C-IV GOTM 09 Results & Congratulations

[Lexad]

As this is the mix of samples from 9 different distributions, you'd better not make far-reaching conclusions. However, we see once again the support that most things in this world are ~normally distributed

 

[bio_hazard]

Not really We do have a database, and the technology is well capable of determining which players have reached which milestones. What sort of badges would you suggest, for what levels of play?

I just ran a few database queries for Civ4 vanilla. here's the bell curve of best scores:

0 - 100: 4 players
100 - 250: 8 players
250 - 500: 9 Players
500 - 1K: 14 players
1k - 2.5K: 33 players
2.5K - 5K: 127 players
5K - 10K: 241 players
10k - 25K: 365 players
25k - 50k: 234 players
50K - 100K: 198 players
100K - 250K: 79 players
>= 250K: 3 players

Total: 1315 players

Well, if I had to design this (and again, I'm not unhappy with the current shields system), I'd keep it simple by NOT breaking out separate lists for different difficulty levels, map types, etc. The HOF data are heavily parsed, and I don't see how we need that for GOTM.

I basically like the bins you listed above, although with a few modifications. For example, start the first cutoff >=5k. Maybe 5k, 10k, 25k, 50k, 75k(?), 100k, 150k(?), 200k, >250k.

I'll leave it to others if they want to make names for these milestones.

Then, in the results announcements, a quick statement like:

"Reaching new GOTM score milestones:

5k: A,B,C
10k: D, E
25k: F

etc.

And if we're in the mood for more pats on the back, how about welcomes to new players and/or listing first victories?

 

[civ_steve]

This is somewhat reminiscent of the Player Community concept that Phillip_martin started for Civ3 GOTM. This assigned 'levels' based on # of GOTM's submitted:

Discussion Thread Link

Link to Display

It is somewhat out of date

 

[da_Vinci]

I did have that thought myself, whether the Adventurer scoring penalty should be applied if the object is to score the lowest. OTOH, if the Shield is awarded to the lowest base score, the penalty might not apply to that, perhaps only to Global Player Points and final score. (I'm just not sure off the top of my head.)

I'm also not sure about the CIV scoring system and whether it works well or poorly as a competition to achieve the lowest winning score.

GOTM 9 is an example of the adventurer shield dilemma:

AndrewN gets the dom shield with an 1844 dom, 6044 base, 49,810 final at adventurer.

blastoidstalker has an 1849 dom with a 5446 base, 50,170 final at contender.

Who should get the dom shield? If we add 15% to AndrewN, we get 57,281 (final before the adventurer penalty), so by either base score or unpenalized final score, blastiodstalker would get the shield in this comparison. Using the penalized final scores, it goes to AndrewN.

But if we go by base score, we now must give it to nurla, with a dom in 1740 for 3760 base, 65,858 on contender.

Yet another but, lowest base may be problematic too, as Cactus Pete's silver medal dom in 1664 for 115,263 final had a base of 4680, which is the third lowest dom base score !! (unless I missed something in the table).

Where are the aspirin? Seriously, since lowest base might go to the upper tier players, perhaps the shields, if retained, need to go to the lowest unpenalized score. That way, the adventurers are still eligible, but don't have an advantage for this award.

What say you, mods?

dV

 

[AlanH]

Sounds like a plan. I must say, speaking purely for myself, I never took those shields very seriously, so I don't really feel strongly about how they are assigned.

Also, no one should be making a career out of playing at Adventurer level. It was only ever introduced as a "shallow end" for newcomers to the GOTM to get their feet wet. Keeping the arm bands on over a lot of games isn't in the spirit of that class.

 

[da_Vinci]

Also, no one should be making a career out of playing at Adventurer level. It was only ever introduced as a "shallow end" for newcomers to the GOTM to get their feet wet. Keeping the arm bands on over a lot of games isn't in the spirit of that class.

An interseting topic. In theory, one could spend a lot of time using adventurer if one never made it into the top half of a GOTM list.

My approach is as follows: once I win a particular difficulty level using adventurer, I won't use adventurer for that level or lower. I got slaughtered in GOTM 10 on immortal with adventurer, and won in GOTM 11 on monarch using adventurer. So I would use contender for monarch or less from here on out, but if GOTM 12 is an emperor, I would plan to use adventurer for that one. When I prove that I can win that level with adventurer, I would play that level on contender therafter.

(Of course, if I win a game on a level outside of GOTM, that also would cause me to stop playing adventurer on that level. But since playing the common GOTMs is so much fun because of the forum, that is likely to be all of my future play [darn that eat, sleep and pay bills thing!]).

Does that fit the spirit of the class?

dV

 

[da_Vinci]

As this is the mix of samples from 9 different distributions, you'd better not make far-reaching conclusions. However, we see once again the support that most things in this world are ~normally distributed

I'd say that most things in the world exhibit "central tendency" (the fancy statistical term for central values being more common than the extreme values), but in my experience many are sufficiently skewed to make them different from normal (and perhaps make medians better indicators of central tendency than means, but now I am in stats speak).

More on the best scores distribution to follow ...

dV

 

[da_Vinci]

I just ran a few database queries for Civ4 vanilla. here's the bell curve of best scores:

I'm not sure that this distribution is as bellish as the display below might suggest, given that the x-axis bins are of unequal size. By estimating average players per 1 K interval, I get the following:


0 - 100: 4 players
100 - 250: 8 players
250 - 500: 9 Players
500 - 1K: 14 players 35 in first 1 K
1k - 2.5K: 33 players 33/1.5K = 22 / K
2.5K - 5K: 127 players 127/2.5K = 51 / K
5K - 10K: 241 players 241/5k = 48 / K
10k - 25K: 365 players 365/15K = 24 / K
25k - 50k: 234 players 234/25K = 9 / K
50K - 100K: 198 players 198/50K = 4 / K
100K - 250K: 79 players 79/150K = <1 / K
>= 250K: 3 players

Total: 1315 players

Perhaps more of a speed bump than a bell (although it could be one of those very flat normal distributions, I suppose).

But laid out my way, it appears that the mean and median scores are lower than one might expect on first glance at the uneven bin presentation.

dV

 

[da_Vinci]

Well, if I had to design this (and again, I'm not unhappy with the current shields system), I'd keep it simple by NOT breaking out separate lists for different difficulty levels, map types, etc. The HOF data are heavily parsed, and I don't see how we need that for GOTM.

I basically like the bins you listed above, although with a few modifications. For example, start the first cutoff >=5k. Maybe 5k, 10k, 25k, 50k, 75k(?), 100k, 150k(?), 200k, >250k.

I'll leave it to others if they want to make names for these milestones.

Then, in the results announcements, a quick statement like:

"Reaching new GOTM score milestones:

5k: A,B,C
10k: D, E
25k: F

etc.

And if we're in the mood for more pats on the back, how about welcomes to new players and/or listing first victories?

@bio_hazard: I like the way you think! Much better than my highest score after a given turn idea (just scrap that one).

We could use the 9 levels of difficulty as "ranks" that GOTMers can earn, each defined by the 9 cutpoints you stated above (5K = Settler ... 250,000 = Deity). As you suggest, we could announce who has earned a new rank in a particular GOTM

To address AlanH's concern about people using adventurer too much, we could decide that only contender or challenger results can earn ranks.

This could be in addition to the shields, although I think the issue of basing shields on unpenalized scores (or not giving them to adventurer) needs to be resolved.

dV

 

[AlanH]

I tried to make the x-axis bins of roughly equal size on a log scale, given that the score distribution is logarithmic. They aren't, of course, but I wasn't going to delve into my ancient stock of school books for my log tables to work out the correct intervals for two intermediate points between 1 and 10.

PS "Concern" is way too strong a word for the low level interest I can raise for a debate about low scores

 

[Thrallia]

Also, no one should be making a career out of playing at Adventurer level. It was only ever introduced as a "shallow end" for newcomers to the GOTM to get their feet wet. Keeping the arm bands on over a lot of games isn't in the spirit of that class.

well, I think if anyone has made a career out of playing at Adventurer level, that's me...I've only submitted 1 GOTM on contender, the others(6?) that I've submitted were all adventurer...however, I didn't win until my 4th(monarch level) adventurer save(didn't even win the prince contender GOTM I played). At that point I decided not to play adventurer for monarch or lower difficulty. I then lost my 5th and won my 6th, which was GOTM10 Immortal, so now I'm not going to be playing adventurer anymore at all(especially since I found out I was in the top half on GOTM9 despite losing).

I'd say that despite my making a 'career' of playing the adventurer saves, it didn't really have much impact on the rest of the competition until I got good enough from my experience that I was able to win in GOTM10, and now I'm settling down for a long, satisfying career of contender submissions, with GOTM11 hopefully being my first contender victory.

oh yeah, and I like the idea of announcing scoring milestones as has been suggested by bio_hazard and the player ranking levels by highest scoring range that da vinci suggested.

 

[DynamicSpirit]

Where are the aspirin? Seriously, since lowest base might go to the upper tier players, perhaps the shields, if retained, need to go to the lowest unpenalized score. That way, the adventurers are still eligible, but don't have an advantage for this award.

This is starting to sound a bit like the scoring debate coming back again, although in a slightly different form.

Awarding lowest-scores to lowest unpenalized scores was something that crossed my mind too yesterday. I think the problem with it is it's going to cause a lot of confusion, and every resultset is going to be accompanied by 'why did X not Y win that award' type queries (we're already getting that to some extent with the 'fastest' awards. I think it'll get a lot worse if different awards are calculated on different final scores).

In terms of adding new awards/etc. My own view is that the way the final score is calculated is so badly broken anyway that, frankly, I don't take much notice of it, and I'd be uneasy at adding more awards based on that final score without fixing the final score first (but that's another discussion that's been had elsewhere). Also, I think there's a danger in adding too many awards/mentions, that you end up with a situation where half the people who entered will win something, and that seems to devalue the fact of getting an award. However, if you want to add something, my initial thought would be something like 'highest scoring newcomer' and 'biggest improvement' (ie. who's improved their score most compared to anything they've previously scored).

 

[da_Vinci]

Awarding lowest-scores to lowest unpenalized scores was something that crossed my mind too yesterday. I think the problem with it is it's going to cause a lot of confusion, and every resultset is going to be accompanied by 'why did X not Y win that award' type queries (we're already getting that to some extent with the 'fastest' awards. I think it'll get a lot worse if different awards are calculated on different final scores).

The whole shield debate hinges, of course, on whether anyone really cares about getting them or not. Maybe it is all moot? If they are not something that people intentionally aim for, then perhaps it does not matter. But if there are some players doing what civ_steve suggested, and planning their play around the shields, then I think we ought to get this right.

To avoid confusion of using unpenalized scores, one could make adventurer submission ineligible for the shields. It would perhaps encourage more rapid migration out of adventurer class. (But don't wake up AlanH to discuss this!)

If a switch to unpenalized scores for shields were clearly stated on the awards page, perhaps the "why X vs Y" queries would not be as bad as you expect.

In terms of adding new awards/etc. My own view is that the way the final score is calculated is so badly broken anyway that, frankly, I don't take much notice of it, and I'd be uneasy at adding more awards based on that final score without fixing the final score first (but that's another discussion that's been had elsewhere). Also, I think there's a danger in adding too many awards/mentions, that you end up with a situation where half the people who entered will win something, and that seems to devalue the fact of getting an award. However, if you want to add something, my initial thought would be something like 'highest scoring newcomer' and 'biggest improvement' (ie. who's improved their score most compared to anything they've previously scored).

Highest scoring newcomer and biggest improvement are good ideas. I think that a series of ranks that players earn and hold (a different idea than an award for the game) would settle into a small number of mentions after a few games, as people's ranks reach their level of skill. Future promotions would then require some large leaps in performance, and these would come less often.

But this next step up in skill is perhaps the one thing that IS within everyone's reach.

Might need different names for ranks other than game difficulties, as they are already used for the forum "ranks" (for number of posts).

dV

 

[bio_hazard]

Is it fair to exclude Adventurer submissions from all awards- just to make it simple? Heck, I usually lose even with the extra help, so it wouldn't make much difference to me! If this results in people moving away from Adventurer, that's probably not a bad thing.

The only exceptions to this might be the ambulances- I'm all for keeping these...

I agree with DynamicSpirit that we shouldn't swamp GOTM with too many awards. If we add something for lower-ranked players, we could drop the shields.

@da_vinci- your milestone names are way too logical! I like them.

 

[jesusin]

The scoring system is so poorly designed that I don&#8217;t think that ranks based on score are a good idea.


Back on topic, thanks again to the staff for the GOTM experience. You guys make my life happier.

Thanks also to the spoiler writers. For those who didn&#8217;t write about their games, please consider: what&#8217;s the point in participating in GOTM if you are not going to write about it? If you write, someone will learn from your ideas (and your mistakes). Most importantly, if you write you will have expressed your ideas in a clear way, so it will be much more easier for you to compare your thoughts to others, thus learning other ideas and also learning about your weak points.
Do you want to become a better player? It is easy, write-up your games and then re-read your write-up after reading those of other people.


Thank you Vynd, Thrallia, godotnut and others for your kind words about my diplo bronze medal. But, you know? it wasn&#8217;t really diplo, it was a military-self voted victory. In addition, I was aiming at a fast diplo, not at score, so it has been an accident. It just happens that diplo requires population and the score system is biased towards population.

@Godotnut: Top player? Me? You must be kidding (and modesty has nothing to do with this). I was trying for the diplo award and I have been defeated by almost two centuries! Lawrence is a top player, not me. When I look at the results, I can&#8217;t take my eyes off that 14th century Conquest. Obormot is a top player. Hey man, look at yourself, you won an award this game, who&#8217;s the top player here?
Did you read my spoiler? It was the first time I used U.S. I have so many things to learn yet&#8230;

@AU_Armageddon: It is a pity what happened in your game. Now, when you say &#8220;Taking my fricking cow and going home&#8221; you are not talking about THIS, are you? ;-)
http://forums.civfanatics.com/showpost.php?p=3281362&postcount=21

 

[Conquistador 63]

The scoring system is so poorly designed that I don’t think that ranks based on score are a good idea.

That might be true, but OTOH there are the combined and speed ranks provided by staff, so we all get our chance to select the best of them all (that is, the one in which we are placed higher ). J/K, but isn't that beautiful that even an average (or below) player like me could be ranked #10 in this game if speed or combined are the chosen criteria?

Congrats to all players, let's keep the ball rolling.

 

[godotnut]

Count me in on the "scoring system stinks" camp.

With the required HOF mod, we now have the potential to adjust it fairly for all. It could be an option in the settings--"GOTM scoring" or something like that. Simply increasing the score value of wonders would be a good start (to help balance builder strategies against the overwhelmingly favored aggessive ones). But that's another topic. And I'm not complaining. As Conquistador 63 said, the new and sortable results page helps a ton (much deserved hat tip to DaveMcW).

@Jesusin:

Well, you crushed my highest scoring diplo game by about 30,000, so you're doing something right. If you don't feel right about the title "top player" comment, perhaps we can at least agree that you should at least adjust the difficulty level indicator that accompanies your avatar up from "Warlord"!

 

[jesusin]

Oh, that! It is not a difficulty level indicator but a number of posts indicator. I have changed it now anyway. I usually play Emperor or Inmortal when opponents are random.

 

[Jove]

Yeah, I agree the in-game score doesn't have a whole lot of meaning for comparison vs. other players. But there's really no harm in enjoying a (totally unexpected) high rank, right?

 

[godotnut]

Oh, that! It is not a difficulty level indicator but a number of posts indicator. I have changed it now anyway. I usually play Emperor or Inmortal when opponents are random.

Ah, now I see. I always wondered about that.