Poll 2: which of these changes would you like to see for the Comanche Rider?

[JamesNinelives]

Please refer to this poll over the previous one! Sorry for the confusion.

I couldn't edit the poll options after making the poll so I'm creating a new one taking into account the feedback recieved. The 'steal strategic' option now clarifies how that mechanic works, and I've limited the number of choices that one can pick to two. I also removed the options that were not discussed in the comments (CS/RCS change) to focus on the leading suggestions. The poll is otherwise identical (if I've done it right). @Stalker0 @pineappledan @looorg hopefully this useful.

I also talked to my Mum (statistician) about the arguments for and against forced choice vs. multiple responses and she more or less said it doesn't matter that much in this context. She emphasised having clear options and being careful in how you analyse the data.

Original poll description: "Based on discussion in the Shoshone thread, most people seem to agree that the Shoshone UU - the Comance Rider - could use improvement. There are several other ideas proposed but I thought it would be good to focus on one things that we can agree on to begin with! Multiple options can be chosen."

 

[Gidoza]

I voted for other - pillaging increases growth of your cities for a time (Stealing women!).

 

[looorg]

I voted for other - pillaging increases growth of your cities for a time (Stealing women!).

In general it's almost interesting to note that this concept isn't more viable or common among unique units. But I guess it's people don't want to touch on those aspects in the game anymore with slaves and such. I guess the units that get "food" could or should also be interpreted perhaps as taking some people. Who knows. It's a valid option for a promotion or ability if it fits in with some civilization or unit.

From a statistics or analytic perspective the options are fine for the most part if the options are just clear. It's options that give room to interpretation or are vague that are problematic. The other problem as noted is that we usually don't have that many people that vote in them. I don't even bother to vote in all of them, mostly cause at some point it seems like everything is a poll and it's polling for the sake of polling and one just wants to see what sticks or something. Polling-fatigue in that regard is not uncommon.

After the discussion with Pineapple I'm willing to try that idea. I think as noted it will be a source of abuse from the player side and I'm not sure the AI will be able to take advantage over it in the same regard. But as I player I will most definitely do, no war will be completed without pillaging all strat resources in the enemy land.

(1) No horse requirement. It's a late unit so I guess horses shouldn't be an issue at the time. So from that point it might make sense. There should be a natural ample supply of horses in that regard.
The other aspect is if you have multiples of them this would be odd imo if you also include (3) as then you should never have a horse shortage problem either. So having a horse cost shouldn't be an issue.

(2) or (3). I don't think I would want to see both of them or them stacking or anything. I refer to the previous thread here. I think it will be bonkers and overpowered but I'm willing to try it, certainly so when it's no longer a matter of removing resources from the victim but just gaining them yourself. (2) has in that regard been done but lots of other civs, also lots of policies etc offer something similar in that regard so perhaps it's not that unique or interesting to have yet another pillage ability.
(3) will just lead to so many more options and buildings in that regard, you can ensure all the factories, trainstations/seaports, power buildings, agribusiness (which is normally the dumping ground for all your excess horses that you don't use anymore) ... you would ever want or need.

(4) is another powerspike option. Everything you get something early as a spike it's an invitation to this is your room for war and aggression.

As I see it's the questions is really yes or no towards (1) and/or (2) or (3) and/or (4).

I'd be down with trying (3) in that regard. The others are not really all that interesting I think as options. If you want a somewhat unique and powerful ability that is.

 

[Stalker0]

I also talked to my Mum (statistician) about the arguments for and against forced choice vs. multiple responses and she more or less said it doesn't matter that much in this context. She emphasised having clear options and being careful in how you analyse the data.

Can she provide how to calculate margin of error rates for such a poll, as I truly do not know how to do it.

 

[Rekk]

(3) will just lead to so many more options and buildings in that regard, you can ensure all the factories, trainstations/seaports, power buildings, agribusiness (which is normally the dumping ground for all your excess horses that you don't use anymore) ... you would ever want or need

I don't think you'll be able to come close to Russia's resources with this UU even if you destroy your diplomatic standing.

 

[JamesNinelives]

I'm honestly fine with any of the options. It looks like moving the unit earlier or letting them steal strategics is the most popular (the same trend was visible in the previous version of this poll) so IMO it's between them. Mechanically stealing strategics is interesting and not something that other civs can do so I'm leaning towards that.

I don't think it's particularly abusable. War may be profitable but so is trade in peacetime. And the Shoshone aren't particularly inclined to be at war all the time, their kit leans towards early expansion and then mainly defensive play IMO.

Can she provide how to calculate margin of error rates for such a poll, as I truly do not know how to do it.

I talked to her about statistical error and she said is that there are different kinds. So I think it depends on what you want to know. To be honest, I don't think there really is any one method that will tell you the whole picture. Sorry :/.

In general it's almost interesting to note that this concept isn't more viable or common among unique units.

The Iroqouis UU has an ability called 'Mourning War' in which they would sometimes capture people from other nations to replace men lost in battle. The Ottoman Janissary were originally children taken from the families of conquered peoples and then trained to be soldiers. The Bandienteres, while best known as heroes who brough back riches and tales of adventure, were originally sent out to capture native people and bring them back as slaves. And the Vikings very much have a pillage-based mechanic which represents stealing among other things. So it's not like we're shying away from this - it's in the game already.

pillaging increases growth of your cities for a time (Stealing women!)

I think giving the Shoshone specifically an ability about 'stealing women' is both ahistorical and also kind of insulting in the context of what real-world relations between the colonisers and the indigenous peoples looked like. Maybe give it to the English lol.

I don't even bother to vote in all of them, mostly cause at some point it seems like everything is a poll and it's polling for the sake of polling and one just wants to see what sticks or something. Polling-fatigue in that regard is not uncommon.

That is why I made sure we had discussed the issue in the relevant thread before creating a poll. If you check the Shoshone discussion I think we got as far as we were going to get without something a bit more actionable.

 

[chicorbeef]

About the poll itself, I think there is an issue in that it doesn't clarify whether people are voting for either moving to rifling/strategic pillaging or for both of them. But it's tricky, because they're not necessarily mutually exclusive and people may be forced to choose if only one option were available in a situation where they wouldn't want to choose.

Personally, I would be fine with either change, moving to Rifling or pillaging strategics.

 

[Stalker0]

I talked to her about statistical error and she said is that there are different kinds. So I think it depends on what you want to know. To be honest, I don't think there really is any one method that will tell you the whole picture. Sorry :/.

I mean fundamentally I need to know that if Option A gets 60%, and option B gets 40%.... does that mean that A is actually "chosen" or is the error rate high enough that you cannot really say A or B is a preferred option?

If we cannot determine that, then this poll has no meaning.

 

[azum4roll]

With 2 options standing out here, maybe we should start a new poll with just "Both", "Option A", "Option B" and "Neither".

 

[JamesNinelives]

I mean fundamentally I need to know that if Option A gets 60%, and option B gets 40%.... does that mean that A is actually "chosen" or is the error rate high enough that you cannot really say A or B is a preferred option?

If we cannot determine that, then this poll has no meaning.

The poll isn't intended to determine the outcome on it's own, it's just a resource to help us decide as a group which direction to head in. Let's look at what we can (and can't) interpret from what we're seeing.

For example, from the conversation I've read here and in the Shoshone forum the majority of people already agree that some change is needed. There isn't really much contention there, but it wasn't clear what change to make. That's why I made a poll. And I think it's reasonable to say that the results so far suggest two of the options are more popular than the others.

The next question becomes: do we want one of these two options, or do we want both? What I've read suggests to me that more people are in favour of one rather than both. I encourage people to make their case if they disagree though!

So far Pdan made a strong case for resource pillaging without moving the unit earlier, and the main objection against it was related to a miscommunication on my part. There is still some concern that it might be exploitable by the player. So I would say it's not clear which of the leading options the community prefers and that bears further discussion .

There is also some feeback relating to poll fatigue, so I'm inclined not to make another poll on the same subject if we might reasonably be able to work things out just by talking about it. In my view the topics from here are: A) one or both and B) if one, which one? That's my read on things, at least.

 

[psparky]

The pillage resources idea doesn't feel right to me. I think the notion of units/buildings needing resources is an issue of supply. Horse units can't heal if there isn't a supply of horses. You need to maintain the supply by keeping control of the resource tiles, by trade deals or keeping city-states allies. This permanent acquisition of a resource doesn't fit the model.

Also, I don't like the possibility of clinching a monopoly through this mechanism.

 

[Delvemor]

I voted for yields on pillage and gaining a copy of a strategic resource on pillage. Unlocking them earlier also makes sense. To be honest I don't think it would be OP to add all 3 options to the Comanche Riders. The Shoshone are very fun to play in the early game and are tougher to conquer than most AI but are otherwise on the underpowered side. If you look at Russia and Morocco, they both have a better unique Cavalry and their kit scale much better overall. Even with the 3 buffs applied to Comanche Riders, I would still consider Russia and Morocco to be stronger civs. On top of that, many people play with ancient ruins disabled.

 

[LarryAR]

“No strategics (horses) required to build” makes sense. My reasoning is that a civilization could not develop a unique unit if it does not have those units to develop. That is, if horses weren't on their property they would find a way to get them.

“Can steal strategic resources…” Sure, the Comanche nation was a tough warrior society skilled and known for raiding using their unique units.
“Unlocks earlier (e.g. Rifling)” If given a third choice this would be it because Comanche Riders were lighter equipped than the other horse units except ancient horsemen and some lancers.

Note that my choices are based on a historical view. Playability becomes a crapshoot with the more tweaks you make.

 

[JamesNinelives]

I think everyone has had a chance to participate. My instinct here is to trial either the top result (unlock earlier) or the two top results given they are close (unlock earlier and gain strategics on pillage) to see how it plays.

 

[Stalker0]

I think everyone has had a chance to participate. My instinct here is to trial either the top result (unlock earlier) or the two top results given they are close (unlock earlier and gain strategics on pillage) to see how it plays.

So lets go back to the question of "how do we interpret the poll"? Because I can't tell how many people actually voted, we have to make some assumptions. Lets look at the two extremes:

1) Everyone voted for one choice: In this model, we assume each vote was from a unique person (aka almost no one voted for multiple options). In this model, we get these results:

Total Voters: 55
Margin of Error: ~+- 8%

• No Strategic: 7.3%
• Yield on Pillage: 18.2%
• Steal Resource: 32.7%
• Unlock Earlier: 36.4%
• Other: 2%
• None: 4%

With this margin of error we get into the following potential scenarios:

• Unlock Earlier = Steal Resource
• Steal Resource = Yield on Pillage
• Yield on Pillage = No Strategic = Other = None

So....its pretty murky. Only Unlock Earlier comes out a little ahead of most results (2%), we cannot say it beats Steal Resources, but its the only one we can say that beats all the rest of the results.


2) Everyone Voted Multiple Times: In this model we assume most of the results were selected multiple times by one voter. We assume the "None" and "Other" result is distinctive, aka that a person who selected none or other did not select any other options.

Total Voters: 23
Margin of Error: +- 20%

• No Strategic: 17.4%
• Yield on Pillage: 43.5%
• Steal Resource: 78.3%
• Unlock Earlier: 87.0%
• Other: 4.35%
• None: 8.79%

With this one, we have the following results (I think, again, multichoice margin of errors is not my forte):

• No Strategic = Yield on Pillage = Other = None
• Yield on Pillage = Steal Resources
• Steal Resource = Unlock Earlier

So with this method we get similar results. Steal Resources and Unlock Earlier are indeterminate, and yield on pillage is in the running. The real answer is going to lie somewhere in this spectrum....as some people chose 1 vote, and some chose multiple.



My personal recommendation. Do one more poll with 3 options (Yield on Pillage, Steal Resources, and Unlock Earlier).... one choice each. That will hopefully remove the last bit of haze and get us a clear result.

 

[JamesNinelives]

Do one more poll with 3 options (Yield on Pillage, Steal Resources, and Unlock Earlier).... one choice each.

Based purely on the stats, yes. I think we can get a little more information from reading the comments, and also referring to some of the discussion in other threads (the previous poll and the Shoshone specific poll). Also the Discord if there's relevant stuff there. I do get a similar conclusion though - it's not clear between Unlock Earlier and Steal Resources. Yield on pillage has some support but for the sake of decision making, I think it's fair to go with the more popular options even if the margin isn't huge. For me the main uncertainly is a) one or both and b) if one, which one.

I'm wary of poll fatigue though :/.

I thought we might do a trial run for one version because that tends to get a lot of feedback. E.g. do both and see how it plays. If it's OP that should be clear pretty quickly and then we have some actual experiences to go on rather than just conjecture. If it's not OP and nobody objects them we have our solution!

I do want to ping @pineappledan because he mentioned not wanting to do both for reasons related when strategics become available. Feedback welcome!

 

[Recursive]

Mechanically stealing strategics is interesting and not something that other civs can do so I'm leaning towards that.

Interesting, yes - but someone would need to do the coding work for that if the mechanic doesn't already exist.