XOTM award and ranking overhaul?

Here is the post that got me started, from the GOTM 19 results & congrats thread:

da_Vinci said:

Maybe it is the awards system that needs a change. The score in civ is really a prosperity ranking. Base score is pure prosperity (hello cow ), final score is time-adjusted prosperity. We have three medals for most prosperous civ (time-adjusted prospertity), one award for most prosperous civ regardless of time (cow), then awards for fastest win by type. And the low awards, which are another issue altogether.

Medals seem to be on a higher tier than awards, as we give first, second and third for prosperity ... does that really make sense?

Prosperity is in part governed by victory type as well, but that is not a factor in the medals.

So ... maybe we should only have one award for prosperity, but within each victory type? So there would be a fastest diplo, and a highest scoring diplo, fastest conquest and a highest scoring conquest, etc. No medals at all.

Or to go to another extreme, have 1st to 3rd each for speed and prosperity within each victory condition (hmm ... five active victory conditions, since time is a default condition, two types per condition is 10, three medals for each ... 30 medals is a bit too much, I think). OK, scrap that.

We could institute a no medal, two awards per victory type system in a new BtS GOTM if everyone likes that ... seems late to "undo" medals in GOTM or WOTM.

But, for the older series ... why not add awards for highest prosperity (score) by victory type? We could add those retroactively to the older games.

For award symbols, just use the existing symbols, add a clock face icon to it for speed award, and a coin stack icon to it for the prosperity award.

Last issue, how does that affect the eptathlon? Currently what we have is a speed eptathlon (plus the cow, I suppose). Adding prosperity awards leaves that intact, and now there is a score eptathlon (remember, score = prosperity), which is all score awards plus the cow). And also, now a Grand Eptathlon: all speed and score awards, plus cow. Maybe that's a Decathlon ... or what is it for eleven?

Seems kind of silly to have awards for lowest scoring win by type (most efficient use of your civ awards?) but no highest scoring win by type (most prosperous civ).

I suppose that if high score (prosperity) in any victory type is inevitably the result of warmongering and then a detour to the other victory type, then perhaps the score by type awards would just be redundant warmonger awards. Thoughts?

Click to expand...

From there, I have some new thoughts:

First, an easy one. We have global rankings for score (prosperity), speed, and a combined rank that is the better of speed and score. Seems it would not be hard to add a fourth table, that is based on the sum of speed and score points for a particular game (sum and average will rank the same, so save the division step). So each would have a Speed rank, a Score rank, a BOSS rank (Best Of Speed and Score) and a SOSS rank (Sum of Speed and Score), BOSS and SOSS being two variants of combined.

The sum of the two will reward those who have good balance between speed and score, in a way that the better of the two does not.

Then there is the classic question of whether we should index the score points to the highest score of that victory condition, as we do with the speed rank.

Problem there seems to be that you would have five players with 100 score points per game ... how to award the medals?

The solution is to no longer give the medals for highest score, but give the three medals for the highest sum of score and speed points after indexing to the max in each. This is illustrated in the attached spreadsheet (well, without the indexing of score feature as I took this off the current global tables).

Since this means the medals are not a pure score item any more, we should then add awards for highest scoring victory in each victory type, to go with the fastest. Something for the scoremongers to shoot at.

This system of score awards, speed awards, and medals based on a blend, both indexed to max performance within a victory type, might be the best system to use going forward (fairest to all victory conditions).

If we wanted to make a change, we are in position to implement it in WOTM 11 and GOTM 21, each the first of a new decade of XOTM.

Anyone else like this idea?

dV

Addendum: in addition to more fairness to the various victory conditions, it gives us a few new elite targets to shoot for:

The 150+ club for having a sum of 150 or higher (indexed score + indexed speed) ... it can never exceed 200.
The 125+ club (sum 125 or higher)
The 100+ club (sum 100 or higher ... unless that is too easy once score is indexed to victory type).

dV

From GOTM19 results thread, from the gold medal winner:

Cactus Pete said:

Indeed . . . Balbes and Lexad played superior games. The scoring system could use an adjustment.

Click to expand...

It is clear for me that the system needs some adjustement.

I would keep it simple:
-The most points wins an award (the current gold medal).
-Fastest conquest.
-Fastest domination.
-Fastest cultural.
-Fastest spaceship.
-Fastest diplomatic.

And nothing else. In other words, I would take out the silver and copper medals and the cow.


Three minor notes:
- Combined ranks are not interesting. You play one game either for points or for fastest finish, so a combined rank makes not much sense. In order to be high in the combined you would need to play always warmonger games.
-I would propose a fastest among all Victory conditions super-award. It is the most natural way of play: given this starting position, what is the fastest way to victory? This idea of mine has been rebuffed in the past because you would need to play always warmonger games.
-100000 points for every winner is not perfect. The GOTM19 Conquest awarded game has much less value that many of the other awards... it would be close to impossible to fix an "expected victory date for the winner" for each victory condition, though.

Obormot's idea of replacing cow in Eptathlon requirements with gold medal sounds good to me.

@dV

Thanks to put some creativity in the current system. I also think that it has to be improved. However, I am not convinced by your first (easy) proposal. As I see it, the problem lies in the fact you based your ranking system on two other rankings (score and speed) that are both far from perfect :

- Current score penalizes totally non-military victories
- Speed score is too primitive for a game called "Civilization" (for instance, the fastest Conquest looks like a barbarian victory and as far as I know "barbarian" is the contrary of "civilized").

Adding the two scores produces some strange mix that will make the results more random, because you won't be able to optimize your result since it depends of the distribution of the other results (for instance, at some point you won't know if you have to put on speed or let you score grow because the good choice will be function of the other players result).

Nevertheless, your post has the merit to address one important need: a global ranking that will deserve fairly all victory conditions. Until now, I think the best (or less bad) proposal is still DaveMcW combined ranking.

If we want to go further, I think it's necessary to address the score formula itself. It's a significant work and we already had a discussion on it. As an interesting protocol for it, here is a post due to Obermot:

"First of all we need to agree, what should be rewarded for each victory condition, select appropriate variables vi and construct a base score function B(v1, v2, ... vn) for each VC. This is a relatively easy thing to do.

The difficult part is comparing different VCs and comparing milked games to fast games. For that purpose we need to construct benchmark functions B0(t) which would represent how B(v(t)) increases as the time goes in a "perfect" game with current settings. It should depend on the type of the map, difficulty level, leader traits, UU, climate settings, AI rivals, and the quality of land near the start, and maybe something else that I forgot. This makes balancing incredibly difficult.

Then the final score at victory date t* would be S(t*,v) = const*B(v)/B0(t*)."

My feeling is that, because of the way the game in general and its scoring system in particular work, warmongering will almost always result in the highest scores. This is regardless of what victory condition you're aiming at. For example, the highest scoring Diplomatic wins I've seen have been the "backdoor domination" types where someone simply votes themselves into office after conquering (and growing) their way to a huge population.

Since score-based awards will always come down to warmongering, with or without score milking, I think it makes the most sense to leave them as they are. Replacing them with awards for "highest scoring X victory" will just encourage weird hybrid strategies, where people fight their way almost to a domination victory then hold back so they can win by spaceship or culture instead. Imho, that seems like a bad way to go.

jesusin said:

I would keep it simple:
-The most points wins an award (the current gold medal).
-Fastest conquest.
-Fastest domination.
-Fastest cultural.
-Fastest spaceship.
-Fastest diplomatic.

And nothing else.

Click to expand...

After looking more closely at the GOTM 19 results, in that game the fastest diplo, space, conquest, and cultural were also the highest scoring in that category. Might not hold for conquest all the time, but may be the case for the (relatively) peaceful victories. Have to look at other game to see if this holds up.

In which case separate speed and score awards for those conditions are redundant.

Also, my idea of summed scaled scores and speeds for the medals still produced a four-way tie at the top with sum of 200 for best of each victory other than domination. So that did not solve the medal problem either. If fastest space, diplo and cultural always gets highest score, then my idea produces a three way tie for gold every time among those three!

Of course, these are results when only speed by category is rewarded ... if we rewarded score by category, would we then see more late backdoor diplos for the diplo score award, and also "backdoor" space and cultural for the points? And would those awards just go to warmongers in sheeps clothing?

jesusin said:

Combined ranks are not interesting. You play one game either for points or for fastest finish, so a combined rank makes not much sense. In order to be high in the combined you would need to play always warmonger games.

Click to expand...

If you mean that the best combined for all victory types would be warmonger games with a final detour to a different victory condition, then I would have to agree. A combined speed and score calculation for individual game awards may create more warmongering, not less.

What do you think about combined rankings on the global list over the nine recent games? There, since you might play some for speed and some for score, a combined makes sense, IMHO.

jesusin said:

-I would propose a fastest among all Victory conditions super-award. It is the most natural way of play: given this starting position, what is the fastest way to victory? This idea of mine has been rebuffed in the past because you would need to play always warmonger games.

Click to expand...

That would guarantee that one person always gets two awards ... does that really make sense? Would fastest overall ever not be a conquest?

jesusin said:

-100000 points for every winner is not perfect. The GOTM19 Conquest awarded game has much less value that many of the other awards... it would be close to impossible to fix an "expected victory date for the winner" for each victory condition, though.

Click to expand...

The fact that the top game in each condition is tied for the top at 200.000 ruins my idea ... and the 100.000 reference does that.

Ultimately, I was looking for a way to make more peaceful victories more competitive for the medals. In reality, there may be no way to do that effectively if medals are for score. Scaling to victory type generates a huge tie for speed, score, or combined.

So perhaps it is enevitable that score awards (medals) are the province of the warmongers in most cases. Since that can be conquest or domination, maybe three of them still makes sense.

Maybe current system is like Churchill said about democracy: the worst system, except for all the others.

Perhaps the addition of Sum of Speed/Score table to global rankings is the one survivor of my list of suggestions?

dV

edit: attaching a spreadsheet illustrating the massive tie for gold with combined scaled score+speed.

First of all we need to agree what requirement the scoring/awards system should meet, to make the game the most fun for the players (this is our goal, right?). In my opinion these are the criteria for such a system:

1. We should NOT reward people for doing things that are tedious and not fun, like milking.

2. We should NOT reward people for doing things that aren't a natural thing to do, for example not winning the game when it is possible to win, but delaying the victory. Also we should not penalise players for making good decisions (I remember someone suggested penalising slavery for example). Using good strategies should be rewarded, using bad strategies should be penalised.

3. The scoring should be "fair", i.e. reward different victroy goals equally. But on the other hand we should only reward popular victory goals: that is high score and fast victories of each kind. There is no point in adding extra awards like high scoring culture wins that nobody will play for.

4. Here is another one that nobody mentioned yet: there should actually be some competition. Now, for example, if I play a GOTM, I am 90% sure that I'll win a medal or an award of some kind. And that is true for all the best players here. Just look at the results of the last few games, you won't be able to find a single entry by Cactus Pete, Balbes, Grey Cardinal, etc. that didn't win an award of some kind. That is one of the reasons I am not an active player now, winning all the time gets boring too. So I am absolutely agains ADDING new awards and ratings. We should simplify the system and remove unnecesarry awards rather then make things even more confusing.

So based on all that I think here are some improvements we can do:

1. Remove the cow award or at least remove it from the reqirements for the epthatlon. Nobody plays for the cow just for fun, people either win it by accident or to get closer to the epthatlon. It is boring and time consuming to milk to 2050AD.

2. Reward high scoring games and fast victory games equally: remove the silver & bronze medals, but make the gold medal required for the eptathlon instead. Remove the score & speed ratings and leave only the combined one.

3. Adjust the scoring system. As da_Vicnci said, score is a measure of prosperity adjusted by the victory date. But the current base score is not a good measure of prosperity: it depends mostly on land and population. A better base score would be a function of GNP, production and food, since any action that a player does in his game are either geared towards improving his position (i.e. increasing GNP, productionb or food), or towards actually winning the game. Developing the actual formula may be tricky though.

4. The most radical suggestion: just adopt the SGOTM system with unique goal for each game and awards for the 3 best players.

Personaly I would just do #1 and #2 for now, since #3 & #4 are too radical and hard to balance.

Not much surviving of my initial ideas

Vynd (among others) correctly points out that highest score for a victory type will be a warmonger game, and it will have the unnatural quality of delaying the win, as Obormot noted.

So high score per victory makes little sense on further thought. It won't expand the list of award winners beyond the current list of regulars, so why bother?

The idea of combined speed + score for medals was an attempt to make medals more available to other victory types, without redesigning the score function. Probably can't be done without a complex set of functions. So scrap that.

Medals will be the warmonger's territory ... so the question raised by jesusin and Obormot is do we need three, or just one? That may still be open for debate.

Kill or keep the cow? I would agree that for a top player, the cow is a real yawn. But for a begining to intermediate player, the cow may be the only award within reach (in part because the good players ignore it! ). And for a new player, chasing the cow may teach some techniques that can be applied earlier to improve later results. So I would say keep the cow, but remove it from eptathlon. Whether to replace it with gold medal or not could still be debated.

EEO correctly notes that the global ranks are based on imperfect functions ... but better ones may be hard to come by. He also states that summed score and speed is both strange and random. Since I don't advocate replacing best of speed/score with sum of speed/score, but just adding a fourth table, what's the harm in that? If someone's play style achieves a balance between empire size and victory speed, is that noteworthy? Or, maybe it will not turn out to be different from the other rankings.

Obormot would remove speed and score rankings and just keep the combined one. Why not keep all three (or even add my SOSS), and everyone can judge for themselves which ranking method they feel is most important. Consider swimming: some swim freestyle, some backstoke, and some swim individual medley.

In any event, minor tweaking rather than major overhaul would describe the change, if any, that would make sense.

dv

The silver and bronze medals have their merits. Please don't remove them. I'm very proud of my silver/bronze medals and they represent an acknowledgment to players who sometimes manages to play at the top of their performance, even if that is not enough to win the gold. They are very encouraging.

The Eptathlon award is a proof (or indication ) that a player can master many aspects of the game. It requires a special skill to win the gold, although I don't really know how much it differs from the fastest domination + milking (no offense intended to those who play for score). Making the gold medal a prerequisite for the Eptathlon would make the award harder to win, but does it really add any value?

I'm happy to remove the cow reward from the Eptathlon, since I find that award boring. (Again, I understand that it takes skill to win the cow.)

I like that awards are given to the fastest wins. This gives a reference to all the other players and indicates what strategies are successful. Personally I'm more impressed by faster victories than higher scores, and I'm sure most players can acknowledge both fast victories and high scores.

Obormot said:

1. We should NOT reward people for doing things that are tedious and not fun, like milking.

Click to expand...

OK. But I fear there is no universal definition for "tedious" and "not fun". One player could think that warmongering is tedious and the other player that 's the only way to have fun ...

Obormot said:

2. We should NOT reward people for doing things that aren't a natural thing to do, for example not winning the game when it is possible to win, but delaying the victory.

Click to expand...

You see things like a pragmatic chess player. I know this, because I was thinking so before Civ 4: firstly winning and secondly ASAP. But if you consider the situation with some detachment, you realize that Civ4 has more dimensions, and that there is something more than winning quickly, it is winning brightly ... Therefore, it makes sense to delay the victory in order to realize something greater than your empire currently has (for instance achieving a wonder, researching some tech or finishing a conquest). I feel it clearly within the spirit of the game if the score formula rewards such a behavior.

Obormot said:

Also we should not penalise players for making good decisions (I remember someone suggested penalising slavery for example).

Click to expand...

Something strange in your logic. A decision is good according to some rules and some objectives. As soon as you change a little bit of rules or objectives, what was a good decision can obviously become a bad decision. If we follow your logic, we shoud never change anything.

For the case of pop rushing (of course, I was the one suggesting it was too powerful) have a look at the post of AU_Armageddon in GOTM 20 First Spoiler: http://forums.civfanatics.com/showpost.php?p=5647693&postcount=7

So, the game designers think too that poprushing is too powerful and they changed it in BTS. QED

Obormot said:

3. The scoring should be "fair", i.e. reward different victory goals equally. There is no point in adding extra awards like high scoring culture wins that nobody will play for.

Click to expand...

Here, I agree absolutely

Obormot said:

4. Here is another one that nobody mentioned yet: there should actually be some competition. Now, for example, if I play a GOTM, I am 90% sure that I'll win a medal or an award of some kind.

Click to expand...

I understand. I think the best we can do is having only one competition based on an only one (good) score formula.

EEO said:

OK. But I fear there is no universal definition for "tedious" and "not fun". One player could think that warmongering is tedious and the other player that 's the only way to have fun ...

Click to expand...

True, but still there are things that almost nobody likes to do.

EEO said:

You see things like a pragmatic chess player. I know this, because I was thinking so before Civ 4: firstly winning and secondly ASAP. But if you consider the situation with some detachment, you realize that Civ4 has more dimensions, and that there is something more than winning quickly, it is winning brightly ... Therefore, it makes sense to delay the victory in order to realize something greater than your empire currently has (for instance achieving a wonder, researching some tech or finishing a conquest). I feel it clearly within the spirit of the game if the score formula rewards such a behavior.

Click to expand...

Well, GOTM is a competition, so unfortunately it is almost inevitable that people play more pragmatic then they would have played in their "private" games (at least the players who care about the score). That is true for any kind of sport. Personaly I find that the most simplistic way to win is also the most "beautifull" in about 80% of the time, but that is a matter of taste.

EEO said:

Something strange in your logic. A decision is good according to some rules and some objectives. As soon as you change a little bit of rules or objectives, what was a good decision can obviously become a bad decision. If we follow your logic, we shoud never change anything.

Click to expand...

My definition of "natural" is independant of the scoring system:

1. When you still don't have a clear advantage over the enemies, you "build up", i.e. increase your research capacity and/or production. That is done by means of strategic wars, building wonders and improvements and increasing population. The main direction of developement may vary depending on the chosen victory condition.

2. Once you have a clear advantage you finish the game with your chosen victory condition as soon as you can. Here you may even sacrifice some of the potential aqquired in the previous stage, because "it doesn't matter anymore".

I understand that some people may find it amusing to continue with stage one developement, even having a clear lead. But again, I don't think it is wise to reward such play in a competition game.

EEO said:

For the case of pop rushing (of course, I was the one suggesting it was too powerful) have a look at the post of AU_Armageddon in GOTM 20 First Spoiler: http://forums.civfanatics.com/showpo...93&postcount=7

So, the game designers think too that poprushing is too powerful and they changed it in BTS. QED

Click to expand...

It is great that the developers track the discussions in the community and nurf the strategies that are overpowered. I'll propably return to GOTM and play at least a couple of games once BtS is released to see how it feals. However I think its wrong to fix such things by adjusting the scoring system. A better solution would be to incorporate the changes done to slavery into the HOF mod for the vanilla game (like it was done with the whipping bug fix after the warlords were released).

EEO said:

I understand. I think the best we can do is having only one competition based on an only one (good) score formula.

Click to expand...

I think you should just go on and try to create an alternative scoring formula. A "natural" scoring system should be based on research potential and production rather then land and population. The conquerors will have high production but low (negative) GNP, space victories will have very high production and GNP, but a late date and diplomatic victories will be somewhere inbetween. It will still be hard to find a balance.

Erkon said:

Making the gold medal a prerequisite for the Eptathlon would make the award harder to win, but does it really add any value?

Click to expand...

Well, Eptathlon is supposed to be difficult. If we remove the cow it will leave us with only 5 awards required for the statue, plus with some luck it is possible to get some of the awards for nothing (together with a shield).

I think the gold medal is a good replacement for the cow because the cow award seems like a vestige of attempts to fix the scoring system back in the first year of the Civ3 gotm. People didn't like the in-game scoring so the Jason scoring system took effect. The cow was kept around out of, I dunno, nostalgia for the in-game score? To placate in-game aficionados? It seemed like a good idea at the time, but since Civ4 even the Firaxians are duplicitous in their scoring: the in-game is there to look at, but the Final Score is automatically calculated. Why bother? Why not have a single score? Yeah I know, what's the formula etc...

Of course there's the contrary viewpoint that advises against killing your cows. It could live on in the space occupied by phrases like 'all your base are belong to us'. And Cricket.

I agree with every word of Obormot's first post (and I suppose that implies that I disagree with most of what EEO said ). Every single word.

da_Vinci said:

After looking more closely at the GOTM 19 results, in that game the fastest diplo, space, conquest, and cultural were also the highest scoring in that category.

Click to expand...

That does no happen every game. This was a very domination-oriented game. Usually gold is a domination and fastest domination is a different game.

da_Vinci said:

What do you think about combined rankings on the global list over the nine recent games? There, since you might play some for speed and some for score, a combined makes sense, IMHO.

Click to expand...

Yes, it makes sense.

jesusin said:

I agree with every word of Obormot's first post (and I suppose that implies that I disagree with most of what EEO said ).

Click to expand...

Jesusin, are you sure you read all the previous posts in this thread ?

Because, if you did, you would have realized that Obormot and I agree on many fundamental points:

- Need of fair scoring system according to the different Victory Conditions

- Importance of maintaining competition by not multiplying medals/awards

- Trying to limit the necessity of tedious actions to get a high score

On the other hand, we have a different taste according to the "speed" aspect of the game. Obormot never sees a reason for rewarding a delayed victory, where I see it within the spirit of the game if it allows a major achievement ...

This exchange with Obormot was very interesting for me, because both of us were aware of the part of rationale and the part of personal taste in our points of view.

I regret that you missed these nuances.

Don't you think you are too disappointed by having had several second ranks in the awards race and consequently that you don't like to see the second (and third) place in score rewarded ?

Please, take some detachment for your bad luck and argue at the level you play the GOTMs.

EEO said:

I regret that you missed these nuances.

Don't you think you are too disappointed by having had several second ranks in the awards race and consequently that you don't like to see the second (and third) place in score rewarded ?

Click to expand...

I am sorry I included that unfortunate comment about your opinion. You are right, your opinion and Obormot's have a lot in common. Please, accept my excuses.

My second fastest results have nothing to do with my position. I have had as many second-fastest as I have had awards.
I don't want to see score given more importance than speed. I want to be an Heptatlete, but I won't play for the cow, even if that was the last achievement needed.

EEO said:

For the case of pop rushing (of course, I was the one suggesting it was too powerful) have a look at the post of AU_Armageddon in GOTM 20 First Spoiler: http://forums.civfanatics.com/showpost.php?p=5647693&postcount=7

So, the game designers think too that poprushing is too powerful and they changed it in BTS. QED

Click to expand...

Can you tell us what it says, for those of us who don't want to read the spoilers yet?

A comment from a more casual player...

Can I request the system to be kept just as simple and straightforward as it is. Eg, awards for score; awards for time; not rankings based on 3 * e^phi(score) + (2 * sqrt(dScore/dTime) )*, or worse still a Duckworth-Lewis system.

As Obormot mentions, there are only a handful of regular award-winners. One reason is because shockingly some of us don't actually care to put that much effort in. However, we still like to see how our scores and rankings relate to those who do put the effort in, without having to deal with bizarrely concocted schemes. The score and the date are both printed on the screen in-game, so those are pretty convenient; calculations based on them aren't so handy.

Eg, if the rankings are based off speed, I can -- without any effort -- see how it will influence the result. ("Hmm, if I put of that war for 40 turns, it'll cost me 40 turns of speed" vs "If I put off that war for 40 turns, it'll cost me... where's my book of logarithms??") In other words, if it's simple, casual players like me might be persuaded to compete a little more so Obormot doesn't get so bored with us patzers; if it's a concocted scheme, I don't think I could be persuaded to read the scheme let alone try to compete to it.

* No that isn't intended to be a real formula!

** A point to remember is we're not trying to be the fastest / highest-scoring; we're trying to trade off where we think we'll place ("somewhere in the top half is ok" or "going for that top 10 spot") with how much fun gets sucked out of the game by optimisation.

DaviddesJ said:

Can you tell us what it says, for those of us who don't want to read the spoilers yet?

Click to expand...

It sais that certain tactics that were "overpowered" in vanilla/warlord like slavery, or the Alphabet beeline will be fixed in BtS:

1. Slavery is nurfed (I don't know how exactly, maybe someone who reads the pre-release info regulary can tell us?).

2. You can only trade techs that you researched yourself.

3. Catapults don't withdraw from combat anymore.

whb said:

Can I request the system to be kept just as simple and straightforward as it is. Eg, awards for score; awards for time; not rankings based on 3 * e^phi(score) + (2 * sqrt(dScore/dTime) )*, or worse still a Duckworth-Lewis system.

Click to expand...

I agree with you. I also like having one clear goal when playing. That's one of the reasons I have always been a fan of a pure speed system. BTW, the built in score is actually already way more complex then your formula.

Obormot said:

2. You can only trade techs that you researched yourself.

Click to expand...

My impression from other threads is that this is a game option in BTS, not the default.

I had not heard earlier about slavery changes. It will be interesting to see what they are.