The Power of the Luxury Slider

[nbarclay]

On another thread, someone asked about the luxury slider. I view the proper use of the luxury slider as critical to making the most of my nation until I reach a point where I can purchase large numbers of luxuries.

The way the luxury slider works is that a certain percentage of each city's gold goes toward making that city's citizens happy. For each gold spent, one citizen goes from unhappy to content or from content to happy.

What this means is that the slider has different effects on different cities. At settings around 10% or 20% early in the game, it may even be that only the largest, wealthiest cities are diverting gold to happiness at all. Conversely, it takes very high luxury rates to have any effect at all on the most highly corrupt cities. Maximizing the value of the slider requires paying attention to how it affects various cities, and requires keeping the cities that are paying for happiness big enough to get an advantage out of doing so.

Now consider the ramifications if your capital is on a river. You pay one gold on the happiness slider to keep each extra citizen happy, but if the citizen is working a roaded tile along a river, he brings in 2 gold. So keeping the capital big and ratcheting up the luxury slider to 10% or 20% can be quite profitable, and that's not even counting the production the extra citizen adds to the city. Yet at that level, small, corrupt, distant cities won't be wasting gold on happiness to support (relatively worthless) extra population at all. (Indeed, especially at 10%, the capital may be the only city affected by the slider!) You get the benefit of using the happiness slider in the cities where it's worthwhile without really affecting the other cities.

And even without rivers, the happiness slider can operate at least close to break-even financially (assuming you build enough roads) while providing a non-trivial production boost. The only losses of gold come from having corruption nibble away at extra citizens' efforts to pay for their existence and from cities that arent' big enough to translate the happiness boost into extra population. With proper planning, it should usually if not almost always be possible to get better than one extra production per net gold lost. Contrast that to the two-gold-per-shield upgrade cost without Leonardo's, or even the one-gold-per-shield upgrade cost with Leonardo's, and it's a good deal - especially since the production can be used for anything rather than just for units.

Also consider the situation of a size three city that's finishing work on a settler on Regent level or higher if no luxuries are available. Without the luxury slider, the city loses two food per turn to support a specialist who provides no food and no production. With the luxury slider, if mined grassland or irrigated plains are available, the citizen can feed himself so the city keeps growing and can contribute to production as well. The settler most likely gets finished sooner, and the city gets big enough to build its next settler and all the ones after that sooner as well. That difference can be extremely powerful.

My other big trick with the luxury slider is that if I start with two or three food bonus tiles by my capital (or maybe a nearby city), I'll usually turn it into a "settler pump" with a granary. Using the luxury slider lets me support a large enough population that I am (1) always working all the food bonus tiles and (2) able to produce settlers as quickly as my city can grow the necessary people. If there's plenty of room for new cities, a city that can pump out a settler every four or six turns can be immensely useful, but the size required for that is greater than can be supported without either multiple luxuries or using the luxury slider (especially on higher levels).

Finally, note that the luxury slider can become even more valuable with lots of marketplaces and/or under Republic or Democracy. Of course if you have enough luxuries and happiness improvements to max out your city sizes without using the luxury slider, there's no need to touch it. But if not, marketplaces or representative governments allow the extra citizens the luxury slider makes possible to pay for their own contentment and then some while at the same time adding to the nation's production. In such situations, if most cities can't reach full size otherwise (at least without entertainers), using the luxury slider should be practically a no-brainer.

I hope these thoughts prove useful (and I hope there aren't too many typos from typing it so late ).

Nathan Barclay

[Arrian]

Nathan,

Great post. I discovered the power of the luxury slider a couple of months ago, due to someone's (yours, perhaps?) post here on 'poly. It can really help you in the early game to keep up with the AI in expansion and production.

Growth is good. Growth allows for more commerce, more production and more cities. Growth requires happiness. Simple equation, really, but it took me a while to get it through my head that using the luxury slider would not cost me a ton of money to help 1 or 2 cities.

In the game I'm playing now, I've been running 10% luxuries pretty much since 1000bc. I'm now up to 4 luxuries, Sistine & Bach, so I'm gonng shut them down.... or maybe not.

One thing you didn't mention (probably because you seemed to be focusing on the early game) is WLTKD. WLTKD can be really helpful for developing cities that are "on the bubble." By this I mean that they are losing about 50% of their production to waste, but have potential to be really nice cities. They lack improvements because they have low production. You have two options: rush buy everything, or rush buy enough (say a courthouse & either a marketplace or cathedral) to get them into WLTKD. This will boost their shield production noticeably. You want these cities to continue to growth, though, and the danger of them falling out of WLTKD due to growth is high, because they still aren't production powerhouses. Anyway, I find that 10% luxuries often keeps them in WLTKD so they can contribute shields to the improvements they need to eventually become productive cities.

-Arrian

[Gastric ReFlux]

This is a good explanation of why using the luxury slider is often better than hiring entertainers, or capping the growth rate of cities. Entertainers are about the most useless form of Civ3 population and you lose out on a tile that can be giving food, shields, and commerce.

There is also an indirect benefit of gearing play towards supporting cities with large populations. To do it, you must be more aggressive in securing luxuries and building marketplaces to keep the citizens happy. This more aggressive attitude can prevent the nothing-to-do, stagnation syndrome which leads to a planless and drifting type of play. But if you know you need to get that luxury, bang on that AI civ to secure those spices, you'll have to form a plan. Having a plan is one of the primary elements of beating the Civ3 AI.

[Theseus]

Great post, and great point about WLTKD.

The best way to learn to use both sliders, IMO, is to play around with them.

I think most people assume there's a 1:1 relationship between money, research, and happiness... NOT TRUE AT ALL!!

Due to all sorts of subtle factors, tweaking the sliders can have very meaningful positive impacts at very low cost. For instance, Nathan did not give specifics, but I do the same thing in the early game in building a settler factory on a river... I'll some times have that baby at 6-7 pop, and luxuries at 50-60% (for a turn or two).

I think about the sliders almost every turn, or at least every few.

[jshelr]

Thanks for the tips. If I know how to put stars on threads, I'd put a few here!

Is there a way to play with the f1 screen to see "what if" results from adjusting the lux slider or are you burrowing into the individual city screens to see what is going on??

[Theseus]

I do it from F1... I typically am looking for research duration, gold generated (or the deficit), the happiness status of citizens, and building duration.

If I see cities with 1-2 content cits (as opposed to happy), I'll dive into that city and see if there are any happiness improvers I can build.

BTW, one of the things I like to do in F1 is play with the order of cities, often ranking by production.

[XOR]

Luxuries at 10% seem to cost me about 300gpt, while buying someones luxury costs me one of each of my own luxuries + something like 1000g. And that's without extortion, because I can always attack the AI and get anything I want in exchange for nothing else than peace, I would like to know if such tactic would work against human players.

How many cities would a human player have to lose to before he surrenders his luxury, some tech, and his maps in exchange for peace? Would it suffice to just threaten him with a dozen nukes? I can imagine "I have 11 more nukes from where that one came, now give me 330gpt so I can move up the luxury slider"

[nbarclay]

Originally posted by jshelr
Is there a way to play with the f1 screen to see "what if" results from adjusting the lux slider or are you burrowing into the individual city screens to see what is going on??

When you move the slider on the F1 screen, the "citizens" on the right of the screen change accordingly. The icons indicate moods (although the differences are subtle enough to be hard to read), and the citizens are grouped by mood. Moving the mouse pointer over a citizen brings up a little text box indicating the citizen's mood. (Actually, "over" isn't quite right; it's offset a bit, but if you play around a little, you can get a feel for it.)

Also note that barring something like war weariness or the loss of a luxury, unless I'm forgetting something, there are only two times when a city will start to riot. One is when a city that was already borderline on happiness grows. The other is when MPs leave a city where they were needed to keep order. So if you can remember which cities are borderline, watching for those conditions can actually be easier than hitting F1 all the time.

By the way, under Despotism, Monarchy, or Communism, MPs can be used to in conjunction with the slider to even things out a bit. Put any units you cans pare from the potential front lines (up to the limit on how many are useful) in the cities that become unhappy first and you don't run into the choice of having to move the slider or use entertainers as quickly. (Of course if you maintain garrisons as a matter of principle, you may not have to choose of which cities to have troops in and which not to, but I tend to keep a rather lean military and focus on growth and building city improvements.)

Nathan

[nbarclay]

Originally posted by Arrian
One thing you didn't mention (probably because you seemed to be focusing on the early game) is WLTKD. WLTKD can be really helpful for developing cities that are "on the bubble." By this I mean that they are losing about 50% of their production to waste, but have potential to be really nice cities. They lack improvements because they have low production. You have two options: rush buy everything, or rush buy enough (say a courthouse & either a marketplace or cathedral) to get them into WLTKD. This will boost their shield production noticeably. You want these cities to continue to growth, though, and the danger of them falling out of WLTKD due to growth is high, because they still aren't production powerhouses. Anyway, I find that 10% luxuries often keeps them in WLTKD so they can contribute shields to the improvements they need to eventually become productive cities.

-Arrian

Have you tried to analyze how using the luxury slider to let distant, corrupt cities grow without losing WLT?D compares with keeping the gold and using it to rush improvements? My usual approach (mid to late game) is to hold distant cities at whatever size they can reach without losing WLT?D by working hills/mountains or using specialists. I don't get the extra production from the extra size, but I don't drain off gold either. If I can research in four turns and still turn a profit, I'll use extra gold to rush buy to give the distant cities a boost, generally focusing on courthouses, marketplaces, and police stations (if available) first. I'm curious regarding how your idea would compare with that.

One potential advantage I see to keeping the gold and rush buying is that rush buying can focus more squarely on the neediest cities. In contrast, a city far from the core may be lucky to get even one point of happiness (unhappy to content or content to happy) from 10% on the luxury slider, and one point may or may not make a difference in how large a population can be supported.

Nathan

[Arrian]

Nathan,

I was thinking of the benifits once the initial rushbuying is done. For instance, a city "on the bubble" will get a temple, courthouse and marketplace bought for it. If it needs other things (harbor, aqueduct) asap, I will buy those too. But once that's done, the city becomes viable, and does generate enough gold to remain happy. The WLTKD bonus helps with shield output so I don't need to buy things anymore... at least not from (almost) scratch. I too will slow down a city's growth via working hills/mts if I'm worried it will drop out of WLTKD.

I have not actually done a study of the relative costs, and it's entirely possible that I'm going about this in the wrong way. However, I tend to run 10-20% luxuries late in the game to make absolutely sure my people are happy. I want to see 95% or better for my approval rating once I'm in the industrial age. So my decision to run luxury spending isn't based upon the "bubble" cities' development, but rather my civ's overall happiness. The benifit of WLTKD in far flung (but not totally hopeless) cities is a nice bonus.

In any case, I'm not talking about using the slider for the benifit of peripheral, high-corruption cities. Those will likely never bring in more than 2 or 3 uncorrupted commerce, and thus 10-20% does nothing for them.

-Arrian

[vmxa1]

I have become a convert to the lux slider. In the very late mop up phase, I will jack it up as you do not need all your incoming.

[vulture]

I was persuaded to give some more thought to the use of the luxury slider early in the game, courtesy of this thread. A little bit of playing around with the numbers surprised me - under certain (fairly common) circumstances, it is much better to let cities grow and control their happiness with the luxury slider than it is to limit their growth to what can be maintained without the slider.

The basic condition for this to work is that you are up to date with building your roads. If your cities are large enough so that they have to work undeveloped tiles, then I think you do better to produce workers for a while until you are ahead of the game with road building.

So for now, assume that every extra square you get to work produces 1 (or more) gold, and if you are up to date with tile improvements, at least 1 shield (flood plains not included, but they produce at least 2 gold when roaded).

The conclusions here don't depend on how many free content citizens you get, how many happiness improvements or how many luxuries you have (or not very much anyway), so for the sake of argument I'll consider cities with 1 content citizen and no improvements. What is the ideal size for this city? As per my usual modus operandi, I present a table First column is city size. Then there are two pairs of columns. For the first one each tile (aside from the city center) produces 1 trade / turn. For the second pair, each tile produces 2 trade / turn. In each pair, the first column is the required luxury rate (in %) to avoid disorder (with no entertainers), and the second column is the 'excess' trade produced per turn (i.e. the amount of trade that isn't spent on luxuries, and so is available for science or the treasury).

1 - 00 3 - 00 4
2 - 40 3 - 30 5
3 - 40 3 - 30 6
4 - 50 3 - 30 7
5 - 60 3 - 40 8
6 - 70 3 - 40 9

and so on. The trend is pretty obvious. If each extra tile you get to work produces 1 trade then you lose absolutely nothing by letting your cities grow and pumping up the luxury rate, and what you gain is the shields produced by the extra tiles being worked. If the extra tiles you work are worth 2 trade per turn, then a larger city means more production *and* more gold coming in to your treasury (or spent on science) regardless of how high the slider has to go (well, since it goes in steps of 10%, you lose out when you go above 90% obviously...).

But the basic point is that as long as all your tiles are roaded (or have rivers), a size 6 city with 70% luxuries is more productive than a size 1 city with 0% luxuries, particularly once you hit Republic.

Sadly you can't set the luxury slider seperately for each city. And this complicates things a little. In the column one scenario (1 trade per tile), any luxury produced in a city that doesn't need luxuries means that you no longer break even in terms of generating gold when considered across your whole empire. If we consider (column 1 scenario again) two size 6 cities, one with 1 naturally content citizen and one with 6 naturally content citizens (due to happiness improvements) then working the 5 extra tiles from the first city costs us 5 gold per turn from the second one, since it is generating 5 luxuries that don't do anything (well, in this scenario, they generate WLTKD, but that's incidental). OTOH if you've been developing tiles properly, then each extra tile worked will generate at least 1 shield, so what you are getting for your 5 gold loss is at least 5 more shields. Compared to the cost of rush buying, that is still an investment, although obviously whether you want to do that will be very situation dependent.

In the second scenario with 2 trade per tile, even this setup doesn't cost you any money, and lets you get the shield production from the extra tiles. The break even point for the 2 trade per tile scenario is when you have more 'useful' luxuries than 'useless' ones (the 5 in the unhappy city are useful, the 5 in the happy city are useless, net result, we break even in gold revenues) - ignoring the extra shield production.

So, as a conclusion, for your core cities which are all at similar levels of happiness improvements and size, it is almost always better to let cities grow as large as they can and to control happiness with the luxury slider, since even in depsotism it will gain you production at no cost in gold/science AS LONG AS the extra tiles you work produce at least 2 gold or 1 gold and 1 shield, and as long as you are generating more 'useful' luxuries than 'useless' ones (which should be the case as long as your core cities are at a similar level of developement). And what makes this work better than suggested here is (as mentioned earlier in the thread) cities produce luxuries more or less according to their size, which is more or less according to need. So unless you go over the top, the useful luxuries should far outnumber the useless ones.



I was surprised, even if no-one else is. And AU 106 may put this to the test, since I expect happiness is going to be a big issue for most people.

[Harovan]

In my current game (huge/16/emperor as India) there was only 1 luxury resource on my continent. As in the early game overseas trade connections were impossible, I had a serious problem concerning my city growth. So I raised the luxury slider up even to 30%. This gave my (better developed) core cities a happiness boost and basically the opportunity to grow up to size 10-11. Thank to this fact, I was able to build a palace (hard way, w/o a GL) in much less than 100 turns. Of course, due to the money spent for luxuries I fell slightly behind in tech. I compensated this by active exploring, map selling, and by making the first overseas contact. This one gave me 3 medieval techs, more than 1000 gold and some gpt. However, when the new Palace was ready in 970AD, I still was slightly behind. But my large investment in luxuries had paid. My P/FP axis was optimal, the improved commercial trait gave me a very good economy and I caught up quickly. I got even in tech around the verge of the Industrial age.

[vulture]

Originally posted by Sir Ralph
In my current game (huge/16/emperor as India) there was only 1 luxury resource on my continent. As in the early game overseas trade connections were impossible, I had a serious problem concerning my city growth. So I raised the luxury slider up even to 30%. This gave my (better developed) core cities a happiness boost and basically the opportunity to grow up to size 10-11. Thank to this fact, I was able to build a palace (hard way, w/o a GL) in much less than 100 turns. Of course, due to the money spent for luxuries I fell slightly behind in tech.

What I'm saying (or claiming) is that, in a republic, with enough roads built everywhere, your science will actually be higher with size 10 cities and 30% luxuries than with size 6 or 7 cities and no luxuries. Especially with market places. Higher luxuries equals more science, once your cities grow a bit. Counter-intuitive, but true (I think).

[Harovan]

True, but unfortunately most of my cities were in mountains, plains and deserts and unable to grow without a massive tile improvement, which I was unable to do so quick. So only a couple of cities could grow large and I fell back in science anyway.

You can see the terrain in my 1550AD savegame, in the "AU mod" thread.