The GOTM Scoring System

[Piscator]

Originally Posted by Obormot
A factor of 6 or even 3-4 would break the system completely in favour of milked culture wins. A factor of about 1.5 sounds more reasonable, but we need to see a milked culture game first, before we can come up with a more precise number.

Agreed. I was just throwing in numbers for the sake of argument. I have little idea what a sensible figure would be.

I don't think a system with weight factors would work very well (as I put in the difficulty thread, but was out of place). It opens up more opportunities for score abuse, than reducing them (as was the point of the discussion in the first place).

A system with setting baselines for fastest finishes and deducing scores from that (that's the Jason system? I'm not familiar with it) would work better imho, but the disadvantages would be that it would be less meaningful to play on score and that certain maptypes would always favor certain victory conditions (in regards to beating the baseline).

As the gotm is now, you can play with your goals set on score or on fastest victories. Especially the latter already is much more appealing because of DaveMCW's great speed ranking. Would you favor a system only set on fastest scores, that would make the gotm less versatile.

Of course, there is a bit of a problem with the current base scoring system (as I pointed out before). Would that be addressed 'somehow', I'm sure that the people who play for highest score instead of fastest victory, will still continue to do so. It's one of the goals of the gotm, so it's only logical.

Part of the problem (as AlanH) put it, is how much stock people put in highests score rewards, fastest victory rewards and the feeling 'how much' people deserve it (in terms of a best game or not).

Let's make it clear that I'm solely playing for fastest victories, so I tend to find 'fastest victory' better. From the perspective of a 'highest score' player that might not be so. I think the current discussion tends to be a bit discrediting to that and should also have a bit more input from people who favor that goal (did we scare those away here? )

btw, if you look at the 'highest score' players they have a great speed ranking as well (obviously because speed already is part of scoring).

Anyway, it's quite a difficult topic and it's even more difficult to concoct a system that addresses the wishes of all. But I saw a lot of creative and interesting ideas.

 

[drkodos]

I tend to agree with that. But I'm not sure that a 1700AD diplomatic victory should blow away an 1800AD diplomatic victory, no matter what.

So the USSR won the "game" by launching Sputnik before the Americans could get John Glenn into space?

They beat the United States to the moon as well by landing an unmanned probe there in 1959.



The point: There is no scoring system that will be panacea. Unlike other games (such as Chess or Pinnochle), Civ does not provide a clean winning condition. It awards points for different facets of play and there are multiple conditions that all provide a "win".

In most games of competition the only thing that matters is the final score. There is one winner. There is one goal.

Civilization is anethema to creating a cogent and fair scoring system because its winning conditions are too open-ended and subjective.

The only way to gain some clarity would be to have each GotM have specific goals that all players would be aiming for. When you have different players aiming for different goals, it is neigh impossible to make sane or adequate comparisons. Unless there is a unified goal, there will never be a system that works.

Take a look at other competitions and show me one legitimate event where people can choose the way they want to win.

It does not exist, and that is the nature of the problem.

 

[Mad Professor]

The point: There is no scoring system that will be panacea. Unlike other games (such as Chess or Pinnochle), Civ does not provide a clean winning condition. It awards points for different facets of play and there are multiple conditions that all provide a "win".

And that is a very good point, and a strength of the game. It's not just one way of doing things that can become boring. The game provides a variety of ways of achieving a "win". Trying to compare them is very difficult indeed. Comparing cultural victories with other cultural victories, conquests with other conquessts is valid and gives us a good "feel" for the relative merits of those games. Comparing different victory methods does not give us this feel, and I think it is exptremely difficult to find a scoring method that will give that to us.

The only way to gain some clarity would be to have each GotM have specific goals that all players would be aiming for. When you have different players aiming for different goals, it is neigh impossible to make sane or adequate comparisons. Unless there is a unified goal, there will never be a system that works.

This would do it, though I wouldn't like to see this. I prefer to let players choose their victory methods, and just live with the fact that we can't very accurately compare the different victory methods. SGOTM2 has a specific single victory method allowable, and that's good, I'm enjoying it, and it woujld be interesting in GOTM's too, but as I said, I prefer the variety even better than that. I don't find it so important to compare my game against everyone else's. I'll compare my game against others who chose the same victory condition, and be happy with that.

 

[Frederiksberg]

I don't really see the benefit of comparing different victory types. To me it's like trying to decide who's the best jumper by multiplying the high jumps by 4 and then comparing that with the long jumps. I don't think such an idea would be very popular at the Olympics . So IMO there should either be a single fixed victory condition in all GOTMs or seperate awards for all victory conditions and no "overall" awards. Why should one victory type be favored over others?

Regarding the use of score as opposed to speed I would go for the simple solution and only use speed as the performance measure. Points could then be used to distinguish between equally fast victories like the goal score is used in many sports. Why dream up a complex scoring system when there is a simple solution (i.e. speed) that can be intuitively understood by anyone? I don't think that many people fully understand the scoring system used in Civ 4 and I don't think they should bother to do so. How do you think an athlete would react if he didn't know exactly how the winner of a competition is determined? Would you tell the long jumper that he has three jumps and then the result is determined by some unknown formula?

The speed performance measure is allready used in the SGOTMs and with a lot of success I would say. Now GOTM just need to follow....

 

[AlanH]

Now GOTM just need to follow....

I'd say we already recognise speed with significantly more awards than we do score. Are you aware of (a) the Five Fastest Victory awards and (b) the Eptathlon award for achieving all of them? That's more speed awards per game than medals, and we now recognise both speed and score in the Global Rankings as well.

 

[Frederiksberg]

I'd say we already recognise speed with significantly more awards than we do score. Are you aware of (a) the Five Fastest Victory awards and (b) the Eptathlon award for achieving all of them? That's more speed awards per game than medals, and we now recognise both speed and score in the Global Rankings as well.

I don't have a full overview of the awards - probably because I have no ambition of getting any . I do, however, think that it would be more fun to have fewer awards and a ranking system that doesn't try to mix different victory conditions. A clear cut goal (speed) and a ranking system that is easy to understand is what I would like as a spectator.

 

[grangerm]

If you want gotm to be about player skill the solution is clear - just use speed rankings.

I'm with you. Can we ever compare a fast military with a fast non-military fairly? I think any system is gonna have its weaknesses that players can exploit, as many people have noted. It would be nice to have a "well-played" cultural win be scored like a "well-played" domination win, but it's too subjective of a topic.

The only way to gain some clarity would be to have each GotM have specific goals that all players would be aiming for. When you have different players aiming for different goals, it is neigh impossible to make sane or adequate comparisons. Unless there is a unified goal, there will never be a system that works.

I agree. As far as I am concerned, the unified goal is what the speed rankings are all about.

Let's face it, we all download the same starting save and go off in our different directions playing what amounts to different games. Speed compares people playing the same game. When I play a cultural game, I don't care that I finish 137th in score. I compare my game to the other people who played for cultural. Why do we have to go change the scoring system, when there's a perfectly legit comparison in speed already calculated thanks to the staff and DaveMcW. Is this about vindicating that how I choose to win is better than how you choose to win?

If anything, speed is skewed towards non-military victories because less people play those victories (or at least that's my perception) and the good players going for medals play military style.

 

[Cactus Pete]

One significant problem with putting all the emphasis on speed is that, to get the fastest finish when playing against players of roughly equal ability, you will have to take risks; and if you lose the gamble, you just wait for next month.

Excessive risk taking is really not good game play (and certainly not the best way to develop and rule over a civilization). Scoring that rewards a combination of a fast finish along with strong growth and development is appropriate. We just need to tweak the system a bit to reduce the advantages of milking and, hopefully, to equalize victory conditions.

 

[malekithe]

Excessive risk taking is really not good game play (and certainly not the best way to develop and rule over a civilization). Scoring that rewards a combination of a fast finish along with strong growth and development is appropriate. We just need to tweak the system a bit to reduce the advantages of milking and, hopefully, to equalize victory conditions.

I'm curious how you define milking. In my opinion, it is anytime you choose to delay victory in order to increase your score. The simplest solution to this is to only award fastest finishes.

Excessive risk taking will almost never pan out as desired on high difficulty levels. Reasonable risk taking does, however, and is what separates great players from good ones.

 

[DaviddesJ]

Take a look at other competitions and show me one legitimate event where people can choose the way they want to win.

Many sports award MVP awards. You have to compare players at very different positions to decide who gets the award.

I'm not advocating that for GOTM, though.

 

[DynamicSpirit]

I'm curious how you define milking. In my opinion, it is anytime you choose to delay victory in order to increase your score. The simplest solution to this is to only award fastest finishes.

I think milking covers more than that. I see it as being about taking actions that're arguably bad, in the sense that they (a) requires little skill or thought, (b) have nothing to do with helping towards your victory, and (c) do nothing to help your civ and would actually damage it in the medium or long term, but you do it because it happens to give you a brief score boost just timed to coincide with your victory.

I'd be surprised if many people deliberately delay victory to improve their score. It may work on some occasions (being about to build the hanging gardens is an obvious one), but I think usually the rapid decline in the early-finish bonus with each turn that passes tends to reduce the score, making a delay in your victory not worthwhile.

 

[grangerm]

Many sports award MVP awards.

And there's never any controversy there...

 

[malekithe]

I think milking covers more than that. I see it as being about taking actions that're arguably bad, in the sense that they (a) requires little skill or thought, (b) have nothing to do with helping towards your victory, and (c) do nothing to help your civ and would actually damage it in the medium or long term, but you do it because it happens to give you a brief score boost just timed to coincide with your victory.

I'd be surprised if many people deliberately delay victory to improve their score. It may work on some occasions (being about to build the hanging gardens is an obvious one), but I think usually the rapid decline in the early-finish bonus with each turn that passes tends to reduce the score, making a delay in your victory not worthwhile.

a) Setting up a game to be effectively milked, at least on a reasonable difficulty level, is non-trivial and definitely requires thought
b) Agree there
c) I do many things during the course of the game that might be long-term detrimental to my civilization. Conquering faster than I could reasonably hope to support, beelining for Mass Media at the expesne of all other techs, and shutting off research to provide money for upgrades in the last war of the game would all hurt my civ in the long term, but are good for winning the game.

The key to milking is conquering enough territory fast enough and setting up a large enough food surplus that you can grow your population fast enough to compensate for the later victory date. I have, once or twice, come within a few percentage points of domination, and then proceeded to delay victory another 100-200 years, increasing my score by a wide margin. Any effectively milked game will abuse the rapid population growth that can be had through early biology and a delayed conquest or domination victory.

 

[Dagnabit]

I think the majority of us are saying the same thing. The scoring system by its nature is subjective, needing to take into acount multiple factors. No matter how you adjust the multpliers you really can not make any useful comparisons between A fast spacerace victory and a fast cultural victory. Again (IMHO) the true GOTM laurals should go to:

• Highest Score
• Earliest Conquest
• Earliest Domination
• Earliest Cultural
• Earliest Diplo
• Earliest Space
• Highest 2050

Like was said earlier with subjectivity not a part of the outcome staff time should be reduced which means more time for more GOTMs!!!

 

[AlanH]

Again (IMHO) the true GOTM laurals should go to:

• Highest Score
• Earliest Conquest
• Earliest Domination
• Earliest Cultural
• Earliest Diplo
• Earliest Space
• Highest 2050

We already deliver this list of awards.

 

[malekithe]

We already deliver this list of awards.

Quite true. However, the speed awards, by virtue of being in the second post of the results thread and because there are 3 score awards, seem to be getting second billing. I understand there is a legacy of giving three score awards, but I just don't think it is the kind of accomplishment which merits awarding second and third places. The question I ask myself is, "Why do my second fastest cultural or conquest victories not get an award, but somebody's second highest scoring game does?" Is score that important?

 

[lilnev]

Well count me with the folks who believe speed is a better measure than score, and comparisons are only meaningful within each victory condition. The lovely sortable page gives me all that info, so I don't really care if the scoring system is changed. I suppose it would be nice if the hoopla in the forum posts was distributed a bit more according to those principles (if the three highest scores are displayed, perhaps the 3 fastest finishes for each victory condition should be displayed as well). But mostly I'm playing for the fun and for the discussion in the spoiler threads, and to compare my game to those of others who pursued the same victory.

peace,
lilnev

 

[Frederiksberg]

I would even go so far as to get completely rid of the score except as a way of distinguishing between two equally fast victories of same type. What I hear most people complaining about is the "unfair" way that the score is calculated (and does anybody know exactly how to calculate it ). Why not cut the Gordian Knot and remove it. That would leave 6 speed rankings corresponding to the 6 different possible victories. There would still be plenty of options for computing global cumulative rankings - also across victory conditions.

Rephrasing a quote from the game: A ranking system is not complete when there is nothing more to add but when there is nothing left you can remove

 

[Spoony]

Ah the scoring debate is back, and better than ever!

If it is possible to balance out the scoring in a satisfactory way then this is a great idea. It will take someone with techy knowledge to be able to put the ideas into action. So one of the first things that the movement for a better system could do with is a willing designer.

Many have commented on finishing date as a replacement to score. I can't help but think if this was fully implemented then it would just be too identicle to the gauntlet set by HOF staff. The GOTM is aiming at something different, or so I believe.

At the end of all this is the GOTM staff have a large amount of work to do here. Dave MCW has a good system for comparing results of different VC's.


I do agree with those that call for the demise of the 2nd & 3rd score reefs as it is unfair to all the other 2nd and 3rds for other awards that go unnoticed.

 

[wwassme]

I voted "no" to changing the scoring system.

After reading through all of the above thoughtful posts, it seems to me that it would be effectively impossible to come up with a scoring system that balances scores to allow a comparison of victories of different types or that (most) everyone would agree to. In addition, I feel it is up to Firaxis, not the GOTM staff, to tweak the scoring system if that is needed.

We all start with the same game save. It is the choice we each make as to the victory condition to pursue and the thousands of decisions along the way (along with the RNG's results) that ultimately decide the final score. I don't see one victory condition as "better" than any other. I choose my victory condition based on what I perceive as best for that leader, map type and conditions, subject to revision as the game progresses.

It is easy to compare victories of the same victory type, either by score or fastest finish using the current scoring system. I would be in favor of adding a category of awards for the highest score in each victory type in addition to the current awards for fastest finish, which I expect would be simple for the GOTM staff to do. Once a player determines the victory type he or she is playing for, he or she can decide to either go for fastest finish or highest score in that category (balancing the increase in score from any "milking" with the decrease in score for a later finish). To me, winning one of those awards would be more satisfying than winning a fastest finish medal since they provide a much more meaningful comparison of skill given the unbalanced scoring system than a comparison across vastly different victory conditions.

Comparing scores across victory types is much more problematic. Obviously, Firaxis' method of computing final score favors domination victories because of the emphasis on speed of finish and population, both of which tend to be higher for a domination victory due to its inherent fit with those emphasized factors. I certainly think the score computation method could be improved in many of the ways discussed by others in this thread.

I'm perfectly comfortable knowing that if I don't play for a domination victory, I'm not likely to win a highest overall score medal (which I'm not likely to win anyway). I have no objection to keeping the top score medals for those who want to play for them by maximizing their scores within the rules of the game and the limitations of the scoring system. I play for the fun of it. My personal goals are to improve my game and to finish in the top 25%, which the Civfantics forum and particularly the GOTM forum has helped greatly. I think the GOTM staff does a great service to the Civ community by providing a common game for us all to play and discuss.