<!--quoteo(post=1829944:date=Feb 5 2011, 09:41 PM:name=ScardyBob)--><div class='quotetop'>QUOTE (ScardyBob @ Feb 5 2011, 09:41 PM) <a href="index.php?act=findpost&pid=1829944"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Anyone know a good way to convert JSON to XML (or another spreadsheet readable format)? I'd like to play around with this data a bit, but I haven't been able to find a good way to get the data into Excel.<!--QuoteEnd--></div><!--QuoteEEnd-->
excel can read .CSV files (Comma Separated Values). It doesn't look like the JSON file would be too hard to translate that into .CSV.
Parse it using javascript/PHP and write it to a file seems like all it would take.
<!--quoteo(post=1829814:date=Feb 5 2011, 12:07 AM:name=MurphyIdiot)--><div class='quotetop'>QUOTE (MurphyIdiot @ Feb 5 2011, 12:07 AM) <a href="index.php?act=findpost&pid=1829814"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->It isn't difficult. I am just not sure what our CPU/bandwidth limits will be like. I can probably add an option to return X number of results or have pages of results.<!--QuoteEnd--></div><!--QuoteEEnd--> More than 100 games would be nice. The pokeball or just numbers works for me. Straight up win percentages is good enough for me.
I would like displaying collection periods. That'd let me know if I'm looking at new data or not when I check in.
<!--quoteo(post=1829993:date=Feb 6 2011, 04:34 AM:name=mattoX)--><div class='quotetop'>QUOTE (mattoX @ Feb 6 2011, 04:34 AM) <a href="index.php?act=findpost&pid=1829993"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->The concept is here, I think that's the main thing... I'm sure when things improve there will be an increase in accuracy.<!--QuoteEnd--></div><!--QuoteEEnd--> True true.
I think all games in beta should upload stats like this. Or maybe all games... oooooh.... data.... the theorycrafting implications are enormous!
<!--quoteo(post=1829958:date=Feb 6 2011, 02:13 AM:name=Tomgupper)--><div class='quotetop'>QUOTE (Tomgupper @ Feb 6 2011, 02:13 AM) <a href="index.php?act=findpost&pid=1829958"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->I had a few minutes spare this morning, so I whipped this up :
Pretty simple for the time being, but it's nice to have some data to play with! I'll keep working on it in my spare time!
- Size of circle indicates length of the game - Green = aliens win - Blue = marines win - Mouseover for some details (date, map etc)
Toggle between arrange by winner / version down the bottom (edit : I just noticed the dates are totally wrong! I must be parsing them incorrectly, let me see what I can do about that) (edit : Dates fixed!)<!--QuoteEnd--></div><!--QuoteEEnd-->
I really like that.... nice way of showing it, very fluid too :)
but with regards to balance 'issues' - I would say there are a few minor tweaks to be made (on both teams) but theres nothing majorly unbalancing..
I have always loved how balanced the game feels, it depends mostly on the players, even if marines are 'weaker' than aliens, you get a good marine team and it wont matter..
Maybe it should be per player? thats the only way to get best balance, keep a track of how that player is doing, if they are dooing poor increse that players armour and damage, if they are doing good decrese..
Yay, data. I threw together a python script which grabs the latest data and then plots the number of wins per team as a function of round duration, eg:
<!--quoteo(post=1829811:date=Feb 4 2011, 09:53 PM:name=6john)--><div class='quotetop'>QUOTE (6john @ Feb 4 2011, 09:53 PM) <a href="index.php?act=findpost&pid=1829811"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->The main problem I see with this is that currently only the last 100 games are played, so if there's a group of aliens that played 10 games in a row and were exceptionally good, that sways the data by 10%
Would it be difficult/cumbersome to allow reporting on a larger group than 100?<!--QuoteEnd--></div><!--QuoteEEnd-->
No. Not if they are using app engine.
They would simply request all the data, process it and then throw it in memcache. If the data in Memcache isn't fresh (aka it is older than UW would like it to be or someone decided to refresh it themselves) the there would be another hit to the datastore.
Another idea is to setup a chron to have the system refresh itself every hour or so and then display to the user whatever is still in the memcache.
READING the data, if properly cached, shouldn't really be an issue. The data doesn't change fast enough and there are only so many queries you can possibly do (and there will probably be a lot of overlap between requests)
I think that upgrades done and structures builded should be used in statistics to know what make a team stronger than the other at a given time in the game even if we can guess it (fade/flametrower)
No answer on my suggestions? I won’t be able to do anything like that with JS.
To further extend my previous suggestions, you could/should/may also set the <!--c1--><div class='codetop'>CODE</div><div class='codemain'><!--ec1-->Access-Control-Allow-Origin<!--c2--></div><!--ec2--> HTML-header to <!--c1--><div class='codetop'>CODE</div><div class='codemain'><!--ec1-->*<!--c2--></div><!--ec2-->. See <a href="https://developer.mozilla.org/En/HTTP_Access_Control#Simple_requests" target="_blank">https://developer.mozilla.org/En/HTTP_Acces...Simple_requests</a>
<a href="http://flo.starshine-online.com/ns2/" target="_blank">http://flo.starshine-online.com/ns2/</a> Another update, the number of game played per hour of the day. (with the start of the game fixing the hour, not the end)
It would be nice if such information was recorded:
# time of research So we know when marines researched flamethrowers. This would allow you to filter by games where flamethrower was researched, and then see what % of those games marines won. And you could see how it affected the length of games, whether flamethrowers increased the length of games or decreased it. If they increased the length of games and didn't help the marines win, then they probably aren't balanced/designed well.
# number of players at the start of the game, number of players at the end of the game, overall average number of players in the game It might not be terribly useful to know that 90% of 4 player games are won by aliens so it would be nice to filter those out.
While I do like the idea, be careful about automated data collection. Too many bots making too many requests, too often, end up causing the server strain they were attempting to avoid in the first place. That being said, I'd be surprised they didn't expect this sort of reaction to the tiny data set.
I'm now not the only one who's mentioned a key concept here: Add a layer of separation in between your external web server requests and your internal JSON queries.
neat idea, but completely unnecessary as of now. still missing alien classes, jet packs, heavy armor, among other things. and are those 1v1 games with someone checking out the new marine/alien commander goodies while the opponent goes to town on the base/hive included? i personally wouldn't take any of that data seriously until after launch.
<!--quoteo(post=1829849:date=Feb 5 2011, 11:40 PM:name=Kissaki)--><div class='quotetop'>QUOTE (Kissaki @ Feb 5 2011, 11:40 PM) <a href="index.php?act=findpost&pid=1829849"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->######, FIFTH REWRITE! -.- I failed with disabled numlock, it always navigated me back -.-
Anyway, here goes again:
<b>Raw JSON returns as content-type text/html, but as it’s pure JSON, could you fix the content-type to JSON?
Also, you may want to make it callback-able, for (easy) JS inclusion.</b>
If you need help on any of that contact me.<!--QuoteEnd--></div><!--QuoteEEnd-->
I'm writing an Android app with animated and interactive graphs, did the brunt of the work tonight and is all working, but not quite ready to show off yet, I'll get it up tomorrow :)
The development time for this game is starting to kiddin me!
Guys make this game playable ... nobody is interested in balancing percentages for a "game" with countless laggs and crashes ... My motivation to keep ns2 in eyes is nearly dead ... looking forward to nuclear dawn.
<!--quoteo(post=1830186:date=Feb 7 2011, 11:35 AM:name=StreifenHirnchen)--><div class='quotetop'>QUOTE (StreifenHirnchen @ Feb 7 2011, 11:35 AM) <a href="index.php?act=findpost&pid=1830186"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->The development time for this game is starting to kiddin me!
Guys make this game playable ... nobody is interested in balancing percentages for a "game" with countless laggs and crashes ... My motivation to keep ns2 in eyes is nearly dead ... looking forward to nuclear dawn.
Rest in pieces - NS , sad but true story ;(<!--QuoteEnd--></div><!--QuoteEEnd--> I know what you feel, i guess the whole community is a bit frustrated but they are just a few guys and they want make a great game, there tiny details in the mod already. It takes time to polish all, in my mind they could make a ns2 source (which is just a port with better graphics, physics and stuff) and then a ns3 with their spark engine what they did now, so they have all the time and money from ns2 source to develop ns3 (if valve don't want to much money for the engine). Then start pre-order after its mostly beta. I will not "talk it nice" so my mind is ns2 don't realy feel good as ns1, it feels a bit mainstream, but its not done yet, thats the problem, you don't have the nice voices of aliens*, marines(*structure under attack), the onos, the heavy and jetpacks and of course the chambers and gorge is a bit useless without building a harvester. But dynamic infestation comming soon and some fixes, so it keeps the fans "on the line". Just have a bit patience, you have the game i guess, try it some days later. Friends of me said they will only play the game if its finished, they have the game already, they try it 5 minues and its not playable, so they stop playing it. If you don't like the actual "state" just wait and don't be mad. Its fully moddable, i realy hope for a ns1 mod, because NS1 is still the best damn mod., and with new graphics and physics and dynamic infestation (just for the optic) it will be the bomb. I love the chamber building gorge, that is what a gorge feel, a builder! Not the lousy hydra spamming fool ^^
<!--quoteo(post=1830186:date=Feb 7 2011, 08:35 PM:name=StreifenHirnchen)--><div class='quotetop'>QUOTE (StreifenHirnchen @ Feb 7 2011, 08:35 PM) <a href="index.php?act=findpost&pid=1830186"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Guys make this game playable ... nobody is interested in balancing percentages for a "game" with countless laggs and crashes ... My motivation to keep ns2 in eyes is nearly dead ... looking forward to nuclear dawn.<!--QuoteEnd--></div><!--QuoteEEnd-->Different developers have different jobs. You wouldn't ask Cory to fix engine bugs, and you wouldn't ask Max to draw up concept art.
<a href="http://flo.starshine-online.com/ns2/" target="_blank">http://flo.starshine-online.com/ns2/</a> I've added 2 more graphs : - The first one is the number of game played depending of the day of the week (useless for now because i collect data since late saturday). - The second one is the % of game played depending of the length and the map (the top filter is affecting this graph too).
With a 20mins step, you can see that rockdown and tram have approx the same % of game length.
<!--quoteo(post=1830099:date=Feb 6 2011, 11:23 PM:name=RaydenUni)--><div class='quotetop'>QUOTE (RaydenUni @ Feb 6 2011, 11:23 PM) <a href="index.php?act=findpost&pid=1830099"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->It would be nice if such information was recorded:
# time of research So we know when marines researched flamethrowers. This would allow you to filter by games where flamethrower was researched, and then see what % of those games marines won. And you could see how it affected the length of games, whether flamethrowers increased the length of games or decreased it. If they increased the length of games and didn't help the marines win, then they probably aren't balanced/designed well.
# number of players at the start of the game, number of players at the end of the game, overall average number of players in the game It might not be terribly useful to know that 90% of 4 player games are won by aliens so it would be nice to filter those out.
# when <b>hives go up and hives go down</b>
# resource income over time<!--QuoteEnd--></div><!--QuoteEEnd-->
tide comes in, and tide goes out sun comes up, sun goes down never a miscommunication.
<!--quoteo(post=1829821:date=Feb 4 2011, 09:31 PM:name=Raza.)--><div class='quotetop'>QUOTE (Raza. @ Feb 4 2011, 09:31 PM) <a href="index.php?act=findpost&pid=1829821"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Does it record the number of players?<!--QuoteEnd--></div><!--QuoteEEnd-->
It currently doesn't but that is a good idea. I'll put that on my to do list.
<!--quoteo(post=1829829:date=Feb 4 2011, 11:33 PM:name=Revenge)--><div class='quotetop'>QUOTE (Revenge @ Feb 4 2011, 11:33 PM) <a href="index.php?act=findpost&pid=1829829"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Have you tried experimenting with <a href="http://www.varnish-cache.org/" target="_blank">varnishing</a> the output?<!--QuoteEnd--></div><!--QuoteEEnd--> <!--quoteo(post=1829869:date=Feb 5 2011, 09:55 AM:name=Vic)--><div class='quotetop'>QUOTE (Vic @ Feb 5 2011, 09:55 AM) <a href="index.php?act=findpost&pid=1829869"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->How about caching them for half an hour when a filtered request is made?<!--QuoteEnd--></div><!--QuoteEEnd-->
App Engine has support for such a mechanism. I will look into it as well as allowing the user to query data "pages".
<!--quoteo(post=1829849:date=Feb 5 2011, 05:40 AM:name=Kissaki)--><div class='quotetop'>QUOTE (Kissaki @ Feb 5 2011, 05:40 AM) <a href="index.php?act=findpost&pid=1829849"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Raw JSON returns as content-type text/html, but as it’s pure JSON, could you fix the content-type to JSON?
Also, you may want to make it callback-able, for (easy) JS inclusion.
...you could/should/may also set the <!--c1--><div class='codetop'>CODE</div><div class='codemain'><!--ec1-->Access-Control-Allow-Origin<!--c2--></div><!--ec2--> HTML-header to <!--c1--><div class='codetop'>CODE</div><div class='codemain'><!--ec1-->*<!--c2--></div><!--ec2-->.<!--QuoteEnd--></div><!--QuoteEEnd-->
Added to the list.
<!--quoteo(post=1829877:date=Feb 5 2011, 10:19 AM:name=Racer1)--><div class='quotetop'>QUOTE (Racer1 @ Feb 5 2011, 10:19 AM) <a href="index.php?act=findpost&pid=1829877"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Also, would like the option to specify both minimum and maximum game time, to make it easy to filter out very short games.<!--QuoteEnd--></div><!--QuoteEEnd-->
Added.
<!--quoteo(post=1829958:date=Feb 5 2011, 07:13 PM:name=Tomgupper)--><div class='quotetop'>QUOTE (Tomgupper @ Feb 5 2011, 07:13 PM) <a href="index.php?act=findpost&pid=1829958"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->I had a few minutes spare this morning, so I whipped this up :<!--QuoteEnd--></div><!--QuoteEEnd-->
Very cool!
<!--quoteo(post=1830006:date=Feb 6 2011, 05:04 AM:name=a_roobs)--><div class='quotetop'>QUOTE (a_roobs @ Feb 6 2011, 05:04 AM) <a href="index.php?act=findpost&pid=1830006"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Yay, data. I threw together a python script which grabs the latest data and then plots the number of wins per team as a function of round duration, eg:<!--QuoteEnd--></div><!--QuoteEEnd-->
Rad! These are just the sort of things I was hoping to see :)
<!--quoteo(post=1830099:date=Feb 6 2011, 04:23 PM:name=RaydenUni)--><div class='quotetop'>QUOTE (RaydenUni @ Feb 6 2011, 04:23 PM) <a href="index.php?act=findpost&pid=1830099"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->It would be nice if such information was recorded:<!--QuoteEnd--></div><!--QuoteEEnd-->
Good ideas. Added.
<!--quoteo(post=1830129:date=Feb 6 2011, 07:18 PM:name=DJPenguin)--><div class='quotetop'>QUOTE (DJPenguin @ Feb 6 2011, 07:18 PM) <a href="index.php?act=findpost&pid=1830129"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->neat idea, but completely unnecessary as of now.<!--QuoteEnd--></div><!--QuoteEEnd-->
It didn't take me too long to add and was a nice little break from normal development. I think it will prove to be more and more useful as time goes on and more data is collected and more people figure out ways to visualize and interpret the data.
PlasmaJoin Date: 2003-04-26Member: 15855Members, Constellation, Squad Five Blue
edited February 2011
Heat map data definitely sounds cool (for where people die often) for each map.
I think if you tracked metadata associated with each player's death, game balance can be studied and we can see whether armor upgrades really have any benefit, or whether gorge's are most vulnerable to players with flame throwers, etc.
Each time a player is killed, perhaps you should also log along with that kill: * The armor/HP amount each player had before the fight/encounter * What armor upgrades were applied * What weapon each player had * The race of the alien
This way you could see how effective (or not) armor upgrades are, or which weapon wins best against different races etc over time.
<!--quoteo(post=1830315:date=Feb 7 2011, 09:21 PM:name=MurphyIdiot)--><div class='quotetop'>QUOTE (MurphyIdiot @ Feb 7 2011, 09:21 PM) <a href="index.php?act=findpost&pid=1830315"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Good ideas. Added.<!--QuoteEnd--></div><!--QuoteEEnd--> When you say "added", it's added in a todolist or added right now in the json ? :D
ScardyBobScardyBobJoin Date: 2009-11-25Member: 69528Forum Admins, Forum Moderators, NS2 Playtester, Squad Five Blue, Reinforced - Shadow, WC 2013 - Shadow
<!--quoteo(post=1830298:date=Feb 7 2011, 11:46 AM:name=Melatonin)--><div class='quotetop'>QUOTE (Melatonin @ Feb 7 2011, 11:46 AM) <a href="index.php?act=findpost&pid=1830298"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->tide comes in, and tide goes out sun comes up, sun goes down never a miscommunication.<!--QuoteEnd--></div><!--QuoteEEnd-->
Ok, Bill O!
From <a href="http://flo.starshine-online.com/ns2/" target="_blank">http://flo.starshine-online.com/ns2/</a> seems to be a decent correlation between map length and marine victories. I'd suspect that it takes marines longer to win rather than long matches lead to marine victories, though.
<!--quoteo(post=1830360:date=Feb 7 2011, 05:06 PM:name=Dalai-Lama)--><div class='quotetop'>QUOTE (Dalai-Lama @ Feb 7 2011, 05:06 PM) <a href="index.php?act=findpost&pid=1830360"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->When you say "added", it's added in a todolist or added right now in the json ? :D<!--QuoteEnd--></div><!--QuoteEEnd-->
Comments
excel can read .CSV files (Comma Separated Values). It doesn't look like the JSON file would be too hard to translate that into .CSV.
Parse it using javascript/PHP and write it to a file seems like all it would take.
<!--quoteo(post=1829814:date=Feb 5 2011, 12:07 AM:name=MurphyIdiot)--><div class='quotetop'>QUOTE (MurphyIdiot @ Feb 5 2011, 12:07 AM) <a href="index.php?act=findpost&pid=1829814"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->It isn't difficult. I am just not sure what our CPU/bandwidth limits will be like. I can probably add an option to return X number of results or have pages of results.<!--QuoteEnd--></div><!--QuoteEEnd-->
More than 100 games would be nice. The pokeball or just numbers works for me. Straight up win percentages is good enough for me.
I would like displaying collection periods. That'd let me know if I'm looking at new data or not when I check in.
True true.
I think all games in beta should upload stats like this. Or maybe all games... oooooh.... data.... the theorycrafting implications are enormous!
I'll be in my bunk.
<a href="http://tenfiftyfour.com/axperiment/experiments/05_ns2/" target="_blank">http://tenfiftyfour.com/axperiment/experiments/05_ns2/</a>
Pretty simple for the time being, but it's nice to have some data to play with! I'll keep working on it in my spare time!
- Size of circle indicates length of the game
- Green = aliens win
- Blue = marines win
- Mouseover for some details (date, map etc)
Toggle between arrange by winner / version down the bottom
(edit : I just noticed the dates are totally wrong! I must be parsing them incorrectly, let me see what I can do about that)
(edit : Dates fixed!)<!--QuoteEnd--></div><!--QuoteEEnd-->
I really like that.... nice way of showing it, very fluid too :)
but with regards to balance 'issues' - I would say there are a few minor tweaks to be made (on both teams) but theres nothing majorly unbalancing..
I have always loved how balanced the game feels, it depends mostly on the players, even if marines are 'weaker' than aliens, you get a good marine team and it wont matter..
Maybe it should be per player? thats the only way to get best balance, keep a track of how that player is doing, if they are dooing poor increse that players armour and damage, if they are doing good decrese..
Just an idea :)
<img src="http://fcostin.github.com/lolgraphs/imbalance_versus_time.png" border="0" class="linked-image" />
<a href="http://github.com/fcostin/lolgraphs/blob/ns2_stats/main.py" target="_blank">here's the python script itself</a>
Would it be difficult/cumbersome to allow reporting on a larger group than 100?<!--QuoteEnd--></div><!--QuoteEEnd-->
No. Not if they are using app engine.
They would simply request all the data, process it and then throw it in memcache. If the data in Memcache isn't fresh (aka it is older than UW would like it to be or someone decided to refresh it themselves) the there would be another hit to the datastore.
Another idea is to setup a chron to have the system refresh itself every hour or so and then display to the user whatever is still in the memcache.
READING the data, if properly cached, shouldn't really be an issue. The data doesn't change fast enough and there are only so many queries you can possibly do (and there will probably be a lot of overlap between requests)
I'm starting to collect data, this is just a first graph showing the % of victories for marines depending to the length of the game.
PS : sorry for my english, so if you can correct my spelling :)
I'm starting to collect data, this is just a first graph showing the % of victories for marines depending to the length of the game.
PS : sorry for my english, so if you can correct my spelling :)<!--QuoteEnd--></div><!--QuoteEEnd-->
very informative, nice one.
- Version
- Map
- Minutes step
So with the 20 mins step, you can see easily that the much longer is the game, better are the chance for marines to win.
I won’t be able to do anything like that with JS.
To further extend my previous suggestions, you could/should/may also set the <!--c1--><div class='codetop'>CODE</div><div class='codemain'><!--ec1-->Access-Control-Allow-Origin<!--c2--></div><!--ec2--> HTML-header to <!--c1--><div class='codetop'>CODE</div><div class='codemain'><!--ec1-->*<!--c2--></div><!--ec2-->.
See <a href="https://developer.mozilla.org/En/HTTP_Access_Control#Simple_requests" target="_blank">https://developer.mozilla.org/En/HTTP_Acces...Simple_requests</a>
Another update, the number of game played per hour of the day.
(with the start of the game fixing the hour, not the end)
<a href="http://ns2.unfine.de/" target="_blank">http://ns2.unfine.de/</a>
(These are site stats only!)
Collect the data is a really good idea. Might start doing that as well now..
# time of research
So we know when marines researched flamethrowers. This would allow you to filter by games where flamethrower was researched, and then see what % of those games marines won. And you could see how it affected the length of games, whether flamethrowers increased the length of games or decreased it. If they increased the length of games and didn't help the marines win, then they probably aren't balanced/designed well.
# number of players at the start of the game, number of players at the end of the game, overall average number of players in the game
It might not be terribly useful to know that 90% of 4 player games are won by aliens so it would be nice to filter those out.
# when hives go up and hives go down
# resource income over time
I'm now not the only one who's mentioned a key concept here: Add a layer of separation in between your external web server requests and your internal JSON queries.
Anyway, here goes again:
<b>Raw JSON returns as content-type text/html, but as it’s pure JSON, could you fix the content-type to JSON?
Also, you may want to make it callback-able, for (easy) JS inclusion.</b>
If you need help on any of that contact me.<!--QuoteEnd--></div><!--QuoteEEnd-->
+ 1
Guys make this game playable ... nobody is interested in balancing percentages for a "game" with countless laggs and crashes ...
My motivation to keep ns2 in eyes is nearly dead ... looking forward to nuclear dawn.
Rest in pieces - NS , sad but true story ;(
Guys make this game playable ... nobody is interested in balancing percentages for a "game" with countless laggs and crashes ...
My motivation to keep ns2 in eyes is nearly dead ... looking forward to nuclear dawn.
Rest in pieces - NS , sad but true story ;(<!--QuoteEnd--></div><!--QuoteEEnd-->
I know what you feel, i guess the whole community is a bit frustrated but they are just a few guys and they want make a great game, there tiny details in the mod already.
It takes time to polish all, in my mind they could make a ns2 source (which is just a port with better graphics, physics and stuff) and then a ns3 with their spark engine what they did now, so they have all the time and money from ns2 source to develop ns3 (if valve don't want to much money for the engine).
Then start pre-order after its mostly beta.
I will not "talk it nice" so my mind is ns2 don't realy feel good as ns1, it feels a bit mainstream, but its not done yet, thats the problem, you don't have the nice voices of aliens*, marines(*structure under attack), the onos, the heavy and jetpacks and of course the chambers and gorge is a bit useless without building a harvester.
But dynamic infestation comming soon and some fixes, so it keeps the fans "on the line".
Just have a bit patience, you have the game i guess, try it some days later.
Friends of me said they will only play the game if its finished, they have the game already, they try it 5 minues and its not playable, so they stop playing it.
If you don't like the actual "state" just wait and don't be mad.
Its fully moddable, i realy hope for a ns1 mod, because NS1 is still the best damn mod., and with new graphics and physics and dynamic infestation (just for the optic) it will be the bomb.
I love the chamber building gorge, that is what a gorge feel, a builder!
Not the lousy hydra spamming fool ^^
My motivation to keep ns2 in eyes is nearly dead ... looking forward to nuclear dawn.<!--QuoteEnd--></div><!--QuoteEEnd-->Different developers have different jobs. You wouldn't ask Cory to fix engine bugs, and you wouldn't ask Max to draw up concept art.
I've added 2 more graphs :
- The first one is the number of game played depending of the day of the week (useless for now because i collect data since late saturday).
- The second one is the % of game played depending of the length and the map (the top filter is affecting this graph too).
With a 20mins step, you can see that rockdown and tram have approx the same % of game length.
# time of research
So we know when marines researched flamethrowers. This would allow you to filter by games where flamethrower was researched, and then see what % of those games marines won. And you could see how it affected the length of games, whether flamethrowers increased the length of games or decreased it. If they increased the length of games and didn't help the marines win, then they probably aren't balanced/designed well.
# number of players at the start of the game, number of players at the end of the game, overall average number of players in the game
It might not be terribly useful to know that 90% of 4 player games are won by aliens so it would be nice to filter those out.
# when <b>hives go up and hives go down</b>
# resource income over time<!--QuoteEnd--></div><!--QuoteEEnd-->
tide comes in, and tide goes out
sun comes up, sun goes down
never a miscommunication.
It currently doesn't but that is a good idea. I'll put that on my to do list.
<!--quoteo(post=1829829:date=Feb 4 2011, 11:33 PM:name=Revenge)--><div class='quotetop'>QUOTE (Revenge @ Feb 4 2011, 11:33 PM) <a href="index.php?act=findpost&pid=1829829"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Have you tried experimenting with <a href="http://www.varnish-cache.org/" target="_blank">varnishing</a> the output?<!--QuoteEnd--></div><!--QuoteEEnd-->
<!--quoteo(post=1829869:date=Feb 5 2011, 09:55 AM:name=Vic)--><div class='quotetop'>QUOTE (Vic @ Feb 5 2011, 09:55 AM) <a href="index.php?act=findpost&pid=1829869"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->How about caching them for half an hour when a filtered request is made?<!--QuoteEnd--></div><!--QuoteEEnd-->
App Engine has support for such a mechanism. I will look into it as well as allowing the user to query data "pages".
<!--quoteo(post=1829849:date=Feb 5 2011, 05:40 AM:name=Kissaki)--><div class='quotetop'>QUOTE (Kissaki @ Feb 5 2011, 05:40 AM) <a href="index.php?act=findpost&pid=1829849"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Raw JSON returns as content-type text/html, but as it’s pure JSON, could you fix the content-type to JSON?
Also, you may want to make it callback-able, for (easy) JS inclusion.
...you could/should/may also set the <!--c1--><div class='codetop'>CODE</div><div class='codemain'><!--ec1-->Access-Control-Allow-Origin<!--c2--></div><!--ec2--> HTML-header to <!--c1--><div class='codetop'>CODE</div><div class='codemain'><!--ec1-->*<!--c2--></div><!--ec2-->.<!--QuoteEnd--></div><!--QuoteEEnd-->
Added to the list.
<!--quoteo(post=1829877:date=Feb 5 2011, 10:19 AM:name=Racer1)--><div class='quotetop'>QUOTE (Racer1 @ Feb 5 2011, 10:19 AM) <a href="index.php?act=findpost&pid=1829877"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Also, would like the option to specify both minimum and maximum game time, to make it easy to filter out very short games.<!--QuoteEnd--></div><!--QuoteEEnd-->
Added.
<!--quoteo(post=1829958:date=Feb 5 2011, 07:13 PM:name=Tomgupper)--><div class='quotetop'>QUOTE (Tomgupper @ Feb 5 2011, 07:13 PM) <a href="index.php?act=findpost&pid=1829958"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->I had a few minutes spare this morning, so I whipped this up :<!--QuoteEnd--></div><!--QuoteEEnd-->
Very cool!
<!--quoteo(post=1830006:date=Feb 6 2011, 05:04 AM:name=a_roobs)--><div class='quotetop'>QUOTE (a_roobs @ Feb 6 2011, 05:04 AM) <a href="index.php?act=findpost&pid=1830006"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Yay, data. I threw together a python script which grabs the latest data and then plots the number of wins per team as a function of round duration, eg:<!--QuoteEnd--></div><!--QuoteEEnd-->
Awesome!
<!--quoteo(post=1830039:date=Feb 6 2011, 08:17 AM:name=Dalai-Lama)--><div class='quotetop'>QUOTE (Dalai-Lama @ Feb 6 2011, 08:17 AM) <a href="index.php?act=findpost&pid=1830039"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec--><a href="http://flo.starshine-online.com/ns2/" target="_blank">http://flo.starshine-online.com/ns2/</a><!--QuoteEnd--></div><!--QuoteEEnd-->
Rad! These are just the sort of things I was hoping to see :)
<!--quoteo(post=1830099:date=Feb 6 2011, 04:23 PM:name=RaydenUni)--><div class='quotetop'>QUOTE (RaydenUni @ Feb 6 2011, 04:23 PM) <a href="index.php?act=findpost&pid=1830099"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->It would be nice if such information was recorded:<!--QuoteEnd--></div><!--QuoteEEnd-->
Good ideas. Added.
<!--quoteo(post=1830129:date=Feb 6 2011, 07:18 PM:name=DJPenguin)--><div class='quotetop'>QUOTE (DJPenguin @ Feb 6 2011, 07:18 PM) <a href="index.php?act=findpost&pid=1830129"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->neat idea, but completely unnecessary as of now.<!--QuoteEnd--></div><!--QuoteEEnd-->
It didn't take me too long to add and was a nice little break from normal development. I think it will prove to be more and more useful as time goes on and more data is collected and more people figure out ways to visualize and interpret the data.
Thanks for all the comments everyone!
I think if you tracked metadata associated with each player's death, game balance can be studied and we can see whether armor upgrades really have any benefit, or whether gorge's are most vulnerable to players with flame throwers, etc.
Each time a player is killed, perhaps you should also log along with that kill:
* The armor/HP amount each player had before the fight/encounter
* What armor upgrades were applied
* What weapon each player had
* The race of the alien
This way you could see how effective (or not) armor upgrades are, or which weapon wins best against different races etc over time.
When you say "added", it's added in a todolist or added right now in the json ? :D
sun comes up, sun goes down
never a miscommunication.<!--QuoteEnd--></div><!--QuoteEEnd-->
Ok, Bill O!
From <a href="http://flo.starshine-online.com/ns2/" target="_blank">http://flo.starshine-online.com/ns2/</a> seems to be a decent correlation between map length and marine victories. I'd suspect that it takes marines longer to win rather than long matches lead to marine victories, though.
To do list :)
I am fast, but not that fast...