Civ 3 Tournament Proposal
- Thread starter Grey Fox
- Start date
in each division there is 4 game 2 of each level i guess.
I think you don'T need a poll and are able to make a good decision
you should realease 6 game at the same time ( one for each difficulti level. )
this meant 2 per division.
then after the time end up for those 2 release 6 more. 1 for each difficultie level .
I think you don'T need a poll and are able to make a good decision
you should realease 6 game at the same time ( one for each difficulti level. )
this meant 2 per division.
then after the time end up for those 2 release 6 more. 1 for each difficultie level .
Falcon02
General
How about this, we set up at least 3 people to act as administrators for the game (Gray Fox included), settle on a point system and set up a start date for April 29th (unless later would be better).
Once we have a group of Admins, than Gray Fox's Summer job won't effect a thing, and avoid any problems that could occure in the future because of admin, unavailiblity.
Also, we can always change the scoring system after we go through the first round of games if need be. Since we are making a scoring system from scratch we need to run it through a full scaled experiment to see if it's effective, and then change it if need be.
Once we have a group of Admins, than Gray Fox's Summer job won't effect a thing, and avoid any problems that could occure in the future because of admin, unavailiblity.
Also, we can always change the scoring system after we go through the first round of games if need be. Since we are making a scoring system from scratch we need to run it through a full scaled experiment to see if it's effective, and then change it if need be.
I think Aeson would like to help, he said he had the time.
Maybe you could be the third Falcon02? You are one of the creators of this after all...
Aeson's scoring system might need some tweaking but it sure are a good system!
We also need to decide how the games should be released. One every 2 Weeks but every game gets a time limit of a month?
And should we play 1 Tiny, 1 Small, 1 Standard and 1 Large?
Maybe this order is the best:
Large-->Tiny-->Standard-->Small
Or some variation of that...
Newly Joined players.
Veteran
- mwright3030
- Jaygatz
Regular
- tdawk
- momarch
- Monster Zero
- Genia4
- Chris85
Maybe this order is the best:
Large-->Tiny-->Standard-->Small
Or some variation of that...
Newly Joined players.
Veteran
- mwright3030
- Jaygatz
Regular
- tdawk
- momarch
- Monster Zero
- Genia4
- Chris85
Falcon02
General
I'll be williing to help however, I can, but I'm not sure how much time I can devote to this, at least until after the AP tests. (I've got teachers piling homework on me to get me ready.... YUCK)
But, sure, I'll help anyway I can.
But, sure, I'll help anyway I can.
Plux
Emperor
If you give the tiny game only 2 weeks, and release it last, we should have all results in in 2 months. Seems best to me...
Bildbert
Chieftain
Hmm. In that case, I must have misunderstod the idea of the tournament. Quote from Grey Fox´s original post:
I interpreted this as getting a high score wasn't relevant, except when it's the goal itself.
Not necessarily. Building up territory and population isn't a goal in itself, it's actually just a means to achieve the real goal, i.e. victory. A player who is good at building an empire with lots of territory and a large population is of course more likely to reach victory faster.
Valid skill areas? Again, this depends on what the tournament is all about. Really, if I want to play in a competion where scoring is important, I enter the GOTM. I thought this tournament would be different, but if the proposed scoring system will be used, it'll just be an extended GOTM. And that's really not my cup of tea, since it in some ways rewards "milking".
Again, I propose a simpler scoring system where achieving the victory condition is everything. 100 points to the player who reaches the goal fastest, 50 to the runner-up, 25 to #3 and so on. The exact number of points can of course be discussed, but you get the idea.
I think this still allows for a large variety of playing styles.
Simple. Highest number of total points after four games takes the top spot for the season.
/Bildbert
SirPleb
Shaken, not stirred.
I don't agree.
Any tournament has a scoring system of course, and the objective is to get the highest score. So the question is not of "a competition where scoring is important". It is only a question of how the score is measured and the goals implied by that.
The proposed scoring formula will not reward milking. It may or may not need tuning after the first round to make that completely true. We'll know better what the right MapSize values should be after trying them. But the principle is clear - it is intended that a milked game will not be able to out-score a game which reaches the specified goal at an early date.
The scoring formula will slightly reward acheiving the specified goal "better" as well as earlier, where better is defined in the game's scoring terms. I.e. having more land and happy citizens is better. I think this generally won't be a large factor - most players who reach the primary goal around the earliest possible date are likely to have about the same game score. But if someone can acheive the goal a few turns later, with a much "better" Civ, should they be rewarded for that? It can be argued both ways. The scoring formula will slightly reward it and as a result should make the game more interesting. It does not seem to me that the formula will ever reward a player for deferring the goal. Once the goal is in hand for a player, it will never make sense to play a few turns more to get a higher score - the formula won't work that way, delaying the win will reduce the score.
Note that the impact of the game score is larger the smaller the map. This is a very important consideration, especially with regard to conquest or domination goals. On a tiny map, the game score component is fairly large. It can afford to be large since the scoring potential of a tiny map is small - it still won't encourage milking. The advantage of making it large in this case is that it can offset early gambles to some degree. On a tiny map, early high risk moves can have a large impact on the conquest/domination date. Luck becomes a large factor. By making the game score a higher percentage of the total on tiny maps, there is some offset which can reward less risky openings which build a bit first.
On a larger map the impact of game score is smaller. E.g. on a standard 70% water, the game score can only account for 20% of the tournament score. On this type of map the reward for early risk taking is not as high, but the potential imbalance from milking is high. The lower factor for game score should handle this.
Another factor to consider: How does one compare losing games? Sometimes the goal and map may work out so that many people don't achieve the primary goal. How do we then determine who did better? The suggested scoring formula handles this problem. It may not be a perfect answer but it is useable.
Caldasar
The Stoneman or LoqDoc
If u need an admin more for the tournament i can help.. (i wanna do something... )
(i wanna do something... )dikwhit
Warlord
Okay yet another day, yet anotherpage of posts. yet still nada has been accomplished. Hey guys...... whomevers running this, whatever the scoring, lets get this show started. The vast majority of us I would imagine are like me, lets play some damn games.
Cant decide on a scoring system? Get a game out for play and we got 100 peoples worth of games to tweak a scoring system.
Its not gonna be perfect to start with, nothing ever is, but if we dont start it, we cant make it better.
One cant build a better mousetrap by talking. so my suggestion is :
1)make a game
2)post it
3) set a short submit by date
4)use aesons score system to see how it works in practicallity
5)tweak accordingly
assuming this all works as planned then start the league play
Cant decide on a scoring system? Get a game out for play and we got 100 peoples worth of games to tweak a scoring system.
Its not gonna be perfect to start with, nothing ever is, but if we dont start it, we cant make it better.
One cant build a better mousetrap by talking. so my suggestion is :
1)make a game
2)post it
3) set a short submit by date
4)use aesons score system to see how it works in practicallity
5)tweak accordingly
assuming this all works as planned then start the league play
- Joined
- Nov 16, 2001
- Messages
- 2,686
When I said we need a way to compare results across 4 games I meant how well each player did in comparison to the others. At first glance points do seem to do this, but giving points soley on placement is too simplistic to actually differentiate between players in all circumstances. It also cuts out valid comparisons between placements in different games.
For instance consider a circumstance where 2 players each 'win' 2 of the games in the season. Each places 2nd in the games the other wins. Who actually did better in the season isn't covered if this is the only method for comparison. Even though the placement was equal, the actual performance might not have been. One of the players could have far outplayed 2nd place in both their wins, and just narrowly lost when they came in 2nd. The concept of 'by how much' is important in an accurate scoring system IMO.
That was the idea behind a 'batting average' based on comparison to the best finish. It should almost never work out to a tie, as each person's percentage will have a variable set of divisors (fastest date and highest score).
Because each map is different, the 'max' percentage will be determined by the best played game. This should make the 'standard' more even between maps, though it isn't perfect I admit. In the case of a tie on one map (quite unlikely), there will be differentation through the composite of all 4. I'm not that good with statistics (especially with variable fractions), but the chance of a tie across all four maps should be virtually nonexistant.
Another aspect is how those who don't place have their results treated. A placement system, regardless of the points given, will give very simplistic results with little comparison value. If I come in 10th place what does that mean, other than 9 people were better than me? How much better isn't covered. I may have performed at 95% in a very tightly packed race, or at 80% with a runaway winner, it isn't specified. Under the percentage scoring system (once balanced properly), everyone will get an accurate comparison value. If the GOTM is any indication, most people in these honor system type tournaments are playing for this comparison.
I understand the aversion to milking that most people have. The Space Race examples that I used should be considered the best case scenario (other than Highest Score) for milking. The 'earlier' victory conditions of Conquest and Domination will make the date percentage drop even more pronounced for a 2050 milked game.
Cultural is a special case because it is so highly dependant on map size. I would suggest we only have a Cultural victory condition on the Standard or Large maps. On a Tiny or Small map there just isn't enough room to build the cities needed to get to 100k culture early enough to combat milking.
For instance consider a circumstance where 2 players each 'win' 2 of the games in the season. Each places 2nd in the games the other wins. Who actually did better in the season isn't covered if this is the only method for comparison. Even though the placement was equal, the actual performance might not have been. One of the players could have far outplayed 2nd place in both their wins, and just narrowly lost when they came in 2nd. The concept of 'by how much' is important in an accurate scoring system IMO.
That was the idea behind a 'batting average' based on comparison to the best finish. It should almost never work out to a tie, as each person's percentage will have a variable set of divisors (fastest date and highest score).
Because each map is different, the 'max' percentage will be determined by the best played game. This should make the 'standard' more even between maps, though it isn't perfect I admit. In the case of a tie on one map (quite unlikely), there will be differentation through the composite of all 4. I'm not that good with statistics (especially with variable fractions), but the chance of a tie across all four maps should be virtually nonexistant.
Another aspect is how those who don't place have their results treated. A placement system, regardless of the points given, will give very simplistic results with little comparison value. If I come in 10th place what does that mean, other than 9 people were better than me? How much better isn't covered. I may have performed at 95% in a very tightly packed race, or at 80% with a runaway winner, it isn't specified. Under the percentage scoring system (once balanced properly), everyone will get an accurate comparison value. If the GOTM is any indication, most people in these honor system type tournaments are playing for this comparison.
I understand the aversion to milking that most people have. The Space Race examples that I used should be considered the best case scenario (other than Highest Score) for milking. The 'earlier' victory conditions of Conquest and Domination will make the date percentage drop even more pronounced for a 2050 milked game.
Cultural is a special case because it is so highly dependant on map size. I would suggest we only have a Cultural victory condition on the Standard or Large maps. On a Tiny or Small map there just isn't enough room to build the cities needed to get to 100k culture early enough to combat milking.
- Joined
- Nov 16, 2001
- Messages
- 2,686
Dikwhit,
I understand your desire to get things rolling. We had originally thought it would take some time to get the support. The last few days have had a ton of signups which is part of the delay. The scoring system really doesn't have anything to do with it, it's just been a popular discussion topic until the real delays can be dealt with. You're right that until we get some results we can't really be sure about how the system will apply.
We just need a bit of time to get all the administration responsibilities delegated. The number of signups was much more and faster than we expected!
It is about time that we finalize the map/difficulty/victory condition rotation though.
Tournament:
Standard-->Small-->Large-->Tiny
I think this is a good rotation and we should go with it. Gives us the most time to play the Large map.
Difficulty level should flip every other time through.
Standard/Low-->Small/High-->Large/Low-->Tiny/High
Standard/High-->Small/Low-->Large/High-->Tiny/Low
We should play each map type with a rotation of victory conditions. I would suggest that Cultural be limited to Large and Standard though, as it becomes very difficult to high 100k on the smaller maps. To offset this, Highest Score should be limited to Small and Tiny maps. This also has the benefit of making the 'milking' games shorter.
EDIT: Switched Large and Tiny to the end, mixed up the seasons so no consecutive games with the same victory condition.
Season 1:
Standard/Low/Domination
Small/High/Conquest
Large/Low/Culture
Tiny/High/Score
Season 2:
Standard/High/Spaceship
Small/Low/Score
Large/High/Domination
Tiny/Low/Conquest
Season 3:
Standard/Low/Culture
Small/High/Spaceship
Large/Low/Score
Tiny/High/Domination
Season 4:
Standard/High/Score
Small/Low/Spaceship
Large/High/Culture
Tiny/Low/Domination
Season 5:
Standard/Low/Culture
Small/High/Domination
Large/Low/Conquest
Tiny/High/Spaceship
Season 6:
Standard/High/Domination
Small/Low/Conquest
Large/High/Spaceship
Tiny/Low/Score
Season 7:
Standard/Low/Culture
Small/High/Score
Large/Low/Spaceship
Tiny/High/Domination
Season 8:
Standard/High/Score
Small/Low/Domination
Large/High/Culture
Tiny/Low/Conquest
Season 9:
Standard/Low/Spaceship
Small/High/Score
Large/Low/Culture
Tiny/High/Conquest
Season 10:
Standard/High/Cultural
Small/Low/Domination
Large/High/Conquest
Tiny/Low/Spaceship
At this point the rotation would start again. It's not perfect, but it gives us 8 of each victory condition.
I understand your desire to get things rolling. We had originally thought it would take some time to get the support. The last few days have had a ton of signups which is part of the delay. The scoring system really doesn't have anything to do with it, it's just been a popular discussion topic until the real delays can be dealt with. You're right that until we get some results we can't really be sure about how the system will apply.
We just need a bit of time to get all the administration responsibilities delegated. The number of signups was much more and faster than we expected!
It is about time that we finalize the map/difficulty/victory condition rotation though.
Tournament:
Standard-->Small-->Large-->Tiny
I think this is a good rotation and we should go with it. Gives us the most time to play the Large map.
Difficulty level should flip every other time through.
Standard/Low-->Small/High-->Large/Low-->Tiny/High
Standard/High-->Small/Low-->Large/High-->Tiny/Low
We should play each map type with a rotation of victory conditions. I would suggest that Cultural be limited to Large and Standard though, as it becomes very difficult to high 100k on the smaller maps. To offset this, Highest Score should be limited to Small and Tiny maps. This also has the benefit of making the 'milking' games shorter.
EDIT: Switched Large and Tiny to the end, mixed up the seasons so no consecutive games with the same victory condition.
Season 1:
Standard/Low/Domination
Small/High/Conquest
Large/Low/Culture
Tiny/High/Score
Season 2:
Standard/High/Spaceship
Small/Low/Score
Large/High/Domination
Tiny/Low/Conquest
Season 3:
Standard/Low/Culture
Small/High/Spaceship
Large/Low/Score
Tiny/High/Domination
Season 4:
Standard/High/Score
Small/Low/Spaceship
Large/High/Culture
Tiny/Low/Domination
Season 5:
Standard/Low/Culture
Small/High/Domination
Large/Low/Conquest
Tiny/High/Spaceship
Season 6:
Standard/High/Domination
Small/Low/Conquest
Large/High/Spaceship
Tiny/Low/Score
Season 7:
Standard/Low/Culture
Small/High/Score
Large/Low/Spaceship
Tiny/High/Domination
Season 8:
Standard/High/Score
Small/Low/Domination
Large/High/Culture
Tiny/Low/Conquest
Season 9:
Standard/Low/Spaceship
Small/High/Score
Large/Low/Culture
Tiny/High/Conquest
Season 10:
Standard/High/Cultural
Small/Low/Domination
Large/High/Conquest
Tiny/Low/Spaceship
At this point the rotation would start again. It's not perfect, but it gives us 8 of each victory condition.
dikwhit
Warlord
works for me. Im not real concerned about the rules, til exploits are found.Then the real debates will start
. I'm just not the most patient individual.
will try to keep the voice of discontent to a managable minimum
. I'm just not the most patient individual.will try to keep the voice of discontent to a managable minimum
Plux
Emperor
I would switch the tiny and small games, so we can shorten the time-limit on tiny to two weeks, and finish the first competiton in two months. Then, starting games is best on fridays, submission deadlines on mondays.
I guess next friday (april 26) would be a good competitionstarter, leaves some room for admiminstrative work, and sets everybody up for the weekend. So, as for the dates:
How about this?
(maybe nicer to switch large & standard, save the best for (almost) last)
I guess next friday (april 26) would be a good competitionstarter, leaves some room for admiminstrative work, and sets everybody up for the weekend. So, as for the dates:
Code:
start finish
Large 26/4 27/5
Small 10/5 10/6
Standard 24/5 24/6
Tiny 7/6 24/6How about this?
(maybe nicer to switch large & standard, save the best for (almost) last)
Falcon02
General
Aeson has provided a layout for game structure and a scoring system. Plux has made a good "tweak" to that system, putting the large twoards the end, (leaves the ones who like Big ones in anticipation, same with having the tiny as last for those who like quick small games).
The start date really doesn't matter to me, I can wait however long it takes. But so long as no one disagrees with tis system, I say we now need to set a date, and verify who all Admins of the game are.
Another suggestion we might want to do is to split up the divisions after the first round. The Top half play only the higher difficulty level for that division, and the Bottom half would only play the lower difficulty level, and then there would be more levels of play and difficulty transistions wouldn't be so extreme as they could be. (skipping emporer and going straight to Deity) This will avoid everybody having to resign up, and the admins from having to resort out everybody.
The start date really doesn't matter to me, I can wait however long it takes. But so long as no one disagrees with tis system, I say we now need to set a date, and verify who all Admins of the game are.
Another suggestion we might want to do is to split up the divisions after the first round. The Top half play only the higher difficulty level for that division, and the Bottom half would only play the lower difficulty level, and then there would be more levels of play and difficulty transistions wouldn't be so extreme as they could be. (skipping emporer and going straight to Deity) This will avoid everybody having to resign up, and the admins from having to resort out everybody.
Similar threads
