88.99%

Tactical advice, How-to, Post-mortem, etc.
Post Reply
badpazzword
Posts: 3627
Joined: Thu Jan 12, 2006 8:00 am
Contact:

Post by badpazzword »

That is the minimum percentage of games the outcome of which can be predicted successfully just by looking at who is playing, without considering who of those is the commander. We are speaking of ALL games with two teams that made it to the ASGS database, whatever the length, whatever the number of players, whatever the type, including even draws.

Discuss.

(EDITed for clarity)
Last edited by badpazzword on Thu Oct 23, 2008 12:37 pm, edited 1 time in total.
Have gaming questions? Get expert answers! Image Image
Raveen
Posts: 9104
Joined: Wed Mar 16, 2005 8:00 am
Location: Birmingham, UK
Contact:

Post by Raveen »

Is it a different value for SGs over PUGs?

Also you need to take into account that the stronger com gets stacked more so Strong com = strong team.
ImageImage
Spidey: Can't think of a reason I'd need to know anything
Malicious Wraith
Posts: 3170
Joined: Sun Jul 16, 2006 12:51 am

Post by Malicious Wraith »

You also need to factor in that the vast majority of coms are incompetent to varying degree's, so this number is not reasonably reflected.

Me versus Vlymoxyd is certainly a game deciding comm differential, provided I get a team that can at least do what the $#@! I tell them too.
Unknown wrote:[Just want] to play some games before Alleg dies for good.
I don't want that time to be a @#(!-storm of hate and schadenfreude.
IG: Liquid_Mamba / Fedman
sgt_baker
Posts: 1510
Joined: Wed Oct 20, 2004 7:00 am
Location: London, UK.
Contact:

Post by sgt_baker »

Raveen wrote:QUOTE (Raveen @ Oct 23 2008, 01:27 PM) Is it a different value for SGs over PUGs?
Thanks for reminding me that I still need to write the SG detection code and squad rating system. :)

QUOTE Also you need to take into account that the stronger com gets stacked more so Strong com = strong team.[/quote]

There is certainly evidence for a correlation between team skill and commander skill. I do actually have a scatter graph knocking around here somewhere which displays the effect. I'll link it when I've found/regenerated it.

B

Edit: Hrmph. So it turns out that I re-purposed the scatter graph code to do something completely different at some point. Time to undo my doings. 'yx
Last edited by sgt_baker on Thu Oct 23, 2008 1:10 pm, edited 1 time in total.
Image
Granary Sergeant Baker - Special Bread Service (Wurf - 13th Oct 2011)
sgt_baker
Posts: 1510
Joined: Wed Oct 20, 2004 7:00 am
Location: London, UK.
Contact:

Post by sgt_baker »

Scatter plot of Team 0 skill vs Team 1 skill. Positive values mean team 0 was more skilled. Negative values indicate that team 1 was more skilled. Red dots indicate team 0 win, blue team 1 win.

If you don't understand what this graph is telling us, don't lose any sleep over it. All it demonstrates is that better comms attract better teams.
Last edited by sgt_baker on Thu Oct 23, 2008 1:25 pm, edited 1 time in total.
Image
Granary Sergeant Baker - Special Bread Service (Wurf - 13th Oct 2011)
Grimmwolf_GB
Posts: 3711
Joined: Wed Jul 02, 2003 7:00 am
Location: Germany
Contact:

Post by Grimmwolf_GB »

This post looks misplaced. Was it supposed to be in gameplay?
badpazzword
Posts: 3627
Joined: Thu Jan 12, 2006 8:00 am
Contact:

Post by badpazzword »

Yes. My point is, isn't that percentage too high? Does it not affect the enjoyability of the game? Is it really fun to have a game where "the stack" (for every value of stack, including pretty even games) wins 88.99% of the time? Is it possible that map, factions and settings can only be found in the remaining 10.11% percent? Should we do anything about it?

Or am I just misreading the data?


EDIT: I do not have answers. Just questions.
Last edited by badpazzword on Thu Oct 23, 2008 1:53 pm, edited 1 time in total.
Have gaming questions? Get expert answers! Image Image
Evincar
Posts: 1139
Joined: Fri Mar 02, 2007 4:00 am
Location: The darkest side of the sun

Post by Evincar »

once alleg skill is implemented, enforced autobalance might actually be a possible, community supported option?
Classifiable up to Trolleomorphism.
sgt_baker
Posts: 1510
Joined: Wed Oct 20, 2004 7:00 am
Location: London, UK.
Contact:

Post by sgt_baker »

Our pre-launch autobalancer is certainly very effective. Post-lauch AB is still a work in progress. Blancing skill and numbers whilst also keeping everyone happy (and accounting for people joining/dropping etc etc) turns out to be a particularly gnarly problem. Fortunately post-lauch AB is a distinct work effort and separate to the principal drive towards implementation of the new rating system as a whole.

As of today I finalised the C# implementation of AS, and only minor changed will need to be reflected in the SQL version to sync them up. This leaves us in a position where we will be able to calculate ranks natively in ASGS. Once both TE and I are satisfied with progress to that point, I can begin work with DogBones on getting pre-launcg autobalance implemented in AllSrv. Post launch AB might take a while longer, since we really don't have an effective algorithm at the moment. Sorry, but I'm working on it. :)

B
Image
Granary Sergeant Baker - Special Bread Service (Wurf - 13th Oct 2011)
spideycw
Posts: 7512
Joined: Sun Jul 06, 2003 7:00 am

Post by spideycw »

I think you could get a higher percentage by taking a look at who is commanding on what Map with what Factions with what Factions than 88.99% personally
I'm sorry I don't remember any of it. For you the day spideycw graced your squad with utter destruction was the most important day of your life. But for me, it was Sunday
Idanmel wrote:QUOTE (Idanmel @ Mar 19 2012, 05:54 AM) I am ashamed for all the drama I caused, I have much to learn on how to behave when things don't go my way.

My apologies.
Post Reply