This game shows an obvious case of Winner Keeps Winning syndrome. What this means is that the players who are doing better get benefits that help them do even better. Examples:

-Winning team has more flags and therefore more bonuses

-Winning team gets more kills and therefore gets more gold

-Winning team spends less time dead and therefore gets more XP and even more gold

You then gain levels and equipment that the losing team doesn't have, giving you a bigger advantage than you already had, and continue to win. This can get very pronounced in some games, and it's just not fun for the losing team. They feel like they're fighting a hopeless battle, and they are! So how to combat it? I've got a few ideas.

As a general rule, I don't want each mechanic to be incredibly pronounced. Each should gently push the losing team to equality in its own way. These are intended to have greatest effect in casual games where one team has a significant advantage over the other while minimizing impact on higher-skill games that are more even.

Desperation: In which the death timer is increased or decreased based upon how many flags the team holds at the time of death. If the player's team controls the map, then the death timer is increased by 30%. If the player's team holds no flags, it's reduced by 30%. This will make the final holdout harder on the winning team, but not enough to make it impossible. Even if they don't win then and there, the decreased death timer gives them a chance at more gold from DG kills. More importantly, it gives the losing team a better opportunity to fend off the winning team and perhaps get a few flags back while they're dead.

Learning From Past Mistakes: In which a bonus to experience is conferred based upon how many times you've died this game. Both the magnitude and duration would scale based on the number of deaths. Say +5% experience for 15 seconds for each time you've died up to a maximum of +50% for 2.5 minutes if you've died 10 times. This gives the losing team a chance to catch up in levels. More even games involving more skilled players will see less of this since less people die in general.

Bounty: In which the losing team will hopefully make use of their other benefits and actually kill one of the opposing DGs. DGs will gain a bounty of 200 gold for each level above their target for each DG they kill. To prevent suiciding, the bounty will be split amongst the other team if a DG with bounty is killed by NPCs or structures. You cannot gain negative bounty. Example: In a 3vs3 game, a level 10 DG kills a level 7 and gains a bounty of 600 gold. He is then foolishly killed by a tower, so 200 gold is given to each of the other team's DGs. In even games, not much gold will be gained from bounties.

Alternatively, you could devalue DG kills that are of a lower level than you so the winning team doesn't end up with a huge gold advantage in the first place. I think a good scale would be 1 below = 90%, 2 below = 70%, 3 below = 40%, 4 below = 0% (that might be a bit too quick ). Or you could do both so the losing team can also try to make up for their experience deficit with some items from the extra gold they get.

Really, the only one that worries me from these 3 is Desperation because it will also affect games that are evenly-matched. The problem with Desperation is that, in Conquest, flags will inevitably be captured no matter what. It's generally what happens before victory, and I don't want it to elongate games too much. Any input on reducing its impact would be great, but I believe there should be something to help losing teams who are getting overrun, though the good old "they will crack under wave after wave of myself" defense that desperation kind of promotes leads to feeding.

Numbers are, of course, always capable of being changed. These are just what I've come up with for now. It's not the numbers that are important at this point, it's the mechanics behind them. When should we reward one team for sucking and penalize the other for being too good and how?

Any input would be great.