Page 1 of 1

A Statistical Experiment

PostPosted: Mon May 19, 2014 3:26 am
by kjeopardy
I'm curious as to the win percentage of each of the 4 sides: rebels, imps, droids, and clones. I'm assuming the Battlefront Server Software does not log such information.

Would anyone be interested in helping keep a record -- for like a week maybe -- of the wins/losses of each game? I'm not sure how the details of logging would work (maybe a google doc), but there's no need to get into that without my seeing if anyone would even be interested in helping.

We don't have to record every single game -- but we should try to get enough so that our results at least mean something. I don't really know statistics, and am not sure how many games would make our conclusions "significant" in the precise mathematical sense, but I'd just like to get a rough idea.

My hypothesis is that the CIS and Rebs win more than their enemies -- but by how much is the question I guess.

So, if there's any interest I'd try to put that together. We only need maybe 7-8 people who play regularly to be willing to help, but it would be more productive if some of the people were in different time zones, play in the morning, etc.

GameSpy has given me a deadline for this little project :mrgreen:, so let me know!

Re: A Statistical Experiment

PostPosted: Mon May 19, 2014 3:54 am
by Duel of Fates
I have been playing for too long. From my experience, it is pretty evenly split as far as winning sides ratios. It usually comes down to experienced players working together that will turn the tide. Good tactics, experienced players, and teamwork on both sides usually means a good close game.

Interesting theory, but a little late to prove one way or the other. Since gamespy will be shutting down soon.

Re: A Statistical Experiment

PostPosted: Mon May 19, 2014 3:56 am
by kjeopardy
I've always felt a bit of a CIS/rebs bent --

But yes, I guess it has to do more with the luck of where the better players get sent.

And I guess there really isn't time to do it -- something like this may really take a month or two to really give anything solid. Ah well, maybe if the game survives...

Re: A Statistical Experiment

PostPosted: Mon May 19, 2014 5:36 am
by Duel of Fates
crossing fingers :whistling:

Re: A Statistical Experiment

PostPosted: Mon May 19, 2014 5:47 am
by WD-40
Even if you do this over 2 years, I believe as Duel said. Having good players is also only so good as team work, but especially good tactics (ie capturing CPs). Having just one sniper sneaking off unseen at high speed for instance to capture a badly needed CP on the other side of a map, while the enemy is overly focused on crushing the only remaining CP can be a game changer. But unique advantages of players such as those that fly to enable quick shortcuts to high ground and CPs is not available to opponents despite level of skills. Some flyers would rather shoot at the enemy instead of flying to get a CP. Then there's those notorious Engineers who don't repair the badly needed health/ammo droids that would either save the day or at least the CP, or have the opportunity to blow up an enemy's droids, knowing they are about to die anyway. Unless you a had ALL good consistent players, ALL the time, it would be tough to analyze this.

Re: A Statistical Experiment

PostPosted: Mon May 19, 2014 4:57 pm
by Mandalore
If you're talking pub [poo], it's pretty random.

In terms of tier list style conversation, when it's the best vs the best played to the utmost extent then rebels tend to be stronger as does CIS.

Re: A Statistical Experiment

PostPosted: Mon May 19, 2014 10:01 pm
by Darth Crater
I suspect that, at least for some maps, the map layout will overshadow any difference between the factions themselves. And that's not even counting things like the double reinforcement loss that happens in... was it GCW Death Star?

Re: A Statistical Experiment

PostPosted: Mon May 19, 2014 11:15 pm
by kjeopardy
What do you mean double-reinforcement loss?