HOF Challenge Series II Discussion

[Sun Tzu Wu]

Discussion continued from Game 2 discussion thread:

Can i improve my result with further submissons? My first culture game was pretty noob one.

The main purpose of the Challenge Series is self improvement via multiple Game submissions.

You can submit as Games as you like at each Difficulty level you choose to compete at.

Edit: Only your best All your games at each Difficulty level will show in the Results page until the Difficulty level Normalization has been implemented. Thereafter only your best Difficulty level Normalized Game will appear in the Results.

This topic should be continued in the main Challenge Series II discussion thread.

Sun Tzu Wu

 

[sanabas]

I definitely think handicapping the games by a certain number of turns in either direction is the way to go.

And personally, I'd be less concerned about getting the balance exactly right, and more concerned with making sure any bias is towards the harder levels.

Not too concerned with where I finish, I'm just playing the games for a bit of fun, I'll probably play at Prince or so for most of them, and post non-competetive finishes.

Now that is a stern warning to Players who are playing well under their Skill (Difficulty) Level. Players planning to submit Settler Difficulty who can easily Win at Prince or higher take heed of this advice. I've added my own emphasis via the DarkRed font.

Meh. I still haven't played enough to know where I can easily win. Prince is certainly easy, but I've hardly played a real game at monarch or higher. (I just looked, I've submitted 9 real prince games, and not a thing higher than that. Therefore I've never even tried playing monarch. I should fix that...) But even if I can win on deity, nothing wrong with having some fun and seeing how fast I can beat one of these challenges at settler level. Fun is the main reason I play, after all.

 

[Dynamic]

Now that is a stern warning to Players who are playing well under their Skill (Difficulty) Level. Players planning to submit Settler Difficulty who can easily Win at Prince or higher take heed of this advice.

Anybody can win low level game, but it's not so easy to win such game really fast.
OK, it looks like +/- bonus turn system is a single available. But I think it will degenerate to 3 level turnament: Settler-Prince-Deity, IMHO.
Well, let's try it.

 

[Sun Tzu Wu]

^I've finished one game but haven't submitted it because I want to wait to see if I can do it at a higher level. It was good practise, though. Geez... who knew you only need 6 cities on a small map to make all the cathedrals in your legend cities?

One can get by with 4 Cities, since the Great Artist Farm shouldn't need any Cathedrals at all; Great Artists will Create Great Works in the Great Artist Farm City which are not multiplied by any Cultural Multipliers. Just build Cathedrals in two of your three Culture Cities. Hermitage usually goes into the Capital and if every works well, the Capital will go Legendary on the Win turn without any Great Works required. The other non-GP Farm will probably need 1-2 GA bombs. The Great Artist Farm will need all the other GA bombs.

The above works best at Quick speed, but Normal speed should work OK as well; just not competitive with Quick speed.

This post should be in the thread for Game #2 and a copy of this post will appear there.

Sun Tzu Wu

 

[cabert]

IF SOMEONE MANAGES GAME1 AT DEITY, it would be worth so much bragging rights i can't think of anything below a permanent announcement to show how deep we all bow

 

[Dynamic]

It looks like in the main Challenge table points are summarized for all entries in one game.
For example, smitsk00 got 16 points (9+7) for #2 and #4 in game2.

 

[Denniz]

It looks like in the main Challenge table points are summarized for all entries in one game.
For example, smitsk00 got 16 points (9+7) for #2 and #4 in game2.

yeah, I was working on showing multiple games per player on the individual game tables. Picking the best of the two involves the difficulty adjustment. I am working on that. I suppose I should do something temporary for the overall page in the meantime.

 

[Denniz]

Added adjusted rank column (Adj#) to game pages. Currently it only serves to prevent double counting anyone in the overall rankings (as reported above) who submits more than one difficulty while still showing those games.

 

[thbrown81]

Personally, I think that it will be very difficult to normalize scores between difficulties for the challenge games since there is a very limited amount of data to compare the games. Using Challenge I as an example, there are only 28 submitted HOF games that are Ancient Start, Large, Epic Conquest Victories. The fastest finish dates for S/C/W were virtually indistinguishable with top dates of 2525, 2525 and 2325 BC respectively (all BabaBrian). The second fastest games were more varied, with dates of 745 BC, 475 BC and 50 AD. However, the jump to Noble is significant, as the fastest Noble finish was 475 BC and BabaBrian's fastest finish was 1070 AD. The fastest Prince and Monarch finishes were also indistinguishable at 1268 AD and 1280 AD. After that, we have an even tougher comparison, because the two fastest Emperor finishes were 200 and 400 AD. So if we analyzed this purely statistically, we would conclude that Emperor is easier than Prince and Monarch.

Continuing with Challenge I as an example, starting with Copper and a lot of trees in your BFC has a bigger impact on winning than the difficulty level up to Noble. At S/C/W, a reasonably competent player should be able to run the continent with axes before the AI gets to Long Bows. However, once you get to Noble it becomes much more difficult, so 10 turns or any other static scale doesn't really illustrate the difference between Warlord and Noble. The distinction between exceptional players also becomes much more apparent at Noble, since running the continent with axes at Noble would be a remarkable accomplisment based on the data.

It would seem most reasonable to create a Bronze (Settler-Warlord), Silver (Noble-Monarch) and Gold (Emperor+) division for each category, with some normalization added within each category since (as others have pointed out) it is much easier to compare a Noble/ Monarch game than a Settler/ Deity game.

By the way, if you think I've forgotten to include data and Immortal and Deity finishes for HOF Ancient/ Large/ Epic Conquest VCs, I haven't. No one's ever submitted one. Not even an Inca game. So I agree, using Gahndi to run a continent on Deity (or even Immortal) without tripping the dom limits or razing would have to be the most remarkable Civ IV feat of all time.

 

[Sun Tzu Wu]

Personally, I think that it will be very difficult to normalize scores between difficulties for the challenge games since there is a very limited amount of data to compare the games. Using Challenge I as an example, there are only 28 submitted HOF games that are Ancient Start, Large, Epic Conquest Victories. The fastest finish dates for S/C/W were virtually indistinguishable with top dates of 2525, 2525 and 2325 BC respectively (all BabaBrian). The second fastest games were more varied, with dates of 745 BC, 475 BC and 50 AD. However, the jump to Noble is significant, as the fastest Noble finish was 475 BC and BabaBrian's fastest finish was 1070 AD. The fastest Prince and Monarch finishes were also indistinguishable at 1268 AD and 1280 AD. After that, we have an even tougher comparison, because the two fastest Emperor finishes were 200 and 400 AD. So if we analyzed this purely statistically, we would conclude that Emperor is easier than Prince and Monarch.

Continuing with Challenge I as an example, starting with Copper and a lot of trees in your BFC has a bigger impact on winning than the difficulty level up to Noble. At S/C/W, a reasonably competent player should be able to run the continent with axes before the AI gets to Long Bows. However, once you get to Noble it becomes much more difficult, so 10 turns or any other static scale doesn't really illustrate the difference between Warlord and Noble. The distinction between exceptional players also becomes much more apparent at Noble, since running the continent with axes at Noble would be a remarkable accomplisment based on the data.

It would seem most reasonable to create a Bronze (Settler-Warlord), Silver (Noble-Monarch) and Gold (Emperor+) division for each category, with some normalization added within each category since (as others have pointed out) it is much easier to compare a Noble/ Monarch game than a Settler/ Deity game.

By the way, if you think I've forgotten to include data and Immortal and Deity finishes for HOF Ancient/ Large/ Epic Conquest VCs, I haven't. No one's ever submitted one. Not even an Inca game. So I agree, using Gahndi to run a continent on Deity (or even Immortal) without tripping the dom limits or razing would have to be the most remarkable Civ IV feat of all time.

The data available to base Difficulty level Normalization may be very sparse, especially on Large and Huge Maps that are usually longer and more difficult to Win, but some Players systems don't have the computing resources to handle Large and Huge Maps. For Challenge Series II, three Games are on Large or Huge Maps (two of which are very difficult Conquest or Domination Games). Personally, I would have focused on more modest Map Sizes like Small and Standard.

Even with the sparse data for Difficulty level Normalization, we should still attempt to do it. We will learn far more from each other playing and discussing play at many different Difficulty levels than if we all just talk to people playing at the same Difficulty level that we do. Personally, I've learned to have more respect for people playing at Settler level who try to Win the earliest Date possible.

It's not so important that we nail down Difficulty level Normalization with the first try. What is so exciting and exhilarating is that we are even trying to do so at all. We will learn at lot about Game Strategy at different Difficulty levels that we could never do isolated in just 1-2 Difficulty levels over time. For example, HOF Gauntlet have always been single Difficulty level events and although they are wonderful for learning Strategy at that Difficulty level. HOF Challenge Series II really continues at multiple Difficulty levels what the HOF Gauntlet does so well at a single Difficulty level. Help us learn to Play the Game better!

I don't think we want a three communities of three Difficulty levels each. We want one community of nine Difficulty levels. We can dare to compare. As long as we don't take the results more seriously that the data its based on, we will have a Golden Age of Civilization IV, before Civilization V arrives.

Sun Tzu Wu

 

[ParadigmShifter]

I don't think we want a three communities of three Difficulty levels each. We want one community of nine Difficulty levels. We can dare to compare. As long as we don't take the results more seriously that the data its based on, we will have a Golden Age of Civilization IV, before Civilization V arrives.

Sun Tzu Wu

Are you Obama?

 

[Fluroscent]

Like it was pointed out by myself and others in the pre-challenge discussion, this type of series will be difficult to apply any non-arbitrary ranking system to. Coupled with the fact that the element of difficulty is entirely optional with ALL difficulty levels allowed, and the name challenge-series, I think the best way to salvage some element of competition is to apply a ranking model as simple as possible: Rank by difficulty first, then by finish date

This is by no means optimal, but it should encourage people to complete the games at a higher level, and thus actually be a challenge. This system will also not require different formulae for the different games, which, frankly, would be a mess.

The series as it is now might get a higher number of submissions, but if most of those are at settler level... Sure, settler level without huts is interesting, and about time to be given a competitive platform, but it could have been handled with gauntlets instead. The lowest levels have no place in a challenge series.

Just my 2c -.- Not going to participate any way until large/huge maps are abolished, but as one of the few who actually completed the first series, I do feel that I am entitled to bring my criticism to the table.

 

[TheMeInTeam]

So if we analyzed this purely statistically, we would conclude that Emperor is easier than Prince and Monarch.

If we did it purely statistically, the error from such a small sample size would be too ridiculous to make any useful conclusions.

Are you Obama?

Scary thought, although I think STW has a better track record of keeping promises, at least in civ.

 

[ParadigmShifter]

If we did it purely statistically, the error from such a small sample size would be too ridiculous to make any useful conclusions.

Agreed.

Scary thought, although I think STW has a better track record of keeping promises, at least in civ.

Meeeooow! Cut your nails you'll scratch someone's eye out girlfriend!!

 

[TheMeInTeam]

Agreed.



Meeeooow! Cut your nails you'll scratch someone's eye out girlfriend!!

Heh, well the party system + method of securing votes here pretty much forces such practices, so it's nothing personal. People who actually do keep promises rather than giving into party pressure (or ignore party pressure to make impossible promises in order to win) are filtered out by their respective parties long, long before they reach presidential candidacy.

3rd party guys lack enough backing and if they don't make those promises, people don't vote for them, which is probably why the parties so much as filter out people who won't make them.

The only thing that bothered me was the # of people who somehow thought this time would be different, ignoring every sign to the contrary regardless of who won the office.

It's a bit of a sad reality, but at least it isn't deity without vassals/city razing attempting conquest.

 

[ParadigmShifter]

Whoever you vote for, you always end up with a politician.

I think George Bush tried to invade Sumeria with no city razing. It went badly.

 

[thbrown81]

If we did it purely statistically, the error from such a small sample size would be too ridiculous to make any useful conclusions.

Does this mean you have a problem with a 1000 year standard of deviation?

 

[smitsk00]

It looks like in the main Challenge table points are summarized for all entries in one game.
For example, smitsk00 got 16 points (9+7) for #2 and #4 in game2.

This may be the only time I ever show up at the top of a list like this I am sure the upcoming adjustment will put me back in my place.;

 

[TheMeInTeam]

Does this mean you have a problem with a 1000 year standard of deviation?

Of course not .

But it wouldn't be useful, and it'd have to be on turns, not years, anyway.

 

[Dynamic]

Denniz, have you some progress for the Level Adjusting Formula?