Problem of unbalanced teams (skills) at the start

[Hicks]

I have been playing for quite some time and after a break of more than a year I was hoping this problem would be fixed. More in detail, often at the beginning of a new game, the teams (which are probably chosen randomly from the game) are strongly unbalanced and I don't mean the number of players in one team compared to the other but I'm talking about the ability of individual players who make up one team over the other. All this besides being frustrating for a team spoils the games because many players of a team, the heavily disadvantaged one, leave the server. I think you wanted to use your resources for other projects, neglecting the aspects that you think are of little importance. A game should be fun for everyone, not frustrating. Many other players think like me too. So the question is, are you going to consider and fix the problem or not? As a programmer I say that it could take very little even just to improve the situation a bit because I realize that having two perfectly balanced teams is impossible.

[TheOlsenTwins]

In my opinion, it boils down to separate areas:

AFK at the start of the game

Putting everyone that is AFK in a third invisible team would already be a big improvement.

Wrong team allocation
Sometimes people are moved to the wrong team that has a lower amount of players. Maybe this happens due to score difference, but IMHO this should only happen after a significant score difference. Maybe the game could track the score per minute of each player internally and use this as a ranking instead of the total score. Especially in long games there can be a huge difference.

Use the average of the VP of both Teams as starting VP

In long games, players that join the worse team, get even more penalized by getting less starting VP. (as they get the average VP of their team) In the long run, this also makes it harder for a team that is far behind in VP to catch up and make a comeback.

In my opinion, the average VP of both teams should be used as the base for the calculation for the VP of a new joining player.

Edited May 11, 2023 by TheOlsenTwins

[Silv]

Pubs are inherently unbalanceable. When you join a pub you hope to end up in a nice match but there are no guarantees.  Match making is an extremely hard problem that AAA games producers struggle with. Pugs are kinda balanced but it's a very different way to play (you get told what team you're on, what your role is, there's a commander from the start, and you are very much expected to listen to the commander).

The game however isn't about individual player abilities as much as it may seem so when you get your head shot off for the umpteened time because of some super sniper. I've seen teams that seem to have better players lose plenty of times. On the other hand I can count on one hand the times teams were so utterly unbalanced that the game wasn't winnable for the losing side.

If you look at why teams win it's generally a combination of two things: coordination and morale. Teams that act like a team rather than a collection of individuals, supporting each other and making coordinated attacks (e.g. rushes), have a massive force multiplier. On the other hand teams where half the people have already decided they're at a disadvantage are pretty much guaranteed to lose, internal chatter about imbalance often does more than any enemy action.

These things can't be caught in code. Kills, repairs, etc can but are a pretty crappy indicator when it comes down to it. Someone who sits in base all game but prevents three sneak attacks that would have destroyed a building will have a lousy score and VP but probably did more to win the game than a frontline tanker.

Regarding the specific suggestions:

• Moving people to a special team is one of those things that sounds easy but it would probably require a lot of effort due to all kinds of internal dependencies and consequences. Just to highlight one: with the current player screen you can't see spectators so doing this would lead to server slots being occupied without that being visible from inside the game.
• I didn't really understand the point about player allocation and server settings play a role here anyway (there are several different algorithms that a server owner can choose) but as a general point having a new player assigned to you doesn't need to be an advantage. It happens all the time that a team is behind on players, a few unknown players come in filling it back up and then a known strong player gets assigned to the other side.
• Using the average of both teams would mean that someone joining later into a team that's behind on points will have more VP than the average player who earned it "the hard way". Is that fair? Furthermore it's not uncommon for a team that's behind in overall game state to be ahead in VP so this might actually be counterproductive. A typical reason for such an effect is that a vehicle factory or air strip was destroyed causing the team to kill tanks but not have their own tanks killed, tanks are great sources of VP.

(Just to be clear, I may have a TA staff tag but I don't work on these kinds of things so my opinion is worth as much as that of an average player here.)

[isupreme]

The feeling of unbalance is a real negative.    Early harvester kills help create this feeling.

Could the Harvester not give  VP?     i.e.   you just deny the enemy credits when you kill the harvester, but you do not gain VP.

Maybe that would throw the game off balance tho?

[Silv]

Early game harvester kills don't give that much VP. The VPs gained by killing harvester is 2 x the number of trips the harv has taken (including the current one) since its last destruction, up to a maximum of 40. Killing the harv on its first trip gives 2 VP to all team members. It's nice but 2 VP is something you can also get just by repping a tank for a few seconds and is but1/50th of the VPs you need to level to veteran.

The much bigger impact of killing the harv early is economy denial. In particular the first harv dump is necessary for a team to get access to heavier classes and vehicles. Delaying that harv dump while having it yourself creates an imbalance in strength which can be used to gain map control and set the stage for further economy denial.

Protecting the harv is pretty easy if you have a halfway competent team, the harv is pretty easy to rep (gives good cover to reps) and it's closer to your base than the enemy's so it's easy to send reinforcements to protect the harv but any attacker that goes down will be out of the fight for much longer.

In the early game taking an economical lead is one of the primary objectives along with gaining map control. You're supposed to take out the enemy harv and protect your own. If your team ignores the harv and instead goes straight to fighting the enemy or their base and gets their economy harmed that's the logical consequence of a badly chosen strategy.

At the same time even without harvester dumps a team still gets economy and teams can and do fight back from such a position all the time.

[Hicks]

Thanks for the reply, I take the liberty of quoting what you say and letting you know my point of view, perhaps with some constructive ideas.

On 5/11/2023 at 6:33 PM, TheOlsenTwins said:

In my opinion, the average VP of both teams should be used as the base for the calculation for the VP of a new joining player.

You mean to use the VPs from the last game for picking teams in the next game? In this case the algorithm doesn't take into account players joining the game in progress, but it would still be a great start... I totally agree.

On 5/12/2023 at 3:24 PM, Silv said:

Pubs are inherently unbalanceable. When you join a pub you hope to end up in a nice match but there are no guarantees.  Match making is an extremely hard problem that AAA games producers struggle with. Pugs are kinda balanced but it's a very different way to play (you get told what team you're on, what your role is, there's a commander from the start, and you are very much expected to listen to the commander).

I also think that the Pugs are more balanced, but I think they are not for everyone, at least not for me. The reason is simple, I don't speak English well and I have great difficulty understanding spoken English and I use google translate to write in English! In any case, I understand that they are only played once a week, correct me if I'm wrong, and that there are only public servers. Perhaps it would be useful, in public servers, to have the possibility of doing a short warm-up in which one could distribute oneself among teams, perhaps letting only two people choose, just like in the Pugs. Or introducing the possibility of creating clans directly from the game, a bit like what happened with the classic Renegade where you also had a clan lobby and in this way you would also have the possibility to organize Clan War, it would be really very interesting.

Then I wanted to suggest two new votes that would be really useful in all games, I don't know if it's correct to write it here, maybe if it isn't tell me and I'll open a new thread.

• Often I play games with few players, for example 2v2. In this case often someone proposes not to kill the base until a minimum number of players is reached, for example to wait until it's 10v10. If so, in theory they should wait for the minimum number of players to actually start playing. In practice most of the time the players don't wait for that limit and this destabilizes the game itself. I understand very well that the ultimate goal of the game is to destroy the opponent's base but a broken deal most of the time leads to real quarrels between the players. These situations could be avoided by introducing an ad hoc vote so that the structures are not damaged, obviously there must also be a way to restore the destructibility of the structures.
• Another ambiguous situation is with superweapons (nuke/ion). I was playing a 2v2 game the other day and some of them nuked to win. I asked him why he used the nuke in a sparsely populated game and he replied that the nuke is provided by the developers and therefore he felt free to use it in that game and in all subsequent ones. His speech is undoubtedly correct, of course (but this is my point of view) superweapons ruin sparsely populated games. Next I was playing a 10v10 game and I nuke myself destroying the WF, at which point all the players taunted me for ruining the game! Again it would be useful to have a vote to enable/disable superweapons to avoid unpleasant situations where a game is ruined. Before you remind me, I know very well that superweapons are a tactical tool and that they have always been included in the game since the classic Renegade.

[NodSaibot]

23 minutes ago, Hicks said:

You mean to use the VPs from the last game for picking teams in the next game? In this case the algorithm doesn't take into account players joining the game in progress, but it would still be a great start... I totally agree.

I believe he means the initial VP you have when joining a game that is already going.

[isupreme]

Silv,   Thank you for the informaton about how Harvester VP works.   I did not realize it was low at the start.

Hicks, your most recent post included the idea to create votes for very small population games to regulate the use of beacons.   I have often wondered why we do not have the ability to make these votes.

[Silv]

Regarding beacons that can be set by server owners. If a beacon limit is configured the game will only allow buying beacons once enough players have connected (though they remain available even after players drop again). This is not enabled on the public TA servers but anyone hosting their own server can use it.

[Silv]

As to why such votes don't exist, usually it's one of two reasons: none of the programmers considers it to be a good use of their free time (remember we're all volunteers) or it was discussed internally and the team thinks it's not a good idea.

In this case it seems like a nice idea but as usual the devil is in the detail. Say there's a vote to turn off building damage. The server empties and new people come in who didn't see the vote. Buildings are still indestructible. People post bug reports. Devs waste time figuring out what's going on. Not desirable.

Say there's a mitigation that says "Buildings are indestructible", people who aren't experienced with the game still won't know how to turn it off, still generating support requests. But this already makes the change more complicated (sending a message is easy, but it would need to be sent only every so often and not when a building just happens to be in the line of fire, good luck getting it right).

Say the mitigation explains how to do votes in sufficient detail for anyone not familiar with the game to find the voting menu. This requires very careful phrasing and a substantial potion of new players would probably still both find it to be too confusing and be annoyed by it.

I know for every objection there's an answer but these kinds of things tend to become sufficiently complex with all the consequences and mitigations that people just go "forget about it". Also the people who are actually good at these kinds of changes have quite a bit of work on their plate already seeing that Firestorm is still in development and the team is really rather eager to get it released (I know it may look from the outside that nothing is happening, from the inside it's a very different matter).

Compare this to the alternative: social pressure and kick votes which don't require any code change.

[Hicks]

5 hours ago, Silv said:

I know for every objection there's an answer but these kinds of things tend to become sufficiently complex with all the consequences and mitigations that people just go "forget about it". Also the people who are actually good at these kinds of changes have quite a bit of work on their plate already seeing that Firestorm is still in development and the team is really rather eager to get it released (I know it may look from the outside that nothing is happening, from the inside it's a very different matter).

 

On 5/11/2023 at 5:28 PM, Hicks said:

I think you wanted to use your resources for other projects, neglecting the aspects that you think are of little importance.

 

5 hours ago, Silv said:

In this case it seems like a nice idea but as usual the devil is in the detail. Say there's a vote to turn off building damage. The server empties and new people come in who didn't see the vote. Buildings are still indestructible. People post bug reports. Devs waste time figuring out what's going on. Not desirable.

We have all been new to this game or others and we all know that the details are discovered over time. A rookie will always be helped by someone with more experience or more simply, as I did, "stealing with his eyes". A periodic on-screen warning I think is the best solution in the case of votes (no base kill & no nuke/ion), in order to inform the players that the server state is set with those two flags and then... it's also a quick way to get newbies interested in grades

[NodSaibot]

I'm open to making changes or discussing changes within the team, just need to know what people would want. However I'm not really in favour of adding a vote type of option. We could add a playercount minimum for beacons on Official Server, though.

Regarding VP, I've always felt it was weird that new players can join and be higher VP than the lower members, but I'm not really sure how to fix that. If you always pushed the bottom players up, next time it did that it would just build on the last one and would add a natural VP progression over time, which I don't think is a good idea.

[Hicks]

11 hours ago, NodSaibot said:

I'm open to making changes or discussing changes within the team, just need to know what people would want. However I'm not really in favour of adding a vote type of option.

I'm glad you're open to small changes even if they are about votes. Like all things to be implemented, you can't know exactly what people think until they are implemented, I as an average player and aware of how the community thinks about it tell you that implementing these two votes is the right thing, in in any case, going back to a previous situation is never that complicated (I know because in my job I often have to program even on complicated projects). If and when these votes are implemented I will continue to write in this thread to show how the community has reacted.