Warrior General Games

Originally posted by brady2moss2endzone
Oh, cool then

I bet most people on here have invites to the testing site, LOL, im just a beginner to you lot, just wanna learn about the technical side of GLB and how it works

Lol, you joined before me. I don't think there are more then 20-30 testers out of the tens of thousands active GLB users, tbh.

Originally posted by Staz
Any thoughts about uploading random players from the live server and having them as test subjects? Give them random names, so people don't know who they are, and assign them to random teams. Would pull players from the Pro Leagues and World League, so we know what upper leveled players would be. Then, also use players in the 40-50 levels, and 20-30 levels. Would be able to determine whether the performance drop off is 'reasonable'.

This would also dispute any claims that the test builds just aren't "realistic" and could fix the "It works on the test server, but not the live server" issues.

Afaik, we can't lift things from the live server and place them on the test server. We'd have to recreate the player/AI/team, etc. I have used builds close to actual players before while testing some bug issues, but not exact replicas.

Originally posted by kardiackids2007
Lol, you joined before me. I don't think there are more then 20-30 testers out of the tens of thousands active GLB users, tbh.

LOL, Yeah but ive never looked at how the game works 'behind the scenes' and how the technical side of it works

Originally posted by cwisler
Originally posted by Staz

Any thoughts about uploading random players from the live server and having them as test subjects? Give them random names, so people don't know who they are, and assign them to random teams. Would pull players from the Pro Leagues and World League, so we know what upper leveled players would be. Then, also use players in the 40-50 levels, and 20-30 levels. Would be able to determine whether the performance drop off is 'reasonable'.

This would also dispute any claims that the test builds just aren't "realistic" and could fix the "It works on the test server, but not the live server" issues.

Afaik, we can't lift things from the live server and place them on the test server. We'd have to recreate the player/AI/team, etc. I have used builds close to actual players before while testing some bug issues, but not exact replicas.

Bort can pull directly from the live server.

Originally posted by Adderfist
Originally posted by cwisler

Originally posted by Staz


Any thoughts about uploading random players from the live server and having them as test subjects? Give them random names, so people don't know who they are, and assign them to random teams. Would pull players from the Pro Leagues and World League, so we know what upper leveled players would be. Then, also use players in the 40-50 levels, and 20-30 levels. Would be able to determine whether the performance drop off is 'reasonable'.

This would also dispute any claims that the test builds just aren't "realistic" and could fix the "It works on the test server, but not the live server" issues.

Afaik, we can't lift things from the live server and place them on the test server. We'd have to recreate the player/AI/team, etc. I have used builds close to actual players before while testing some bug issues, but not exact replicas.

Bort can pull directly from the live server.

And he won't do it because of the shitstorm that would follow it immediately.

He's said that's he's pulled players onto his personal developer box in the past, but that he will not put people on the test server.

Originally posted by pottsman
And he won't do it because of the shitstorm that would follow it immediately.

He's said that's he's pulled players onto his personal developer box in the past, but that he will not put people on the test server.

Perhaps if Bort imported the entire roster of all the pro leagues and World League, hid the builds, and ran about 10-15 tests overnight without the help of the testers, he could get some results?

This would eliminate the 'cheating' cries because the testers themselves wouldn't be involved in any of this. The results would be public, so people could see exactly how things went, and the build would be private. This would prevent guys from saying "hey, this worked, so I'll build my guy like that". To me, that seems like a solid idea, as it removes all of the 'shitstorm' ammo from the equation.

Keeping the builds private, games public, and testers out of the game planning seems solid. Of course, they'd be asked to watch game 'film' and point out bugs and weird 'behavior'.

On top of that, any hope to include the general user base in watching these test games? Possibly have 5 testers run two games each, and then post the links to the games on the test forum and ask for feedback?

Originally posted by Staz
Originally posted by pottsman

And he won't do it because of the shitstorm that would follow it immediately.

He's said that's he's pulled players onto his personal developer box in the past, but that he will not put people on the test server.

Perhaps if Bort imported the entire roster of all the pro leagues and World League, hid the builds, and ran about 10-15 tests overnight without the help of the testers, he could get some results?

This would eliminate the 'cheating' cries because the testers themselves wouldn't be involved in any of this. The results would be public, so people could see exactly how things went, and the build would be private. This would prevent guys from saying "hey, this worked, so I'll build my guy like that". To me, that seems like a solid idea, as it removes all of the 'shitstorm' ammo from the equation.

Keeping the builds private, games public, and testers out of the game planning seems solid. Of course, they'd be asked to watch game 'film' and point out bugs and weird 'behavior'.

On top of that, any hope to include the general user base in watching these test games? Possibly have 5 testers run two games each, and then post the links to the games on the test forum and ask for feedback?

Keeping the builds private is going to cause all sorts of problems. Without access to the builds, we have no idea what is overpowered/underpowered. Knowing the builds is a crucial part to viewing a test, or to at least have an idea about why a player did something. Testing without showing builds to anyone is the same as just running a normal WL game. Bort knows, and again, he's said he does that on his own private developer box, but he does not do it on the test server where we can watch.

Fair enough. I'm just trying to figure out a way to import normal builds from the game to test these things. The testers all have stated that building an entire roster is a huge pain in the ass, which makes perfect sense. There has to be a way to use real players on the test server and do so in a fair and public manner to reduce the "omg cheaterrrsss" threads. Of course, there will always be individuals who just like to stir up drama, but there has to be a way to do things in a fair and public way.

Originally posted by Staz
Fair enough. I'm just trying to figure out a way to import normal builds from the game to test these things. The testers all have stated that building an entire roster is a huge pain in the ass, which makes perfect sense. There has to be a way to use real players on the test server and do so in a fair and public manner to reduce the "omg cheaterrrsss" threads. Of course, there will always be individuals who just like to stir up drama, but there has to be a way to do things in a fair and public way.

Hazy made a team of what he thought was good builds, but he didn't know that equipment didn't transfer over if we reset teams, fml

Oh jeez. We need a volunteer to spend the next week developing two teams of Level 72 players, two teams of Level 64 players, two teams of Level 56, two teams of Level 48, two teams of Level 40, two teams of Level 32 and two teams of Level 24 players.


Or maybe get 4 testers to do that. 14 teams each, let's say 42 players each team. Hm. 714 players for each tester. Shouldn't be tough.

-A little bit about the ineffectiveness of pass rushing (brief discussion on why they are high on the test server, and non-existent on the live server...needs more discussion)
Anything to do with the fact that DLs on the tes server have higher attributes/better builds than the OLs, whereas on the live server itīs the other way round? seems pretty obvious to me.
Edit: just read stazī post from page 1 re this and totally agree. More testing should be done with real players.

-Discussion about increasing the energy drain, making stamina more valuable.
This sounds very suspicous to me. If I were Bort, I would def take the time to check the teams/players of the guys pushing for this to see whether they stand to gain from it. Ie: any testers out there whose team(s)/player(s) have high stamina?


And I forgot: thanks a lot, DB, much appreciated.

Originally posted by Staz
Oh jeez. We need a volunteer to spend the next week developing two teams of Level 72 players, two teams of Level 64 players, two teams of Level 56, two teams of Level 48, two teams of Level 40, two teams of Level 32 and two teams of Level 24 players.


Or maybe get 4 testers to do that. 14 teams each, let's say 42 players each team. Hm. 714 players for each tester. Shouldn't be tough.

I have loads of time on GLB in the next week or 2.......so i could help

I just need some basic info on creating and developing the players........once i get it on how it basically works, i can get right on it

If you need any help of course........

Originally posted by Kevin Smith
-Discussion about increasing the energy drain, making stamina more valuable.
This sounds very suspicous to me. If I were Bort, I would def take the time to check the teams/players of the guys pushing for this to see whether they stand to gain from it. Ie: any testers out there whose team(s)/player(s) have high stamina?


Actually, that seems to be a Bort-bias. He was active in the Project Epic Suggestions PF last night, tossing out a bunch of different ideas on things (check Suggestions for a few quotes, and the next Test Server update should include some of them), but at one point or another he commented on something with "Stamina needs to be more important, not less". I think it was in a thread about re-working/adding some VAs. But in general, I think he wants the stamina, not some random tester with staminators. Though if someone has those, of course they'd agree.

New smilies are coming too!

Originally posted by kalkmanc
New smilies are coming too!

maybe we could get a smiley of Bort's picture