|
Post by countlieberkuhn on Apr 21, 2013 6:54:27 GMT -8
Not sure if it's a formatting error when loaded online, but I've noticed that sometimes things aren't being shown as continuous/seperate lines on your playerdata.txt. Liike
Blue6=1|1|1|1|1|1|1|3| NORMALCenter=1|1|1|1|1|1|1|3|NORMAL
Depending on how Bob's coded the text file read-ins, that may be problematic if it's the same on the text file, particularly if he's using the end of line as a indicator for moving onto the next player. I personally work on a 'one-object-per-line' basis when doing this sorta thing as it makes it way easier to work with, so keep an eye on that.
I'm not sure if it would be what causes the errors you're seeing though.
As for Bob's idea about ball control increasing your shot value, I quite like that idea. No team ever seems to hold the ball for more than a few turns anyway, so I don't think it'd ever get too ridiculous, even with a pass-heavy team, but it would reward teamwork and the like more than just having a couple of star players. I'm with Wyvern in thinking that this is probably a better thing to prioritise than the other idea about the prevalent package for each player, although that's also a cool one.
For all rounder, maybe you could have it so that when a player has to take an action, there's a smallish chance (say, 20-25%) that the appropriate stat for the situation will be increased by an amount (maybe 10% of the base stat +2?). It would emphasise that all rounders are supposed to be prepared for any situation, and would give them a chance to get a reasonable boost in a particular stat to make up for the fact that none of their individual stats are likely to be too stellar, since this package would need to be top of the list.
|
|
|
Post by Bob on Apr 21, 2013 9:09:00 GMT -8
Count: It's an online thing and it's pissing me off. Of course I have each person on their own line. It'd be ugly and annoying otherwise. And yet, I guess it comes out as one line. Ugh. That bit of confusion helped nothing. So yes, each person on each line. I'm even breaking up teams by putting the team name up top. Like:
Commies: Red1statsstatsstatsstats Red2statsstatsstatsstats Blahblahblah
Blues: Blue1statsstatsstatsstats Blue2statsstatsstatsstats Blahblahblah
Fleck: Whoa. That output you're getting is WEIRD. I don't know how RightDef is able to block something a jillion times! Very confusing. Does that happen a lot? I haven't seen it in any of my outputs, so I'm curious about that.
I am still struggling with this weird home team advantage thing. I really can't figure out why they have any advantage. The shooting aspect seems to be working perfectly. It calculates the correct distance, uses the correct goalie's stats, and all the functions in the program are bi-directional, so there isn't any "if on this team do one thing, on this team do another." Only thing that's unique to a team is coding their movements, but I THINK that's working fine, too.
By the way, I decided for my shot bonus, the bonus goes up as follows: +1 Survive a tackle +1 Make a pass through a defender (+1 for each defender) +3 Complete a pass
And yes, as Count mentioned, teams don't seem to hold the ball for terribly long anyway. I also slowed down the ball carrier slightly (he moves 90% the distance of everybody else), so more encounters are happening. There were too many "1 on Defense" situations for me and I wanted to see more "2 on Defense" or even "4 on Defense" since that CAN happen, especially if you're not passing.
The ideas for passive abilities I am taking into consideration, but until we resolve this home team bias, I'm waiting on anything new. I wouldn't mind starting up a test tournament though and seeing if the home team really has a big bias in practice. I at least found that when I made a superior team, even when all players were still level 10, the superior team would still win on the visitor's side, so I'm not entirely convinced of a bias, but if we see it in a practice tournament or something, then we will look into it.
So what do you all think? Fake tournament next weekend? I can sim it, too, since I seem to have fewer issues than everybody else, though I have no idea why. I'm thinking it'll just be a practice/scrimmage tournament, so we'll wait on doing anything stamina related.
|
|
|
Post by Fleck on Apr 21, 2013 12:35:54 GMT -8
RightDef's Jillion Block Jutsu (as I am now calling it) didn't show up until Version 2. I... have no idea what's going. He must have been training with a new master.
Whatever the issue is that favors the home team, it has to be something relatively minor, unlike the last time we had this problem, because the away teams ARE still winning on occasion, but there is a clear preference for the home team. Also, the super smash goal fests always happen for the home team, never the away. It's always 20-0 and never 0-20. I thought that maybe when I was doing my "1-2-3-4-5-6" thing I might have been pushing a wrong button and making certain players load twice or something, but I have since attempted to deliberately cause problems and your program won't let me. So I have no idea.
The shot bonus thing sounds good though. And sure, let's have a tourney at some point, although I won't be able to sim it if this problem persists.
|
|
|
Post by Bob on Apr 21, 2013 16:01:51 GMT -8
It's funny because I have NEVER seen an issue like RD's JBJ. I don't have the slightest idea what'd cause that. So odd that a computer, trained to do the exact same thing every time, can produce such different results like it is with all of us. I DO see the preference for team one though. Like you said, I AM seeing both teams winning, but blowouts tend to be in favor of team one. At this point, I'm almost certain it isn't a shooting issue, but maybe their movement is different or they're getting some sort of extra boost or something.
And yes, I've rigged the program to keep you from putting somebody on the same position. You'll also find that if you bench two players, then it won't let you bench any more. Gotta pick SOME combination of six players!
Want me to sim this tournament? Seems that things are going best on my computer at the moment. Then someday we can maybe switch off.
|
|
|
Post by wyvernxk7 on Apr 21, 2013 16:05:59 GMT -8
I had a ton of sequential blocks and tackle failures in some of my games too.
I thought that was just part of it all or an indication that a player really sucked at defending.
|
|
|
Post by Bob on Apr 21, 2013 17:56:35 GMT -8
If they're constantly re-engaging, then yes. But if it's like in the output Fleck provided with this happening:
Blue1's ball at (0,1). Shot Bonus: 0 2 on defense. Blue1* shoots! RightDef gets a hand on it! RightDef gets a hand on it! RightDef gets a hand on it! RightDef gets a hand on it! RightDef gets a hand on it! RightDef gets a hand on it! RightDef gets a hand on it! RightDef gets a hand on it! RightDef gets a hand on it! RightDef gets a hand on it! RightDef gets a hand on it! RightDef gets a hand on it! RightDef gets a hand on it! RightDef gets a hand on it! RightDef gets a hand on it! RightDef gets a hand on it! RightDef gets a hand on it! RightDef gets a hand on it! RightDef grabs the ball!
...then it's a glitch.
|
|
|
Post by wyvernxk7 on Apr 21, 2013 18:30:28 GMT -8
Yeah, glitch.
|
|
|
Post by Fleck on Apr 21, 2013 20:42:16 GMT -8
Ya'll are just jealous of my Jillion Block Jutsu. I got friggin Goro on right defense.
|
|
|
Post by Bob on Apr 21, 2013 21:12:38 GMT -8
|
|
|
Post by wyvernxk7 on Apr 22, 2013 12:40:35 GMT -8
So - tournament results in:
I created two teams: Crushas and The Weak Sheets.
Crushas: Gangsta=3|3|3|3|3|5|3|3|HIGH Damien=3|3|3|3|3|3|5|3|HIGH Skullhead=3|3|3|3|5|3|3|3|HIGH Yeti=3|5|3|3|3|3|3|3|HIGH Wifeswapper=3|3|3|5|3|3|3|3|HIGH SUPERNOVA=3|3|3|3|3|3|3|5|HIGH Earthquake=5|3|3|3|3|3|3|3|HIGH Tupac=3|3|3|3|3|5|3|3|HIGH
The Weak Sheets:
Derek=0|3|0|0|0|0|0|0|HIGH Dr. Doctor=0|0|0|0|0|10|0|0|HIGH Your Mother=1|0|1|1|0|0|0|0|HIGH Greg=1|1|0|0|0|0|0|1|HIGH Fantasy Island=0|0|10|0|0|0|0|0|HIGH Skip=0|0|0|0|0|0|0|3|HIGH Dazed=1|1|1|1|1|1|1|3|HIGH Frnnnnn=1|1|1|1|1|1|3|1|HIGH
Game one: TWS-C: 0-4... So I thought, okay, the away team won! That's a good start.
Game two: TWS-C: 0-0 Hm... That's peculiar. Crushas should have gotten some in.
Game three: C-TWS: 44-0 Well, damn.
|
|
|
Post by Bob on Apr 22, 2013 22:05:12 GMT -8
That's...peculiar, yes. I'll continue to search my program for whatever the heck's going on.
|
|
|
Post by Bob on Apr 25, 2013 19:20:36 GMT -8
If you guys want to try a tournament this weekend, then please send me some team data in a PM before Saturday. I'll try to tweak the program and find this weird team 1 bias before simming, but I'd like to at least give a tournament a shot. Since this'll be a kinda test tournament for just some basic fundamentals, I don't mind if you don't want to use your actual team builds (do any of you mind?). I think for the sake of this tournament, it might be nice to allow a re-design following the tournament, just so you don't build a terrible team because you didn't understand some fundamental idea. So PM me the playerdata and teamdata for any teams you'd like to try out. I won't peek at your stats if you don't want me to. I'll do a quick copy/paste and be done with it!
|
|
|
Post by Fleck on Apr 25, 2013 19:57:22 GMT -8
I've no complaint with starting up a test tourney this weekend. But two things I want to say before we proceed:
1: Since we're still testing the program itself, I think the stamina system and backup players should be ignored for the tourney. For one, it adds unnecessary work, switching people around. For two, a tourney is too short to really get the strategic benefit of a well-planned substitution sequence. And for three, I worry that swapping characters around might introduce variables that would screw with the testing process. Since one of the things we're looking for is a home-team bias, it might screw things up if we change too many things mid-tourney.
2: Due to the aforementioned, "don't want too many variables thing" I'm just gonna submit two fairly generic teams, not my actual team builds. Not that my team builds are unusual or anything, but I don't think the ideas I have in mind for my teams would help the testing process any. So I'll post my team stats here when I stop being so lazy.
|
|
|
Post by Fleck on Apr 25, 2013 21:51:00 GMT -8
Stopped being lazy:
Loveless:
LO=Loveless:Center|Genesis|Terra|Lightning|Cecil|Sabin|Vivi|Hope|Marche
Lightning=0|0|0|0|0|5|5|0|LOW Genesis=3|3|3|0|0|0|0|1|LOW Terra=0|0|0|0|0|5|5|0|LOW Sabin=0|3|3|3|0|0|0|1|LOW Cecil=0|3|3|3|0|0|0|1|LOW Vivi=0|0|2|3|0|0|0|5|LOW Hope=0|0|0|0|0|0|0|0|LOW Marche=0|0|0|0|0|0|0|0|LOW
Guests Who:
GW=Guests Who:Wide|Minwu|Edea|Larsa|Leo|Alma|Olan|FuSoYa|Lebreau Larsa=0|0|0|0|2|3|5|0|HIGH Minwu=0|0|0|0|0|0|5|5|HIGH Edea=0|0|0|0|2|3|5|0|HIGH Alma=0|0|5|5|0|0|0|0|HIGH Leo=0|0|5|5|0|0|0|0|HIGH Olan=5|0|0|0|0|0|0|5|HIGH FuSoYa=0|0|0|0|0|0|0|0|HIGH Lebreau=0|0|0|0|0|0|0|0|HIGH
|
|
|
Post by countlieberkuhn on Apr 25, 2013 23:22:01 GMT -8
Here's my teams! I don't mind my actual ones being used, I doubt I'll make much in the way of modifications to them unless they fail spectacularly, I'm personally more interested in the role-playing aspect of it and making the stats match the characters as much as possible. Although, if either of them ended up performing very well, that'd be a bonus SM7=Sublimely Magnificent 7:NORMAL|Balthier|Seifer|Jecht|Edge|Cid_Highwind|Edgar|Laguna|Snow Jecht=1|0|0|1|1|5|2|0|HIGH Balthier=3|4|0|0|3|0|0|0|NORMAL Seifer=0|0|4|1|0|2|3|0|HIGH Cid_Highwind=3|0|2|3|1|0|0|1|NORMAL Edge=2|0|0|2|2|2|2|0|LOW Edgar=2|2|0|2|0|0|2|2|NORMAL Laguna=3|2|0|0|1|0|4|0|LOW Snow=1|1|1|0|1|1|1|4|HIGH MAG=The Magic Missiles:WIDE|Garnet|Rinoa|Rydia|Yuna|Aeris|Lulu|Selphie|Vanille Rydia=2|0|0|0|1|3|4|0|NORMAL Garnet=3|1|2|2|2|0|0|0|LOW Rinoa=2|0|0|0|5|3|0|0|NORMAL Aeris=2|3|3|0|2|0|0|0|NORMAL Yuna=1|2|0|2|1|0|0|4|LOW Lulu=0|4|4|0|0|0|0|2|NORMAL Selphie=4|0|0|0|2|2|2|0|HIGH Vanille=2|2|2|2|1|0|0|1|NORMAL
|
|