Jump to content


It's not the bots, it's your decisions that make or break a game.


  • Please log in to reply
108 replies to this topic

losttwo #21 Posted 10 April 2019 - 06:04 AM

    which way do we go?

  • Community Ace
  • 1700 battles
  • 13,154
  • [S-S-G] S-S-G
  • Member since:
    05-15-2012

View PostLMG, on 09 April 2019 - 09:24 PM, said:

 

There were several problems with capturing the airfield:

  1. I was on a Me P.1099 B-2, which aside of having the usual issues of being a GAA, is equipped with long-range quad 30s that are horribly unreliable in defense, let alone offense. So shooting planes on it is out of the question.
  2. It was a contested Forward Airstrip. A neutral Forward Airstrip needs you to kill most ground targets to cap. If a single ally dies, the sector cannot be captured solely by ground targets and you'll need to go shooting planes, something my plane cannot do reliably. And wouldn't you know it, the first thing my bots did was die there :)
  3. Being on a ground pounder, my main job is to capture sectors. Doubly so if I'm the only human ground pounder in my team, let alone the only human in the entire match. If I'm not getting the caps, no one is, and in this scenario my bots certainly did not spread out like the enemy bots did.

 

Considering the above issues, the best case scenario in case I would have gone for the airfield would be that we cap it and the enemy gets the garrisons, all 4 of them. The next move would be to secure the garrisons, but considering my bot's performance so far, chances are they'd end up losing the airfield before I get a Garrison, resulting in superiority. I could not try to defend the airfield due to the inherent weaknesses of my plane, and because my team needed caps and would not go get them on their own. At that point we can't reasonably catch back up without a superiority of our own, something my bots are clearly incapable of. It's a case of "I cannot do everything on my own" :facepalm:

 

Lets not forget the map with the 1 rocket base in the center and each side has a Plant and Airfield( repairable )

Not once are the bots smart enough to capture and hold the rocket base, then again neither are 95% of the player base.

Every time I get that map and the full bot teams, regardless of what I am flying, it will be a loss.

ObiWan is my only hope.

At least with humans there is a 50-50 chance of a win.

The potatoes are just dumb enough to keep the bots busy.

 

 



Bobby_Tables #22 Posted 10 April 2019 - 06:21 AM

    Second Lieutenant

  • Member
  • 5828 battles
  • 1,674
  • [-DOW-] -DOW-
  • Member since:
    06-16-2014

View Postlosttwo, on 10 April 2019 - 12:04 AM, said:

 

The potatoes are just dumb enough to keep the bots busy.

 

Us taters gotta feed the bots somethin', elsewise, they'll starve.

 



GonerNL #23 Posted 10 April 2019 - 08:03 AM

    Second Lieutenant

  • Member
  • 891 battles
  • 1,027
  • Member since:
    01-17-2018

View Postlosttwo, on 10 April 2019 - 03:13 AM, said:

Humans are unpredictable. Bots are predictable.

 

I doubt that ... had some funny battles (mid tier) yesterday where it looked as if the red bots were put on steroids halfway the battle to make up for the lack of quality of their human team mates. They turned into scalp-hunters and took the top 4 spots on the enemy team. And won !!

Edited by GonerNL, 10 April 2019 - 08:04 AM.

Flying on NA and EU servers

TankNoob1960 #24 Posted 10 April 2019 - 09:39 AM

    Senior Master Sergeant

  • Member
  • 4417 battles
  • 271
  • [FK] FK
  • Member since:
    02-05-2016

I think the bot programing in way off base.

They just don't do what they should and what they do, do is not a help.

IMO

Gunny


 

 


losttwo #25 Posted 10 April 2019 - 10:17 AM

    which way do we go?

  • Community Ace
  • 1700 battles
  • 13,154
  • [S-S-G] S-S-G
  • Member since:
    05-15-2012

View PostTankNoob1960, on 10 April 2019 - 04:39 AM, said:

I think the bot programing in way off base.

They just don't do what they should and what they do, do is not a help.

IMO

Gunny

 

Most definitely.

They have not changed very much in the last couple of years.

Although there are still the 3 major types of bots Potato, French Fry, and of course Mrs. Potato-head.

Now they have added a 4th style of bot that acts like a vulture.

 

You can fast forward this video to 2:45 and see a major game changing difference between a BOT and HUMAN.

Yes, bot GA still wont chase you down.



White_Widow18 #26 Posted 10 April 2019 - 10:49 AM

    Second Lieutenant

  • Member
  • 6719 battles
  • 1,168
  • [I_D] I_D
  • Member since:
    10-08-2018

View PostCorvusCorvax, on 09 April 2019 - 10:34 PM, said:

 

Please read the final sentence of my original post.  It was written that way on purpose.

View PostCorvusCorvax, on 09 April 2019 - 12:49 PM, said:

The bots are not the reason that you win or lose.  The choices that you make or don't have much more influence.

 

I read it, I don't entirely agree. The choices the bots make collectively have more impact than the choices you make assuming an all-bot match and if it's part human part bot.. still. the rest of both teams have more collective impact. 23 choice trees between all the bots. Vs... your one set of choices. Statistically speaking, the bots have -far- more impact than you do and far more -opportunity- for impact than you do.. but even assuming the bots are too dumb to take advantage of their opportunities at every turn.. and? There's always the smart bots and by sheer attrition of opportunities they have greater chance to sway the game.

That's my outlook. I realize I have an impact, but I'm not foolish enough to think I can carry an entire game solo. I've tried that. I've had several 7,8, 9, capture matches and still lost because my team was potatoes.

Personally I'm in debate as to who is more likely to make mistakes though.. your average player or the least talented of the bots

Editing to add an amusing all bot match I had this morning:
 
Spoiler

 

Spoiler

 


Edited by White_Widow18, 10 April 2019 - 03:24 PM.

Opinions are like certain body parts. Everyone has them and most of the time they stink.

trikke #27 Posted 10 April 2019 - 02:11 PM

    First Lieutenant

  • Member
  • 2966 battles
  • 2,779
  • [R-A-W] R-A-W
  • Member since:
    01-26-2012

still shocked that the code monkeys think it's a good idea for bot GAs to dogfight at altitude 

 

i don't see it a lot, but enough to remember that other folks besides myself come up with dumb ideas while at work


Spittoon says #smarterpilotswinmore

White_Widow18 #28 Posted 10 April 2019 - 02:16 PM

    Second Lieutenant

  • Member
  • 6719 battles
  • 1,168
  • [I_D] I_D
  • Member since:
    10-08-2018

View Posttrikke, on 10 April 2019 - 08:11 AM, said:

still shocked that the code monkeys think it's a good idea for bot GAs to dogfight at altitude 

 

i don't see it a lot, but enough to remember that other folks besides myself come up with dumb ideas while at work

 

I actually think they got the idea -from- players who do that. I've seen T10 GAA with 6 air kill pips but only 4 or 5 GT kill pips. They -rarely- win a match, but they keep on flying like that anyway.
Opinions are like certain body parts. Everyone has them and most of the time they stink.

CorvusCorvax #29 Posted 10 April 2019 - 05:55 PM

    Captain

  • Member
  • 2948 battles
  • 3,226
  • [JG52] JG52
  • Member since:
    01-26-2012

View PostWhite_Widow18, on 10 April 2019 - 10:49 AM, said:

 

The choices the bots make collectively have more impact than the choices you make assuming an all-bot match and if it's part human part bot..

There is a reason the top players have a 70%+ win percentage.  If bots decided the battles, everybody would have a 50% win rate.

 

In addition, bots don't make choices.  They have their programming.  That's it.  Once you figure out how they react, you can predict what they will do.  For example, have you ever noticed that expert bots seem to know exactly what control input to make in order to make your shots miss?  While at the same time, dragging you out into no-mans-land, and lower altitude?  THey know what control input to make because they are actually seeing it in real time, and countering it.  So, you have to fake them out by doing things to make them do what you want, so they'll be where you need them to be, and when.

 

Ever get in a turn battle with a human?  If they think they have the advantage, they will continue turning on you until they come around onto your tail (unless you do something to counter.)  Bots will go a few turns, even expert bots, until they see something shiny, and then they'll fly off in pursuit.  Bots can be tenacious, but rarely are they as single-minded as a human - I've had a human follow me all over the map trying to knock me down.  Bots will never do that.  

 

All bots have is code.  Their "decisions" are hardwired, and if you can figure them out, you can win much more often than not (assuming only bots, no humans).

 

And yes, getting Hero of the Sky and losing is easy.  Take a HF, and go and help take the single mining plant on a map.  Now, for the rest of the battle, defend it.  You will lose, because the red team will take the other capture points, eventually. (Sometimes not, because RNGeezus smiled on you.)  You rack up an impressive capture point total, along with shooting down a bunch of bot bombers and GAA (the specific mission of the HF.)  I got a 23k point loss out of one of my German heavies doing exactly that.  My individual performance was good, but my choices were bad.

 

This is why we tell the new players to follow your bot bombers or GAA - freeing them up from screwing around over neutral or red capture points means that they move on to work on different capture points.  Letting them flail around on their own MIGHT give you results, except for potato behavior by bot GAA/bombers.

 

It's not sexy work, but it helps in winning.  A high-altitude duel with a human in a HF is much more fun than escorting bots, but unless you do that duel after the squall line, it's almost a waste of time and effort.



White_Widow18 #30 Posted 10 April 2019 - 06:42 PM

    Second Lieutenant

  • Member
  • 6719 battles
  • 1,168
  • [I_D] I_D
  • Member since:
    10-08-2018

View PostCorvusCorvax, on 10 April 2019 - 11:55 AM, said:

There is a reason the top players have a 70%+ win percentage.  If bots decided the battles, everybody would have a 50% win rate.

 

In addition, bots don't make choices.  They have their programming.  That's it.  Once you figure out how they react, you can predict what they will do.  For example, have you ever noticed that expert bots seem to know exactly what control input to make in order to make your shots miss?  While at the same time, dragging you out into no-mans-land, and lower altitude?  THey know what control input to make because they are actually seeing it in real time, and countering it.  So, you have to fake them out by doing things to make them do what you want, so they'll be where you need them to be, and when.

 

Ever get in a turn battle with a human?  If they think they have the advantage, they will continue turning on you until they come around onto your tail (unless you do something to counter.)  Bots will go a few turns, even expert bots, until they see something shiny, and then they'll fly off in pursuit.  Bots can be tenacious, but rarely are they as single-minded as a human - I've had a human follow me all over the map trying to knock me down.  Bots will never do that.  

 

All bots have is code.  Their "decisions" are hardwired, and if you can figure them out, you can win much more often than not (assuming only bots, no humans).

 

And yes, getting Hero of the Sky and losing is easy.  Take a HF, and go and help take the single mining plant on a map.  Now, for the rest of the battle, defend it.  You will lose, because the red team will take the other capture points, eventually. (Sometimes not, because RNGeezus smiled on you.)  You rack up an impressive capture point total, along with shooting down a bunch of bot bombers and GAA (the specific mission of the HF.)  I got a 23k point loss out of one of my German heavies doing exactly that.  My individual performance was good, but my choices were bad.

 

This is why we tell the new players to follow your bot bombers or GAA - freeing them up from screwing around over neutral or red capture points means that they move on to work on different capture points.  Letting them flail around on their own MIGHT give you results, except for potato behavior by bot GAA/bombers.

 

It's not sexy work, but it helps in winning.  A high-altitude duel with a human in a HF is much more fun than escorting bots, but unless you do that duel after the squall line, it's almost a waste of time and effort.

 

Every bot has either a set of algorithms that give it a set of parameter based decisions it can make. Complexity of that determines how "smart" it is. 

Humans function on the same basic level. Humans do have an unpredictability factor often lacking from most AI bots in most games.. but I've also seen the bots abrogate what you claim are there limits. 2+ minutes of me and another turnfighter locked onto each other in a dogfight in the middle of an airbase while both our teams utterly ignored both of us. I've also experienced bots follow me all over the map trying to knock me down, this has been a point of discussion repeatedly on the forums and many have had similar experiences.

I agree with a LOT of your points, I just don't agree with the totality of your conclusion. That's fine, don't have to agree. Your advice to examine what decisions you made and your strategy and to stop using "bots" or -whatever- as a blanket excuse is solid 100-proof advice. 

Insofar as specific situations, I point to the screenshots I posted. Most of my bots, in fact, scored better than the enemy team's bots. Only one of their bots out-chevroned mine and more of mine got chevrons than theirs. They did not, however, play as efficiently which is why we were down nearly 300 points and only won because of attrition. We won by sheer luck and tenacity as far as I can tell. It literally clicked over 800 -as- I was putting the last shots down range on their last plane. If my bots hadn't made the proper decisions -as well- in those last moments, we would have lost with as much of a deficit as we were at. It wasn't -just- me would be my point I suppose.

I would add another piece to your instrospection advice though: Are you trying to force yourself to play a plane that doesn't fit your style and capacity? I made that mistake. My win-rate at a thousand games was in the 20s. I was obsessed with particular planes and how cool they looked on paper and in the sky. They didn't fit the way _I_ play very well however (as evidenced in my successes with other planes). I thought they did... and several planes I thought would be total turkeys turned out to be some of my favorites. Most of the high win % players have very specific lists of planes they fly and they fly them constantly.. they found a plane they were good at and honed that skill to a razor sharp edge to cut the rest of us down with. I appreciate that as part of the reason they win as well and that is another decision point of a player having an impact on the match.

Like I keep saying, I agree with a lot of your points and even could add a few of my own. I'd even agree with 80-85% of the time the player(s) choice(s) being the primary factor.. I just don't agree it is a singular factor situation at all times, barring you being a literally perfect pilot. Assuming average skill? There's a bunch of things that go into a win/loss that you didn't have an effect on and couldn't have. It -does- give you room to improve things without complaining about the game all the time and should make someone consider their criticisms of OP or Broken before they spout them off and instead consider very heavily if they're just making a wrong choice leading to them feeling that way.

 
Opinions are like certain body parts. Everyone has them and most of the time they stink.

CorvusCorvax #31 Posted 10 April 2019 - 07:39 PM

    Captain

  • Member
  • 2948 battles
  • 3,226
  • [JG52] JG52
  • Member since:
    01-26-2012

View PostWhite_Widow18, on 10 April 2019 - 06:42 PM, said:

 I'd even agree with 80-85% of the time the player(s) choice(s) being the primary factor..


 

I knew we'd get there, eventually. But please note, I never said the player's choices were the singular factor.

 

As a note, I play all types of aircraft, with all flight styles.  If I focused on only one aircraft or flight style, I'd get bored.  While I have finally come to terms with my love of all things German heavy fighter, I can still grab a Kawasaki for TnB fun.  Or go bomb the ever-living carp out of some poor mining plant with my B-32.  Or fly biplanes for grins.  Whatever is fun.

 

 


Edited by CorvusCorvax, 10 April 2019 - 07:39 PM.


FIight #32 Posted 10 April 2019 - 08:50 PM

    Command Chief Master Sergeant

  • Member
  • 4896 battles
  • 744
  • [HVAR] HVAR
  • Member since:
    09-11-2015
Decision making for bots is always hard, and I could do a thesis about Markov Decision Processes based on that

White_Widow18 #33 Posted 10 April 2019 - 09:07 PM

    Second Lieutenant

  • Member
  • 6719 battles
  • 1,168
  • [I_D] I_D
  • Member since:
    10-08-2018

"It's not the bots, it's your decisions that make or break a game."

I'm still going to go with disagreeing with that as it is a blanket statement and point to it as where I got the feeling you were stating it was the singular factor. It was your initial thesis statement as it were and I don't happen to agree with it.  It is one factor in a sea of others, is it a heavy weight in terms of variables that go into a win or a loss? Yes. Is it the -only- variable? We seem to have realized we agreed all along that it is not, I consider that positive dialogue and appreciate that -greatly- given this is the World of Warplanes Forum :teethhappy::D

See you in the skies hopefully soon.. I need something that will actually test the limits of my XF-90. My losses have all been fairly predictable and preventable mistakes on my part along with a few dc/load losses *twitch* I wish they offered a faster internet service at my home.. but I suppose I should just be grateful they finally gave us early 2000s DSL at least :P 

 


Opinions are like certain body parts. Everyone has them and most of the time they stink.

CorvusCorvax #34 Posted 10 April 2019 - 10:31 PM

    Captain

  • Member
  • 2948 battles
  • 3,226
  • [JG52] JG52
  • Member since:
    01-26-2012

View PostWhite_Widow18, on 10 April 2019 - 09:07 PM, said:

See you in the skies hopefully soon.. I need something that will actually test the limits of my XF-90. 

 

It's going to be a matter of who sees who first.  Like almost any aerial battle, the first one to see the other usually has the initial advantage.  The other player may start at an initial disadvantage, but can pull it out, with the right combination of moves.  But at T10, I have two planes - the HGIII and the P.1101.  Each can (and has before) defeat a human-piloted XF-90.  BUT:

 

I have never felt that I had the luxury of concentrating solely on a human opponent, but if I see your name, I'll give in to a little "me" time.  ;)



Chuck_norris10 #35 Posted 11 April 2019 - 12:31 AM

    Major

  • Closed Beta Tester
  • 1276 battles
  • 6,237
  • [300SP] 300SP
  • Member since:
    03-02-2013

I wonder if the ww2 pilots talked like this in the barracks before a mission.:amazed:

Bots die just like humans and lately been flying a lot of lf's and no time to check if its bot or not.

Too busy clearing my desk with the joystick when the furball starts.

I should set up a cam so you could see the performance.


Edited by Chuck_norris10, 11 April 2019 - 12:33 AM.

 

 


 

Chuck_norris10 #36 Posted 11 April 2019 - 12:40 AM

    Major

  • Closed Beta Tester
  • 1276 battles
  • 6,237
  • [300SP] 300SP
  • Member since:
    03-02-2013

Tier v, lots of humans folks.

See ya there tonight!

 

 


Edited by Chuck_norris10, 11 April 2019 - 12:42 AM.

 

 


 

Zaikadi #37 Posted 13 April 2019 - 07:19 PM

    Senior Master Sergeant

  • Member
  • 1346 battles
  • 290
  • [R-A-W] R-A-W
  • Member since:
    12-24-2013

"There is a reason the top players have a 70%+ win percentage.  If bots decided the battles, everybody would have a 50% win rate."

 

While it makes sense that WR would be 50% over time if bots decided the wins In the short term, I have had battles where my bots run to the corners while the enemy bots fly in unison from cap to cap turning the map red with no chance for me to turn the tide. However, over time, this should eventually balance out.

 

I think the prime determiner of success is whether you are in a Flight or not. Take a poll of the top players and see how often they are in Flights. Most of the time that I encounter top players on the winning side of a battle, they are in a flight, and if they are in the only flight in the whole battle, they have little trouble dominating. I don't think the MM takes this into account, and I think it is more of an advantage than which tier the planes are. I would rather have a pair of Flighted Tier Vs than a pair of unFlighted Tier VIs on my team.

 

Another thing to consider is which kinds of planes are involved. Slow T&B light fighters have a big handicap in determining the outcomes of battles, especially if they are the only humans. They can't get across the maps in a timely manner and can't turn caps like a GA or bomber can.

 

I hope forumers of this game are not like those of WoT. When I started WoT, I was usually driving stock tanks with no added equipment and 75% (or even 50%) crews, and I had arrogant forumers say that tomatoes were "detriments" to their teams. Those that fly specialized, fully equipped, Flighted, high performance/OP planes should not blame struggling players without these advantages. 


Edited by Zaikadi, 13 April 2019 - 07:52 PM.

Favorite planes:

Nakajima Ki-84 Hayate  |  Kawasaki Ki-102   |  Kawasaki Ki-61 Hien  |  Bell P-39N1 Airacobra  |  Heinkel He 111 H2  |  Polikarpov I-17  |  Curtiss P-40  |   Mitsubishi A6M5  |  Nakajima A4N  |  Messerschmidt Bf 109 B  |  Messerschmidt Bf 109 E  |  Polikarpov I-16 (e.)  |  Fairey Fantome  |  Me 410 Hornisse


legoboy0401 #38 Posted 13 April 2019 - 07:37 PM

    Second Lieutenant

  • Member
  • 1305 battles
  • 1,502
  • Member since:
    11-30-2013

View PostZaikadi, on 13 April 2019 - 11:19 AM, said:

"There is a reason the top players have a 70%+ win percentage.  If bots decided the battles, everybody would have a 50% win rate."

 

While it makes sense that WR would be 50% over time if bots decided the wins. In the short term, I have had battles where my bots run to the corners while the enemy bots fly in unison from cap to cap turning the map red with no chance for me to turn the tide. However, over time, this should eventually balance out.

 

I think the prime determiner of success is whether you are in a Flight or not. Take a poll of the top players and see how often they are in Flights. Most of the time that I encounter top players on the winning side of a battle, they are in a flight. I don't think the MM takes this into account, and I think it is more of an advantage than which tier the planes are. I would rather have Flighted Tier Vs than unFlighted Tier VIs on my team.

 

Another thing to consider is which kinds of planes are involved. T&B light fighters, especially slow ones, have a big handicap in determining the outcomes of battles, especially if they are the only humans. They can't get across the maps in a timely manner and can't turn caps like a GA or bomber can.

 

I hope players of this game are not like those of WoT. When I started WoT, I was usually driving stock tanks with no added equipment and 75% (or even 50%) crews, and I had arrogant forumers say that tomatoes were "detriments" to their teams. Those that fly specialized, fully equipped, Flighted, high performance/OP planes should not blame struggling players without these advantages. 

 

I've been on both sides of the coin before. When I'm on the "high performance" side, I try to rely on my team as little as possible, because they need to be able to depend on me, not the other way around. When I'm in a stock plane, especially when I'm bottom tier, I have to rely on my team for most things, because there's very little(except hunting GAAs of the same tier) that I can do, and almost nothing I can do efficiently, especially with planes with gun upgrades that have weak stock guns.

 

I have respect for both, but especially for the latter because MM's fickleness can be very hard on stock planes, especially on a day where it decides that a certain tier which your stock plane sits at is going to be permanently bottom tier for the day.


An average player with an average goal: greatness, someday. Today, mediocrity will have to suffice. But no matter, I will always play to the best of my ability, and I will always strive to make the best of my ability just that little bit stronger and better.

 

#Failureisalwaysanoption

 

#Givingupneveris


losttwo #39 Posted 13 April 2019 - 07:42 PM

    which way do we go?

  • Community Ace
  • 1700 battles
  • 13,154
  • [S-S-G] S-S-G
  • Member since:
    05-15-2012
I need bomber kill marks on my IL-10, IL-20 and IL-40

hoom #40 Posted 13 April 2019 - 11:20 PM

    Second Lieutenant

  • Closed Beta Tester
  • 6530 battles
  • 1,304
  • Member since:
    01-26-2012

Block Quote

 Take a HF, and go and help take the single mining plant on a map.  Now, for the rest of the battle, defend it.  You will lose, because the red team will take the other capture points, eventually.

 On single Plant maps this is a near guaranteed win unless there are also Command Centers that'll flip it.

 

1* Plant = 3* normal caps so most variants of that map this is a win.

Most of the time you can rely on your bots to cap one of the other points at least occasionally -> you get a lead.

Even on 2* Plant maps getting one & holding it will often get you the win.


C'est magnifique, mais ce n'est pas le SerB.




1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users