Unit experience in MP

[ragnar]

Greetings,

To improve the MP game, I propose a rather drastic reduction in the rank requirements, but *only* for MP games and skirmishes. The rationale is that right now, the requirements are so high that it doesn't come into play at all.

I think its rare to ever get a unit above "regular" in a MP game ( http://guide.wz2100.net/experience ). For reference, it seems they did a similar overhaul in Supreme Commander with the introduction of the Forged Alliance expansion ( http://supcom.wikia.com/wiki/Veteran_status ). Even now in FA MP, you don't see a lot of level 5 veterans, but it's at least not completely non-existent. Browsing through the forums, it seems there's already different code for handling experience in MP, so the implementation could be fairly easy.

So, let's raise unit experience in MP games from its obscurity. It's in the game, so I reckon it ought to be a part of it, and right now, it just isn't.

regards,
Ragnar

[Olrox]

Yesterday, by using a commander, I've managed to raise it to Veteran, almost elite, in a 2 hourse game against 5 simple AIs.

Is it limited to the current numbers because they're potencies of 2? If so, reducing them by 50% would be insane. But I agree that, if it can be made, a reduction of 20% would be nice to see:

80% reduction for MP would be reasonable, IMO. Values are rounded-up, of course.

ranks.PNG (7.81 KiB) Viewed 9763 times

The point of the reduction is, you don't get any experience from previous games like in campaign. For more intense gameplay, I think it would be great to have this reduction! Maybe even poll that if the devs think it's not going to ruin balance - if it is most probably going to, then let's forget it.

Another option would be to raise the bonuses in MP :rolleyes:

What about making a Newbie ranking for units that didn't kill anything in 2 hours? xD

Hehehehe, would be fun :P

Newbie.PNG (2.05 KiB) Viewed 9762 times

[Zarel]

Browsing through the forums, it seems there's already different code for handling experience in MP, so the implementation could be fairly easy.

Yes, and that code is causing TONS of bugs. I just fixed a commander bug relating to that, and the fix apparently is causing problems in campaign...

Although I think I know how to fix that problem too, and after that, sure, your idea sounds good. How's just halving the experience requirements for everything?

[Olrox]

Although I think I know how to fix that problem too, and after that, sure, your idea sounds good. How's just halving the experience requirements for everything?

I thought it might be too drastic, but we could test that, yeah

[ragnar]

20% , 50%, yeah, we could try that, but I'm pretty sure you need 75% ("green" requiring 1 kill and "hero" 128) to produce an effect in multiplayer games. And with "produce an effect", I mean (slightly) change the way people play. Remember that unlike in a skirmish, when playing an actual human counterpart, clocking up kills is no easy task.

Yesterday, by using a commander, I've managed to raise it to Veteran, almost elite, in a 2 hours game against 5 simple AIs.

That sortof proves my point. You took 2 hours against a swarm of noobs to almost get to elite. Divide the requirements by 4 and you would have had a hero instead. Sounds about right and you know using a commander against human players just means its gonna die quickly.

Alternately, make it a linear instead of exponential progression. That means 4 kills for green, 32 kills for hero. Of course, that may be a little too easy, so double the base number as well (green 8 - hero 64).

[Saberuneko]

I think exponential is good, if not, we would suffer spam of heroes... and that isn't very cool...

I think that reducing in the way to hero = 128 kills is good.

[Olrox]

I think exponential is good, if not, we would suffer spam of heroes... and that isn't very cool...

I think that reducing in the way to hero = 128 kills is good.

Hmmmrlfhf...
But... it's a Hero! It is the highest rank attainable! I think it should be mantained a unique thing. Even in campaign, it is hard to get, let's say, even 5 heroes.

50% is more reasonable, IMO. 128 kills isn't too big of a deal, especially on maps with scavengers.

[Assault Gunner]

Heck, getting one hero unit is hard in campaign, let alone five!(Even counting commanders)

[Saberuneko]

But we are talking about changing that only on MP...

[Olrox]

But we are talking about changing that only on MP...

I know, I know. What I'm saying is that, given the bonuses a Hero unit get, it should be a very hard thing to achieve.
The problem is that in the current state, you have the same requirements as the campaign in a single MP game.

The kills from a campaign run should be taken into account - a player that gives an effort to search & kill most enemy units will hardly get 3 heroes. The average kills from a 2 and a half hour game (which I believe is an average for 8-player game) should be enough to get you with 3 heroes in the best situations, IMO. Longer games might get you more heroes, shorter games might not even get you any heroes (which is VERY common when playing campaign without caring about making your aces retreat or destroying neutral buildings like each scav wall piece and huts to farm kills.

Always remember that a map that has been made with fairly strong scavengers will get you a higher kill count. Especially if you use a small team and retreats to repair often.

My point is, if you are going to get around 15 heroes in a MP match, your enemies will very likely do so. If you are going to get hardly 3, odds are, he's also going to. If it's too easy to get a large number of heroes (I think that 128 is really easy), you won't get that good feeling you have when you see that silver insignia turn into a star.

It's like reducing epic items' drop rate in WoW - items will no longer be epic that way

[Zarel]

Yes, that's true. There's also an issue that high-experience units don't have balance problems in campaign, but in multiplayer, they might be more of an issue.

Might be simpler just to have campaign and skirmish use the same experience system.

[Olrox]

Might be simpler just to have campaign and skirmish use the same experience system.

You mean, using the same system but with requirements reduction?

I imagine it would be simpler, but really think that it should be quicker in MP, due to the drastical differences in duration.

[ragnar]

While glad to see a serious discussion, but I'm, not sure all the points made are valid.

I've found it relatively easy to get a Hero or two in the campaign ...if I want to. However, there's no reason to get misty-eyed about a Hero. Yes, it's great, but if the table is correct, it's only 3.7% more "awesome" than a Special and it requires twice the effort. Usually I stop caring when I have most of my team at Veteran.

To be honest, I really don't see how getting 128 kills and surviving is easy in *any* game.

A real issue though, is that we'd just be promoting all Greens to Regulars, which wouldn't really influence anything except the graphics on the screen. I always go by the assumption that in a game, every effect must have a cause that is either random or by player choice, or a combination of both. Lowering the kill requirements only raises the effect to a level where it is noticeable, it does not necessarily "do" anything.

Hence the suggestion for a linear curve as opposed to an exponential one. An exponential curve may seem attractive because it prevents a situation where Professionals outnumber Rookies. However, first: that may not be all that unrealistic and foremost: it just won't happen. To amass kills, a unit must fight and when it fights, it will get killed.
While in a campaign or AI skirmish you can perhaps micromanage a way around getting killed, there's no way to do that in multiplayer without seriously pulling your attention away from something else. So I'm assuming things will even out and experience from other games (SupCom) suggests this is a fair assumption.
So the trick is to create a reward for a style of play that will amass kills, so that that style of play becomes more attractive. What I remember most from MP games is hordes of mantis cannon hovers. It would be nice to encourage the use of a more robust and durable unit and to discourage zerging, if only a little.

Anyways, perhaps a poll? Candidates:
000 - 004 - 008 - 016 - 032 - 064 - 128 - 256 - 512 Oldskool (100%)
000 - 003 - 006 - 013 - 026 - 051 - 102 - 205 - 410 Olrox (80%)
000 - 002 - 004 - 008 - 016 - 032 - 064 - 128 - 256 Zarel (50%)
000 - 001 - 002 - 004 - 008 - 016 - 032 - 064 - 128 Ragnar (25%)
000 - 008 - 016 - 024 - 032 - 040 - 048 - 056 - 064 Linear
000 - 004 - 008 - 012 - 020 - 032 - 052 - 084 - 136 Fibonacci

All numbers x4 for commanders, of course. Personally, I'm liking the last two the most as the moment

[KukY]

000 - 004 - 008 - 012 - 020 - 032 - 052 - 084 - 136 Fibonacci

Looks very good to me.

[Olrox]

Well...
Actually I'd rather have a Fibonacci progression x 1.4, or mantain 50% reduction. Special and Hero units are for epic fights, and are really something to zeal for - Having them with particular heavy repair units & retreat at heavy damage, IMO.
Focusing on farming exp or swarming, or making something between those options, is a choice the player must make. I personally think that 128 kills is very easy to achieve, especially on a 8c game, if you aim for farming kills.
Ah, here's a graphic I've made to better illustrate my idea:

Graphs are nice.

As you can see, the 50% reduction mantains an exponential curve which goes steeper and steeper in a very accentuated rate (especially after "elite"). The gap between Special and Hero is huge, more than 2 classes - 128 kills.

The Fibonacci progression is much more shallow. Too shallow IMO, as the 6 first ranks are in the same class, the seventh barely fits the next class, the eighth rank is on the same class as well, and the hero is the only rank in a distinct class.

If we multiply the Fibonacci curve by 1.4, it turns out to be more steep, but keeps the smoother effect in relation to the exponential curve. The 6 first ranks are on the first class, but veteran is 5 only kills from the next class. Elite fits its own class, being almost on the half of it. Special fits its own (special) class as well, being to too far from the middle of its class as well. On the top of the curve lies the heroes, with under 200 kills (which is reasonable for a unit that has a decent ammount of concentrated kills) and only 30% diverted from its class' median.

I really hope that this helps you to see why I think we should opt either for 1.4xFibonacci or, If someone finds something that compromises it, 50% reduction.

~Olrox