Objective Scoring

[ClaytoniousRex]

This is a spawn from the Victory Conditions thread. This one is specifically (and only) about scoring for the Objective victory condition.

There's a concensus that points earned for controlling an objective should be proportional to the number (and maybe type) of units that are actually sitting on the objective. We're open to that, but first I want to offer some history and explanation of why the scoring currently works the way it does.

Right now, the defender is assumed to control the objective unless the attacker has at least one unit within its radius. When the attacker has anything within the radius, points start peeling off of the defender's score and onto the attacker's over time. When attacker units are cleared out of the area, points move back to the defender at a slower rate.

For a short while during beta, we did make the score depend upon the number of units within the radius exactly as many here are now requesting. So if the attackers only had a single unit near the objective, then they would slowly gain a tiny trickle of points. To gain the maximum possiblke points, everyone on the attacking team had to pile into the objective area. Similarly, to regain lost points at the quickest possible rate, the defending team had to get all of its units into the objective area (since you get more points for more units). Based on that description, you can probably already see the problem.

Here's an example of a good defense of the bridge in Dead Gulch (green is the defending team and red is the attacking team):

The defender is holding the attacker off at a distance. The defender is deployed in what happen to be, at the moment, good tactical positions in order to keep the attacker away from the objective. Notice that the defender is not sitting directly on top of the objective. Sitting directly on any objective is rarely the best way to defend it. But with a scoring system that awards points proportionally with the number of units that are sitting on the objective, the kind of defense pictured here actually penalizes the team. The team is losing points because they are in good tactical positions instead of just piling onto the objective itself in a big heap. This is even more true if you look at the same figure above but now imagine that green is the attacking team, which has already pushed the defenders off of the objective. So now the roles are reversed - now that green has taken the objective they are defending it against counterattacks from the defending team. If they deployed as pictured above, they would be bleeding points. They would be penalized for not all being in a big pile on the objective itself.

When you only get maximum points for being on the objective, then this deployment is the only one that gives you the max points:

Not a very effective arrangement. "Let's all huddle right on the bridge and just try to hang on for dear life while the other team comes at us in wave after wave!"

The other reason that scoring works the way it does now is simply because of the definition of "defense". We sometimes complain that the attacker is winning the game when he only has a single Paladin sitting near the objective. Let's imagine that your commanding officer has ordered you to "defend the bridge". He comes back a few hours later to survey the situation.

He asks you "Is the bridge secure as I ordered?"

"Yeah, except for this one enemy armored vehicle sitting on it. Besides that minor detail, the bridge is secure."

What kind of look does he give you at that point? If the enemy has breached your defense and is physically present inside of the perimeter that you were ordered to defend, then have you succeeded in defending that perimeter? It's secure because only one enemy squad or vehicle is inside of it? No, it's not, and that's why defenders do and should lose their points due to a single enemy unit's presence on the objective. However, from the attacker's perspective, it's quite a different conversation from the CO. The attacking team's CO would not consider the objective taken if only a single attacking unit had made it into the objective area and that's where the current point system really falls down.

Now, having said all of that, we definitely see where everyone is coming from with the current problems. So I laid out that history and explanation not in order to say everything's fine right now, but so that whatever we do now won't repeat the mistakes of the past.

There clearly is room for some kind of relationship between the amount of force occupying the objective and the points earned for occupying it, but without going "all the way" - some (probably low) number of occupying units should be the maximum that are considered. And, for that matter, the types of units should count. So, for example, having either 2 MBT's, 3 Apollos, or 4 infantry squads on the objective should fully count for maximum points. Any more than that do not increase the total score or the scoring rate. Having only 1 MBT, or only 2 infantry squads, gives a team half as many points. This way we do not encourage the entire team to pile onto the Objective, but they do need to commit at least some decent sized "holding force" to the objective itself. I think this solves both of the issues raised above as well as those raised in other threads.

Thoughts?

[ September 14, 2006, 11:50 PM: Message edited by: ClaytoniousRex ]

[aittam]

Ok is it possible that a scoring system proportional to each side unit in place to kick in when the first enemy unit enter the objective? so that i'll start with my 1000 points, defend from a tactical position w/o loosing points

is it possible that enemy units score points only if they are not crippled past combat readiness (unable to fire and/or move) and if they don't have los of my units?

because most of the time a crippled vheicle is not easy to pick from a dead one (and that's fine) and in some maps a shrike can hide in creeks where is probably stuck w/o beeing able to contribute to the fight and never the less score points

[ClaytoniousRex]

Sure, that makes sense Aittam. We would work total combat effectivess into the scoring of occupiers. So in my example above of 2 MBT's scoring full points, the system would actually be saying "2 mobile vehicles with thick armor and working 120mm guns" to arrive at that conclusion. So if one of them was unable to fire, its contributing score would be far, far less (for example).

And I think you mean they should have LOS to the objective in order to score, right?

[Alexander SquidLord Williams]

Combat effectiveness as a modifier to score rate seems reasonable; done right, it covers extra points for various hardware in the zone (ie. if you derive a generic "firepower rating" for elements in the zone, a Thor and an Apollo 120mm are both going to have a higher value than a Shrike; even an immobilized but still fire-able Apollo will be more valuable than an mobile ion Paladin). Working up the formulea will be a bit of a pain, but that's why you get the big bucks, right?

I'm pretty sure aittam means LoS to his elements, not to the objective. Being under observation implies there are still elements active in the perimeter, so I understand what he means by suggesting it, that the attacker needs to really have control of the site to get maximum scoring, but I think that scaling by "relative element value" probably covers the issue just as well.

[aittam]

I mean that if an attcaking vheicle is hidden inside the objective area but without LOS of it, so that it can not attcak defending units should score points (i was thinking at your example of the secure bridge)

[Alexander SquidLord Williams]

Originally posted by aittam:

I mean that if an attcaking vheicle is hidden inside the objective area but without LOS of it, so that it can not attcak defending units should score points (i was thinking at your example of the secure bridge)

Hmmm, I'm trying to think of any of the maps where you could be in the Objective zone but not LoS to it. Since it's defined as an area of ground rather than as LoS to a point, I think if you're "in the zone," by definition you have LoS to the soil under your feet.

[aittam]

just mentioned that because somebody was complaining about an hidden spot where bots seem to like to hide

[pilgrim]

Originally posted by ClaytoniousRex:

The attacking team's CO would not consider the objective taken if only a single attacking unit had made it into the objective area

If the attackers goal is to "take/seize/control the bridge", then let's examine the satisfaction level of the attacking CO for the following cases:

(Let's define "group" = "2 mobile vehicles with thick armor and working 120mm guns")

0. No attacking groups on the bridge, 1 defending group on the bridge. Completely unsuccsessful for the attacker.

1. 1 attacking group on the bridge, but also 1 defending group on the bridge. In this case, it seems that neither side controls the bridge, so it's "contested control". For the attacker, this is certainly better than the defender having full control, but no CO would say that the job is done.

2. One attacking group on the bridge, no defending groups on the bridge. Now, the attacker has full control of the objective.

So, the situations could result in the following point scheme:

</font><blockquote>code:</font><hr /><pre style="font-size:x-small; font-family: monospace;">Situation Attacker Points Defender Points

--------- --------------- ---------------

0 None Max

1 25% of Max None

2 Max None</pre>

[yllamana]

Yes, pilgrim. I think the main silliness that people object to in the current system is one vehicle inside an objective that is covered in defenders scoring points at a tremendous rate. The best example I can think of is that map with the objective point on the hill in the middle (greenish+desertish map, ion bases in northwest and southeast). If the teams are mostly bots then the attackers nearly always win because the defenders, while they may have superior force on the hill, will never clear all the attacking forces out of the small area to the south of the hill that they don't have LoS to.

With respect to the C.O. reaction thing, I think that depends entirely on what our abstract objective really is. Taking the bridge over Dead Gulch where we decided we need it to move armoured assets between the two bases without dropship support or whatever it was in the other thread, the engagement basically should never end until one side or the other gives up (or the bridge is destroyed). On the other hand, if the objective was to, I don't know, loot something from that point, or to "cause disruption to enemy forces," then maybe having one attacker nearby being fine would make sense.

Maybe what we really need is a more clearly-defined victory condition for each side on any given scenario...

Failing that I think the relative ratio of force strength near the objective is the most important. If you have uncontested control of the objective then your score should go up faster. Maybe mixing in some ratios of distances to the objective too would help (so closer units count for more, within limits).

Then it's not your raw quantity of units on the objective that counts, but the quality of your control of the objective. If you just have one rifle squad sitting on it then that's fine if nobody else is nearby. You have full control and the score should reflect that. If you have a rifle squad on the objective but an enemy rifle squad is closing on you, your control of it isn't so good (it's superior because their squad is further away, but not much). If you're both near the objective then it's the same.

I think that's a lot more intuitable than the current thing.

[Bnej]

Usually a realistic objective might be "take that hill", but the actual intent would be: "don't allow the enemy to shoot us from that hill, and make it so we can shoot at their positions from there".

Or "take that bridge", would actually mean "make it so we can move vehicles safely across that bridge" - the area that "controls" the bridge would be somewhere else entirely.

So, probably, you're having problems finding a solution to the wrong problem. The problem isn't "how should the scoring be?", it's "what should the objectives be?" and "how can you tell they are accomplished?"

Now IMO, a bridge objective where you are aiming to cross should be determined by scoring the number of defenders who can fire on the bridge (Have LOS and weapons with range capability on that bridge - 20mm @ 2000m is worth less than 120mm @ 1200m). To secure the bridge, the attacker would have to reduce the enemy's ability to fire on units crossing the bridge by controlling overwatching positions, rather than occupying the bridge directly.

Because, in modern warfare, sitting on an objective is nothing, it's all about who can shoot at what, where, how soon and how hard.

[ClaytoniousRex]

That's true, Bnej, but it really does depend on the specific story of the scenario in question. The LOS objective as you've described is probably better done as its own victory condition so we have both types available for scenario authors to use as appropriate.

For 1.1.6, the normal Objective game type now uses force ratios for awarding points on the objective(s). So far, I think it's really nice.

[aittam]

anyways Bnej proposal isn't hard at all, is just a matter to move the scoring areas in the right strategic places

[__Yossarian0815[jby]]

Hmmm, realism this, realism that. from a game fun point of view it´s important that an objective can be taken and retaken inside half an hour. (I havent´t played v116 yet so this isn´t meant to be criticism!)

@victory conditions: would it be possible for the scenario deginger to define exactly what territory needs to be occupied? (and this would be cleartly visible in the tac map) i.e in the level with the ridge in the middle, the designer could say its important to hold the northern slope only or somefink.

[Neutrino 123]

I would like to reiterate and expand upon my previously mentioned scoring suggestion.

First of all, I still think that the attacker and defender should have identical victory conditions.

“They only have one vehicle here and we have five!” might be an undesirable response to “Have you successfully defended the objective?”. However, “Well, we have one hiding vehicle there and they have five!” is an even more unsatisfactory response to “Have you successfully captured the objective?”

Thus, I think that ratios are the best thing to determine objective control. I will refer to units as “points” sometimes here for reasons that will become clear in due time...

The absolute number of units near the objective should not matter. If one side has only one unit near the objective, while the other side has nothing, then full control should be awarded. This should prevent objective clumping in nearly all circumstances.

However, when two sides have points at an objective, the control should be based on the number of points each side has. If one side has more points then the other side, then the side with more point should get points according to the following formula:

Rate of point gain = (normal gain)*(# of points in majority side - # of points in minority side)/(total # of points)

Finally, so that it is important to completely clear an objective of enemy units, not just have a majority there, the maximum number of victory points one could gain from controlling an objective would be:

Maximum victory points = (normal maximum)*(# of points in majority side - # of points in minority side)/(total # of points)

To rebalance the scenarios, the number of each unit type that the defender ahs could be reduced by ¼ or so.

Now for the points. Each unit should have an individual point value, which would count for objective capturing AND kill points.

That’s right. I am suggesting that unit kills play an important role in scoring. In certain scenarios, they rarely will have an effect, but this only happens if the two sides are very close at the end of an objective game, or each side ties in a building objective or flag capture game.

I believe it should be possible to win an area objective game even if the opponent has full control over the objective at the end based on a greatly superior kill ratio. “Pyrrhic” victories are not really victories at all.

We could keep the 1000 point value for the objective games, and maybe ramp up the flag capturing value to 200 or 500 (or somewhere in between).

Each unit could have a different number of points. For example, here are some numbers that may be roughly balanced for a 1000 point objective scenario:

Thor-120 50

Thor-ion 40

Thor-mortar 40

Apollo-120 35

Apollo-20 25

Apollo-ion 30

Apollo-mortar 30

Paladin-76 25

Paladin-20 20

Paladin-ion 25

Paladin-mortar 25

Paladin-ATGM 20

Hermes/Bacchus 35

Cutter 35

Command Vehicle 50

Shrike 15

Hurricane 50

Tempest 35

Turret 5

Infantry specialist 2

Infantry rifleman 1

Sensor 1

Sensor jammer 1

Viper 150

Dropship 100

So for example, if one side has the objective, but loses the equivalent of 20 MORE Thors then the enemy, they will still lose the game, though probably not by much (could different ‘levels’ of victory be introduced?)

While I’m at it, I may as well make some more suggestions in other areas:

The command vehicle and to a lesser extent, the Cutter, seem to be VERY resistant to penetrating shots. This is surprising for the command vehicle since it is so cramped, so it should be quite vulnerable to penetrating damage, but it seems to much for even the cutter (in most instances). For example, I put 50 76mm AP rounds into various points of a cutter’s side armor, and then put 5ATGMs into it, but all I did was immobilize it :eek: . It committed suicide before I could give it the finishing blow...

I think making these things a little more vulnerable won’t hurt...

Thors really need to be more fearsome. Bump up their side armor so that it’s immune to 20mm at all ranges and angles!

I think this has been mentioned before, but 76 and 20 millimeter HE rounds are useless against units. I even prefer 76mm AP for killing infantry at close range...

I still don’t think the 76mm is powerful enough. It flies and penetrates well, but it doesn’t do enough damage. I speak not on an absolute scale, but compared to the 20mm. A three round burst of 20mm will often kill a unit from the side, but it takes a similar number of 76mm rounds to do the same thing! A 76mm round should have at least 50x the mass of the 20mm, unless it is odd-shaped or a sabot, but even then it should be much more. The 20mm round’s greater velocity should not be sufficient to give it roughly equal killing ability, especially since we know that the 76mm penetrates more armor in the first place.

Wasn’t the 76mm supposed to “overpenetrate”, or was that changed?

Overall, I recommend that the 76mm retain its current penetration, but gain improved damaging ability. If this unbalances units too much (right now I think that the 76 unit is still weaker in game terms then others), then its rate of fire should be reduced.

[Alexander SquidLord Williams]

So for example, if one side has the objective, but loses the equivalent of 20 MORE Thors then the enemy, they will still lose the game, though probably not by much (could different ‘levels’ of victory be introduced?)

I think this is a bad idea. Not the "levels of victory" idea, per se, though that's hard to crunch so that you actually consistently see some variation, but because you're designing from a "realism / logic" point of view and not a "designer / game dynamic" point of view.

To wit:

If, as the Attacker, I don't have to go into the Objective area, why would I ever do so? It's dangerous, the Defender has vast fixed emplacement advantage (ion towers, other pre-positioned assets), and the advantage of terrain, in most cases. If I can possibly get a scenario win out of sitting at range and sniping, trusting to 120mm HEAT and AP crit strikes to pop targets, why wouldn't I? There's no advantage to risking my assets like that.

Likewise, for the Defender, there's no reason to go hunting those Attacking assault snipers at range, because doing so puts your forces at even greater risk, rather than just digging holes and berms and hunkering down. They're not coming in after you, and you're not going out after them.

It becomes a game of turtling, because there's no advantage to not doing so.

Not really an Objective game at that point; really an anti-objective game.

Thus, the Objective points have to be worth more than 2-3x the points you can get from kills, total. Kills them become a discriminator, but not the discriminator.

I think making these things a little more vulnerable won’t hurt...

It's hard to get people to drive a Cutter now. Slow, no real weaponry, and it doesn't do that much of interest. It requires an escort of elements, which can be hard to coordinate. Make it any more tissue-paper, and you can just write off capturing outposts at all; it just won't be worth it.

The Mercury is best plopped down between buildings or over a hill with a Jammer, and it's best use is just calling down turrets. Which, basically, are damned useless, because the bots put rounds through them before they can fire even once, even ignoring nearby Apollos or 76mm Paladins to do so. Without the ability to pop off mine zones in front of attackers, the only thing the Merc is good for in a real sense is dropping, placing arty and/or smoke, then extracting. A commander in a Merc is effectively a pretty wasted asset other than that, save in the late period when the Attackers are waving into the Objective in Paladins and Shrikes, having depleted heavier assets. Then the 20mm can be brutal.

Thors really need to be more fearsome. Bump up their side armor so that it’s immune to 20mm at all ranges and angles! [smile]

Uh, no, please. Especially for the Attackers, who generally have to go head on with Thors on the Objective out of the gate. I don't mind that 20mm can pop a hole in the back of a Thor turret. It's hard enough to get a decent flank a lot of the time.

I still don’t think the 76mm is powerful enough.

This I'll agree with. The 76mm either needs a 3 round burst mode, or to do more damage overall per round. I'd actually prefer the burst. And the HE needs a limited but definite explosive radius, since at present it just does nothing worth discussing to infantry nor buildings.

[aittam]

Originally posted by __Yossarian0815[jby]:

sorry for the OT but Yossarian are you playing BF1942 on the moongamers servers lately?

[ClaytoniousRex]

Neutrino, that's how the scoring is working now except that it doesn't clamp to a maximum number of points. This is for the reasons outlined by Alex: the objective has to remain the focus for that game type.

The Mercury and Cutter do, indeed, seem to absorb an awful lot of hits at the moment. Will get on the test range and see what I can find out...

jby, do you mean some kind of irregular shape for the objective area?

As for the 76mm, we've been very close for a long time now to making it overpenetrate. This alone would probably give it that little bit of extra "ummph" that it needs.

[Alexander SquidLord Williams]

jby, do you mean some kind of irregular shape for the objective area?

Actually, if we could define areas with PNG masks, that might be interesting, if only because normalizing from 0-255 (pixel range) to 1-100 (percentage), you could theoretically determine how many points a given zone is worth holding, meaning it might reasonably be worth setting up an Objective game where holding the center position is worth 100% of points-per-second and the ion outposts are worth 25%. Holding both outposts wouldn't be a winning position, but it could certainly force the Defender to fight for them.

As for the 76mm, we've been very close for a long time now to making it overpenetrate. This alone would probably give it that little bit of extra "ummph" that it needs.

I think this needs to happen. As stands, given the choice, the 20mm chassis are always the better choices, not only because they do more overall damage but because they're just scarier. I don't worry overmuch at the thud-thud of a 76, but the chatter of a 20mm makes me hustle.

[konstantine]

RE: the 76mm. Two things would improve this weapon without giving it uber status: one, a canister round. Two, a very limited one-shot kill capacity versus other Paladins. Say, a very well-aimed shot to the side, low and perfectly center.

[poesel]

Yes, a canister round instead of the useless HE would make it an infantry killer.

[konstantine]

Actually come to think of it a more modern or futuristic anti-infantry round would incorporate flechettes (the IDF used them in Lebanon--5000 flechettes per 120mm round!). But pretty much the same thing.

[Neutrino 123]

ClaytoniousRex, I am not sure what you mean by “clamping”. In the area objective battles, isn’t the maximum points one could get clamped at 1000?

Alexander SquidLord Williams, if the attacker adopts a stand-off policy, then there is no reason the defender could not do the same, with the added advantage of base terrain and laser turrets. Thus, the attacker could try to achieve a far better kill ratio under these conditions, or actually try to secure the base while to having massive losses, the success of which would then give a significant advantage.

Basically, what I want to see in area objective battles is for kills to matter, so if one side does a LOT better then the other side in kills, but still loses the objective, they might be able to eek out a victory. I don’t want to think, “Oh, that computer is on the other side of the map. Killing it would be counterproductive since then it might drop closer and immediately be a threat,” or “I will leave that partly disabled guy right there, since killing him would make the computer drop a mobile units, which could be dangerous.”

I’ve found the Mercury to be very effective, since it can basically serve as a tracked, fast 20mm vehicle. Sure, it does not have the speed of the 20mm Paladin, but it tracks and low profile make up for that. There is no need to additionally make is super-hard to destroy. As a side note, did you know that artillery is much more accurate if you call with a line of sight? HE artillery usually don’t do much except to very light vehicles, but a well placed EMP barrage can turn the tide of a game.

I’ve found it to be very easy to get flank shots against Thors with the 20mm. At any rate, Thors are only around for the first part of the game, at which time both sides have things that can kill the Thor from the front, much less the side. If the 76mm did more damage, it would be a good choice to conduct flank attacks against the Thor.