Stalker0's State of the Mod - 1/11/2022

[Stalker0]

Another year, and another chance to look at the wonderful mod that is Vox Populi. As I do periodically, I will look at where I think the mod lies, and where I think its good to move forward on.

My notes are going to be brief on this one, as I'm focusing on more high level topics than nitty gritties.

The Roll of Polls
This last year saw the importance of polls as a factor in mod decision making. Polls were used to both get takes on certain ideas but also to drive forward changes in the system.

I would say the results were mixed. We had some real great polls that helped clear through the cobwebs of things like religious changes, but we also saw people frustrated by the frequency of polls as well as the statistical weaknesses of using polling with such a small user base. Polls also have a tendency to slow down change, and for those in community that prefer the mod remain "experimental and high change", this can be annoying.

Ultimately, I think there is still a divide on the effectiveness of polls and how they should be used in the future.

I feel partially responsible for the rise of polling this year, and have remained an advocate of polling. I even promised G a document several months ago to give my thoughts on a more crystalized version of polling and change management. The reason I haven't is simple....I too have been conflicted as to how that process should move forward and how fast it should operate. Ultimately are polls a force for good or ill?

To that end I have another thread I will be posting shortly, to give my latest idea a voice, and to see if there is some consensus to move forward with it. That will be coming soon.

Culture Victory
I do think there is some consensus that CV remains too easy a victory. The concern is an inheritance from the last major changes to tourism. I think the changes remain excellent ones, so this is not a change to the model itself, just some further tweaking of tourism numbers.

Late Game - Should it be changed?
A lot of discussion in the last couple of months is around the late game. I think there is a general consensus that the late game is a bit lackluster. You get to a point where for the most part your going for your VC, you press end turn a lot, and you either win or you lose. There aren't a lot of buildings to build, and in fact its usually best to just shift into processes.

That said....it can be well argued this is an inherent part of the 4x model, which is common in many many games beyond Civ 5. To "fix" it, would likely require something much more fundamental than a new building, or a tweak to a VC. It would require rethinking things such as:

• Should weak civs eventually "go away" in the game, leaving the Top X civs? This could be complete removal, vassal to a higher power, etc.
• Should VCs be removed, added, or reimagined?
• Should the function of late game buildings be reimagined? Instead of providing yields, should they alter fundamental rules of the game, or provide big penalties to rival nations?

In other words, do we want to expand the project into a real "solving" of the 4x late game issue, by introducing some fairly radical changes?

My view: This is well out of scope of the current mod....but this mod has long gone past its original scope years ago when it was mainly a "balance mod". VP has added all sorts of new gameplay and content over the original game. I think it comes down to the modders, is it time to just polish up the gem and call it done....or are you eager for a new frontier to conquer? The late game is probably one of the last places we could consider some really experimental adjustments to gameplay.

Spies
We haven't talked about spies in a long while. The entire spy system was changed radically many months ago, we had several rounds of updates and bug fixes, and then we stopped talking about it.

My honest assessment at this point.... I don't think the new system is superior to the old one, and in fact still has several weaknesses. There remains a lot of fiddliness with the system, and further I'm noticing a lot more King/Emperor players complaining of large AI tech leads. I think part of the reason is....the new spies are just plain weaker than the old ones. Spies were one of the key rubberbands in the game, allowing a human player to catch up to a strong AI science economy. Without that strong rubberband, there aren't that many ways to gain tech parity (tech trading being the other big one).

So I feel its been 1 step forward, 2 step backward here.

 

[chicorbeef]

Spies could use some adjustment. I feel like there should be more the spies can do even at level 1, and that spy events could trigger more often, but be less powerful. Stealing giant sums of gold every 20-30 turns just to be able to eventually be able to steal giant sums of science every 20-30 turns makes spies feel like a yield dump more than a dynamic tool to actively scout/sabotage opponents.

I personally still don't think CV needs to be as obscure as it currently is and I think the tenet requirement should still be changed to the equivalent policy requirement (beating a dead horse at this point but I'll keep beating it until it's in the grave).

 

[Rekk]

How frequently are the AI trying to stop other players from reaching their victory conditions? Generally this means war ... and military units are the majority of what information era entails anyway.

Otherwise we need brainstorming other ways to attack players without war.

 

[Rekk]

King/Emperor players complaining of large AI tech leads. I think part of the reason is....the new spies are just plain weaker than the old ones. Spies were one of the key rubberbands in the game, allowing a human player to catch up to a strong AI science economy

There are a bunch of things in this paragraph that conflict with the goal of the community patch:
A) the human player shouldn't be catching up to the AI player, they should be competing with each other at all stages
B) there shouldn't really be a mechanic that, the results of which, the player gets better use out of than the AI

AI and humans used to be pretty close to parity in techs a couple of years ago (humans could pull ahead in the middle eras but AI would claw its way back up later on). Maybe ABC bonuses aren't working as well as we hoped when they were first introduced, with all of the AI improvements/changed since then.

I think effort spent on ironing out these issues would be well worth it.

 

[balparmak]

I think this thread was needed, thanks for putting in the effort. I agree with chicorbeef on the CV tenet requirement, IIRC the majority of the posters also agreed on its thread, with the only downsides being a somewhat minor culture cost difference between tenets and policies, and the necessity to tweak the AI. I think that change would open up the CV and simply make it more fun, which sometimes gets overlooked as people tend to focus on balance and numbers. I also agree that the spy system is wonky, but it isn't beyond saving.

Making weaker civs with no possibility of winning subservient to majors at the late game is a very interesting idea, ideology to some extent does this but I wouldn't mind if we have power blocs/poles in the game with the leading civs practically making all the foreign policy&security decisions for others. That said, it would be better to have this as a gameoption or modmod until it's thoroughly tested, both for function and popularity. I think a nuke-rework is also one of the ways to improve the late game, and was recently discussed on the discord.

I wouldn't mind integrating more stuff from modmods, I know the whole point of them is to provide alternatives but there are cases when the modmod is too good to left out. ENW comes to mind, it's very popular already, and at least some basic features (two-tiered rockets & supercarriers) from that should've been integrated to VP by now in my opinion. There are also very popular UI modmods, and it would save some clicks if VP had them integrated. I think at this point VP could use some more flavor and polish, and borrowing/repurposing stuff from established mods is the logical way to go. Maybe going forward this could be something to ask in the polls.

I think unhappiness needs a thorough tweaking, and that's probably the most upstanding issue for me now. The system is fine, the end-result is also fine (as in how manageable&influential unhappiness as a whole), but for a lot of players Poverty unhappiness is always a few times more than the others, I frankly cannot remember when was the last time I cared for other sources of unhappiness.

 

[usadefcon1]

Spying has to be the #1 issue right now, it's glaring how unpolished it is compared to the rest of the game.

Personally I think spy yields need to be based on the target city's yields, minus security level.

Spy actions should be percentage based on what the spy finds when they get there (if the city has scientist-citizens then tech stealing should have a higher chance of being successful, etc...)

Show the player what the odds are and then have the player make the choice of what action to persue.

 

[usadefcon1]

An example on the defensive side would be

I have found an enemy spy in our city, what should I do?

A) Kill the enemy spy immediately (70% of eliminating the spy, 10% chance of finding their origin, 20% they escape)

B) Track the enemy spy (50% of finding their origin, 50% they escape)

C) Bribe them, costs gold (80% of finding their origin, 20% they escape)

 

[HeroOfTheSword]

I agree with many others here about the spy system. When I first played it I thought it was fun but now I feel it's a giant pain in the behind. It's just too random. Plus the Steal Gold that doesn't actually steal any gold is ludicrous especially when you compare it to the Disrupt Productivity which does work but is way too op in my opinion.

I'm playing a game as England right now and pretty much every 4 turns my capital is getting whacked with Productivity Sabotage which is practically making the game unplayable at this point. I've had to use IGE a number of times now to counteract it, which I hate using because I end up overusing it. Like I was building a wonder, 4 turns to completion, got sabotaged and that jumped to 12! I think that's ridiculous altogether. One or two for a large city with a constabulary and a spy thwarting 50% makes a lot more sense than the crazy numbers this system is coming up with! To be honest I'd nearly turn the whole system off in the setup except I'm playing England and that's one their supposed special abilities, which begs another question; if the system isn't working properly, is it really fair to have part of a Civ's unique ability centered around it to begin with? Like I think it's a pretty weak UA particularly when England's UA should be a lot better considering the historical background of the nation.

In saying all that, I do think the spy system has great potential and I can see why it was introduced. It just needs more tweaking probably.

As for the late game, I never get that far but I think your ideas are fantastic and maybe I'd be more inclined to play those eras if the game play was switched away from yields to some other system.

 

[Kim Dong Un]

I won't get too detailed here, but a few things that come to mind:

• Two other VC - I know there was a recent attempt at this that fell through, which is unfortunate because I really feel that incorporating an economic / religious VC would go a long way in spicing up the game in general. This would obviously be a massive task with a myriad of moving parts requiring many balances and tweaks across the board; probably why it was abandoned.

• Concerning the early game, I only play a modified King difficulty, but it seems the AI are consistently ahead in tech regardless of me focusing science with civs like Babylon, Korea, etc.; when I hit around 10 techs, I won't be last, but there's always a few civs several techs ahead in the early/mid teens. Milae's difficulty mod makes this better to an extent, but it's still noticeable.

• Trade screen is much improved, but strategics are still being undervalued most times. It gets more apparent as the game progresses and the more valuable ones (aluminum, uranium, etc.) come online.

• Spies are better, but still need refinement. We want minimal micro, but I think most mission duration is still currently too long. I like that tech stealing is gone outright, but the fun is diminished when my lvl3 thief still takes 30+ turns to nab those science yields. I've utilized a few different options in a couple recent games, but overall I still mostly use spies the same as always, by sticking them in strategic CS and leaving them for most of the game.

 

[phantomaxl1207]

I think we need more Spies.

The biggest problem with Science victory to me, is that you need every single technology in the game. There should be a more defined direction in Info Era on what VC you are going for. Domination has to decide on whether it wants Stealth Bombers or GDR's. Diplomatic and Cultural go the top route as Globalization and Internet are close to each other.

 

[Solic]

I think we need more Spies.

The biggest problem with Science victory to me, is that you need every single technology in the game. There should be a more defined direction in Info Era on what VC you are going for. Domination has to decide on whether it wants Stealth Bombers or GDR's. Diplomatic and Cultural go the top route as Globalization and Internet are close to each other.

Funny, I think we would actually need less spies. Reducing the number of spies could make their actions and consequences more impactful and make it less micro-eye. It'd also ensure, you really need to choose who you are going after. I think spies should actually be restored to really hurting your opponent as well instead of the duplication of gold/science we have going on now. We have other hurtful actions like pillaging/production loss, so there wasn't even made a great split. I think the spy city state quest could also use a change in maybe having a spy in a certain civ for a number of turns. Now I typically take a spy and make it do the shortest event possible (hurt production usually) having no real thought into it. Counterspies could also have more meaningful decisions in my opinion. I'd like to see either a high chance of killing enemy spies/an even higher chance of identifying enemy spies/increased CS for the city/tourism increase/general yield increase. Some other new actions could also spruce them up. Maybe policy finisher locked ones like:

Statecraft: When stationed in a City state, gain the action to remove an embassy of a player (allows for a peaceful diplomatic offensive weapon and balances the choices for great diplomats a bit. Keeps the choice space for great diplomats into the lategame and nerfs embassies more in comparison too)
Fealty: Build a religious building that gives out holy city level pressure of your religion (only useable when you have a holy city yourself (ensuring you have a religion)) and/or Inquisitors can no longer be used on this city and/or convert this city to your religion.
Artistry: Steal the highest great people points of this city and add them to your capital.

On the ideology spy tenets:
Autocracy: Same as Statecraft, except if you already have Statecraft's spy action, then you do not destroy the embassy, but steal it instead and gain ownership of it. (Fits with diplo victory goal of Autocracy)
Freedom: When stationed in a City state, it places a great diplomat embassy even if there is already a different one. (Fits with diplo victory goal and placing spies in city states)
Order: Actually fully block actions if you kill spies using counterspies. (Fits with placing counterspies) and/or give them a tourism counterspy option to make Order more culture oriented than it currently is.

---
I agree with other points made in OP. I think Culture victory is the easiest victory now and I would love if the economic/religious victory conditions concept became a reality. It would automatically fill up late game a bit more with at least more religious buildings and maybe some more gold oriented ones. Religious victory (how it was pitched) also forces you to keep tabs on other civs more in the lategame to try and thwart them from winning before you.

 

[Closed Sky]

Funny, apart from some minor things, I consider the mod fairly complete and balanced. I wouldn't say we need any overhaul of late game. I would focus on reworking poverty, or those small things like making creativity +1 culture and faith for every three followers, so it's usable again, or nerfing pagodas and orders so they no longer do far more than their alternatives, just to allow more choice, as was done before with founders.

However, I view spies rework as a failure. Gazebo had good initial idea with many dynamic actions every few turns. But as of now, they are messy and the same as before rework, with the difference being you get science instead a tech. You need to put a spy, get him to do production sabotage in few turns, which does nothing important, and reassing to get science. It would be much better and faster if you could just select science from the beggining with fresh spy with just doubled duration. Same thing, much clearer. I'm againt any more spy management and nuance as they basically are primitive instant science event.

I would suggest either reverting or symplifying the system or, if it's simple enough, scrapping spies altogether and making them a GP gained from civil servants along diplomats. That will served a niche for civil servants and diplomatic buildings for non statecraft plays. Great spy would travel to enemy city with 5 moves, no terrain penalty, and could be used to incite a revolt and lower defense for some turns, kill some citizens, steal a great work, or sabotage production of world wonder. Statecraft would gain +50% generation and maybe 25% more effects. Science and gold would disappear from spy play, that would be good, in favor of more dynamic stuff. Maybe some small +science and gold added in few places, for example +2 science on caravansary which is very niche and +2 gold to every chancellary or university added to printing press national wonder.

 

[youngsteve]

I certainly don't agree about having more spies, as they are a real nusinance with the few you have.

Same thing happened to me like Herofthe word playing Arabia. I was absolutely crippled with production disruption dispite having a spy in city & other protection, never able to build this wonder. I was going to turn espionage off after this, but still use it.

I also think spies should be more useful with CS. I know they disrupt elections, but the results seem paltry in comparison, & some civ soon sends a couple of diplomats of the converbuilt to counter this. I would much rather you could have spies/diplomats more powerful in CS & remove all the tedium of building diplomats. This is going to be contraversial but I would like the option to have more than one embassy in a CS, which would make GD more relevant in later game, instead of just a massive boost in relations. Give more votes in Congress, & be more like cold war with opposing philosophies.

I also agree with Kim Dong Un, in how far the other civs get in tech in the early game. I only play on Prince level & notice this. Understood if playing a high level.

I also find the late game such a bore, lose interest once in Modern era. Gaining science victory is such a drag, having to complete the tech map.

 

[Closed Sky]

Two other VC - I know there was a recent attempt at this that fell through, which is unfortunate because I really feel that incorporating an economic / religious VC would go a long way in spicing up the game in general. This would obviously be a massive task with a myriad of moving parts requiring many balances and tweaks across the board; probably why it was abandoned.

As much as it is unexciting, you can play like that with just house rules, assuming you win by being first in some of demographics for an entire era or having all the cities/players converted to your religion. Victory screen would be nice but interface is a house of cards to mod. I don't see why it would require any tweaks though. I agree it would be great but prefer keep it cosmetic addition.

 

[looorg]

More stuff to do in the late game, at least more buildings. More to do. Even if you are not culture victory bound there is not that much to do in the very late game (last and second to last era). It's mostly building a few buildings and then many units and wars. Votes. If not at war it's mostly that all the cities will go into process mode and then it's just hitting next turn until the next vote or some spy action need adjustment. It's hard to get past that all the meaningful and interesting aspects of the game happen in the first two eras or so. The rest is the slog towards the end somehow.

In that regard perhaps the spies should be more or actions should be faster or something. As it is now I find myself back in the old 'set and forget' mode as far as spies are basically concerned. I run basically three actions -- steal gold, steal science, make forgeries. Every once in a while I set a spy for something quick to gain vision or do a city-state quest, if at war and the city can't be bombarded to submission the aggressive actions can help. But it's somewhat hard to time them as in they take many turns to build up and then you should try and time that with the war effort etc. In some regard I think the sabotage actions should be like coup -- you get there, give the spy some C4 or whatever (depending on age) and then let them go. Mortality rate high and if is instant back to base no matter what. I'm not sure if the game speed makes this worse or not but a standard interesting actions is usually about 40 turns or so to complete on marathon, hence the set and forget mentality I guess.

That said I don't want more spies. It's very annoying in the sort of mid-late game when the AI spies on you and it becomes very hard to get something built in the capital and some other cities due to the near constant production sabotages. The gold stealing and such I don't really care about anymore since it doesn't do or change anything for me -- besides the AI having more money.

Also if you want more spies just go statecraft and you'll have more then you need really. That said I think I do actually prefer this new spy system to the old system. It does feel a bit more meaningful somehow, or at least a bit more options. If anything I would probably like the actions to be faster even at slower speeds thereby creating more interaction with them. That is I guess overall the thing I would like more of, on slower game speed I want the votes in congress and the spy actions to be faster and more frequent. Perhaps as much as having the times cut in half or so. More interaction; 40-50 turns between things is just to long for interaction to be meaningful in that regard.

There are a few of the actions tho which I sort of wish had more leave-town once completed. Such as the kidnap specialists. Since it then tries to do the same action again afterwards only for it to fail since you can't do it while you still have the event going. So that action should kickback to base once completed. It can just sit there and do nothing until the event is over and then do it again but it's somewhat silly.

 

[Hinin]

I would suggest either reverting or symplifying the system or, if it's simple enough, scrapping spies altogether and making them a GP gained from civil servants along diplomats. That will served a niche for civil servants and diplomatic buildings for non statecraft plays. Great spy would travel to enemy city with 5 moves, no terrain penalty, and could be used to incite a revolt and lower defense for some turns, kill some citizens, steal a great work, or sabotage production of world wonder. Statecraft would gain +50% generation and maybe 25% more effects. Science and gold would disappear from spy play, that would be good, in favor of more dynamic stuff. Maybe some small +science and gold added in few places, for example +2 science on caravansary which is very niche and +2 gold to every chancellary or university added to printing press national wonder.

I would actually be for an hybrid system : keep the few spies (not too many) in the spy window to establish "information networks", "support movement", "counter-spying" etc (some passive effects more similar to the current diplomats) while giving you points towards "Great Spies" that would be used for major effects (steal technologies, destroy the defensive buildings of a city, give vision of the entire enemy empire for a few turns, provoke a coup etc).

This would allow us to keep the spy system, while giving it more specific tiers and power spikes : less micromanagement most of the time, but precise moments when you can make the system have drastic effects.

With a system like this, we could even hope to make spying relevant before the renaissance era : humans didn't wait for gunpowder to spy on each others.

 

[randomnub]

I wouldn't mind going back to the original spy system either (or have an option to), it felt dynamic enough without acting as a minigame on its own; late game spies cost way too much micro to be an enjoyable mechanic if you ask me.

 

[Hinin]

I wouldn't mind going back to the original spy system either (or have an option to), it felt dynamic enough without acting as a minigame on its own; late game spies cost way too much micro to be an enjoyable mechanic if you ask me.

The previous spy system is dead : the code was a mess, and was removed. The new code is more optimized, and offers more options in the long term, but we have to decide what to do with it.

As I said, I'm in favor of lowering the general micromanagement of the spy system by making minor spies more about passive effects that your choose whenever you want, instead of once every few turns, but with the addition of ponctual great spies able to make important plays.

 

[Milae]

Thanks for trying to push things forward with polls and discussions and stuff man. My response to your points here:

Culture Victory: I agree it's too easy. I suggest removing the instant tourism yields off buildings and giving them tourism per turn instead. That would mean that you actually need to rely on historic events and gradual tourism increase rather than the strategy right now which is to stack tourism to other player modifiers (using musicians etc) and then build zoos/stadiums in every city for like 5k+ tourism each.


Lategame: Main thing people usually mention about lategame is the turn times getting too long because of too many units. Unfortunately to solve this we'd have to try making armies or something which is a pretty massive change. Also to steal another thing from Humankind I think their War Support mechanic has potential. I actually have an idea to try making a mode where humans can get the AI to do their troop movements for them so that could help but would also need a lot more work on the AI and interface for that.

Other things we can improve for lategame though:

- We have done some good work on improving the World Congress AI and Diplomacy so modifying/adding proposals can help to spice up the endgame with more options and ways to bring people down diplomatically.

- I've said this before but I think tourism should have more roles outside of the victory condition. All the other Victory conditions help you as you work towards them and IRL if the people of another country love your culture you do get a lot of benefits from that.

- Related to tourism is Revolutions. There is that screen in the culture menu which shows this but having cultural influence doesn't ever seem to do more than just add a couple of unhappiness. I honestly have no clue how this system even works and also I can't remember the last time I saw a City-State break off from another Civ like they used to do sometimes. Giving these mechanics some love could add some extra dynamism to lategame.

- Tech costs are a bit weird as well. Costs increase massively through Industrial and Modern but then barely at all from Atomic onwards, even though science increase quite a lot here. I think this contributes to some of the massive tech disparities you get lategame since the first civ to get to Atomic can power through Atomic and Info so quickly while the others are still struggling through Modern. Also the lategame does feel like quite a blur as you get so many techs and there are the bonus techs/Scientists you get from some of the Wonders as well that just push you further ahead. Atomic Era could probably do with a bit of love as well since there are very few wonders/buildings in there, mostly just tanks and Nukes, while there is far more useful stuff in info era.

- Military AI seems to fall off a bit too. They rarely build many tanks and seem to either build 0 anti air or like 90% of their units are AA. They also can be quite timid when they are actually in a war even though they have a lot of units and use nukes very sporadically. I would say this about the AI in general though that particularly on high difficulty they should just throw more units at cities since they may as well use the numbers advantage that they have for something.


Spies: I agree that the new system doesn't feel that great atm. But I disagree about the previous being better. Perhaps it was simpler and more powerful before but I think free techs from spies with a fairly random chance was not a good way of doing tech catch up. There is the KNOWN_TECH_MODIFIER which currently is 10% if every other civ knows a tech so that could be increased or other mechanics could be expanded on such as tourism/trade routes for this.
My only suggestion would be to have an xp bar for the spies which fills up and then each mission gives a certain shown amount of xp for success. This would make levelling spies more transparent and then we can decide what should be buffed/nerfed about spies with full knowledge of how they actually work.

 

[Fluffball]

For the end game i feel that the AI simply needs more consitent focus in the latter half of the game or to be stronger so the end is uncertain and the players continue to be challenged during the victory run.

Memorable end games for me where i played to victory or loss include.
Science victory with India as main competitor. India was on another continent so i could not do much but try to out pace them. It was really down to the wire who would finish first and i just beat India with me having to employ every tactic i could to slow them down while maximising my own progress.

Culture loss against Brazil. Brazil was relatively isolated and also well advanced in science so again i had limited options to counter them apart from trying to limit their cultural inflence and maximising mine. I was a handful of turns away from victory when they won.

Diplomatic victory against Askia. There were two main continents and some large islands big enough for a couple of cities each as well as a smaller islands bridging the two oceans between the main continents. I had a relatively quiet diplomatic early game until we gained the tech to cross oceans and then the real fun began.
By the time i explored the other continent Askia had already conquered most of it and there was a race between various civs to colonise the bridging islands while Askia finished off his continent.
Askia then began to look towards us with some heavy fighting over the colonies in particular. With me and Askia eventually dividing up the colonies between us and us both having a sphere of colonies. It then became an arms race with wars breaking out as one side got a tech advantage which often involved huge naval battles around the colonies and in the late game us trading nuclear strikes against each others navies to defend those colonies.
Askia was tearing away for a culture victory while i decided to focus on a diplomatic victory as i had a lot of votes already and Askia was also at least equal in science.
Askia seemed very adept at stealing city states from me and got me decolonised a couple of times just before the next DoW and would often conquer or try to conquer my friendly city states so it was very dificult for me to build enough votes while i tried to build up an invasion force as a back up for a potential invasion to take his capital as we were the last two owning our own capitals.
It came down to around 10 turns before Askia was due to win a cultural victory i got my one chance to get the diplomatic victory and as a back up i had the bulk of my navy, masses of paratroops, numerous aircraft carriers, nuclear subs armed with nukes and a large invasion force sitting off his coast for a last gasp, do or die suicide run to his capital if i didn't get it. Askia stole a couple of my city states on his turn and proposed decolonisation again but i managed to get just enough back block decolonisation and then to scrape the victory vote.
As a test i decided to play out the invasion afterwards as it seemd like fun to do and also to see if i would have won and i wouldn't have, it just turned into a huge nuclear apocalypse.


Most games end up obviously won long before victory is possible either because the you have simply overtaken the AI or it can seem like it is going to challenge but then never commits to finishing, science victory being the obvious case where AI can be first to start it yet when i have won it they have barely made any further progress.
It rarely seems to push it's advantage militarily either. One of my memorable AI fail games was a game which seemed like it might pan out similar to the Askia example above but with the mongols instead. This time i was playing warmonger though and all the way up to indistrial era me and the mongols were slowly conquering our way toward each other with me expecting it to end up with just the two of left in a final epic war but instead they just seemed to stall at the industrial era and when the war came i had destroyers, battleships, fighters and heavy bombers while they were still using ironclads and had a couple of tri-planes.