Difficulty Levels Revealed

from Stardock Forums
What difficulty levels do players, and their empires, really prefer? The Spymaster's latest report reveals the truth about a large sample of the current Metaverse spanning 520 players in the 13 top empires.

Gerakken's toughness scale is not meant to measure maximum difficulty level played, but the current preferred levels of play based on recent games. The last 3 entries on a player's record are checked against a scale, averaged, and rounded back to the nearest category. Losses, anomolous scores, and scores from old version games that did not report difficulty are ignored. The preferred sample is 3 games, but if a player has only one or two valid scores, the record is still processed.

The scoring scale is based upon game difficulty:

1 Cakewalk
2 Easy
3 Simple
4 Beginner
5 Normal
6 Challenging
7 Tough
8 Painful
9 Crippling
10 Masochistic

To use a fictitious player record as an example, say the last three games were normal (rated 5), painful (rated 8), and crippling (rated 9). Those add up to 22 points. The average would be 22/3=7.33 rounded down to 7. The player rating is tough (7) on the scale.

422 player records out of 520 contained eligible scores. The scores are listed as a composite of all players combined and are also broken down by empires. Empires are rated on a similar toughness skill to players. The overall rating of each valid player record in the empire is averaged together. Empires with less than 10 usable player records are not listed in the empire breakdowns because the sample is too small to get a real empire toughness, but their players still appear in the composite.

Here are the empire breakdowns, in ascending order of empire toughness:

CDG Creations League
Cakewalk : 0
Easy : 0
Simple : 5
Beginner : 2
Normal : 2
Challenging: 1
Tough : 0
Painful : 0
Crippling : 0
Masochistic: 0
Toughness rating: 3.90

Canadian Star Federation
Cakewalk : 5
Easy : 3
Simple : 9
Beginner :12
Normal :15
Challenging: 7
Tough : 1
Painful : 3
Crippling : 0
Masochistic: 1
Toughness rating: 4.36

Jedi Republic
Cakewalk : 1
Easy : 1
Simple : 3
Beginner : 7
Normal : 3
Challenging: 0
Tough : 1
Painful : 4
Crippling : 0
Masochistic: 0
Toughness rating: 4.70

Apolyton
Cakewalk : 2
Easy : 4
Simple :10
Beginner :18
Normal :15
Challenging: 6
Tough : 4
Painful : 8
Crippling : 3
Masochistic: 2
Toughness rating: 5.03

ANZAC
Cakewalk : 4
Easy : 0
Simple : 2
Beginner : 5
Normal : 8
Challenging: 3
Tough : 1
Painful : 1
Crippling : 4
Masochistic: 1
Toughness rating: 5.14

European Alliance
Cakewalk : 3
Easy : 5
Simple : 8
Beginner :11
Normal :16
Challenging: 9
Tough : 5
Painful :10
Crippling : 4
Masochistic: 1
Toughness rating: 5.22

GROSS
Cakewalk : 2
Easy : 1
Simple : 6
Beginner : 2
Normal : 7
Challenging: 8
Tough : 4
Painful : 7
Crippling : 4
Masochistic: 1
Toughness rating: 5.79

Gamma Squad
Cakewalk : 0
Easy : 2
Simple : 1
Beginner : 3
Normal : 5
Challenging: 2
Tough : 0
Painful : 5
Crippling : 3
Masochistic: 0
Toughness rating: 5.86

MWC
Cakewalk : 0
Easy : 1
Simple : 4
Beginner : 3
Normal :14
Challenging: 6
Tough : 5
Painful : 6
Crippling : 6
Masochistic: 0
Toughness rating: 5.97

The Fellowship
Cakewalk : 0
Easy : 0
Simple : 1
Beginner : 3
Normal : 7
Challenging: 7
Tough : 4
Painful : 7
Crippling : 9
Masochistic: 0
Toughness rating: 6.76

The composite rating of all 422 eligible players in all 13 top empires:
Cakewalk :17
Easy :17
Simple :50
Beginner :70
Normal :92
Challenging:52
Tough :29
Painful :54
Crippling :35
Masochistic: 6
Toughness rating: 5.36

Anything less than 4.5 on the empire toughness scale represent empires whose players are laid back about difficulty level. A 4.5 to 5.5 represent empires who take a middle of the road approach to difficulty. Anything above 5.5 represents an empire that is leaning toward harder games.

In my research, I have found 4 general types of player records:
1. The traditional climb of starting low and levelling off at some point higher on the scale.
2. The peak and drop, where a player starts low, climbs to a certain height, and then drops back down to a comfortable difficulty level.
3. A straight line: the player starts at a certain difficulty and rarely deviates from it.
4. A lab rat/wild record scenario where difficulty changes very often.

I feel this is a fair scale no matter what the pattern of the record. It does not exclude newer players with few scores on the board, but the three game sample also does not discriminate against older players with long records.

I may make empire difficulty breakdown a standard feature in my Major Empires thread to better help players pin down whether any of those empires are right for them. For the moment, though, I just present my work here for evaluation and commentary. May these numbers prove useful.

Edit: Attempted chart formatting.
[Message Edited]
23,933 views 134 replies
Reply #1 from Stardock Forums Top
Good stuff..


In my research, I have found 4 general types of player records:
1. The traditional climb of starting low and levelling off at some point higher on the scale.
2. The peak and drop, where a player starts low, climbs to a certain height, and then drops back down to a comfortable difficulty level.
3. A straight line: the player starts at a certain difficulty and rarely deviates from it.
4. A lab rat/wild record scenario where difficulty changes very often.
End of quote

Are these 4 player types listed in majority order or just the way you thought of them?

Reply #2 from Stardock Forums Top
Very interesting stats, Gerakken.
Kudos to you for analysing all this for us. :CONGRAT:


Curtains - Customize Windows with additional styles
Reply #3 from Stardock Forums Top
Are these 4 player types listed in majority order or just the way you thought of them?
End of quote


They are listed very much in the order they appeared in. The traditional start low, climg higher, then stay there was by far the most common model. This is probably why the composite was over 5.00: most people try to climb up the ladder toward crippling instead of staying at normal. People seem to have enough sense to avoid maso. Not only did I detect only 6 eligible players who currently win maso as their level of choice, but the admitted losses of those attempting maso where few. (I should note that I did see several losses submitted in my checking, but a maso loss was still rare.)
Reply #4 from Stardock Forums Top
Stats say what you tell them to say. What I see is that Apolyton has the most Maso players in your stats :).
Reply #5 from Stardock Forums Top
Stats say what you tell them to say. What I see is that Apolyton has the most Maso players in your stats

You also have 72 players listed, The other empires, far fewer.

And with 42 players we have 5 crippling or above to match your 5.

I'd say you were doing the cherry picking much more than Gunner:
Reply #6 from Stardock Forums Top
once you get your planets, im finding the game pretty easy and straight forward. the "painfull" game i just played wasnt any harder then the "normal" and "challenging" game if you ignore the first dozen turns in the beginning of the game.
Reply #8 from Stardock Forums Top
crippling is my favourite - it keeps the interest up, and you don't have to have a meltdown to win.

good stuff, loyal gerakken.
Reply #10 from Stardock Forums Top
Stats say what you tell them to say. What I see is that Apolyton has the most Maso players in your stats.
End of quote


That's a fact. Hope to see more Apolyton Masochists back here on the board once the Xpack is out.
Reply #11 from Stardock Forums Top
Gerakken,
very nice work. It is worth pointing out though that these are not all active players!!! The difficulty levels have been fine tuned a lot over the past few months. In particular the hardest two levels have been increased dramatically since version 1.00. It might be interesting to see how things change if only active players (maybe version 1.04 or later) are compared.

Paul.

~SDC~
Reply #12 from Stardock Forums Top
I'd say you were doing the cherry picking much more than Gunner:
End of quote


Of course I was picking cherries. We aren't that active, and when we are, the scores somehow don't get on the metaverse...
Reply #13 from Stardock Forums Top
Looking at some scores, I have noticed "Suicidal" as a difficulty level for 1.05 in one on my empire's scores (Dark Assassin's). What is that setting? It is not on your list of levels. I hope the Terran Federation may be included in this calculation soon, when we quailfy.
-Zen
Reply #14 from Stardock Forums Top
It is worth pointing out though that these are not all active players!!!
End of quote


Very true. But the goal at the time was not only to make a composite, but to cover the larger empires as well. Doing what you ask would be a monumental task and the focus would no longer be on empires. To do this truly right, taking only 1.05 and greater games would be better and then one would have to use individual players taken right from the Meta list from top to bottom instead of using empire rosters. Any volunteers?

If I get incredibly bored or have extreme writer's block I might take a whack at it. But then people would scream that I only took say, the top 200 on the Meta. Granted, they would be the most active people and almost certain to be using solely newer version games by now. At least in my original numbers some old dogs got their day. :)
Looking at some scores, I have noticed "Suicidal" as a difficulty level for 1.05
End of quote


That is a known bug. There is no valid suicidal level. It was never properly implemented (that I know of, at least). Sometimes a cakewalk game is listed as suicidal.
Reply #15 from Stardock Forums Top
Gerakken,

Is there a way you can use your statistics to show us how the scores range in each of these levels(assumeing 1.05 of course). So, for example, if I can consistantly score 6k as a beginner that equates to a 8k at normal and 10.5k as a challenging, etc. I know a lot depends on things like techs researched and so on, but a nice range would suffice.

B
Reply #16 from Stardock Forums Top
Is there a way you can use your statistics to show us how the scores range in each of these levels
End of quote


Nope, no way to do that. I have already explained elsewhere that there are simply too many unknowns and too wide of a variance between scores to make a comparable rating. And, to make it worse, difficulty seems to be a lesser modifier compared to everything else in the mix except map size. (Map size is always a monster factor.)

Now I present a truly grand event. Since I was bored, had writer's block, and because Solitair asked, I am going to present a breakdown of the top 10, 25, 50, 100, and 250 players on the Meta as of now using version 1.05 or later:

Top 10 Players
Cakewalk : 0
Easy : 0
Simple : 0
Beginner : 3
Normal : 0
Challenging: 1
Tough : 0
Painful : 3
Crippling : 2
Masochistic: 1
Toughness : 7.00

Top 25 Players
Cakewalk : 0
Easy : 0
Simple : 1
Beginner : 4
Normal : 4
Challenging: 3
Tough : 0
Painful : 5
Crippling : 7
Masochistic: 1
Toughness : 6.80

Top 50 Players
Cakewalk : 1
Easy : 0
Simple : 1
Beginner : 8
Normal : 7
Challenging: 4
Tough : 3
Painful :11
Crippling :14
Masochistic:1
Toughness : 6.96

Top 100 Players
Cakewalk : 1
Easy : 0
Simple : 4
Beginner :13
Normal :13
Challenging:11
Tough : 8
Painful :20
Crippling :28
Masochistic: 2
Toughness : 6.84

Top 250 Players
Cakewalk : 2
Easy : 3
Simple :11
Beginner :26
Normal :42
Challenging:36
Tough :24
Painful :53
Crippling :48
Masochistic: 5
Toughness : 6.58

The five maso players in the top 250 are:
Andreas Finn
Arkdoz
LDiCesare
Peter Berger
Staffa

That was some serious Spymaster training. Enjoy.
[Message Edited]
Reply #17 from Stardock Forums Top
very interesting thread. one would think SD keeps these stats when games are submitted.
Reply #18 from Stardock Forums Top
Top 250? Why don't you lie down, Gerakken. Here ya go. Take a few deep breaths and just relax on that soft leather couch. Some men in white uniforms will stop by soon and take you on a lovely vacation. ;)
Reply #19 from Stardock Forums Top
one would think SD keeps these stats when games are submitted.
End of quote


That is precisely the problem: we think and they know. So I am doing a small task to fill in some of the blanks with hard figures. Of course, if anyone up there wants spill the contents of their database on a regular basis, I would be perfectly fine with it! :)
Top 250? Why don't you lie down, Gerakken.
End of quote


Spymaster training can be hard, but I have to earn my booze money somehow. Otherwise, I would have to go without and suffer real extensive periods of sobriety. Thank you for visiting, Kit. And is that a straightjacket I see hidden behind your back? ;)
Reply #20 from Stardock Forums Top
Take a rest now friend and fry yourself up some KFD and pop a cold one. It'll do you(and us) some good.

B
Reply #21 from Stardock Forums Top
This thread got stickyed! Someone in Stardock must really like your work, Gerakken. :)
Reply #22 from Stardock Forums Top
I wonder if we will ever get to see some statistics from the main Metaverse database. That would be terrific!

Stuff like favourite alignments, map sizes, difficulties, victory types, etc.
Reply #23 from Stardock Forums Top
Wow! Looks like "they" are onto me. I got to hide! LOL Thanks to whomever.

Another scan might be in order in the next few days. The current two distinct camps of extreme strategy in the Metaverse will be interesting to watch. There is the old school theory of real big maps at the hardest difficulty one can manage (like gigantic maso, but more likely to be a gigantic painful/crippling for the majority of real hard core difficulty mongers) to get a series of large scores at a relatively slow rate, and the EBZ cranking theory, a refined and updated version of an old school strategy that emphasizes small maps on low difficulty levels to generate fast games. The faster, simpler games generate less individual score, but are designed to overwhelm the slower, harder players with sheer numbers of games. Right now, the prevalent thought among top players, whether it is top 10 or top 250, is "harder is better."
Reply #24 from Stardock Forums Top
that a straightjacket I see
End of quote


It's an "I love me" jacket - excellent for hugging yourself!

Alex


~SDC~
Reply #25 from Stardock Forums Top
I think Gerakken actually works for Stardock.

He has been planted, to stir up interest and rivalry. ;p
[Message Edited]