Sereglach, on 21 August 2018 - 01:16 PM, said:
*Quote to link back to feedback off OP notes but before listening to podcast. Still pertinent*
Ok, listened to the podcast, and have more to say.
Most of my original concerns about participation still stand. However, one point confuses me a bit: Does the whole "favor tracking system" revolve around earning LP/RP, or is it going to be separate? Does it mean that Freelancers dropping into Faction Play will be earning LP for factions based on who they side with for an event? If so that actually would alleviate a great deal of my rewards concerns for FP and could certainly incentivize me to play the mode again. However, the biggest reason for the confusion is that you (Paul) later go on to state that the Loyalists are the ones earning LP and that the design doc is basically written for Loyalists. Again, I think the biggest solution for Freelancers is earning some LP for whoever they're working for and some RP for basically playing as a solo Merc . . . still rewarded (the LP/RP rewards that actually matter), just at a slower rate than a pure Merc or pure Loyalist. Bonus XP is still pretty rubbish for many/most veteran solo players, and combined with gating (see below) it's even less worthwhile for any player.
Moving on, here's a thought for preventing the units going "Lets all side with faction A to stomp faction B, guarantee our success and guarantee great rewards!" Limit the signups. Take the population for FP and effectively split it half by force. When an event triggers, tally the populations of involved loyalist first, then close off sign-ups to whomever has the higher population. When someone tries to side with the faction that has the higher population, they get a message along the lines of, "Sorry, Merc, but we don't need any more forces, we've already filled all open contracts at this time" for the IS, or for clanners something like, "Your unit has failed to submit a suitable bid for this trial; and you have forfeit participation under our banner at this time." This, by fiat, forces them to join the other side . . . and as a Merc, they should be going where the money is. Once the population disparity has shifted the opposite direction, the ability to signup also shifts with it. If the population is somehow exactly even, or maybe within a certain % of their opponent, then the next unit to sign up can go wherever it wants.
In turn, it also helps bolster the forces of factions with particularly low populations during "their" events, because they'd automatically be the first open for recruitment. In addition, using your new more "interactive" faction loyalty/reputation system, you theoretically could even clump appropriate populations together from the beginning. If Davion is fighting Liao for the wedding event, for example, then Steiner is automatically clumped into the Davion team and Kurita is automatically clumped into the Liao team. IF you're already acknowledging how factions will react on an event by event basis, can't that automatically decide the starting populations of the factions, and then open "event recruitment" from there?
Artillery Strikes (Long Toms) in drop zones: I see a HUGE issue in times when you need to wait 1-2 drop cycles in order to have your team regroup for a push/counterattack. Your own team would be penalized for attempting to regroup. It seems like a very tricky issue to deal with and maintain a sense of balance. If something like the Artillery is used, then I think it'd need to be approached very carefully. It also might not be needed if the matchmaking improvements pan out in reasonable fashion.
Rewards Kickers: There's a modest list I can think of off the top of my head.
- Heat Damage from Flamers, counted like physical damage. You're giving up physical damage potential to help the team through the "crowd control" function of heat damage.
- AMS missiles destroyed, counted like physical damage. Some mechs (2-3 AMS mechs) give up potentially significant damage potential (up to 4.5 tons+ on the 3xAMS Uller, for example) to provide extra protection for the team.
- Count capture zone rewards by % bar moved instead of time ticks. Currently mass people in a capture zone only count for a few tics, and a mech with capture acceleration (you just make this a serious SDR-5V quirk, for example) gains less tics of the reward despite the investment in capping faster.
- Underdog bonus (killing mechs from larger weight classes than yourself) as a light mech reward comparable to "Brawling" from the heavier weight classes.
- ECM Protection for giving coverage to allies . . . bonus counts tics similar to "Protected" or "Lance in Formation" while at least one other mech in your ECM coverage is in combat.
- ECM Jamming . . . smaller reward given as the opposite of "ECM Counter" by entering into jamming range of enemy mechs and keeping them within range for X seconds.
Bonus for Answering Call To Arms: Again, give Freelancers adequate LP and/or RP rewards (the real FP reward paths) and solo players will be more inclined to participate in general. However, that said, giving people a potential 10% extra c-bills or something because they're answering a last second call to fill a match isn't a bad idea. Just make sure that it is clearly communicated to the player.
C-bill Transfers Among Players: Make it a rank controlled function (allows unit leader to designate WHO can give out c-bills) and require a certain number of matches in FP with c-bill controlling ranks (at least several hours of investment in matches).
Community Driven Planet Lore: Players submit lore data to PGI directly, PGI vets data, approved data gets put into the little planetary data box that is currently only filled with prior planet ownership box (maybe needs a tab for lore).
Gating Faction Play: Absolutely . . . at a bare minimum at LEAST make someone play through their cadet bonus matches. Preferably, though, they should probably have at least one set of drop decks (1 Incursion and 1 Scouting) of their own mastered mechs. I'm a casual gamer and even I fully understand and respect gated game modes. Whether it's Smite, Paladins, League of Legends, DOTA, or others, gating the "hardcore/ranked" game mode until a player at least understands the game and has a reasonable "roster" for the more competitive game modes is fully understandable. It's basically becoming an industry norm (especially for games with Ranked or e-sports aspirations), so I don't think it'd upset people if PGI implemented such a requirement. Besides, it'll also prevent some new players (like those who refuse to read warnings . . . the whole "Windows Installer Syndrome" you joked about) from accidentally jumping into a game mode that'll quickly ruin their "new player experience" and drive them from the game.