Qsc20-Spain Results & Strategy Discussion

[DaveMcW]

For my personal calculations, I value gold at 0.25 shields in the ancient age. This explains why mass upgrading is such a powerful strategy - you get double value for your gold.

But for QSC scoring purposes, it will probably be easier to value gold in your treasury at 55%, the same value as technologies.

 

[Moonsinger]

Originally posted by DaveMcW
For my personal calculations, I value gold at 0.25 shields in the ancient age. This explains why mass upgrading is such a powerful strategy - you get double value for your gold.

And with Leo's Workshop down the road, you will get at least 3 or 4 times more for your gold with mass upgrading.

 

[Aeson]

I think gold has indirect applications that make it more valuable than if you just consider the direct transfer of gold to shields in upgrading or rushing. Gold is very flexible, and irreplaceable in many applications. Production alone can't max out efficiency, neither can gold alone, but together they can.

One is that you can use gold to save shields in badly timed builds. Sometimes just a few shields short rushed can mean the difference of a full turn of production. I've short rushed a shield in some instances that counted for 30. In those cases I paid 8g for 30 shields. Normally these types of situations in the QSC period don't count for nearly that much, but they do occur and you need the gold to take advantage of them. Gold accumulated in the QSC period is just as useful outside the QSC period as anything else.

Upgrades aren't just powerful because you get shields 2 for 1 (or 1 for 1 with Leo's), but also because you can have your new units ready to go the turn after discovering the proper tech. Using straight production, this wouldn't be possible. A well timed mass upgrade can get new units out faster than production alone can, and it can be a rather big difference. Knights for example usually take 5-10 turns for a decent sized city of the period to build. That's 5-10 extra turns of the Knights window of opportunity, increasing the value of the production put into them as well as the gold. Upgrading also keeps shields invested into units from losing value until they are only worth their disband price. Upgrading a Horse to a Knight isn't just adding 40 shields of value, but also maintaining 23 shields (I don't think I've ever actually disbanded a Horseman... 7 shields right?) of value as well.

Diplomacy and research both benefit from gold. There are times where a deficit needs to be run, and gold in the treasury is the only way to do that. Trading opportunities sometimes require large sums of gold to pay off.

Basing the value of gold on how it converts to shields really isn't any more fair than basing the value of production on how it converts to gold (wealth). Both have their functions. In a pinch (or a good opportunity) one can be replaced by another at a low efficiency (in a direct sense), but that efficiency isn't their value.

 

[TedJackson]

Originally posted by Aeson
Basing the value of gold on how it converts to shields really isn't any more fair than basing the value of production on how it converts to gold (wealth). Both have their functions. In a pinch (or a good opportunity) one can be replaced by another at a low efficiency (in a direct sense), but that efficiency isn't their value.

Agreed!


Ted

 

[DaviddesJ]

Originally posted by Aeson
I've short rushed a shield in some instances that counted for 30. In those cases I paid 8g for 30 shields.

This is a red herring, because everyone can do this, whether they have 200 gold in their treasury or 2000. Sure, it would be a big disadvantage to be flat broke, but it doesn't capture the marginal value of the large amounts of gold that most people accumulate in the QSC.

Upgrades aren't just powerful because you get shields 2 for 1 (or 1 for 1 with Leo's), but also because you can have your new units ready to go the turn after discovering the proper tech.

Many (and soon I think it will be most) players are deliberately delaying technologies, or avoiding connecting resources, or disconnecting resources, so that they can build cheaper units and then upgrade them. Those are the upgrades that I'm talking about, not just upgrading units when you get a new tech.

Basing the value of gold on how it converts to shields really isn't any more fair than basing the value of production on how it converts to gold (wealth).

Producing wealth doesn't really have any value, certainly not in the QSC period. The goal of QSC scoring is to measure the strength of the position for the rest of the game. If, as a practical matter, the best use of the gold is upgrading units or rushing improvements, and that's how people actually use it, then that's how it should be valued, if you want the QSC score to align as closely as possible with "how good" the position actually is.

 

[Aeson]

This is a red herring, because everyone can do this, whether they have 200 gold in their treasury or 2000.

Everyone can use a Cow to grow faster too. It doesn't mean Cows aren't worthwhile. When placing value on gold, it doesn't matter if everyone can do it or not... it takes gold to do, and so you have to take that ability of gold into consideration when evaluating it.

Many (and soon I think it will be most) players are deliberately delaying technologies, or avoiding connecting resources, or disconnecting resources, so that they can build cheaper units and then upgrade them. Those are the upgrades that I'm talking about, not just upgrading units when you get a new tech.

That really only proves the value of gold is pretty high. If I disconnect my Iron, build some Horses, and upgrade them to Knights, I only do so if I can get the Knights out faster and more efficiently that way then by normal production/gold rushes. You can usually get units out faster and more efficiently using resource disconnect and upgrades, and so players do it.

The benefits of getting your units out faster still hold up, even if the upgrades aren't done the moment the tech is discovered.

If, as a practical matter, the best use of the gold is upgrading units or rushing improvements, and that's how people actually use it, then that's how it should be valued, if you want the QSC score to align as closely as possible with "how good" the position actually is.

You have to take into account that when upgrading a unit, you aren't just buying the upgrade. You're buying turn advantage with each unit, and you're keeping spent shields from becoming obsolete or degrading to their disband value.

As an extreme example, how much are 20 Knights worth for the 5 turns before an AI gets Feudalism or hooks up an Iron source? 10, 20 cities, more surviving elites for chances at a Leader, and/or 5-10 Knights that would have died attacking Pikes to do the same thing? And if the difference is gold, then gold has bought all that new territory, increased unit survival rates, and in many cases been the deciding factor in the game. It's an extreme example, but one that I've used in about half the Civ games I've played to blow things wide open.

Indirect benefits of turn advantage are very hard to evaluate, and will vary from game to game, situation to situation, but I think it makes up most of what gold itself is worth. It means everything is available sooner (units, tech, improvements), which makes everything else easier.

 

[DaviddesJ]

Once again, the goal of QSC scoring is (or should be) to determine whether one position is "better" or "worse" than another. Is it better to have 20 swordsmen and 400 gold in treasury, or is it better to have 20 warriors and 1200 gold in treasury? The answer is that the former is generally somewhat better: it's more flexible to have already upgraded the warriors, than to have to upgrade them later. And the latter player is going to do this upgrade sooner or later. But the QSC scoring pretends that the latter is better, because it gives an artificially high value to gold in the treasury.

If it makes you feel better, I'll rephrase my suggestion as "units should be worth 2*shield value". Everything you say supports the notion that having units right now is worth more than getting them later. But the QSC values getting them later more than having them now. Whether you correct that by making the units worth more, or the gold worth less, is just a matter of semantics.

 

[DaviddesJ]

Originally posted by Aeson
When placing value on gold, it doesn't matter if everyone can do it or not... it takes gold to do, and so you have to take that ability of gold into consideration when evaluating it.

If you're valuing something that requires only a small amount of gold, then assigning that same value to large quantities of gold, far beyond what you need for the occasional short rush, is disproportionate. Perhaps you should give 1 point for the first 100 or 200 gold, but 0.5 points for the next 1000 gold, and 0.25 points for additional gold beyond that. There's certainly some sort of diminishing returns.

 

[Aeson]

And the latter player is going to do this upgrade sooner or later.

The problem is that not everyone's plans revolve around upgrading to Swordsmen. I've played lots of games where I never had more than a couple Swords, if any, and it can work out very well to just focus on getting to Knights (or Cavalry) and getting as many of them upgraded as fast as possible. It's not better than an early Sword rush, or worse, it's map dependant.

A player who is saving up to upgrade to Knights (or Cavalry) may very well be in a 'better' situation than one who used their cash on Swords. That's something that varies game to game, and by playstyle. You can't have Knights in most cases in the QSC, but you can have Horses/Chariots and cash building up for them.

If a player upgrades to Swordsmen, there should be collateral points scored due to what those Swords accomplished. The Swords are worth more than the points they directly receive. Cities taken, leaders, promotions, ect. If they didn't accomplish anything (or failed to gain more than they lost), then they were a waste in the QSC period, and may very well be a waste for the rest of the game. If the upgrades were just done, then they are worth the same as Warrior+Gold under the rules.

it's more flexible to have already upgraded the warriors

Keeping your gold allows far more flexibility than investing it before being able to make use of that investment. If you have a use for the Swords, upgrade. If not, don't. Upgrading every unit is not always the best course of action.

I would say the 20 warriors and 1200g are worth more if everything else is equal, because everything else shouldn't be equal. The player who upgraded Swords should be scoring more from other sources (territory/pop) which they have aquired with their Swords.

Whether or not they are in a better or worse position depends on what both players are doing. If the player with the gold holds off and has a much better Knight rush because of it, they may very well overcome the Swords early advantage.

------------------------------

I just don't agree that 1g is only worth half of a shield. I think it depends on the manner in which the player uses both, the goal the player has in the game, and in many cases is terrain dependant. There are some gameplay styles and game situations where gold is easily more important than production, others where it's only worth the shields it can buy/upgrade. I don't think it's fair to base it's value on one of the extremes.

A 1:1 ratio I think is fine, because the ratio is going to be variable, and it allows for the fact that gold or production could be the 'main ingredient' to success.

 

[Aeson]

If you're valuing something that requires only a small amount of gold, then assigning that same value to large quantities of gold, far beyond what you need for the occasional short rush, is disproportionate.

I didn't say otherwise. I'm arguing against using a 1:2 (or worse) ratio based on the upgrade or shield rush exchange rates in game. The ability to short rush and actually receive more shields than cost in gold is just one point I've made that shows the value of gold isn't defined soley by those exchange rates, and so shouldn't be scored soley based on them.

There's certainly some sort of diminishing returns.

Certainly. It also applies to production. If you build 30 Warriors and have no cash to upgrade them, it's not very efficient. I very much doubt we should (or could) rate units based on the relative strength of their comrads in arms, even though their worth may be more or less depending on what's around them.

For instance, a Warrior may be worth as much as a Spear if there is no threat to the homeland and all the units are being used for is garrisoning purposes. So should Spears (or any other unit being used soley for garrison purposes in areas that are militarily secure) only be worth the shield cost of a Warrior? That type of analysis is far beyond the scope of the QSC scoring system, and probably impossible to implement. Who knows if in the future that Spear will end up a Mech Inf or long dead... while the Warrior may be upgraded to the Sword that kills an invader that would have killed the Spear, or ends up being disbanded because it serves no purpose. It's the same type of analysis you are applying to gold though.

Gold has the potential to be more useful than shields in many situations. Production has the potential to be more useful than gold. They are both most useful when used together of course, and saying one is more important than the other would be a very difficult thing to prove.

 

[DaviddesJ]

It's better to have the units already upgraded than to have to do so, because when they are upgraded you can use them at a moment's notice, while if you haven't upgraded them yet you have to get them to a barracks (or rush build one), and then get them back to where you want them, all of which takes time.

In theory, it could be advantageous to save your money in case something else is more important than upgrading the units. But in practice this doesn't happen, imho. All of the units are going to get upgraded, the timing is the only question.

In my experience, you'll almost always have considerably more gold even than you need to upgrade all of your units (at least, on the favorable GOTM maps we've seen lately). Rather than a value of 0.5 shields/gold being too high, because you might find something better to do with your gold than upgrading, in my opinion it's really too low, because usually/often you will have more gold than you can use for upgrades, and you'll wish you could do even more upgrades (i.e., you'd like to trade gold for shields at 2:1 if you could). But 0.5 is a reasonable "compromise" value that takes into account various attributes like flexibility.

But if I haven't convinced you by now, I'm not going to. I think certain assumptions by the moderators are "baked into" the scoring, and these create biases, and it would be better if these were eliminated, but if it doesn't happen, it's not the end of the world.

If there's ever any future interest in improving the QSC scoring, it's probably best to try to poll a substantial number of "top" players, as clearly there's no way for any one person to "prove" that a particular set of values is "right". But if you asked a bunch of players, I think you could get a rough consensus.

 

[cracker]

David,

We will probably not see any major revisions to the QSC scoring system for the next 3 to 6 months.

A change in the arbitrary cutoff date from 1000bc is unlikely to occur in any foreseeable future. 1000bc is just as good as any other arbitrary cutoff and was specifically chosen to target the 40 turn tech gambit cutoffs vs move decisions. If this bothered me or had any reasonable effect on play, (which it doesn't), then we would just change the max tech research turns to 39 (but again that is unlikely to happen).

We just did a major scoring update based on a statistical analysis of QSC games over a 4 to 6 month period. This was just before you arrived at the Port Authority Terminal.

We do consult regularly with top performing players to ask them to help review specific performance issues. The process is not designed to seek a concensus of "opinion" among top players but to monitor the impact of the scoring system as reflected in player performance and behavior. In this respect, you get one vote for each game you submit and zero votes for each game that you make up parallel results for. What we look at is how the scoring elements perform across large numbers of player submissions. Not just who makes the most vociferous or eloquent speechs about why they think their personal ox has been gored this week.

In the recent scoring system changes the average value of units was increased by about 20% compared to their raw shield values compared to gold costs. The value was increased by an additional 15-50% by adding scoring bonuses for unit experience.

The only reason technology costs were devalued and that the value of starting technologies were taken to zero, was because we evaluated expected tech trading sequences and determined that in the ancient age players could trade for or obtain for free approximately 6 out of every 10 technologies. This fact and the value of upgrades has nothing to do with the fact that a basic scoring principle of the QSC is:

1 = 1 = 1 food

While there are circumstances where each commoddity may have different trade offs and conversion values, it may be tantamount to running after large, noisy, untamed waterfowl to pursue an instantaneous modification to the scoring system as your your second and third entre in to playing elite.

 

[DaviddesJ]

Originally posted by cracker
Not just who makes the most vociferous or eloquent speechs about why they think their personal ox has been gored this week.

I don't have any "ox". I just have opinions about the scoring. Those opinions are clearly shared by at least some people, although most people don't seem to care (or haven't bothered to express an opinion one way or the other). It's unfair to characterize them as motivated by some sort of self-interest, whatever that could be. I don't have any QSC scores that are significantly affected; to the extent that I do, I'd probably "benefit" more than "lose" by overvaluing gold, and I don't even care if my QSC scores are high or low.

But the distinct lack of interest in this whole question by most QSC players/contributors is even more a sign that this discussion is pointless, than your own lack of interest in it. I'll shut up about it for a while.

 

[cracker]

David,

DO NOT interpret my current satisfaction level with the scoring system as anything remotely related to disinterest.

You are just new to the process and you open the door to the party van and step out with a first broad set of attacks against elements of the scoring that have been discussed and reviewed just before you arrived on the scene.

The community interest level in the scoring system just will not rationally support constant gratuitous change each time someone new arrives on the scene. If you play for 3 or 4 months and look at data across 3 or 4 months of games and then make some thoughtful inputs your opinions will begin to have more weight.

Remember again that you opinion has much less weight than thought full analysis of how the scoring system may be reflecting or effecting player performance.

The debate on the FP scoring is a great example of lack of perspective because in a review of nearly 600 QSC games game we have had TWO total FP builds up to this point in time. There just is not enough information to support any sort of debate on that issue at this point in time.

 

[DaviddesJ]

I meant only that you seem disinterested in having this discussion. I also don't think that anything I've said is an "attack", and I'm sorry that you choose to take it that way. I would hope that people can disagree without attacking each other. And even if I were "right" in every particular (which I doubt I am), I think it's perfectly reasonable for you to say, "We aren't going to make any changes for a while because we just got done making changes, and changing the system all the time is a bad idea." I agree with that too.

As for how the scoring affects play, I think it motivates some players to send out early suicide galleys. Other than that, I don't think it affects play much. A very relevant observation is that many people seem perfectly content to have done 2x40 turn research even though it greatly lowered their QSC scores. This is proof that the QSC formula isn't affecting their decisions. I think most people are playing the QSC period for what they think is the best long-term position, and not to maximize their score, which is a good thing. I also think the goal of the QSC score is, or should be, simply to evaluate various positions relative to one another, and not to "guide" players in a particular direction. But maybe you disagree.

But I can't resist commenting on the FP. We know this: the two players who built the FP clearly thought that it was worth more than 30 points. If they thought it was only worth 30 points (i.e., that the value to their position of having the FP was only as much as the value of having 3 warriors or 1 swordsman), then they would have built a bunch of units instead, which would have been worth a lot more. Even if the players were wrong in thinking that the FP was worth 200 shields, it's hard to believe that its "real value" is only 30 shields; that would imply a monumental lapse in judgment, by a couple of very good players. It also seems decidedly odd for a 98% complete FP to be worth 196 points but a 100% complete FP to be worth 30 points.

(It's also interesting that you give 60 points for a temple even in a situation where you clearly think that the temple isn't a good choice and isn't worth the cost, but you don't do the same for wonders.)

I don't think we need more than two FP builds (or indeed any at all) to come to some logical conclusions of this sort. On the other hand, certainly it doesn't matter very much how many points are awarded for the FP, or wonders in general, just because there aren't very many of them. And if it makes you mad for me to even discuss it, then it's not worth it, and I'll stop.

 

[Qitai]

Well, just to chip in. I do agree with all of DaviddesJ's point so far. The QSC scoring has definitely improve since the last review. But just like the first Jason score introduction, I do find that further review might be necessary to reflect who is truely in the best position in 1000BC. Also, the 40 turn technology thing, shifting it slightly might be the best since we do have alot of players losing out alot of score due to 40 turn gambit.

Maybe a vote will be the best way to see what is the popular opinion on this matter.

 

[cracker]

David and Qitai,

Just to emphasize again, we actively review the scoring system and will come back to the scoring review in 3 or 4 months when we have more and better data.

Our observations about the scoring system so far is that in general terms 40-50% of the players are too new to tune in to what the scorings system may be asking them to do. The QSC is primarily a toll for selfassessment and players that do this end up performing well in the QSC and in the overall game.

On the issues of votes, it is not that we do not want to know player opinions it is just that the scoring system is not designed to be based on a democracy. It is designed to be based on observations of player behavior. Your vote only counts in the end after the data from your behavior has been measured in the system.

The FP issue continues to be particularly illustrative of evolving game play issues. Two data points out of 600 games, Qitai's 30 point displacemnt value included, do not provide any reasonable measure of performance and behavior.

Most of my concern here is just for timing and experience. We need to be dedicating resources to helping players get the most benefit out of the QSC as it is and not going into the 35th episode of second guessing the scoring system. The armchair quarterbacking of the scoring system is a natural process that only truly begins to have a negative effect when it is constant and unrelenting instead of cyclical and based on review of player performance issues.

One challenge that we face is to try an hold the scorings system fairly rigid and constant while we catch up to it with the support documentation and also develop the player support mentoring pool that can get the value out of the qsc and into the hands, hearts, and minds or the players.

Or single greatest problem in the GOTM and QSC games continues to be the impetuous player threshhold problem where new players dive into the game and miss out on all the possible features of the QSC and the game depth. Each month we have inexperienced players who rush through the first 200 turns without giving any thought the resources that exist to help them grasp and play well in the game. Unless we can actively intervene to try and contact and support these players, we know that 60-70% of them will end up frustrated with their own lack of performance/success.

This final element is where we need your active help and leadership in everything you do. Even if you got hit by a bread truck tomorrow, the QSC scoring system will adapt and evolve. Focusing the energy and resolve that you have demonstrated, but focusing it on support for the masses of players who need you help in understanding how the existing score systems provide them a valid performance measure will truly be a greater benefit choice for you and for the comminuty as a whole.

 

[cracker]

I will also add a seperate note that Aeson's comments about short rushing, either by despot pop slamming or through the well timed use of cash, may be one of the most important under discussed issues that intermediate level players can use to improve their game success.

I actively use short rushing and food shifting in almost every turn of play to makes sure that I get almost every units and improvement 1 to 3 turns earlier than my opponents.

Cash is also the primary conduit to shift AI performance in tech research areas and either by funneling it selectively to certain AIs or harvesting it as a cash crop you can alter the game performance in global ways that speed up victory conditions as well as end up making the games more exciting and enjoyable.

Let's talk about short rushing instead of the mechanics of lobbying for additional changes to the scoring system. (at least for a while until the ink is fully dry on the last set of changes).

 

[Txurce]

Yes, let's talk about food shifting and short rushing and gold - whose value in higher-level early-game tech trading is worth more than a partial shield, by the way. How we give and take gold has a huge effect on the research rate. This is a very broad topic which I grasp well enough to think I'm doing what I want, but would benefit from having it discussed more fully.

By food shifting, I assume you mean juggling tiles for fastest growth or production, sometimes sharing them with another city. This is another area where I have a good working sense of the mechanics, but would surely benefit from more discussion.

Pop rushing can mean a few things to me, and I don't functionally understand at least one of them. Using gold in the later game to pay for a worker in one turn, then changing production to a bigger-ticket item and paying for that - all in the same turn - is something i do more and more. (It's particularly relevant to GOTM 21, by the way.)

What I never do, mainly because the rules changed with almost every patch until I was left behind in Civ 1.21, is despot pop slamming. How does it work, and how can it be used effectively.

 

[Mark Cutt]

I'm an analyses/numbers/facts driven person and I do not like to participate to opinion based discussions but this time I'd like to say something I think I understood about the way the Staff manages our community.

I think their objective is to grow a larger and larger community of satisfied civ players. Any other things like QSC or GOTM scoring are intermediate sub-targets needed to achieve the big one.

Therefore I fully understand and support Cracker suggestion not to overreact to every single (supposed) flaw we find in the system but to proceed in a release-like mode, analyzing all the issue, finding as musc as possible global solution and delivering a new system that includes many well tested improvements.

Marco