiss the first proposal without carefully thinking about it. I think that's a shame.
I think both are being discussed ProPain. Some people don't like the simple ladder system, I think it has both good and bad points. I have been looking at scoring systems that would alleviate some of the bad points, this itself will bring new problems. A few people have dismissed the simple ladder system, but that does not mean it is without merit, but that those people are not yet convinced.
2. I think the scoring system can be quite ok but I don't like the fact you can keep adding points to your score. This is what the 'chess system' does and this means

- It favors smaller maps (faster games)
- It favors people with lots of freetime
- It favors people who have a long history of playing in the system. When a new really good player comes along he has to build up score to overtake a lesser player who has played a lot
- You can become no 1 without ever playing the current no 1. although I don't think this last one won't really be a problem I still think that's weird for a ladder system.
The scoring system I last proposed (form mel's chess link) can only give you more points if you win. Losing games will lose you points. If you decide to play people lower than you in ranking to get easy points you don't get very many, and also when you do lose, you lose much more points. This will help balance the game, and I don't think it will favour those who play more, unless they are good enough.

Let's discuss the various systems and try to find one we all like. If you think the first system is best, then you should convince us of it. Personally, after much careful consideration, I think a points based system may be best, although (as always) I am happy to be shown otherwise.
 
When all's said and done, theres a lot more said than done.

Does it really matter?? You can retroscore the games. What matters is playing and the basis for the games.
 
Top Bottom