57 replies to this topic

[Nightbird]

I'm not talking about "video game" theory, but game theory, developed in the field of economics to scientifically describe how humans make decisions.

If you want to read the long version, take a look here: https://plato.stanfo...es/game-theory/

The greatly shortened version of it is, people act rationally in a way that benefits themselves. It is possible to model this decision making in the form of a decision tree: if I decide A, what are my risks and benefits, if I decide B, ditto.

How does this relate to MWO? Let's take the Quick Play (random teams) conquest mode for example, as designed, suppose the choice in front of a player is: A: 12 vs 12 nascar in the center, I will ignore caps, and B: 11 vs 12 nascar in the center, I will try to cap. Let's say decision A has a 50:50 chance of winning or losing. Decision B, based on how MWO Conquest is designed (TTK, cap location, number of points needed for cap victory etc) means the chance of winning reduces to 40:60.

Each player, after experiencing A and B enough times, will choose decision A because it benefits them more.

Of course, MWO QP is not just conquest, but you can try this thought process with any game mode, any map. Should I choose a sniper, brawler or mid-range mechs? Should I NASCAR or snipe or cap or flank? As it is designed, MWO QP favors mid-range nascar over any other tactic.

Well, how do you fix this? Let's go back to QP conquest for a moment. Suppose when the decision tree comes up: A: 12 vs 12 nascar in the center, I will ignore caps, and B: 11 vs 12 nascar in the center, I will try to cap. How do you design the game mode so that B is the better choice? Well, you just need to make it so that B has better than a 50:50 chance of winning, which can be done for example by lowering the number of points needed for cap victory to 300-400 points. The team that devotes more resource AWAY from the center nascar has a BETTER chance of winning, when this is true nascar ends.

my 2 cents

Edited by Nightbird, 03 March 2021 - 10:15 AM.

[MechWarrior5782621]

That's not bad, but it only fixes conquest.

The only way to fix skirmish modes is by changing the maps or the spawn points. Like you said, they need to design the game in such a way as to disincentivize nascaring.

They could put some obstacles on the maps that would make it harder to nascar, and instead reward teams for setting up firing lines in advantageous positions.

In terms of spawn points, one option is to have each team spawn in a straight line, with the assaults up front. That way you wouldn't be able to pick off the other team's assaults that are getting left behind, so there would be no point in nascaring.

[Nightbird]

Fainting Goat, on 03 March 2021 - 10:17 AM, said:

That's not bad, but it only fixes conquest.

The only way to fix skirmish modes is by changing the maps or the spawn points. Like you said, they need to design the game in such a way as to disincentivize nascaring.

They could put some obstacles on the maps that would make it harder to nascar, and instead reward teams for setting up firing lines in advantageous positions.

In terms of spawn points, one option is to have each team spawn in a straight line, with the assaults up front. That way you wouldn't be able to pick off the other team's assaults that are getting left behind, so there would be no point in nascaring.

If you used game theory, you'd realize why your suggestions wouldn't work it's a powerful tool regardless of what you want to do in life

[MechWarrior5782621]

I'm familiar with it. They would work.

Your idea for conquest was a poor solution, I was just being polite. It might prevent nascar but it would make the games too short, & not worth playing. There are better options like making larger maps.

[Nightbird]

Fainting Goat, on 03 March 2021 - 10:32 AM, said:

I'm familiar with it. They would work.

Your idea for conquest was a poor solution, I was just being polite. It might prevent nascar but it would make the games too short, & not worth playing. There are better options like making larger maps.

lol unless your obstacles include an invisible wall to keep teams on two sides, none of your suggestions do anything to reduce nascar.

Edited by Nightbird, 03 March 2021 - 10:40 AM.

[crazytimes]

The problem with speed based modes is there is absolutely no balancing done by speed. At best there's a token weight balance, and if your lights are Urbanmechs and Cougars...

I've seen a couple of dominations lost because one team had a flea, the other had nothing fast and the match finished before any shots fired. Conquest on polar can be victim to that too, it just takes longer to lose.

That is absolutely no fun for anyone. There should be no incentive for lucking out with all the fast mecha on one team, unless matchmaker starts balancing by speed as well.

[Nightbird]

crazytimes, on 03 March 2021 - 10:53 AM, said:

The problem with speed based modes is there is absolutely no balancing done by speed. At best there's a token weight balance, and if your lights are Urbanmechs and Cougars...

I've seen a couple of dominations lost because one team had a flea, the other had nothing fast and the match finished before any shots fired. Conquest on polar can be victim to that too, it just takes longer to lose.

That is absolutely no fun for anyone. There should be no incentive for lucking out with all the fast mecha on one team, unless matchmaker starts balancing by speed as well.

In that case, take out all game modes, keep skirmish, and rename it nascar. You either fix most of the games by sacrificing a few bad games, or sacrifice most games to prevent a few bad ones. Even in conquest, I don't see how a flea can get to your two close caps faster than your heavy mechs.

I also never said a fix for domination, I gave one example for conquest to describe this method of problem solving. Create a situation where having mechs not participating in the center nascar is beneficial to your team, and nascar won't happen anymore. Don't and it will keep happening. Simple right?

[MechWarrior5782621]

Nightbird, on 03 March 2021 - 10:40 AM, said:

lol unless your obstacles include an invisible wall to keep teams on two sides, none of your suggestions do anything to reduce nascar.

Not invisible, but it could be done with huge mountain ridges or ravines that make it physically impossible to nascar.

[Vlad Ward]

I don't see how Objectives or Objective-based play will ever matter without respawns.

Reducing the Conquest point limits may encourage a player to take nearby caps initially, but also decreases the length of the match and forces everyone right back to fighting over Theta. The early advantage is just solidified.

At what point do we say "Screw it" and add periodic dropship landings with fresh Mechs. Not a drop deck, just a shiny new version of the Mech you were piloting before. We've already suspended disbelief by letting these things magic their way back to waxed and polished between matches.

[Gagis]

The most important thing objectives need to do is break stalemates. Like, two teams can't try to kite and avoid engaging each other forever if one of them has capped a Conquest point. That is important, and the reason why Conquest, Domination and Assault are all strictly better than Skirmish.

Incentivising varied tactics is a good secondary goal tho.

[GARION26]

Fainting Goat, on 03 March 2021 - 11:37 AM, said:

Not invisible, but it could be done with huge mountain ridges or ravines that make it physically impossible to nascar.

Actually I was reading the thread and thinking it might be interesting to have say a river of lava with a few bridges and scattered terrain that doesn't necessarily allow people to just ridge peak and snipe (which would just give us polar highlands again). Terra Therma sort of has that - but make the Lava damage dealing without central open areas. A light interested in backstabbing could go to an undefended bridge and sneak in, midrange folks would be able to battle accross, and it would make jump jets useful for more then just pop tarting. On the otherhand I am sure there are plenty of ways that kind of map might be unfun, and we know new maps aren't likely.

[LordNothing]

the problem is if you lower the cap goal in conquest and then boring victories happen without a lot of combat.

also that's not the usual way conquest goes down. if every lance caps the nearest point, converges and then forms a line around theta you can do the tried and true cap 3 and fight. every lance deals with their closest cap, even a couple assaults just walking through a cap on their way out is enough to secure it. sometimes 2 is enough to keep the enemy from cap winning while you kill them. lots of battles get lost from going all in to the fight strategy ignoring the cap situation, and then an industrious flea on the other team wins the match. its when the enemy team has 500+ cap points and there are still 3 mechs on the map that need killing, 2 of them are squirrels, and then you realize that cap is important. 80% fighting and 20% cap is a good balance i think. neglecting the cap completely is a bad strat.

the way to make conquest good is to get the teams split up into little pocket battles. i don't think there is any easy fix for that in mwo. in living legends you have terrain control, with the caps spread out on a larger map, the caps are actual bases with turrets and sometimes repair/rearm facilities. so you get the little pocket battles. a few mechs hold the cap until the threat is dispatched, and then go to the next thing. there are enough tickets and enough time on the clock for the game to play out.

thats why fp conquest plays better because there is time on the clock and respawns and victory goes to the more mobile team on most maps. though some have power positions which can block 3 or more caps from one spot provided your team has enough long range trade capabilities, like frozen city and grim plexus, and those you can have a static offence and win through attrition. i dont find that fun, but its nice to have maps where different strategies work.

[Nightbird]

LordNothing, on 03 March 2021 - 01:19 PM, said:

the problem is if you lower the cap goal in conquest and then boring victories happen without a lot of combat.

also that's not the usual way conquest goes down. if every lance caps the nearest point, converges and then forms a line around theta you can do the tried and true cap 3 and fight. every lance deals with their closest cap, even a couple assaults just walking through a cap on their way out is enough to secure it. sometimes 2 is enough to keep the enemy from cap winning while you kill them. lots of battles get lost from going all in to the fight strategy ignoring the cap situation, and then an industrious flea on the other team wins the match. its when the enemy team has 500+ cap points and there are still 3 mechs on the map that need killing, 2 of them are squirrels, and then you realize that cap is important. 80% fighting and 20% cap is a good balance i think. neglecting the cap completely is a bad strat.

the way to make conquest good is to get the teams split up into little pocket battles. i don't think there is any easy fix for that in mwo. in living legends you have terrain control, with the caps spread out on a larger map, the caps are actual bases with turrets and sometimes repair/rearm facilities. so you get the little pocket battles. a few mechs hold the cap until the threat is dispatched, and then go to the next thing. there are enough tickets and enough time on the clock for the game to play out.

thats why fp conquest plays better because there is time on the clock and respawns and victory goes to the more mobile team on most maps. though some have power positions which can block 3 or more caps from one spot provided your team has enough long range trade capabilities, like frozen city and grim plexus, and those you can have a static offence and win through attrition. i dont find that fun, but its nice to have maps where different strategies work.

Please organize your thoughts more. Currently optimal strat = ignore caps, kill other team, then get caps. Change optimal strat by making it difficult to kill other team before capping. It's an easy to understand and easy to implement change. Other changes are possible but if they are hard then PGI won't bother doing it.

I put forward the problem in the easiest to understand way with the easiest to implement solution. Since it's still too hard to understand and unite behind, I'll give up. GL HF and vrrrmmmm!

[LordNothing]

Fainting Goat, on 03 March 2021 - 10:17 AM, said:

That's not bad, but it only fixes conquest.

The only way to fix skirmish modes is by changing the maps or the spawn points. Like you said, they need to design the game in such a way as to disincentivize nascaring.

They could put some obstacles on the maps that would make it harder to nascar, and instead reward teams for setting up firing lines in advantageous positions.

In terms of spawn points, one option is to have each team spawn in a straight line, with the assaults up front. That way you wouldn't be able to pick off the other team's assaults that are getting left behind, so there would be no point in nascaring.

don't add more obstacles, the maps in this game are already messy obstacle courses with death traps and ratchet corners and fake thoroughfares that force you to backtrack and waste time. and the ground clutter, the horrible ground clutter that can stop a 100 tonner cold.

i think the maps actually need more power positions rather than obstacles. interesting places you want to hang out. nascar stops cold if there is a strategic hill hump position around the bend. hpg has one, caustic has one. but most players dont recognize them as such and nascar through. every map should have no fewer than four power positions where no one position gives advantage over any of the others (best example we have is probably canyon network, its why its so popular in comp).

players enter the game with zero alternative options and so nascar kicks up and once a couple mechs are doing it, everyone does it because they know it will snowball out of control and they will get left behind if they dont. when i spawn in assault nascar is automatic for me, im on rails. power positions give us nexus points where decisions can be made and the game theory can iterate and people get off the rails.

Edited by LordNothing, 03 March 2021 - 01:34 PM.

[Meep Meep]

Changing maps around to make nascar more difficult won't stop nascar because its the natural hunting drive to get behind an enemy so you don't get shot. Humans automatically try to flank just like any predator and since most are right handed we tend to drift that way at the beginning and once the rotation starts you have to go with the flow or get overrun by the other side. The only solution that works I have seen in quick play is someone taking command over voip and getting a firing line set up or simply getting the rotation to go left so both teams end up running into each other instead of after each other.

[Nightbird]

LordNothing, on 03 March 2021 - 01:32 PM, said:

don't add more obstacles, the maps in this game are already messy obstacle courses with death traps and ratchet corners and fake thoroughfares that force you to backtrack and waste time. and the ground clutter, the horrible ground clutter that can stop a 100 tonner cold.

i think the maps actually need more power positions rather than obstacles. interesting places you want to hang out. nascar stops cold if there is a strategic hill hump position around the bend. hpg has one, caustic has one. but most players dont recognize them as such and nascar through. every map should have no fewer than four power positions where no one position gives advantage over any of the others (best example we have is probably canyon network, its why its so popular in comp).

players enter the game with zero alternative options and so nascar kicks up and once a couple mechs are doing it, everyone does it because they know it will snowball out of control and they will get left behind if they dont. when i spawn in assault nascar is automatic for me, im on rails. power positions give us nexus points where decisions can be made and the game theory can iterate and people get off the rails.

Nope, power positions don't work because you need teamwork to take advantage of them. One or two mechs in a power position with the rest of the team nascaring away means you have a 12vs2 against your power position and death. Canyon network, caustic, HPG, grim etc all have plenty of power positions outside the center where if the whole team stops and camps it, the Nascaring team will lose, yet these are the worst Nascar maps. This theory is one which isn't true in practice, try again.

[LordNothing]

Nightbird, on 03 March 2021 - 01:29 PM, said:

Please organize your thoughts more. Currently optimal strat = ignore caps, kill other team, then get caps. Change optimal strat by making it difficult to kill other team before capping. It's an easy to understand and easy to implement change. Other changes are possible but if they are hard then PGI won't bother doing it.

I put forward the problem in the easiest to understand way with the easiest to implement solution. Since it's still too hard to understand and unite behind, I'll give up. GL HF and vrrrmmmm!

its cap 3 and win, and the 3 are in route to the power position on most maps. then you fight, thats how games are won. take hpg, 2 of the caps are right there near spawn, so you can cap them early, and the team is not in danger because they are still inside the walls when this happens. whether or not you take theta, you just bought a lot of extra time to fight.

polar is the polar opposite. there is so much spread between the caps that i think the best strategy is to push the heavy stuff for theta, and send your squirrels wide to get the caps on the other side, then converge back towards theta as they encounter resistance. or if the squirrels feel they have the advantage they can fight and deprive the other team of their squirrels and win the match with time to fight. then circle back to theta where the bulk of the combat happens and join the blood bath.

in any case its a 20:80 cap to fight ratio that has the best chance to win. durring the course of battle the mechs might find themselves unable to out cap the last squirrel. people have torsos and legs missing, might be low on ammo or have lost weapons. and squirrels love to feast on damaged mechs. putting off the cap to the last minute can be disastrous.

organized thoughts? aint got no time for that.

Edited by LordNothing, 03 March 2021 - 01:53 PM.

[Nightbird]

LordNothing, on 03 March 2021 - 01:47 PM, said:

organized thoughts? aint got no time for that.

then I won't waste time explaining where you are wrong...

[The6thMessenger]

Can't resist.

That's just a theory.

A GAME THEORY!

[PocketYoda]

Wouldn't a better option be for conquest, you must destroy/defend the outposts then you'd need to get your mechs to the points to stop the enemies from doing damage, like the points in incursion.. you need a fair bit of damage to destroy those bases..

To me that makes a better Conquest.. and its hard to nascar when you are attacking / defending areas.. Also if you don't defend you lose.

They could even add laser turrets to the defenders bases.

Edited by Samial, 03 March 2021 - 03:41 PM.