HoF Rules: Objective Analysis

[Neuro]

Hopefully, coding changes will eventually fix the exploit of abusing the AI. Some examples of coded changes:

1) Until you get to a certain level of friendliness and/or trust, the AI will only trade GPT for Luxuries. They will, of course, gladly TAKE cash in return for Luxuries or GPT.

2) Open Borders' value should be based on territory size, though not a direct scale. Open Borders into a one-city empire shouldn't be nearly as valuable to the AI as a large multi-city empire.

3) If you ever break a deal in which you received a lump sum for some GPT/Resource, you should take a huge diplo hit and even be denounced. Other civs shouldn't trust you anymore (Civ 3 did this, it amazes me that Civ5 doesn't).

 

[vexing]

I'm now uncertain if it was fixed in this way. Can anyone verify yay or nay? If it has been fixed, whether human or barb pillage the tile wouldn't make a difference. I'm really not sure how the quoted post is relevant.

It does, however, add weight to the enforcement team. If a barb pillages a resource, now the game is questionable on whether it should be accepted; if that resource was traded for 1k gold for example the player may or may not have been trying to protect that resource. Given the nature of barbs in this game, it might be hard to prove deliberate misconduct, and that proof would be necessary or you'd have to reject every game that has traded resources pillaged by barbs. Having what amounts to an instant L due to a tile being pillaged is a rather brute force solution .

Of course, that's only if barbs are enabled. As a chance factor we've evidence they should not be before this point, but this point isn't helping that setting.

it was fixed so a player cannot pillage their own resources. i hadn't realized that at the time of my initial post, i checked after you posted it. i quoted that because you have previously implied developers may have intended people to trade their GPT + luxury for lump gold with the expectation they would then pillage the resource to avoid paying anything.

still, the point remains, on deity it would not be unreasonable to find yourself in a situation where you could, if this were not the rule, have a barbarian end its turn on top of one of your untraded resources, in which case you should immediately trade that resource and ALL your gpt to the AI with the most gold. it's entirely possible to trade 200gpt + luxury resource for 4500 gold, which effectively becomes a trade of 200 gold for 4500 gold.

in regards to barbarians or not, i agree, games would be more consistent (as in less luck involved) with barbarians disabled, as well as ancient ruins disabled.

 

[Martin Alvito]

No, he has not, and I said why in response to him. If you're going to weigh in on my or anyone else's argument on this thread, ignoring parts of said argument damages your position.

He stated an opinion. It's logically consistent and cannot be falsified. You have a different opinion. I happen to agree with him.

Your counterargument is that a better rule exists. That may be the case from the players' perspective, but disabling settings by VC is not a better rule for the enforcement staff. They have the right to try to keep the rules as simple as possible in order to reduce confusion and therefore the headaches they have to deal with.

Lose EVERY time? Not likely. Also, if what you say is true and there is functionally no choice, why did HoF see fit to force people to use it as opposed to taking an advantage? Are raging barbs also banned?

Unless you come up with some radical, brilliant strategy that the rest of us aren't able to see, you will lose every time. You're giving away literally thousands of by turning barbs (and therefore camp quests) off. That's going to yield a multi-turn advantage. You won't get blown out, but there are enough good players that you can't afford a mistake of that magnitude and still take the top spot.

I was under the impression that they patched it such that if you pillaged a resource in active trade it would drop to a negative value and the AI would retain the resource. If not the relative strength of this move goes up dangerously, although it's worth pointing out that it actually can't be stronger than 2 of the examples I gave because they indicate an instant W.

You have one instant-win scenario. It isn't applicable for the HoF because letting the AI conquer is going to be slower than rolling up your sleeves and doing it yourself. If AIs happen to help, that's going to speed things up, but that's an uncontrollable luck factor. A surprise amphibious attack on the last AI will shorten the game, but repeated pillage and resale will shorten it a lot more.

I can't speak to self-pillage because I don't, but when the barbs pillage the resource immediately becomes available for trade again after the improvement is fixed.

Also, as the rule stands it is unclear whether taking large lump sums from an AI and then immediately declaring war is illegal. I would be extremely angry if ANY game doing that were banned under current rules, as it's something the AI can and occasionally does do and is not explicitly banned...we need a lot of clarity with this.

I'd make the argument that anything that isn't expressly forbidden is permissible. We do need a ruling here. Taking huge GPT via capitulation after minimal conquest or taking huge lump sums and declaring are both pretty terrible game flaws. I don't think either breaks the game at the level of self-pillage (these actions have negative consequences and opportunity costs), and so I think we just live with it until Firaxis finally implements a solution. But the staff should get together, discuss and then weigh in.

What the heck does "balanced" resources actually do?

I don't think anyone really knows. If balanced is the only resource setting you want to enable, I don't think it's a huge deal either way. Other resource settings make nutty stuff happen.

But if you think about it, having multiple resource settings available when we can't easily see their impact means that we've included a choice whose implications we don't understand. It makes sense to restrict this setting to a single value as a result.

HoF games are a much more tightly-fixed system than real life or even markets. It should be possible to clearly define a criteria that is guaranteed to make certain actions illegal, and then everything else goes that isn't stated explicitly. IMO HoF needs something of that sort ASAP, and that might merit a separate thread on its own to be sorted out.

I agree with the push for maximum clarity, but it's going to be impossible to remove subjectivity entirely unless we can anticipate all possible scenarios in advance, which seems unlikely. I also agree that "anything that is not expressly forbidden is permissible" should be the standard.

It does, however, add weight to the enforcement team. If a barb pillages a resource, now the game is questionable on whether it should be accepted; if that resource was traded for 1k gold for example the player may or may not have been trying to protect that resource. Given the nature of barbs in this game, it might be hard to prove deliberate misconduct, and that proof would be necessary or you'd have to reject every game that has traded resources pillaged by barbs. Having what amounts to an instant L due to a tile being pillaged is a rather brute force solution .

Whether or not this happened is easily verifiable using Deal History. All you need to do is look for a deal that broke out of sequence, count resources, and see whether or not the player had more trades going than they should at that time. If it's unclear, request the save from the turn when the problematic deal was signed.

As far as instant-L: That's a judgment call. That's one way to handle it, but it's not the only way. If a player shows a pattern of that behavior, that's a problem. If you re-up a deal right after signing it and fixing the luxury, that's a problem. If you make an error and re-sign 27 turns after the original deal instead of 31 once every few submissions, that's different.

The problem here is that any set of standards other than zero tolerance can be gamed. So if the staff chooses any other set of concrete guidelines, they should not reveal them.

 

[vexing]

What the heck does "balanced" resources actually do?

I don't think anyone really knows.

i don't have the game or .lua scripts in front of me, but if you're talking about the strategic balance setting, it guarantees horses, iron, and oil in each players starting area. that's all afaik.

 

[TheMeInTeam]

Your counterargument is that a better rule exists. That may be the case from the players' perspective, but disabling settings by VC is not a better rule for the enforcement staff. They have the right to try to keep the rules as simple as possible in order to reduce confusion and therefore the headaches they have to deal with.

You're still ignoring one of the key reasons I'm calling the rule into question:

Also, is the culture victory actually less formulaic with "no policy saving", or does it simply take longer on average? Just because it's overwhelmingly favorable as an option does not mean it's a bad option! Marathon for example is irrefutably the strongest speed for "best times" in virtually anything, but this doesn't mean it's banned.

Neither you nor he has answered this.

Unless you come up with some radical, brilliant strategy that the rest of us aren't able to see, you will lose every time. You're giving away literally thousands of by turning barbs (and therefore camp quests) off. That's going to yield a multi-turn advantage. You won't get blown out, but there are enough good players that you can't afford a mistake of that magnitude and still take the top spot.

My radical, brilliant strategy is to have unusually good spawn luck while my opponent has terrible spawn luck . This is really nitpicking though. Your point is that barbs add a (random) advantage and I agree with it.

A surprise amphibious attack on the last AI will shorten the game, but repeated pillage and resale will shorten it a lot more.

Yeah, if you could actually do it.

But if you think about it, having multiple resource settings available when we can't easily see their impact means that we've included a choice whose implications we don't understand. It makes sense to restrict this setting to a single value as a result.

There should be a way to determine the consequences of the setting. I'm not advocating fantasy realm or something here, but rather something that would potentially make games *more* consistent, rather than less.

The problem here is that any set of standards other than zero tolerance can be gamed. So if the staff chooses any other set of concrete guidelines, they should not reveal them.

Hiding rules is never the correct answer. Please don't copy one of Firaxis' largest flaws into here.

Note: We're on page 2 and we haven't seen a whole lot of numerical basis supporting the rules. Even now, the best we have is an allusion to the extra multipliers toward culture that SP saving enables, without any clearly defined criteria that would make such a thing actually ban worthy.

 

[da_Vinci]

Note: We're on page 2 and we haven't seen a whole lot of numerical basis supporting the rules. Even now, the best we have is an allusion to the extra multipliers toward culture that SP saving enables, without any clearly defined criteria that would make such a thing actually ban worthy.

Are you looking for evidence that you know (or can reasonably suspect) doesn't exist?

Is the "numerical basis" you desire to support some rule something like "really good player A plays 100 games at fixed settings with item of interest on, and then plays 100 games at same fixed settings with item of interest off, and we measure the magnitude of the difference" (with or without p-values and confidence intervals)? (Edit: actually, he has to randomly alternate between on and off, as his skill improves over the series ... )

I suspect such numerical basis doesn't exist at this point in time, and won't unless someone starts a "Hall of Evidence" series.

But remember that lack of evidence is NOT evidence against ... not found guilty is not the same as found innocent.

Is your position that absent this kind of numerical basis, nothing can be banned?

dV

 

[vexing]

Note: We're on page 2 and we haven't seen a whole lot of numerical basis supporting the rules. Even now, the best we have is an allusion to the extra multipliers toward culture that SP saving enables, without any clearly defined criteria that would make such a thing actually ban worthy.

why would you expect numbers to support matter of preference decisions?

i can give you a number for culture saved, but it's completely irrelevant.
in an ideal game i imagine you'd stop after buying 6 policies, and if you get cristo the time of your 18th policy, you end up saving 4630 culture total, which is 16% of the total culture needed to fill 5 branches.
furthermore the need to manage eras and rush to cristo are reduced.

anyway, none of that is an argument for or against policy saving; policy saving enabled just requires a different strategy. personally i find disabled more challenging and enjoyable, and am therefore glad it is banned. were it a choice, i'd prefer it be another condition that tables were differentiated on (just like map types and game speeds) so culture games with policy saving weren't compared to ones without.

 

[Peets]

Also, is the culture victory actually less formulaic with "no policy saving", or does it simply take longer on average? Just because it's overwhelmingly favorable as an option does not mean it's a bad option! Marathon for example is irrefutably the strongest speed for "best times" in virtually anything, but this doesn't mean it's banned.

Marathon games are separated from other speed games. So I don't see the any problems with it. Do you have another example?

 

[TheMeInTeam]

Are you looking for evidence that you know (or can reasonably suspect) doesn't exist?

Yes. I reasonably expect it doesn't exist, and if it doesn't we have conclusive evidence that the rules need to be changed. Before I advocate that strongly, however, I want to make sure said evidence *actually* does not exist, whatever my suspicions may be. I'd like to have more faith than to see things banned arbitrarily/on a whim.

Is the "numerical basis" you desire to support some rule something like "really good player A plays 100 games at fixed settings with item of interest on, and then plays 100 games at same fixed settings with item of interest off, and we measure the magnitude of the difference" (with or without p-values and confidence intervals)? (Edit: actually, he has to randomly alternate between on and off, as his skill improves over the series ... )

Something along those lines. HoF has some ability to track this it would seem, since Denniz was able to come up with a % of games that had "no barbarians" in civ IV. It would seem the mod has the ability to correlate settings with victories, finish date, or whatever you want based on that, unless he did a by-hand count.

Short of statistical analysis, in some cases we can look at the % bonuses allowed here, or the "this unlocks that which leads to x multiplier by turn y" which in extreme cases could be conclusive evidence on its own. Alvito's example of 1000 gold/turn equaling ~250 /turn is a good example of this requirement. If there were a tactic existing that could consistently give that much gold it would be difficult to overcome it; rarely do cities produce more than 10-20 hammers in the earlygame and as such 1000+ gpt would demonstrate a production spike of a factor of 10 or more in some cases. I wish people would actually hash that out in their arguments, but that's essentially what he said.

But remember that lack of evidence is NOT evidence against ... not found guilty is not the same as found innocent.

It's an irrelevant point. These HoF rules are the equivalent of finding someone guilty without the evidence to support guilt. I'm not trying to prove innocence, I'm asking for proof of asserted guilt!!!

Is your position that absent this kind of numerical basis, nothing can be banned?

Absent some kind of numerical basis (theory derived carefully from in-game #'s/% or submission data tracking) these bans are arbitrary, yes. This sounds odd to some in the abstract sense; but things that get banned in competition should have a very good reason. "I think this might be overpowered" or "this seems gamey" are not valid reasons in a competitive format to ban an action.

We have very strong reason to believe that the rules are in place on the basis of "this seems right" or "I prefer this setting" or things along those lines. Those are not valid reasons to restrict players from using settings; especially not in a format where optimized finish times are the goal.

in an ideal game i imagine you'd stop after buying 6 policies, and if you get cristo the time of your 18th policy, you end up saving 4630 culture total, which is 16% of the total culture needed to fill 5 branches.
furthermore the need to manage eras and rush to cristo are reduced.

You save culture, but forgo some economic advantages that would let you tech quicker in saving SP. This is still going to give storing a lead in the case of culture, but less pronounced than some make it appear.

That doesn't make the option bad, however. It's one use for a goal similar to other strategies like a rifle rush or relying on wonders for scientist GPP. Having SP storing as an option allowed would mean almost everyone would play with it; not unlike say ruins (which aren't banned, btw, despite being much more chancy and having at least as much "game" potential). However, would it add to or hinder the number of potential viable options in a given game? Does it break the game, and make games inconsistent/incomparable to each other? Every rule or ban needs to have some reasoning behind it. What is the REASON that SP storing is so bad it must be banned? In a lot of ways it's like banning ruins, playing on settler, etc. What critical difference separates this setting?

Marathon games are separated from other speed games. So I don't see the any problems with it. Do you have another example?

.

That's not true for current civ V HoF, the rankings include multiple speeds.

Yes, they are "separated". You can separate games via a filter to show anything you like! That would go for SP saving also. How are they incomparable, because of the "front page" difference in civ IV? If you can't think of them for yourself, I'll give you some examples that exist on civ IV HoF within each speed tier:

- No vassal states
- Permanent alliances (!)
- No barbarians
- Aggressive AI (!)
- Always war (!)
- City flipping after conquest

Each of these can have a material impact on the outcome of the game (some are somewhat minor, others are very drastic). Depending on victory condition (much like SP storing), some of these are obviously favored (or negative).

Of course, civ V has some too, many of which are similar to IV.

 

[Martin Alvito]

Neither you nor he has answered this.

I can tell you exactly what you take (and pre-reqs) before Cristo on any difficulty, and it's extremely formulaic: Monarchy, Meritocracy, Constitution, Free Speech, Democracy (not in that order). Policies that effectively yield or don't pay for themselves with sufficiently increased build speeds, and Mandate of Heaven won't pay for itself fast enough. If you have an extremely high risk tolerance, you can go Educated Elite instead of Democracy, but the math on that starts to really suck when you have the option of saving policies.

All you need to figure this out is a calculator, knowledge of your options and the table in the Number Crunching Thread.

If you check out the G-Major I thread, we still haven't settled the SP question over there. We have two candidate solutions. The difference in posted results so far is three turns, and vexing is still trying to win his way AFAIK. I think that Educated Elite is going to require too much good fortune all in one game to pay off, but I could get proven wrong.

My radical, brilliant strategy is to have unusually good spawn luck while my opponent has terrible spawn luck .

That might work if we had events that only lasted a single day.

There should be a way to determine the consequences of the setting. I'm not advocating fantasy realm or something here, but rather something that would potentially make games *more* consistent, rather than less.

If you can get into the code and demonstrate that balanced actually reduces the luck factor, or run a bunch of games and prove it with data, by all means. Otherwise, we need a single solution to the problem and it makes sense from the HoF staff's point of view to have that single solution be the default option.

Hiding rules is never the correct answer. Please don't copy one of Firaxis' largest flaws into here.

Smart people in the discipline of economics have shown mathematically that hiding the rules is the correct answer some of the time. This is one of those times. You cannot compare hiding a computer game's rules from hiding enforcement rules. The implications differ.

 

[vexing]

I can tell you exactly what you take (and pre-reqs) before Cristo on any difficulty, and it's extremely formulaic: Monarchy, Meritocracy, Constitution, Free Speech, Democracy (not in that order).

you don't want to take free speech pre cristo, you gain nothing. you get that as your first post. i suspect tradition, liberty, citizenship, meritocracy, aristocracy, freedom, constitution. one of those gotten via oracle leads to only 6 bought policies. not purchasing 5&6 only saves 300 culture, tradition alone clearly nets more than that.

 

[Martin Alvito]

you don't want to take free speech pre cristo, you gain nothing. you get that as your first post. i suspect tradition, liberty, citizenship, meritocracy, aristocracy, freedom, constitution. one of those gotten via oracle leads to only 6 bought policies. not purchasing 5&6 only saves 300 culture, tradition alone clearly nets more than that.

You want to take enough post-Constitution policies that you want to take FS.

Legalism is critical. At that point it makes sense to take Monarchy as well, because the extra cash means you can activate an ally earlier. The math works out.

Democracy will yield an extra Great Artist for enough +24 turns that it's worth taking the 25% hits on policies.

I think you skip Aristocracy. It doesn't speed up the acquisition of enough Wonders by enough turns to pay for itself, given that you're taking Legalism and Democracy.

 

[TheMeInTeam]

If you can get into the code and demonstrate that balanced actually reduces the luck factor, or run a bunch of games and prove it with data, by all means. Otherwise, we need a single solution to the problem and it makes sense from the HoF staff's point of view to have that single solution be the default option.

Oh? By that logic literally every setting should be the default option right now. No exceptions if that logic applies.

Getting into the code is by far a superior approach in this instance, however. I'm not good at this, but if you tell me how I might do so I'll probably take a look (this forum lost its best code divver in danf5771 over a year ago, and it'd be nice to have a few more people who know how to do it that actually post regularly).

Even so, it's pretty arbitrary to act one way for this setting and completely different ways for other settings. Perhaps HoF should lock ALL settings into defaults? Certainly, that would provide consistency although options like ruins/barbs/etc are fairly clear sources of noise. Although such a rigid system would probably need refining over time, it IS probably a better opening than banning things wily nilly.

Smart people in the discipline of economics have shown mathematically that hiding the rules is the correct answer some of the time. This is one of those times. You cannot compare hiding a computer game's rules from hiding enforcement rules. The implications differ.

Source? I'm actually curious about this from a non-civ standpoint.

Economic rules or enforcement rules? Which rules did the economist conclude could be beneficial to hide?

The implications for hiding enforcement rules in HoF exist, and they are pretty dark. For a video game competition where one is encouraged to push the limits it would really come down to a game-spam issue. Sooner or later people will realize when things get rejected or not. Arbitrarily inducing tedium to find this cutoff? No thank you.

Also, how do you separate a "lucky" guy who got pillaged on turn x from a guy who deliberately allows it? Or do you go back to banning both regardless? Even so, people are going to game this. You and I both know it; it would be foolish not to do so if you want a top spot.

 

[Martin Alvito]

Oh? By that logic literally every setting should be the default option right now. No exceptions if that logic applies.

You're taking this out of context. Remember, we have a problem here where players don't know what the implications of the setting are. Since we don't actually know what the "balanced" setting does with certainty, players can't make a reasoned choice over which setting to pick.

We know the likely implications of changing rainfall, sea level or opponents, so it's fair to leave those choices in the player's hands.

Even so, it's pretty arbitrary to act one way for this setting and completely different ways for other settings. Perhaps HoF should lock ALL settings into defaults? Certainly, that would provide consistency although options like ruins/barbs/etc are fairly clear sources of noise. Although such a rigid system would probably need refining over time, it IS probably a better opening than banning things wily nilly.

All rules are arbitrary. You judge them on the basis of the results they generate. In this case, we want to produce the best gaming experience possible. Consistency is a desirable property in a set of rules because it makes the rules easier to remember...but it is not universally the most desirable property.

Source? I'm actually curious about this from a non-civ standpoint.

You don't want to try to read that stuff if you don't have to. The math is brutal if you're not used to it. I'll walk through a quick example that we're all familiar with as proof that the set exists.

Suppose we have two people: a construction foreman, and a business owner. The foreman oversees a team of workers building a shop for the business owner. The business owner can't be on site all the time to oversee the foreman's actions, and the foreman can save enough money cutting corners to afford a hefty bribe to any third party paid to watch him. However, the owner does have time to spot check the foreman's work.

If the timing of the spot checks is known to the foreman in advance, then he can save money by having his employees cut corners in ways that can be covered up, and be sure that any incriminating evidence is hidden when the owner arrives.

That's bad for the business owner. It's also bad from a societal point of view; cutting corners in construction saves the foreman much less than it ends up costing the business owner in labor to undo the shoddy work and do it right, so productive capacity has essentially been wasted by the foreman's decision.

The business owner can make some progress on this issue by making the spot checks random. At that point, the foreman can only safely cut corners if the evidence can quickly be covered up should the owner arrive. That eliminates a lot of the possible harm the foreman could cause, saving the business owner a lot of money and society a lot of labor put to an unproductive use.

We can conclude that making the enforcement rule non-transparent is better from a societal point of view in this case, and so the set exists.

The implications for hiding enforcement rules in HoF exist, and they are pretty dark. For a video game competition where one is encouraged to push the limits it would really come down to a game-spam issue. Sooner or later people will realize when things get rejected or not. Arbitrarily inducing tedium to find this cutoff? No thank you.

You can live in one of three worlds here:

- The zero tolerance world
- The world where the ruleset is hidden, with the implications you indicated
- The world where we define the tolerances and make them public, resulting in players literally trying to herd the barbs onto the tile on the right turn to get the 300

All of those outcomes suck. The question on the table which one sucks least. The claim I made is that the second world is better than the third one. If you systematically push the boundaries, you risk having the HoF staff push the big red button and disallow all of your work. Since players don't know where the boundary is and the downside risk is large, most of them will behave. The scumbags will get thrown out, and good riddance. At the same time, you don't have to scrap your game just because you forgot and re-upped a single deal three turns early, which is what happens in the zero tolerance world.

Also, how do you separate a "lucky" guy who got pillaged on turn x from a guy who deliberately allows it? Or do you go back to banning both regardless? Even so, people are going to game this. You and I both know it; it would be foolish not to do so if you want a top spot.

It's possible to self-enforce, and I think that's what you have to ask. HoF provides a public record of everybody's games. Go look at the gold medal game I have and root through the completed Deal History. Two deals stick out like sore thumbs. In one case I gave Darius 3 to finish off a Research Agreement. In the other case (turn 79) I got pillaged. Count how much Cotton I have, and how many Cotton deals I signed before 31 turns later. If the two numbers are equivalent, there's a problem.

If you're doing it all the time, or just in most of your really good games, you're going to get caught. You're right that we can't separate out the occasional accident from the occasional malicious instance. That's intractable. But people can't game the system often.

 

[TheMeInTeam]

If you're doing it all the time, or just in most of your really good games, you're going to get caught. You're right that we can't separate out the occasional accident from the occasional malicious instance. That's intractable. But people can't game the system often.

To save quote space, I'll just quote this but this is for all of the above argument:

Since I can't think of a 4th alternative it really does seem like an "audit" system on this exploit mechanic is best. My only concern is whether HoF can do this with each of the things it defines in explicit fashion as "exploit"...and of course if they actually attempt to employ anti-exploits retroactively as the current rule seems to imply with "verify potential exploits".

I think you've hashed out reasonably well the argument for an audit system in enforcement of the exploit clause (assuming HoF mod itself has a means of screening, which one would suspect is the case given their reload + replay of turn detection capability); now we simply have to define the most obviously terrible things and have a subjective clause that states something along the lines of "actions that provide outputs beyond the basic game rules are illegal" (this would cover things like infinite oracle tech bug in IV and the SP swap in V even if not explicitly defined). It doesn't address what happens to games that use grey area exploits not explicitly banned yet (I think you just have to leave them), but it's probably the best the HoF staff can do given the unfortunate condition of the game itself.

So while we haven't justified any of the rules yet, we've re-worked one to the point where it would be in a lot better shape and you've done a lot to show why that is. Now we just have the others...

In this case, we want to produce the best gaming experience possible.

Yes, but the only way to do that is to present REASONS a rule leads to the best gaming experience possible. I don't think we've even done that yet when it comes to HoF's format + setup. I'm not sure HoF has done that, so maybe that's where we should start before approaching each of the remaining rules...

 

[vexing]

Since we don't actually know what the "balanced" setting does with certainty

as i said before, strategic balance guarantees oil, iron, and horses in the starting location. now that i have AssignStartingPlots.lua in front of me:

	-- Add mandatory Iron, Horse, Oil to every start if Strategic Balance option is enabled.
	if self.resource_setting == 5 then
		self:AddStrategicBalanceResources(region_number)
	end

if you look within the AddStrategicBalanceResources function you can see that is all it does. this does slightly change the starting location ratings further down the line as well, but i doubt it's significant.

 

[Martin Alvito]

Honestly, I think that the public will do a better enforcement job on this problem than the HoF staff. If you're both dirty and good, eventually somebody is going to be looking through your games for tips and discover your little secret. It will take a while, but it will happen.

By contrast, games probably aren't going to be reviewed by the same person all the time, and they look at a lot of games, so the pattern will be harder for the staff to spot. I think they'll be able to catch the dirtiest players and give them a 'shape up or ship out' ultimatum easily enough, but the more borderline cases will slip through...for a while.

As far as figuring out reasons, I think that we need to take a step back and decide what the goals are, and how heavily to weight them. The following objectives already appear to be on the table:

- Reduce randomness. We'd all like to be able to play as few games as possible and feel confident that we've wrung everything we can out of a gauntlet or given set of settings. Rerolling maps is tedious. Rerolling half-completed games due to the absence of critical luck factors is more tedious.
- Interesting choices. To paraphrase Soren Johnson's opening to the Civ IV manual, maximizing the number of these is the objective of Civ as a game.
- Minimize pain and suffering for the HoF staff. Some threshold of grief exists at which they will all get tired of us and cease volunteering. As the players, we don't want that.

What else should be on this list? Also, what is more or less important? Once we know what we want and how much we want it, we can make a reasoned evaluation of how to change the rules to get us more of what we want.

Finally, we need to keep in mind that the HoF staff functionally has a veto. We don't have to put their needs first all of the time, but they do have a strong bargaining position.

as i said before, strategic balance guarantees oil, iron, and horses in the starting location.

Is this a good thing? It heavily skews things in Cathy's favor. It probably also means that on the lower levels you want to beeline AH and IW, then get a Worker on the ground ASAP to improve and resell before the AI hooks them up. On higher levels, you probably still go AH early for the hammer(s). Is that better than the default state of affairs?

The more I think about it, the more I think that we need a series of games where we test this stuff. I'm pretty good at predicting the implications of rules changes due to my training, but I'm not flawless.

 

[TheMeInTeam]

Is this a good thing? It heavily skews things in Cathy's favor. It probably also means that on the lower levels you want to beeline AH and IW, then get a Worker on the ground ASAP to improve and resell before the AI hooks them up. On higher levels, you probably still go AH early for the hammer(s). Is that better than the default state of affairs?

People with more patience than others can re-roll maps continuously until they get these resources (2 of which wouldn't take TOO long to discover). If that's all "balanced" does it's hard to make a case against it, even with Russia since anyone could just brute force to get the same result (often w/o the AI having said resources!).

 

[Peets]

.

That's not true for current civ V HoF, the rankings include multiple speeds.

Yes, they are "separated". You can separate games via a filter to show anything you like! That would go for SP saving also. How are they incomparable, because of the "front page" difference in civ IV? If you can't think of them for yourself, I'll give you some examples that exist on civ IV HoF within each speed tier:

- No vassal states
- Permanent alliances (!)
- No barbarians
- Aggressive AI (!)
- Always war (!)
- City flipping after conquest

Each of these can have a material impact on the outcome of the game (some are somewhat minor, others are very drastic). Depending on victory condition (much like SP storing), some of these are obviously favored (or negative).

Of course, civ V has some too, many of which are similar to IV.

As far I can read the Hof tables, you can receive a medal and score for each unique setting. So I still don't see a problem with marathon games.
If you wanted to combine them all in one and then compare then you are right that marathon games gives an advantage. But at the current setting I don't see how it is a problem. I read somewhere that there are around 200000 combinations. If you would add your saving policy it would double the combinations. Ending in a result that it would be even more difficult to have some competition outside the gauntlets.
That shouldn't be the only reason for banning it though.

Indeed Civ IV had some options that would change the game heavy. But in Civ V those who would have an impact are banned atm.

I can't find another example why SP should not be banned.

It is indeed my opinion that SP should be banned as it yours that it shouldn't.

 

[da_Vinci]

It's an irrelevant point. These HoF rules are the equivalent of finding someone guilty without the evidence to support guilt. I'm not trying to prove innocence, I'm asking for proof of asserted guilt!!!

I think the point is highly relevant. It gets to the issue of what do you you with no or limited evidence. You agree that the hundred game randonized trial is not going to exist for most (if not all) questions. So you won't have that empirical evidence.

The absence of that empirical evidence neither proves the innocence or guilt of a proposed bannable exploit. You might subjectively chose a policy of assuming innocence when guilt can't be proven, but that is not proof of innocence.

In the absence of the randomized trial, what are you willing to accept as evidence, and what is your threshold (beyond reasonable doubt, as in criminal law, or preponderance of evidence, as in civil law)?

If someone can make a convincing case that a particular setting, tactic, etc. can reasonably be expected to generate a particular level of abuse of the AI, or of the spirit of the game (1000 gpt on turn 20, for example, in ancient start), or some unfair advantage over other participants (although, why dont they all just use it?), is that enough evidence to ban on without the randomized trial?

Hmm ... What are the infractions that might make a setting or tactic bannable?

1. Its clearly a cheat, in the sense of my earliest post here.
2. It violates the spirit of the game (as if there is a consensus on that ... maybe the moderators get to decide that?)
3. It hugely abuses the AI ... taking candy from a baby (again, how huge is huge?)
4. It give the player using it an unfair advantage over other participants in the competition?

Maybe #4 doesn't belong, as all participants are free to use it? Also, if it makes that much differnece, maybe the response is to say it is a different arena for competing, rather than say let's not play in that arena?

Actually, the separate arena vs. not an arena question might apply to cases 1 -3 above as well. The problem being that too many arenas dilutes the number of competing games in each.

So I think we have two questions:

1) What are the criteria that make something bannable?

2) What evidence is required to decide that a particular setting or tactic meets those criteria, and is thus bannable?

dV