19 Alien wins, 4 marine wins, still counting...
MrChoke
Join Date: 2012-12-13 Member: 174990Members, Reinforced - Gold, Reinforced - Shadow
Join Date: 2012-12-13 Member: 174990Members, Reinforced - Gold, Reinforced - Shadow
Since patch 236 that is the stats, 82% alien wins. Pathetic. Before patch 236, I tracked 83 games, 62 of them Aliens won (74%). But hey, flame this post instead. Keep it coming with all of the L2P, and marines can't shoot and yada yada yada. Statistics simply do not lie and no excuse will make it so. I am pushing this so hard because I love the game and wish it could get fixed. But I have to fight though all the people on this forum giving endless excuses and drowning people like me out.
Get a piece of paper and a pen and leave it by your PC. Track each win. Its not so tough. Unless you want the game to stay at a 75% imbalance, track it. Then post it. Prove to yourself and UWE, this game is broken and needs to be fixed. ASAP. It should have been addressed months ago.
Get a piece of paper and a pen and leave it by your PC. Track each win. Its not so tough. Unless you want the game to stay at a 75% imbalance, track it. Then post it. Prove to yourself and UWE, this game is broken and needs to be fixed. ASAP. It should have been addressed months ago.
Comments
I must be missing something here.... I looked through the whole ns2stats.org page and don't see where it says win/loss ratios for aliens/marines.
It's on the right side of the page. Pie chart.
How so? I know the wall walk hitreg was improved but that's overlapped by the gorge hitreg buff and map changes.
I think there was also a Lerk hitreg buff as well, but I'm not sure. (Anyone know if that's true or not?)
<!--quoteo(post=2062410:date=Jan 17 2013, 08:24 PM:name=Kallistrate)--><div class='quotetop'>QUOTE (Kallistrate @ Jan 17 2013, 08:24 PM) <a href="index.php?act=findpost&pid=2062410"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec--><!--coloro:#FFC0CB--><span style="color:#FFC0CB"><!--/coloro-->It's only early yet, but pub games for builds 236 and 237 show an alien win rate around 60%. Definitely looking like an improvement so far.<!--colorc--></span><!--/colorc--><!--QuoteEnd--></div><!--QuoteEEnd-->
Wasn't the Alien/Marine winrate always around 60%? That'd mean there hasn't been any change.
And the win rate was actually around 70% for aliens before. Again, it's only early and not a huge sample size, but it's an improvement. :P<!--colorc--></span><!--/colorc-->
And the win rate was actually around 70% for aliens before. Again, it's only early and not a huge sample size, but it's an improvement. :P<!--colorc--></span><!--/colorc--><!--QuoteEnd--></div><!--QuoteEEnd-->
Oh wow. That's even worse than I thought @___@. Well, the equalization of W/L is good then.
Get a piece of paper and a pen and leave it by your PC. Track each win. Its not so tough. Unless you want the game to stay at a 75% imbalance, track it. Then post it. Prove to yourself and UWE, this game is broken and needs to be fixed. ASAP. It should have been addressed months ago.<!--QuoteEnd--></div><!--QuoteEEnd-->
You say statistics dont lie? Then why is there the famous saying of "Bullsh1t, Lies & Statistics"?
Statistics by them selves mean nothing, raw uncleansed data is totally unusable for anything other than taking up space.
Anyone who works with historical data and uses it to help predict future events knows this, there are hundreds of variables that made up those results such as;
Map size
Starting locations
players on server
Comm experience
Player experience
Player skill
How long hte game went for.
Tech paths chosen
Rate at which upgrades obtained
How aggressive did your team play
Map control
etc etc
You cant compare data sets unless all the variables match (or atleast 90+%) as each of these variables can contribute to the game being won/lost.
You may well know how to play yourself...but if your team does not individual skill has almost no bearing on the result.
If you could take data on face value I would not have my job...and neither would countless people around the world, weather reports would be more accurate than we see currently (after all they are looking at data). But alas it has to be contextualised and sometimes normalised as some data is an anomoly and not a real data set (ie troll comms recycling).
I know its frustrating being on a side that loses...some servers I have come across it seems one side always wins...thats more due to a skill inbalance and team stacks.
Get a piece of paper and a pen and leave it by your PC. Track each win. Its not so tough. Unless you want the game to stay at a 75% imbalance, track it. Then post it. Prove to yourself and UWE, this game is broken and needs to be fixed. ASAP. It should have been addressed months ago.<!--QuoteEnd--></div><!--QuoteEEnd-->
And just where are you getting your accurate percentage ratio from? Your hand written calculations on a notepad next to your mouse?
But that patch gives gorges tunnels and babblers. I mean idk how it will work but I can imagine what options aliens get with that. Meanwhile all I hear is that marines get a railgun exo while nice, is still a late game option and will require extra research in addition to the normal exo. And it's still an exo in speed, HUD obstruction and the fact that it cannot phase or beacon.
I know there will probably be more surprises to the content patch than just this. I'm just afraid the shake up will just give aliens more of an upper hand. So I just really hope things go well.
As for performance has anyone even gotten an increase? It feels more or less the same for me and in some cases I get less as I tend to test frames in the same area every time a new patch releases.
We'll need thousands of games to determine where the balance truly lies, and any early imbalance can be attributed to people simply not knowing how to deal with the changes, what if the swing towards marines doesn't slow down?
Might be that since the patch there's a lot more gorges with better hit reg and combined with marines being over confidant at killing gorges is what the stats are reflecting for now, if that's the case people playing marine side will adapt.
We shouldn't draw conclusions so soon, even based off NS2stats.
Could go either way, we'll see once the dust has settled.
> result is +-25% more skewed than much more general stats covering many thousands of games
> doesn't think its a learn to play issue
idontthinksotim.jpg
I figured it out.... internet explorer wasn't bringing anything up except for a small box and search field when I went to ns2stats.org... switched to firefox and it shows everything.
<!--quoteo(post=2062420:date=Jan 17 2013, 11:41 PM:name=hakenspit)--><div class='quotetop'>QUOTE (hakenspit @ Jan 17 2013, 11:41 PM) <a href="index.php?act=findpost&pid=2062420"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->You say statistics dont lie? Then why is there the famous saying of "Bullsh1t, Lies & Statistics"?
Statistics by them selves mean nothing, raw uncleansed data is totally unusable for anything other than taking up space.<!--QuoteEnd--></div><!--QuoteEEnd-->
There is valuable information that<u><i> can</i></u> be gleaned out from raw data if that data is sufficiently large enough or is taken from a large enough pool. When large amounts of data is taken, the percentage of anomolies such as troll comming becomes miniscule enough to disregard trying to account for it entirely and it also normalizes some variables such as player skill.
The OP's data though is entirely too small. NS2stats is better.
>STATISTICS
fixd it for you
Ok, but it's still a huge sample. By your logic, every poll ever conducted for anything is grossly inaccurate since the conductor didn't ask every person on the planet.
regarding the topic: it doesn't really matter what the winrate is until you take a step back and understand what causes the winrate to be skewed, then you have something significant. otherwise the thread is simply a waste of time discussing points that are covered in introductory statistics.
The difference is that in a poll bias is reduced by methodologically administering polls, while in NS2Stats case it is by whoever feels like installing the mod on their server, which isn't random. God knows what affect that has on the data, but you probably shouldn't disregard it.
Hahaha...very funny.
To say raw data is unbiased ignores the events that led to the data being there.
An example of this is if you have a big 1 off sale, you will see your profits go through the roof for that week.
Well if you just take that weeks worth of raw data...its terribly biased.
Sorry but nothing is unbiased...raw data or not....there are always factors that have influenced (or biased) the outcome/result.
<!--quoteo(post=2062464:date=Jan 18 2013, 04:08 PM:name=Gliss)--><div class='quotetop'>QUOTE (Gliss @ Jan 18 2013, 04:08 PM) <a href="index.php?act=findpost&pid=2062464"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->regarding the topic: it doesn't really matter what the winrate is until you take a step back and understand what causes the winrate to be skewed, then you have something significant. otherwise the thread is simply a waste of time discussing points that are covered in introductory statistics.<!--QuoteEnd--></div><!--QuoteEEnd-->
Yes, its teh taking a step back and understanding what the data says thats the hard part, people want an easy face value answer and there often isn't one....sure there is an answer but its hidden and not sraight forward to see/find.
I play lerk regularly, and it does seem as if I'm getting hit much more often (despite flying quite evasively) in the latest patch.
Methodic polling introduces bias through the subjective method of the person doing the data recording. Natural polling introduces bias through real life and practical influences.
Which sounds more representative?
it's not just that, but anecdotal evidence is very close to useless in NS2 because it has many RTS elements but a large deciding factor of most games and engagements still comes down to individual skill. it doesn't seem very useful to use "I am shooting skulks better / I am getting hit as a skulk more often" as evidence of hit registry being fixed, for example.