We need more players.

TheKidPunisher
Autonomous Entity
Autonomous Entity
Posts: 563

Re: We need more players.

Post#21 » 02 Aug 2017, 19:36

Jeffrey Harpen wrote:(Funny it was said by the one who didn't get frustrated playing only assaults... :) )


I get frustrated, not you? When you reach league where you thought people rekt you only by having a sniper? (After i see the true power of assault and try to build it)

About lock only ranked to free player, i am not sure that they will pay just for ranked game. But giving them access to private will surely give them the idea to invite their friend...

pier4r
Skynet
Skynet
Posts: 3380

Re: We need more players.

Post#22 » 02 Aug 2017, 20:08

I would really like to see TKP with all classes playing, but I guess you got so used to assault that to reach better levels than your original account it would take time.
http://www.reddit.com/r/Gladiabots/wiki/players/pier4r_nvidia_shield_k1 -> Gladiabots CHAT, stats, insights and more ;

Jeffrey Harpen
Automaton
Automaton
Posts: 119

Re: We need more players.

Post#23 » 02 Aug 2017, 20:44

TheKidPunisher wrote:
Jeffrey Harpen wrote:(Funny it was said by the one who didn't get frustrated playing only assaults... :) )

I get frustrated, not you?

Actually I'm thinking about getting back to assaults with slightly deeper research and some ideas that surely won't give me any advantage, but I want to test. Freeplay and private matches give a perfect opportunity to do this.

TheKidPunisher
Autonomous Entity
Autonomous Entity
Posts: 563

Re: We need more players.

Post#24 » 02 Aug 2017, 21:52

Jeffrey Harpen wrote:Actually I'm thinking about getting back to assaults with slightly deeper research and some ideas that surely won't give me any advantage, but I want to test. Freeplay and private matches give a perfect opportunity to do this.


I will not play that much during my vacations but feel free to private match me if you are bored to play against special bot and see a fight of assaults vs assaults

darktheatre
Script
Script
Posts: 34

Re: We need more players.

Post#25 » 03 Aug 2017, 17:02


User avatar
Ritter Runkel
Neural Network
Neural Network
Posts: 498

Re: We need more players.

Post#26 » 03 Aug 2017, 21:16

+1

TalonZa
Script
Script
Posts: 24

Re: We need more players.

Post#27 » 11 Aug 2017, 00:52

Many things for me make the game not so enjoyable anymore. The game doesnt drive engagement at all. GFX is the only developer so everything takes a looong time. There are no notifications in the game so players aren't coming back often enough. No marketing for the game. Players can hard code AI per map that isnt really AI, just the owner saying do x,y then z. Editing of AI is tedious with no undo or version control. Perhaps some hints while creating AI like "this action will never be reached". Too few players. Games repeat. Ranked skill groups limit the people ever more that you can compete against.

I know above sounds lile ranting/crying. But lots of reasons for players not to come back. Personally Im going to stop playing until I see updates, or find another similar games.

User avatar
0x1de
Automaton
Automaton
Posts: 156

Re: We need more players.

Post#28 » 11 Aug 2017, 03:51

It would be a shame to see you go, because you are a respected opponent.

pier4r
Skynet
Skynet
Posts: 3380

Re: We need more players.

Post#29 » 11 Aug 2017, 09:35

TalonZa wrote:Players can hard code AI per map that isnt really AI, just the owner saying do x,y then z.

and which AI is not doing this? Even alphago has a neural network that is fine tuned by the dev team. I understand what you mean, when I had the course in AI I learned the minmax and I discovered that AI is nothing else than an hardcoded heuristic that computes really fast with neat observations to avoid following wrong paths.

So for my perception it sounds like "I lose because I do not hardcode solutions, if the others wouldn't do so, they will lose against me". It is an excuse. Your AI is nothing different from the others, you also say "dear bot, when you have low shield and everyone is attacking you, please retreat". Is this not an hardcoded behavior? If you wait that your bot gets to know this by its own, it will be always vaporized.

Editing of AI is tedious with no undo or version control.

That editing is tedious, I agree. Version control it is not so easy. I mean, I challenge you to produce a quick framework for versioning just files (not using git or whatever, create yours). Not directories, just files. You will see how not easy it is, and you have only the task of making a version control system, not an entire game + its infrastructure. So I wouldn't expect it. While I would expect a better editor.

Perhaps some hints while creating AI like "this action will never be reached".

Another not easy thing.

Games repeat.

Again, which game doesn't ? Heck life repeats. Don't you sleep every day? Don't you eat every day?

Ranked skill groups limit the people ever more that you can compete against.

What would be the solution, compete against everyone? And where is the fun to beat a 1300 guy when you are 1700? Just to feeling better because you win? "oh look I have a streak of 50 wins against opponents way worse than me!" . Meh.

I know above sounds lile ranting/crying.

Not all the points, but some yes. I really believe you are whining on some points, whatever you say. I also remember your rants about the elo as being unappropriate when you lose (I wonder what could be appropriate, since other systems are even less merciful than elo. Maybe something that discount losses? Once again, Meh).

But lots of reasons for players not to come back.

For what I observed and discussed (on discord especially, that has a lot of inactive players) it boils down to two things:

(a) paywall. Few people will accept to pay upfront a cost of 3$, like a mcdonald burger. With a fremium model I suppose there will be way more willing payers or at least casual players.
(b) difficulty. There is little to hide it. There are hundreds of examples (I do not know about studies, but surely there are) that show that when things gets hard, people quit. Most of the people like "low effort, nice results for one's self esteem". There was someone saying that a lot of MMORPG are successful because they ensure that a player can find its niche place where to win, so the game gives back an easy reward. Here instead there are duels and if people do not put effort in it, they are going to suck, and sucking is frustrating. I guess it is hardwired in humans, don't dotoo often what does not give you back nice results.


For example let's pick your stats. You played only 404 games (pretty a little, if those would be 1500, maybe they would be enough).

So in master you had hard times, but more or less you had balanced games. You got in grandmaster and you got demoted pretty quickly, then again in master you won a bit, and now in GM you are again "cannon fodder". I understand that this is frustrating, but one does not see that one has nothing to lose when he is at the bottom of a league. Just adapt until you crack the enemy and the elo will reward you immediately. You may lose 2,3 points when you lose to a 1900+ , but then you get +15 when you win (it means, 5 defeats covered) or even +2,3 when you draw.
Last edited by pier4r on 11 Aug 2017, 09:59, edited 2 times in total.
http://www.reddit.com/r/Gladiabots/wiki/players/pier4r_nvidia_shield_k1 -> Gladiabots CHAT, stats, insights and more ;

pier4r
Skynet
Skynet
Posts: 3380

Re: We need more players.

Post#30 » 11 Aug 2017, 09:56

Once again about difficulty. A lot of games are based on grinding, for this they are appealing.

There may be a better player than you, but you play 7 hours more a week than him for weeks, you get more experience for your character (or a crew, in world of tanks), better equipment and boom, you equal his skills.

In modern combat 5 (for android) I am floating around 1700 in strength (the max is 4000) because after a while you see people that are unable to aim, but have so many (paid or not) additional gadgets that shoot for them, that I don't care. They are mostly using "allowed aimbots". And so they are rewarded.

There is no grinding in chess or gladiabots (or checkers, or line of actions, or whatever mind game). If you suck, and you play 7 hours more per week, if you don't play to improve but just to grind, you are going to suck 7 hours more per week.

Same with quake duels (Or warsow, painkiller and what not). You get in the arena, if you are going to play without putting much effort, you are going to hit a wall of difficulty and suck. In quake you need to improve your aim, keep the time of the powerups, develop tactics and so on. If you see top level games, out of 10 minutes, they may not shoot at each other for 8 minutes, because it is all tactical positioning to control more map/powerups than the other player and the end score is like 3 kills to 0 for this or that player.

If a player ends up in a bad tactical position, the score may be 50-0 .

So in gladiabots either one willingly stays in a league (like you decide to stay in master league) to have a bit of balance, or one has to swallow the defeats and improve.
http://www.reddit.com/r/Gladiabots/wiki/players/pier4r_nvidia_shield_k1 -> Gladiabots CHAT, stats, insights and more ;

TalonZa
Script
Script
Posts: 24

Re: We need more players.

Post#31 » 11 Aug 2017, 20:12

0x1de wrote:It would be a shame to see you go, because you are a respected opponent.


I just need a break. Pier is right, I dont have time to spend on tweaking the AI per map so its made me frustrated as Im probably at ceiling without that. I get murdered in GM and do very well in M.

User avatar
Ritter Runkel
Neural Network
Neural Network
Posts: 498

Re: We need more players.

Post#32 » 11 Aug 2017, 21:08

Only a tip for the ones entering GM. There is a very active telegram group with almost all top 20 player inside. There we analyse matches together and have an open discussion about tactics, coding and ideas.

There is even a gladiabots simulator to test scenarios. Chris is working on this one, this could help too. Some of my best ideas in my AIs I got during a chat on this group.

pier4r
Skynet
Skynet
Posts: 3380

Re: We need more players.

Post#33 » 11 Aug 2017, 21:25

TalonZa wrote:on tweaking the AI per map so its made me frustrated as Im probably at ceiling without that. I get murdered in GM and do very well in M.


Note:
- a lot of players are generic.

- Miojo (miojo also uses, barely, one tag)
- Buzzefall (max 2211)
- T800
- I believe sollniss
- Ritter almost completely
- I am with decorators and I think mrchris is too.

then you have decorators ( viewtopic.php?f=22&t=126&start=20#p4470 ) that I published to help all those lazy like me that do not want to develop one AI per map.

In short, with decorators I ensure some initial positioning, then I self tag 5, and I execute only the generic AI.
http://www.reddit.com/r/Gladiabots/wiki/players/pier4r_nvidia_shield_k1 -> Gladiabots CHAT, stats, insights and more ;

mcompany
Autonomous Entity
Autonomous Entity
Posts: 872

Re: We need more players.

Post#34 » 12 Aug 2017, 07:05

TalonZa wrote:I know above sounds lile ranting/crying. But lots of reasons for players not to come back. Personally Im going to stop playing until I see updates, or find another similar games.

I'm sorry to see another player leave, especially one that actually was doing pretty well. It's kinda sad that so many people are leaving, especially since very few are at a high level and are active. That said, I do have to do something like what pier did, and make points out on what you said
TalonZa wrote:There are no notifications in the game so players aren't coming back often enough.
It is extremely doubtful this will make players play much more often. Don't get me wrong, there are plenty of times where I'm deploying quickly and am trying to wait specifically for when I get more games. However, I also have a life outside of gladiabots, and so do many other people. If no one is playing, then no one is playing, and I don't think most of the higher level players that actively try to get better will be watching the game so much to always be ready for when the next notification comes.
TalonZa wrote:No marketing for the game.
Honestly with the paywall, I think this is definitely the biggest reason so little new players are coming, which is what we really need.
TalonZa wrote:Players can hard code AI per map that isnt really AI, just the owner saying do x,y then z.
Yeah, as pier said, thete really isn't a real legitimate distinction. However, for now I'm not going to argue about in this topic, because that isn't the focus of this topic. That said, even then, what you say isn't quite true for any map except Mind Game or SYP, and even that's becoming less true. Also, the more features the game gets, the less I think people will do more than simple decorators.
TalonZa wrote:Editing of AI is tedious with no undo or version control.
This is once again that whole thing I like to repeat about requests repeated over and over again. That said, I honestly would rather people continuously complained about these kinda common requests. That way it is definitely more of something that obvious that whatever that is requesting is basically a problem for not already being in the game.
TalonZa wrote:Perhaps some hints while creating AI like "this action will never be reached".
Development time for gfx. Please, don't try to get this to be a thing until after the editor gets better or until the AIs are given more information and freedom.
TalonZa wrote:Too few players.
Considering the title of this thread, I'm pretty sure this is another one of those "I've heard it so many times that it's old" things.
TalonZa wrote:Games repeat.
Honestly, this I believe is that main reason the lack of players is a big problem for players playing. I think it can get pretty demoralizing for me when I lose a certain way against Ritter or Miojo, because I know I don't have a meaningful way to counter what they are doing and thus is stuck losing the same way over and over again in back to back matches, particularly because no one else was playing at the time. And I definitely do know how this can be highly annoying to you
TalonZa wrote:Ranked skill groups limit the people ever more that you can compete against.
First of, freeplay mode. It honestly needs a lobby system for deploying and picking matches, but that's what I think the mode for playing everyone should be.

darktheatre
Script
Script
Posts: 34

Re: We need more players.

Post#35 » 13 Aug 2017, 15:37

Game Repeated.
yes.agree..
dev had planning to make random map generator. so no one bored failed in static map.

pier4r
Skynet
Skynet
Posts: 3380

Re: We need more players.

Post#36 » 15 Aug 2017, 12:17

Pier A, [15.08.17 11:42]
so I watched like 20 replays from matches not checked and I had like 6-7 fixes, 1h 30 min of focus. This is way even with a couple of games (defeats) gladiabots is super intensive.

Pier A, [15.08.17 12:09]
I mean with the low activity one can test less quickly. So the evolution that there was in 1 day maybe is now in 3-4 days (if one deploys every day) but it is also true that: (a) a person can digest that much of information per day, so the marginal utility of more games is not that high (so having less games is not that bad) and (b) thanks to the new leagues the matches are more intense so the likelyhood of hints got by every game is higher.

Pier A, [15.08.17 12:11]
So I would say: on one side one can test less and the evolution of tactics is slowed down, on the other side one has more useful data per game so one needs less games to fix things.
http://www.reddit.com/r/Gladiabots/wiki/players/pier4r_nvidia_shield_k1 -> Gladiabots CHAT, stats, insights and more ;

ai drone
Script
Script
Posts: 35

Re: We need more players.

Post#37 » 16 Aug 2017, 07:52

As a casual player, I think the old way of earning the specialized bots was fine. The only thing I would have liked to see while coming up through the ranks was the ability to learn to use or defend against the specialized bots in the test arena even before earning to use in the career mode as well as the same time counter as in career so one can see if the routines are working fast enough.

When I lost all my ai's I lost that bit of magic that was working so well with around 50 total nodes. I know that I am losing interest in the game due to the fact that I don't have the hours to spend creating the extremely complex ai's that you all seem to be using. I usually have only about an hour at most each day to spend trying to analyse and improve. My ai has never had much more than 50 nodes total. I had stumbled across something that worked great and I got up to 1783 in GM and was refining my ai enough that I was occasionally able to beat the top ranked player at the time.

This could be why we have lost players and maybe are scaring new players away. I think maybe the specialized bots should unlock at the different leagues and when being demoted back to lower leagues those same bots are removed. This would make everyone be on level playing field within the league. I think the inter league matches would still be ok even though the higher league player would have potentially a specialized bot that the lower league player doesn't have as this again would help players learn to defend/attack those specialized bots.

Sorry about rambling. This is a very challenging game that has become too time consuming to be enjoyable since the majority of the time is spent working on the programming instead of the entertainment of the matches.

pier4r
Skynet
Skynet
Posts: 3380

Re: We need more players.

Post#38 » 16 Aug 2017, 08:18

For the boys we had this already and people complain that they cannot get better.

For the ai size , a couple of players experimented and with the right 50 nodes they were able to float in the lower part of the top 20. So I guess you have your chances .

Tinnvec for example kept the ai under 60 nodes
http://www.reddit.com/r/Gladiabots/wiki/players/pier4r_nvidia_shield_k1 -> Gladiabots CHAT, stats, insights and more ;

pier4r
Skynet
Skynet
Posts: 3380

Re: We need more players.

Post#39 » 18 Aug 2017, 14:26

The guy that mostly introduced me to the game 8I had it installed but without playing).

He played in 2016 then stopped until some days ago, losing the old account. (yes I told him about the synch possibility but he was ok buying the game new).

2017-08-18 14_21_02-Telegram.png
2017-08-18 14_21_02-Telegram.png (46.39 KiB) Viewed 1176 times


So he says what many already said. "perhaps it would have been better to have 100 multiplayer matches for free instead of 10"
"One can test better and one gets a proper feeling about the game"
"10 weeks" (this is related to something else)
"If I had not known the game, I wouldn't have bought it after 10 games"

That is quite a feedback for me, knowing the guy.

So once again:
- either giving out some free games per day (from 5 to 10)
- or making the "demo" longer. Maybe even time based, like you have 30 days free after your first game whenever you play a lot or not.
- viewtopic.php?f=5&t=484&p=11288#p11288
http://www.reddit.com/r/Gladiabots/wiki/players/pier4r_nvidia_shield_k1 -> Gladiabots CHAT, stats, insights and more ;

User avatar
TheGnoo
Automaton
Automaton
Posts: 175

Re: We need more players.

Post#40 » 18 Aug 2017, 17:41

pier4r wrote:...


+1

GFX did a great work on tutorials to attract new players, but the buying step is still a bit too rough imo.

Return to “General Discussions”

Who is online

Users browsing this forum: No registered users and 1 guest