Worst Leader?

[PieceOfMind]

feralminded,

I don't think the asymmetry in the way combat mechanics works is without reason. It makes things more interesting in the sense that units that are meant to be defenders (archers, longbows etc.) have a decent chance against attackers, and are themselves bad at attacking because of their often low base strengths.

Apart from the way combat promotions are treated differently, I don't think it's much more confusing. If anything it frequently makes calculations simpler because you just add up all the positive and negative bonuses and apply them to the defender. What would have been nice however, is if Firaxis actually bothered to tell us how at least these modifers worked. In every previous version of Civ there has been a worked example or two in the manual. But, of course, Civ4 and its expansions had the worst manuals in existence.

 

[feralminded]

feralminded,

I don't think the asymmetry in the way combat mechanics works is without reason. It makes things more interesting in the sense that units that are meant to be defenders (archers, longbows etc.) have a decent chance against attackers, and are themselves bad at attacking because of their often low base strengths.

Apart from the way combat promotions are treated differently, I don't think it's much more confusing. If anything it frequently makes calculations simpler because you just add up all the positive and negative bonuses and apply them to the defender. What would have been nice however, is if Firaxis actually bothered to tell us how at least these modifers worked. In every previous version of Civ there has been a worked example or two in the manual. But, of course, Civ4 and its expansions had the worst manuals in existence.

The asymmetry produces ridiculous use cases though (as per the previously stated "don't promote your Praets to city raider" example). Ultimately there's NO reason to do it this way ... not only is it slower (floating point math ), it also misleads users. Honestly if you want defenders to have better odds on defense just make a promotion or innate ability "+50% when defending" or if you want them to be weak when attacking "-25% when attacking". You don't have to cook up an asymmetrical combat system that obscures this from the end user. Again this is not the first time I've seen bad math in video games and I'm sure it won't be the last but it never ceases to annoy me because even a slightly mathematically inclined person could show you the obvious issues with this approach.

Any time you design an interface that rewards non-intuitive behavior, you've designed a poor interface ... regardless of what your intentions are.

 

[PieceOfMind]

Actually the math is done in integers. The only time floating point math is called is for a specific part of the odds calculator where it is unavoidable unless you want heavily rounded intermediate results and hence probabilities that are way off. Unless you want to tell me which floating point math you are referring to?

I don't know why you are calling it bad math. Saying so makes it sound like it is inherently wrong. You just don't like the way it works - that doesn't make it "bad math". I agree there are some oddities to the system (praets benefiting more from Combat I than CR1 in some situations, for example, which I showed less than a week back), but any combat system will have some problems - even a completely symmetric one like you propose.

To be honest, having -25% when attacking on tons of units or +50% when defending is only making things more complicated. I just get the impression you are basing your judgement immediately off the fact the system is not intuitive at first, and ignoring the several issues which go into making a combat system which is fair and gives reasonably expected odds between a huge number of different units.

Having the bonuses always apply to the unit itself would make all city attackers frequently benefit more from the C1 promotion than CR1, meaning you'd have to re-adjust all the CR promotions. Whichever system you use there is always going to be thresholds where it is wise to switch from one promotion type to another because of the base strengths. As it is now, the cases where Combat1 beats CR1 are fairly rare (very high base strength units attacking weak defenders, like praets (which are possibly overpowered anyway) and modern armor).

By the way, I'm currently working on a combat calculator and have been studying the way combat odds works for some time. I know my way around a lot of the relevant code and have tested a lot of different combat results so I have a pretty good idea what I'm talking about, just in case you thought I was coming out of nowhere.

 

[feralminded]

I don't know why you are calling it bad math.

Unintuitive math is bad math whenever the end user is involved in making decisions about it. This is a basic precept of interface design.

Having the bonuses always apply to the unit itself would make all city attackers frequently benefit more from the C1 promotion than CR1, meaning you'd have to re-adjust all the CR promotions.

Huh? How could an honest +25% ever be worse than an honest +10%? A strength 2 unit would become 2.5 vs 2.2, a strength 10 unit would become 12.5 vs 11. That's my point ... such a system is perfectly linear and thus intuitive and predictable. Ultimately you can use whatever system you want if its entirely behind the scenes but the problem here is the current system misleads the end user who DOES interact with it when choosing promotions. If we didn't choose promotions it wouldn't matter because then all that would matter would be the odds calculation ... but we do.

Honestly I think thats my big gripe ... you are asking the user to make a choice on a non-intuitive system. That's wrong and a poor implementation.

 

[PieceOfMind]

You can say it's a poor implementation all you like but unless you can prove the alternative system is free from possibly bigger problems, then you'll get nowhere.

You are describing the problem as not being intutive for the end-user. I'd say the problem is more the documentation and how the effects are described which is very similar to what you are arguing but putting it differently I guess.

For example, I believe they would have been better off describing CR promotions as -20% to defender when attacking cities - but it is more longwinded and I see why they went for simplistic descriptions because most people (not you or I or a several other civfanatics) would never really notice.

My beef is with the documentation (including how promotions are described) - not the actual effects they have in combat.

 

[UWHabs]

If it was documented better, we wouldn't have this stuff.

I still don't see why they don't just add the bonus to whichever unit it pertains to. So, a C1 spearman attacking a C1/shock chariot the battle would be:

4 * 2.1 = 8.4 vs 4 * 1.35 = 5.4 instead of it being 4.4 vs 2.42.

I guess it makes it a much more logarithmic scale, where since you're dividing, you tend to have more skew (in the example above, you get R=1.56 in the first case and R=1.8 in the second, so the way they do it helps the attacker more). There has to be a reason why they devised it this way - I mean, it's not like it's the first system of combat that I would think of, so there must be a reason.

 

[feralminded]

You can say it's a poor implementation all you like but unless you can prove the alternative system is free from possibly bigger problems, then you'll get nowhere.

There are no big problems with linear systems. The only problem with linear systems is that they are ... in fact ... linear and sometimes you need a curve. That said you can actually use linear systems to simulate a curve to a certain extent and this can often time (though not always) be enough to meet the design requirements.

I agree with you that they likely did what they did to assist defenders but my only point is they sacrificed clarity. Even with proper documentation not everyone has that firm of a grasp on math. Honestly if it says +10% in all cases and +25% in a special case ... an end user shouldn't have to parse through (relatively) complex math to make that choice. It should perform precisely as stated and be that simple.

Its just poor interface design, that simple. I do think it is possible to maintain a clear interface AND implement a system that meets their design requirements. As I said if they wanted a unit to have an advantage on defense they simply needed to create a defense promotion or ability. It would have served the same basic purpose and with a variable value could simulate the logarithmic curve their current system produces WITHOUT forcing the user to understand a complex system before making game play decisions.

I mean these are very basic tenets of interface design. You can have all the complexity you want behind the scenes but you NEVER confuse or mislead your user when giving them a choice.

 

[feralminded]

The more I think about it the more I suspect their decision was less based on the curve but more based on a desire to have opposed promotions like CG and CR completely negate each other. Its an intuitive design goal that creates an unintuitive user experience. All too often things like this slip through the cracks I'm afraid (speaking from lots of experience). I'd put my money down that this is what happened.

 

[PieceOfMind]

I mean these are very basic tenets of interface design. You can have all the complexity you want behind the scenes but you NEVER confuse or mislead your user when giving them a choice.

Let me put an example here.

How are you saying this information should be displayed instead? (ignore the extra info with the red line through it)

I think it is pretty simple that it highlights in green the good bonuses for me, and in red the bad bonuses - the ones to the defender.

And as you just said, equal bonuses negate each other - that is in fact a very intuitive feature and you realised it yourself.

From your posts it is hard to decide whether you are challenging the mechanics, the interface, or both.

 

[feralminded]

That's not the interface I object to, its the interface where a user selects promotions for their units and the way those promotions are effectively misleading.

The rule is: Any time you ask a user to make a decision you provide them nothing but facts.

As I said I can understand how they got to where they are ... I could even see the spec sheet ... but I cannot condone them for leaving their users behind in the process. Now they could fix it in one of two ways. One is to actually label them for what they do and clearly explain the combat system ... but this will very likely confuse a lot of players (leading to an ever poorer interface) or they can design the system to meet the user's needs while still retaining the level of playability they desire from the combat engine.

My instincts as a software engineer myself tell me that odds are they went the 1st path and then QA said "this is too confusing, simplify the labels" and never really considered the 2nd path because at that point the engine was written and would be too "hard" to refactor or more importantly rebalance and regression test. This is fine, concessions like this happen all the time in this industry, but that doesn't make it any less sub-optimal and as a consumer doesn't make me any more willing to accept it and not be critical of it.

 

[Arlborn]

You guys just managed to confuse the sht out of me. But then again, it is not that hard to do when numbers are involved!


I am starting to see it though, and it is very weird, not at all like I thought the combat system really worked... Geez, I should have read more over it in the past, but too much calculations in a thread does put me off hehe.


Well, I will stick with the little % number I see in the interface!

 

[PieceOfMind]

Sorry I'm going a bit quotey on you here

That's not the interface I object to, its the interface where a user selects promotions for their units and the way those promotions are effectively misleading.

They're misleading only if you don't know how they work. Not knowing how they work is because of a documenation problem - it was not clearly explained in the manual (well... not even touched by the manual )

The rule is: Any time you ask a user to make a decision you provide them nothing but facts.

Agreed, unless those facts unnecessarly overwhelm the user with information. There are lots of little inconsistences all over the game that are not explicitly documented. Just look at the way it says that Protective leaders get double production speed of walls. In fact they get +100% on the base production, very different when the civ also has stone because it means it would only be one and a half times the speed increase (200% to 300%). I could complain about that in the documenation but I can see why they opted to just describe it as "double speed", even though I still personally disagree with that option. Another example is the way the March promotion is described in the documentation (can heal while moving). Many people mistakenly take that promotion thinking it means their medics will heal units as the medic moves!

I just think the issue with the promotion bonuses is minor or at least comparable to a huge number of other interface issues and I don't see a strong reason to pick this one out in particular as being ridiculously misleading beside all these other problems...

As I said I can understand how they got to where they are ... I could even see the spec sheet ... but I cannot condone them for leaving their users behind in the process. Now they could fix it in one of two ways. One is to actually label them for what they do and clearly explain the combat system ... but this will very likely confuse a lot of players (leading to an ever poorer interface)

Exactly, they simply should have given one or two worked examples in the manual and really there would be very little confusion for the players who wanted to think about the numbers when promoting.

or they can design the system to meet the user's needs while still retaining the level of playability they desire from the combat engine.

That may be easier said then done. I think the playability from the combat engine is a very high standard if you disregard any interface/documentation failings.

My instincts as a software engineer myself tell me that odds are they went the 1st path and then QA said "this is too confusing, simplify the labels" and never really considered the 2nd path because at that point the engine was written and would be too "hard" to refactor or more importantly rebalance and regression test. This is fine, concessions like this happen all the time in this industry, but that doesn't make it any less sub-optimal and as a consumer doesn't make me any more willing to accept it and not be critical of it.

Or they went along the first path and QA said it was too confusing so they decided to simplify the labels knowing that the number of people who would pick out the asymmetry would be few. Actually I think you are the first one I know who has actually criticised the labels as they are - I always thought they were a bit odd but I figured it was to help keep the information from being overwhelming.

I have no idea why you bring the phrase sub-optimal into this dicussion. Sub-optimal from the point of view of the most inspecting high-level players in the most significant minority of the player-base (software engineers who use civfanatics)? Oh please... I hardly think implementing the "optimal" path for these specific people is even worth a moment's thought for the developer, and I doubt they thought that was a sacrifice they needed to make.

Anyway, it's not my intention to start a quote war but I like the debate about the game mechanics. I guess I'm tired too which might be making my responses less coherent . Really I think we have a similar stance on the problem with the interface here - the difference is you think it is more misleading than I would think to call it, at least to the average user (is it 15 year old gamers now?).

 

[feralminded]

No need to apologize ... you're right I am an excessive nit-picker and whenever I find an inaccuracy I tend to go ballistic. Of course I am human and I spew glaring inaccuracies as well but whenever I discover this I go home, cut lines into my arms, and then cry myself to sleep with a bottle of Jack Daniels listening to Emo music.

That said one little nit-pick.

Or they went along the first path and QA said it was too confusing so they decided to simplify the labels knowing that the number of people who would pick out the asymmetry would be few.

You quoted what I said and then followed up with the above statement, which is exactly what I said ... meaning here we agreed even though you thought we disagreed. I do believe the engineers probably designed it, tested it, and then labeled it clearly before delivering it to QA ... who then said "No one understands this crap" ... which led to the current interface. I can't sharp shoot them too much for not looking at it and saying "well gee ... lets redesign from the ground up so we can have a system that is easier to clearly label and then rebalance and regression from there." Honestly "Just apply a label that while inaccurate still gets the idea across" was obviously easier and more sane. My only hope is that in a future version we get an improvement ... either algorithmic or documentation.

 

[elisk]

Getting back to the original discussion here... I like Charlemagne. I don't believe either Pro or Imp are the worst traits. Pro allows you to defend cities with fewer units, which can be good, especially if you are trying to defend captured cities without having to stop building offensive units for your war to build defenders. The Rathaus also cheapens the cost of maintaining those captured cities. Imp allows you to get twice the GGs of a none Imp leader (I think, I haven't done the math). Your stronger defenders from Pro allow you to accumulate XP more quickly, too. I once had a city that gave something like 15 xp to start out with to every unit, with about 10 of those from GG, as Charle. That means you can have landskneckts that can negate the Crossbows as well, and get defenders that start out with LOTS of defensive promotions. In my opinion Charle is an excellent warmongeror if you fight a war in the Medieval times.

 

[bestbrian]

Getting back to the original discussion here... I like Charlemagne. I don't believe either Pro or Imp are the worst traits. Pro allows you to defend cities with fewer units, which can be good, especially if you are trying to defend captured cities without having to stop building offensive units for your war to build defenders. The Rathaus also cheapens the cost of maintaining those captured cities. Imp allows you to get twice the GGs of a none Imp leader (I think, I haven't done the math). Your stronger defenders from Pro allow you to accumulate XP more quickly, too. I once had a city that gave something like 15 xp to start out with to every unit, with about 10 of those from GG, as Charle. That means you can have landskneckts that can negate the Crossbows as well, and get defenders that start out with LOTS of defensive promotions. In my opinion Charle is an excellent warmongeror if you fight a war in the Medieval times.

I like how you bring the conversation back on topic 6 months later.

 

[mechaerik]

I was stunned to see how many people knocked Cathy in the early pages of this thread. Sure, the UU isn't that great (in BtS), and the UB comes a little late, but her traits are awesome. Screw stonehenge and religion; build SETTLERS!!!!


CRE has always been one of my favorite traits, even though I'm more or less a warmonger. Helps a lot in the early game, and helps set up conquered cities easier.

IMP is a warmonger's trait. The fact that the bonus for settlers only applies to hammers is a little disappointing, but still, its a bonus to settlers. GGs are always useful, too.

The only real disadvantage to playing as Cathy is that you won't see her AI.....

 

[AutomatedTeller]

I think Protective gets a bad rap, to be honest - but it's better on games where you'll be doing a lot of fighting in with muskets on, or on games where you'll need the defensive uptick.

Ironically, I suppose on low level games, you can do a valid archer rush with protective earlier than you can an axe rush, so maybe it's good on low level games, too.

 

[elisk]

Oh! Ha ha! I didn't check the dates. Anyway, I think that any trait can be advantageous in some situations. There is a different strategy for each, and different situations where each can be valuable. It is true that some traits are more universally valuable than others. Financial for example.

If I had to pick a worst leader though, I might have to go with Shaka. His UU is faster, but it's not as good versus horses than Spears. It has even chances versus a Horse Archer. His UB is actually quite good, especially in combo with Agg, but that doesn't make up for his not very strong traits (Agg, Exp), especially with Exp not having much synergy with his other traits or his UB. Exp might be good on higher levels, but I play on the lower ones, and I don't find it very good.

 

[The Almighty dF]

Oh! Ha ha! I didn't check the dates. Anyway, I think that any trait can be advantageous in some situations. There is a different strategy for each, and different situations where each can be valuable. It is true that some traits are more universally valuable than others. Financial for example.

If I had to pick a worst leader though, I might have to go with Shaka. His UU is faster, but it's not as good versus horses than Spears. It has even chances versus a Horse Archer. His UB is actually quite good, especially in combo with Agg, but that doesn't make up for his not very strong traits (Agg, Exp), especially with Exp not having much synergy with his other traits or his UB. Exp might be good on higher levels, but I play on the lower ones, and I don't find it very good.

Shaka's an AI leader.
He's hard for a human to play as, because he doesn't fit human goals usually. He does fit -AI- goals. His job is to irritate and destroy. To hold back anyone developing anything.
He's more of a random event than a true opposing leader.

 

[TheMeInTeam]

I think Protective gets a bad rap, to be honest - but it's better on games where you'll be doing a lot of fighting in with muskets on, or on games where you'll need the defensive uptick.

Ironically, I suppose on low level games, you can do a valid archer rush with protective earlier than you can an axe rush, so maybe it's good on low level games, too.

In low level games the warrior rush blows out the archer rush, and is much more likely to hit before the target gets archers (which screw both archers and warriors).

The added defense can be nice, however, if you screw up or draw some bad neighbors.

especially with Exp not having much synergy with his other traits or his UB. Exp might be good on higher levels, but I play on the lower ones, and I don't find it very good.

EXP keeps up with IMP and CRE as an expansion trait (hence its name). It has plenty of synergy with the ikhanda ----> you can afford more cities.

People underestimate this trait because its effects are less obvious than many others. The cheap granary and workers both come into play early, and the granary is the #1 economic building in the game, now available for barely more than a chop. New cities can come online through growth QUICKLY, which means that they can produce additional units sooner than a typical new city. The cheap workers are a boon to get the necessary improvements going in such cities.

The health is icing but it does boil down to +2 food/city that grows. Combining these bonuses is not immaterial, even if they seem weak independently.

The UU is admittedly weak for anything but speed chokes vs opposition w/o copper or playing below one's level, but the top flight (seriously, Ikhanda is top 5) UB compensates it enough to keep shaka well away from the bottom. Would you seriously imply that AGG/PRO with no early UB and a UU in the weakest era for offensive waring is better?

He's hard for a human to play as, because he doesn't fit human goals usually. He does fit -AI- goals. His job is to irritate and destroy. To hold back anyone developing anything.

He's one of the strongest AIs in the game. He's one of the top AIs to cause a human to quit early, and if his war spree gets on a slight role, he can wind up capping enough people to win UN or with enough land to brute force his way to space through city spam. Probably the only way he'd lose in standard high level games is another AI, possibly his own vassal, going culture.