Warrior General Games

Originally posted by Novus
I'm aware most people shoot for 40 to get OTK... been doing that myself for a while.

Catch-22 was building dots with low Confidence at a time when MOST top dot-builders were building dots with low or low-ish Confidence, so his low-Confidence dots could get away with it because they were playing in a low-Confidence environment.

I've seen people explore the effects of Confidence over the years, but your post made me think of something: haven't those tests generally been focused on regular season games and scrims? If Confidence has a larger impact in the playoffs, the only way to test that is by running experiments with playoff games -- which owners don't generally don't want to do because they understandably want to win playoff games. And playoff games are, by their nature, difficult to experiment with -- there's no reliable way to select particular opponents or matchups short of full-on manipulatioin of the standings. And when two teams play each other in the playoffs once, that's it -- there's no way to repeat the experiment, unless you can keep two Plateau'd rosters completely unchanged for a second season in a row, get both teams to meet in the playoffs again, and have them use the same gameplans from the season before.

If Confidence actually does have a larger impact in the playoffs than it does in the regular season, you won't see that impact in regular season games. A squad of otherwise very well built dots with low Confidence could go 16-0 every regular season, and people can point to that and say, "See? There's proof you don't need Confidence." But it's actually just proof that you don't need Confidence in the regular season. And we've all seen teams that regularly dominate in the regular season and regularly fall flat on their face in the playoffs. It's easy to assume it's because they're bad at gameplanning when it matters, and there's something to that, of course -- but what if it's not the only factor?

but catch22 also had insight as to how the SIM worked, which may be why he ignored confidence

Originally posted by reddogrw
but catch22 also had insight as to how the SIM worked, which may be why he ignored confidence

Very true -- but he also had insight into how other agents were building their dots, so he would've known what level of Confidence he could get away with having. And Bort said -- on multiple occasions in the Q&As -- that agents were under-valuing Confidence.

lolconfidence

proof that overconfidence is not good

this cpu team turned it on just before and during the playoffs: https://glb.warriorgeneral.com/game/team.pl?team_id=381

you might be on to something about "playoff sim"

Originally posted by SeattleNiner
proof that overconfidence is not good

I see what you did there.

Forget single attributes for a sec, wasn't there something implemented that makes full cpu teams scale up to near-human-dot levels, to the point where gameplanning and tactics are necessary?


Mostly CPU team https://glb.warriorgeneral.com/game/team.pl?team_id=196
Week 5
Win 24-10 vs Beltway Bandits (human team)

Week 14
Loss 9-12 vs Beltway Bandits (human team)

Two competitive games right? Well..

Playoff week 1
Win 114-0 vs Beltway Bandits


The only difference is that the mostly CPU team lost their last human player before the playoffs, making them full CPU for week 1 of playoffs. I know this because my HB was that last human player and joined Beltway day 32 hoping to make a playoff run




And check this out.. two games against one opponent
Week 13 https://glb.warriorgeneral.com/game/game.pl?game_id=3003670 (mostly CPU team gets blown out)
Playoff week 2 https://glb.warriorgeneral.com/game/game.pl?game_id=3009731 (full CPU team wins in blowout)

I wonder if the penalty that applies to human owned/controlled teams with too many CPU dots works in reverse (becomes a bonus) for teams that are CPU owned/controlled with human dots on them?

Originally posted by .spider.
https://glb.warriorgeneral.com/game/game.pl?game_id=3009626

But they did also beat me both reg season games as well....and no i havent worked on confidence yet....

I guess my only gripe with losing to CPU teams is you cant really "gameplan" against them....

I honestly thought you guys were playing possum so i couldnt pick up on anything with reg season losses

Originally posted by psi
Forget single attributes for a sec, wasn't there something implemented that makes full cpu teams scale up to near-human-dot levels, to the point where gameplanning and tactics are necessary?

All CPU dots scale up or down to the league's average level. That's true for all CPU dots, regardless of the number of human dots or the number of other CPU dots on the roster.

However, what people fail to consider is that for HUMAN-owned teams, if you have more than X number of CPU dots on a HUMAN-owned team (I think X = 10), the CPU dots are penalized. This is to encourage human owners to sign human dots, rather than just buying a team and filling it with CPU dots.

But that penalty only applies to HUMAN-owned teams. If a team is CPU-owned, there is no penalty. That's why from time to time you'll see a CPU-owned team dominating in a near-empty league, and some chucklehead will buy that team hoping to cash in on an easy trophy -- only to watch their once-dominant team lose the next game 200-0. The INSTANT a human buys the team, the CPU dots are penalized.

There's also evidence that if a human owner sells their team and it becomes CPU-owned, their AIs and Packages remain in place while the team is CPU-owned. If the default AIs are good AIs, the team can roll along ownerless and still look decent with CPU dots.

Long-story-short, there's more than one type of "CPU team." Human-owned teams full of CPU dots are terrible. CPU-owned
teams full of CPU dots can be respectable. Owners and Coordinators tend to treat both types of teams as if they're the same, but they're not.

Originally posted by Novus
All CPU dots scale up or down to the league's average level. That's true for all CPU dots, regardless of the number of human dots or the number of other CPU dots on the roster.

However, what people fail to consider is that for HUMAN-owned teams, if you have more than X number of CPU dots on a HUMAN-owned team (I think X = 10), the CPU dots are penalized. This is to encourage human owners to sign human dots, rather than just buying a team and filling it with CPU dots.

But that penalty only applies to HUMAN-owned teams. If a team is CPU-owned, there is no penalty. That's why from time to time you'll see a CPU-owned team dominating in a near-empty league, and some chucklehead will buy that team hoping to cash in on an easy trophy -- only to watch their once-dominant team lose the next game 200-0. The INSTANT a human buys the team, the CPU dots are penalized.

There's also evidence that if a human owner sells their team and it becomes CPU-owned, their AIs and Packages remain in place while the team is CPU-owned. If the default AIs are good AIs, the team can roll along ownerless and still look decent with CPU dots.

Long-story-short, there's more than one type of "CPU team." Human-owned teams full of CPU dots are terrible. CPU-owned
teams full of CPU dots can be respectable. Owners and Coordinators tend to treat both types of teams as if they're the same, but they're not.



Is it that CPU dots scale to the league or the opponent? I've seen both claimed.
I was told to avoid being on teams with 10+ CPU dots.
I play survivor and get screwed often when I fail to see that a CPU team has been bought. And the rare, team gets sold midseason.
And yet occasionally, you see that human owned, CPU filled team destroy a CPU owned team. This also frequently eliminates me in survivor. I've never understood how, since as I understand it, the max CPU penalty is 25%, which is still a huge deficit when 2 teams play that have nothing but CPU players, but one is human owned. I just don't see how game planning can overcome that kind of disadvantage.

Originally posted by Monkey Boy
Is it that CPU dots scale to the league or the opponent? I've seen both claimed.
I've also seen both claimed. But from what I've observed, it seems to scale to the level of the league.

Originally posted by Monkey Boy
I was told to avoid being on teams with 10+ CPU dots.
You should avoid being on human-owned teams with 10+ CPU dots. That's when the penalty kicks in.

Originally posted by Monkey Boy
I play survivor and get screwed often when I fail to see that a CPU team has been bought. And the rare, team gets sold midseason.
And yet occasionally, you see that human owned, CPU filled team destroy a CPU owned team. This also frequently eliminates me in survivor. I've never understood how, since as I understand it, the max CPU penalty is 25%, which is still a huge deficit when 2 teams play that have nothing but CPU players, but one is human owned. I just don't see how game planning can overcome that kind of disadvantage.
It all depends on the AIs that the CPU-owned CPU-filled team is stuck with. If they have bad AIs, that opens a window for the human-owned CPU-filled team to overcome the penalty by using superior tactics.

Come to think of it, tagging could be an issue too. If the CPU-owned team is using AIs that rely on tagging and they play an opponent that the previous human owner had tagged at some point, I imagine the tags would hold up. But if the CPU-owned team plays a brand new opponent that the previous human owner had never had a chance to tag, I imagine that could mess up the AI for the CPU-owned team.

It's always tricky, trying to win Survivor by picking a winner between two CPU-filled rosters. Doesn't stop me from trying, but sometimes I get burned by it.

Back in the day when the sim was fluid we would build our dots with essentially 0 confidence,

As the final season of your dots life became alive you could train confidence. Catch 22s dots were always shit and never did well in big games and that mfer was trash and has 0 rings.

Back to confidence. The champs of the modern era used to go anywhere from 40-50 confidence and then go 15 HOAC and walk into a playoff game with 80 confidence and 140 confidence on the qbs,

So while I think most felt, thought and stated lolconfidence at the same time it appeared to be an optimal move to smash 15 VA and maximize return at the end of a dots Training career

Originally posted by Dpride59
Back in the day when the sim was fluid we would build our dots with essentially 0 confidence,

As the final season of your dots life became alive you could train confidence. Catch 22s dots were always shit and never did well in big games and that mfer was trash and has 0 rings.

Back to confidence. The champs of the modern era used to go anywhere from 40-50 confidence and then go 15 HOAC and walk into a playoff game with 80 confidence and 140 confidence on the qbs,

So while I think most felt, thought and stated lolconfidence at the same time it appeared to be an optimal move to smash 15 VA and maximize return at the end of a dots Training career

Heart Of A Champion... man, I completely forgot that VA even existed.

"This player excels in the playoffs. Each level of Heart of a Champion gives a +2 bonus to confidence in the playoffs."

15 in HOAC means +30 Confidence in the playoffs. Damn.

Thanks Dpride!

Dpride bringin da bombs, yo!