Last night I was dropped into a number of games in progress. We've all been there, and it mostly sucks (because you're almost always on the losing team). But I can think of one way to at least give you a fighting chance to catch up.

I think an algorithm should be in place to fill your super meter up on spawn. You're immediately behind everyone else in terms of charging your super and often get supered several times to start the match without a chance to counter. I think there are 2 options regarding how full your meter should be:

1) Your meter should be as full as it would be had you started the game from the beginning. If you get in the match past a point where your super would have been full, then you should have a full meter. Or,

2) Your meter should be filled to the room average. If the average meter is filled to 80%, you spawn in with an 80% full meter, allowing you to be full rather soon and competitive right away.

What do you think of this idea?

I think we had a discussion similar to this a while back, but we focused more on adjusting the scoring rather than the players abilities. I think this is the better option to even the results of a game. Scoring becomes a ratio of members on each team. This way, all in all, single player versus player battles remain equal. This still may not work, because now the team that has fewer targets can score more points and has more targets to kill. This doesn't really address your point about supers though.

I'd recommend that a disconnected player be replaced by a bot. The bot keeps all your stats/abilities and when the next guy spawns in, he replaces the bot with the super/grenade/melee at the same %.

Third, because you can't distinguish between poor connections and poor sports, you often get a larger percentage of players who mean well but have bad connections all in the same game, which just means the multiplayer experience for those players is going to suck major balls with people joining and dropping all over the place and they may as well quit because they'll never finish a match and it's because you've only matched them with other unstable connections.

I have trouble swallowing this. Poor connections will DC fairly randomly, win, lose, or draw. Poor sports will not DC if they're winning. Seems pretty simple to me. Now, if your sample size is only one DC, then yes, there is no way to tell, but you can develop a profile over a period of time.

True enough. I think that, at least at the time I was working in games (about 10 years ago), bandwidth was an issue, and the game's servers didn't actually keep track if the goings-on during matches beyond who the host was and which players were connected to him. It seems, in retrospect, pretty easy to send some info like the score along with the "this guy disconnected" signal, and maybe that was done (I'm really not sure), but I don't believe we were using player connection metrics in order to determine if disconnects followed patterns or not. I know we talked a *lot* about how hard it was to know if a disconnect was intentional or not, and the impression I was left with was that they just didn't have a good measure of that.

The gist of it is that poor connections (in almost every FPS I've ever played with the exception of Destiny) hurt the ability of that player to do well, which means that player's team suffers as a result of his poor connection and he's more likely to be on a losing team when he eventually drops, even if he didn't drop on purpose. That, and the fact that if you start banning people just for having a bad connection, you're going to get a lot of angry calls to customer service, and a few returned games, and both of those things get expensive. So without pretty close to absolute certainty, any method of detection was deemed inadequate.

Third, because you can't distinguish between poor connections and poor sports, you often get a larger percentage of players who mean well but have bad connections all in the same game, which just means the multiplayer experience for those players is going to suck major balls with people joining and dropping all over the place and they may as well quit because they'll never finish a match and it's because you've only matched them with other unstable connections.

I have trouble swallowing this. Poor connections will DC fairly randomly, win, lose, or draw. Poor sports will not DC if they're winning. Seems pretty simple to me. Now, if your sample size is only one DC, then yes, there is no way to tell, but you can develop a profile over a period of time.

I expect Bungie has been down this (and other) paths, and ultimately doing nothing, allows the team with people not leaving to have a good time, and an unfair advantage -- there's not much they can do anyway to add fun in this situation to one side without taking away fun from the other.

The question of what to do about disconnects mid-game was the single most frustrating problem for us to deal with back when I was QA lead on some FPS games. There are basically three ways someone can get disconnected:

1) Accidental disconnect.
2) Get kicked.
3) Intentional disconnect (or quit/exit from match).

There is no way to detect the difference between an accidental disconnect and an intentional one, which is why the option to leave a match is always included in games (to encourage players to use a method whereby the developers can gather data).

The method currently viewed as "best" by most devs I'm familiar with is to have a matchmaking ranking based on how frequently a player disconnects. You then match players to other players who are likely to behave in a similar fashion. This is not ideal.

First, it creates three distinct pools of players: Those who never disconnect, those who sometimes disconnect, and those who frequently disconnect. You've now limited your matchmaking options by 1/3 for any given player.

Second, this method completely falls apart when two friends from different player pools are in a party together. Do you match to only other teams with similar ratings etc? It becomes exponentially more complicated the more players you have in a single party.

Third, because you can't distinguish between poor connections and poor sports, you often get a larger percentage of players who mean well but have bad connections all in the same game, which just means the multiplayer experience for those players is going to suck major balls with people joining and dropping all over the place and they may as well quit because they'll never finish a match and it's because you've only matched them with other unstable connections.

So yeah, I've spent at least 100 hours in meetings with people trying to solve this particular problem and what we landed on is that the current (above) "solution" is still the best option, and no one is happy with it.

--

Xbox Live actually has an under the hood matching system based on player feedback, so players who routinely get crappy feedback will be matched with other crappy feedback players. I think that's potentially better, but it's also prone to abuse since you can just give crap feedback on everyone if you want (though apparently it checks for this sort of behavior as well). I'm not sure how I feel about it, and I'm also not sure if it does anything at all since I've not noticed anything like "hey, wow, there's no cheaters in this game and haven't been for months." Of course, I have noticed WAY more cheaters on the PS4 than on the Xbone, but that's also totally different games, so who can say if the feedback matchmaking is the reason for that or not?

--

I have no idea if Sony does anything with the player feedback you send. Bungie, I know, adds people who get reported to some kind of list or something, presumably to check and see if they're really cheating and to eventually ban them when the annual purge comes around (note to Bungie: Please increase the frequency of your ban passes to at least monthly).

But you're missing the point, this thread isn't about the people still in the room as much as it is about getting the person being dropped into a match in progress up to speed with the rest of the group. We've all had bad players giving the other team points from dying too much, but giving them more health still helps them die less. Remember, points aren't only taken from the other team (kills) but given to the other team, too (deaths).

I'm not missing the point so much as defending the 5 players that are likely winning. They should not be served a disadvantage, increasing their likelihood of losing, just because an opponent left the game.

I believe your scenario would allow for a player to use his/her super to wipe a team, drop from the game, and immediately have someone replace them with a full super. This is why I'd suggest (my other post) to have the person entering the game come in with the super filled to the amount of the guy that is being replaced.

That's really unlikely. The person leaving would be taking a loss for abandoning the game just so the other people who aren't him can maybe take advantage of 1 more super use assuming someone is matched into the game immediately?

I'm not suggesting they would do it intentionally. But as an active player with a good connection, it's far more likely that my opponent leaves the game, than that I am joined into the middle of one. And when my opponent leaves, likely because he's losing, I don't want the new guy to have any more of an advantage than the guy who left. By all means give the new guy the same grenade and super energy as the guy who left, but never any more.

I actually don't understand this mentality that the new guy should get some greater advantage than the guy who leaves. As far as I'm concerned, the new guy should be as close to the guy who left as possible, that includes skill level and energy.

But you're missing the point, this thread isn't about the people still in the room as much as it is about getting the person being dropped into a match in progress up to speed with the rest of the group. We've all had bad players giving the other team points from dying too much, but giving them more health still helps them die less. Remember, points aren't only taken from the other team (kills) but given to the other team, too (deaths).

I'm not missing the point so much as defending the 5 players that are likely winning. They should not be served a disadvantage, increasing their likelihood of losing, just because an opponent left the game.

I believe your scenario would allow for a player to use his/her super to wipe a team, drop from the game, and immediately have someone replace them with a full super. This is why I'd suggest (my other post) to have the person entering the game come in with the super filled to the amount of the guy that is being replaced.

That's really unlikely. The person leaving would be taking a loss for abandoning the game just so the other people who aren't him can maybe take advantage of 1 more super use assuming someone is matched into the game immediately?

If I'm a Sunsinger and don't use my super, even though it's charged, for a minute or two, am I penalizing the team?

Yes, I think you are punishing your team. Averaging more supers per match, especially successful ones, is clearly beneficial for a team. Not only is it more points, but the orbs enable other players to have supers faster. That's why they keep a stat for average supers per game.

Yes, but they key is using them at the right time. That time might be different in trials vs control, but if you use your super to kill one person, that's a waste. It's better to save it for a time when it can turn the tide, such as to take a control point in a key moment, deny heavy ammo to the other team, win a close round in trials, etc.

Nope.

I've seen a ton of people say "I'm gonna save my super for a time when I can turn the tide/take a control point/kill multiple people" and then die (over and over again) without actually getting a chance to use it.

Certainly if you are faced with the choice of using a super on one person vs using it on 3, take the 3. But if you're looking at 1 kill vs no kills (and I realize it's hard to tell that you're looking at no kills, since it's 1 kill now vs no kills FOR THE REST OF THE GAME, or until you'd be super'd again), you should take the 1 kill.

I personally went through a stretch where I said "I'm going to make sure my super kills more than one person" - and when I compared my overall super kills during that stretch to a similar number of games before that stretch, my overall super kills were much lower (though my kills per use were higher).

tldr: Try to use your super for big things, but don't fret when you only get one kill - it's rarely a waste.

As I mentioned in my reply to Cody's comment, a lot of it comes down to the gametype you are playing. Games like Control have fewer opportunities for a true "tide turner" super move. You might score a few kills, but very rarely will a single super make a game-changing difference.

But in Trials, every single super has the potential to win the round. So you need to be more selective about how you use them

tldr: Try to use your super for big things, but don't fret when you only get one kill - it's rarely a waste.

It's a waste in the sense of opportunity cost. You may get one kill, but then a blade dancer on the other team gets 3 kills because you used your super, and didn't have it to stop him. Net loss for your team. Supers are tide turners, and if you don't use them that way, another good team will.

Of course you can't predict the future exactly, but that's what makes it interesting. Supers are a risk, and can either pay off big, or not pay off at all. If you truly think that 1 kill is the most you'll get… well then go for it.

Of course, sometimes a single kill CAN be a tide turner.

That's the big part right there. I don't think its about the number of kills you land with a super, so much as the results of the kill(s). If it allows you to maintain a dominant position, or get an important kill at just the right time, then it is certainly worth it.

It also depends on the game mode you are playing. In control or clash, I find it is usually best to fire supers quickly and often to keep a constant flow of orbs across your team. But in Trials, you need to be far more careful about picking your moments. The team with the most kills is not always the team that wins. It doesn't matter how many kills you get if you let the enemy revive each other. So I'm not going to fire my Nova bomb to take down a single enemy as he thorns me to death with his teammates in revive range behind him. I'm going to save that super for a time when it makes an impactful difference.

You bring up a good point though. There is strategy in the use of supers, at least for better players. I've noticed that the self rezzing Warlocks are some of the easiest to misuse. Often a guy will wait until his entire team is down to res but this usually just delays the inevitable. Best to wait till the enemies turn their back and are sandwiched between you and your other teammates. Give em a good flanking and take them from the front and the back.

The best time is usually if one or two guys are camping your orb. If you can get up without them expecting (like if they are engaging your team), you can turn the match around much more than if you resurrect solo at the end of a match.

This is why super use is interesting. You have to balance the risk / reward of waiting for a good time to use it. Just like using an hmg over rockets. If you die in the initial volley you waste your heavy, but if you survive you rack up huge kill sprees. I personally try to use my super when the value of using it is greater than just the kills it nets.

Agreed. While it can feel that they are OP at times, there really is a nuance to using them. It really adds an extra layer of tactics that most other shooters don't have. I've wondered if they should be removed from trials, because when I think of competitive gameplay, I think of gunskills, positioning and movement, and communication as the defining qualities of competitive players.

You bring up a good point though. There is strategy in the use of supers, at least for better players. I've noticed that the self rezzing Warlocks are some of the easiest to misuse. Often a guy will wait until his entire team is down to res but this usually just delays the inevitable. Best to wait till the enemies turn their back and are sandwiched between you and your other teammates. Give em a good flanking and take them from the front and the back.

tldr: Try to use your super for big things, but don't fret when you only get one kill - it's rarely a waste.

It's a waste in the sense of opportunity cost. You may get one kill, but then a blade dancer on the other team gets 3 kills because you used your super, and didn't have it to stop him. Net loss for your team. Supers are tide turners, and if you don't use them that way, another good team will.

Of course you can't predict the future exactly, but that's what makes it interesting. Supers are a risk, and can either pay off big, or not pay off at all. If you truly think that 1 kill is the most you'll get… well then go for it.

Of course, sometimes a single kill CAN be a tide turner.

If I'm a Sunsinger and don't use my super, even though it's charged, for a minute or two, am I penalizing the team?

Yes, I think you are punishing your team. Averaging more supers per match, especially successful ones, is clearly beneficial for a team. Not only is it more points, but the orbs enable other players to have supers faster. That's why they keep a stat for average supers per game.

Yes, but they key is using them at the right time. That time might be different in trials vs control, but if you use your super to kill one person, that's a waste. It's better to save it for a time when it can turn the tide, such as to take a control point in a key moment, deny heavy ammo to the other team, win a close round in trials, etc.

Nope.

I've seen a ton of people say "I'm gonna save my super for a time when I can turn the tide/take a control point/kill multiple people" and then die (over and over again) without actually getting a chance to use it.

Certainly if you are faced with the choice of using a super on one person vs using it on 3, take the 3. But if you're looking at 1 kill vs no kills (and I realize it's hard to tell that you're looking at no kills, since it's 1 kill now vs no kills FOR THE REST OF THE GAME, or until you'd be super'd again), you should take the 1 kill.

I personally went through a stretch where I said "I'm going to make sure my super kills more than one person" - and when I compared my overall super kills during that stretch to a similar number of games before that stretch, my overall super kills were much lower (though my kills per use were higher).

tldr: Try to use your super for big things, but don't fret when you only get one kill - it's rarely a waste.

I've seen/heard arguments both ways when watching videos from top players commenting on strategy. Yes waiting for a 3+ kill is great. You make more orbs and get more points for multikills. However you also get extra points for kill streaks, so if your super helps you stay alive you can earn more points that way too. If you can use your super to stop a single guy from capturing a point in Control, that can really be crucial, even if you swap deaths (I see Cody mentioned this but I'm emphasizing more of the single kill here). You may have just stopped the other team from getting those bonus points for having 2 or 3 control points.

I have to agree with Claude that just using it is better than waiting for a moment that may never come. Also, Bubblebros. We make orbs just by activating a super and we can choose when and where to do so. :)

If I'm a Sunsinger and don't use my super, even though it's charged, for a minute or two, am I penalizing the team?

Yes, I think you are punishing your team. Averaging more supers per match, especially successful ones, is clearly beneficial for a team. Not only is it more points, but the orbs enable other players to have supers faster. That's why they keep a stat for average supers per game.

Yes, but they key is using them at the right time. That time might be different in trials vs control, but if you use your super to kill one person, that's a waste. It's better to save it for a time when it can turn the tide, such as to take a control point in a key moment, deny heavy ammo to the other team, win a close round in trials, etc.

Nope.

I've seen a ton of people say "I'm gonna save my super for a time when I can turn the tide/take a control point/kill multiple people" and then die (over and over again) without actually getting a chance to use it.

Certainly if you are faced with the choice of using a super on one person vs using it on 3, take the 3. But if you're looking at 1 kill vs no kills (and I realize it's hard to tell that you're looking at no kills, since it's 1 kill now vs no kills FOR THE REST OF THE GAME, or until you'd be super'd again), you should take the 1 kill.

I personally went through a stretch where I said "I'm going to make sure my super kills more than one person" - and when I compared my overall super kills during that stretch to a similar number of games before that stretch, my overall super kills were much lower (though my kills per use were higher).

tldr: Try to use your super for big things, but don't fret when you only get one kill - it's rarely a waste.