QUOTE (NMS @ May 3 2010, 08:29 AM) <{POST_SNAPBACK}>

You could nerf missiles by having some of the fights take place in systems with asteroids and/or interference. You could also allow point defense, which would encourage people to use at least 2 different weapons.

I just want to point out asteroids wouldn't work. In fact, they would cause Qaanol (or any human observing the fight) to directly alter the outcome if he came close enough to either competitor or their missiles. Asteroids only appear near the player. If the player isn't there, asteroids won't be there. If an asteroid drifts far enough away from the player, it disappears.

I noticed this when working with asteroid miners in HOTS. My version of the mining ship is much slower so it's easier to see this when it happens. If they're chasing an asteroid that's faster than they are, and I'm sitting in more or less one spot, the miner will eventually turn around and target a different asteroid because the previous one disappeared due to me not being close enough to it.

This is why I think the miners in EVN were so slow, too. If they weren't, they'd be constantly following the player because that's where all the asteroids would be. You can test this yourself, you don't even need a plug-in. Load up EVN and get a fast ship. Go to a system you know has miners in it (I went to the Archenar system). When a miner appears, fly away from them and keep them targeted so as to help you know where on radar they are. They will continuously fly in your general direction because that's where the asteroid they're targeting is located.

As a side note, miners will sit and do absolutely nothing if in a system void of asteroids.

QUOTE (Qaanol @ May 3 2010, 12:55 PM) <{POST_SNAPBACK}>

I am under the impression that larger values of MaxIonize mean that as you get ionized the effect it has on your ship is diminished, but it does not actually cap the total ionization you have received.

You should probably test this to make sure. If there's really no cap, ships could end up ionized for a very long time.

QUOTE (Qaanol @ May 3 2010, 12:55 PM) <{POST_SNAPBACK}>

CustSndID.

Oh, make them come out of hypergates. Clever.

QUOTE (Qaanol @ May 3 2010, 12:55 PM) <{POST_SNAPBACK}>

I had thought about capping the points that can be spent per-weapon. What do people think?

Probably a good idea, but people could just buy multiple copies of the weapon or make multiple identical weapons, depending on how those options are priced.

QUOTE (darthkev @ May 3 2010, 03:04 PM) <{POST_SNAPBACK}>

Asteroids only appear near the player.

I should have guessed that that was the case from their movement patterns. I suppose that's out then.

QUOTE (Eugene Chin @ May 3 2010, 02:18 PM) <{POST_SNAPBACK}>

If you don't want long-ranged ships dominating, you need to change the costs to make range proportional to the point cost, not to the square of the point cost.

Indeed, I was aware of this as I wrote the rules, and I considered a variety of options, eventually choosing to make the point-costs as simple as I could for each field. Ideally there ought to be diminishing returns on range. One way would be to charge for range directly, although that makes calculating point costs involve a bit more calculation.

On the other hand, if guided missiles are made sufficiently pricey (perhaps the cost of choosing a guidance type could come in the form of multiplying the point-cost of the “count” field by some value) then any sort of inaccuracy automatically causes a diminishing value to be gained from increasing range.

Also, on the same note as why it made sense to have equal numbers of points invested in count as in speed, it also made sense to have the same number of points invested into reload as into all of the following combined: shield damage, armor damage, ionization, recoil, impact

QUOTE (Qaanol @ May 3 2010, 06:06 PM) <{POST_SNAPBACK}>

Indeed, I was aware of this as I wrote the rules, and I considered a variety of options, eventually choosing to make the point-costs as simple as I could for each field. Ideally there ought to be diminishing returns on range. One way would be to charge for range directly, although that makes calculating point costs involve a bit more calculation.

Another alternative would be to fix the weapon speed for each given guidance type. That way, the players only pay for weapon lifespan, the point investment becomes proportional to the final range, and you could set the values so unguided shots were faster than guided weapons.

QUOTE

Also, on the same note as why it made sense to have equal numbers of points invested in count as in speed, it also made sense to have the same number of points invested into reload as into all of the following combined: shield damage, armor damage, ionization, recoil, impact

May I ask how you arrive at this conclusion?

My own analysis reveals that is it strictly not worthwhile to invest equally into shield and armor damage.

Remember, ships disable at 33% armor and can't fight back after that. That's the minimum for a win. This caused everybody but me to forego the use of armor in favor of shields.

If some sneaky schmuck had entered a contestant that used purely energy damage, and only one unit armor damage, he would have been able to kill the other contestants using far fewer hits (assuming he had a strategy that could hit both of them).

As for recoil and impact, this is only valid if you're not coming under immediate fire. Take the case of my official submission vs Chamrin's submission. Impact could force Chamrin's ship away from Hit_And_Fade, but Chamrin already has shots in the air by that time. The only thing that would help Hit_And_Fade escape that is recoil (and Speed ). Impact would not help there.

If I had diverted points from recoil to impact on Hit_And_Fade, it's entirely possible that Hit_And_Fade could have been hit by the initial volley from Chamrin. If you need a demonstration of this, put Hit_And_Fade up against Hard Snipah, and reduce Hit_And_Fade's recoil by 1. Hard Snipah will hit.

This post has been edited by Eugene Chin : 03 May 2010 - 08:15 PM

QUOTE (Eugene Chin @ May 3 2010, 07:25 PM) <{POST_SNAPBACK}>

Another alternative would be to fix the weapon speed for each given guidance type. That way, the players only pay for weapon lifespan, the point investment becomes proportional to the final range, and you could set the values so unguided shots were faster than guided weapons.

That is a worthy idea.

QUOTE (Eugene Chin @ May 3 2010, 07:25 PM) <{POST_SNAPBACK}>

May I ask how you arrive at this conclusion?

My own analysis reveals that is it strictly not worthwhile to invest equally into shield and armor damage.

Ah, I apologize if I was ambiguous. What I intended to say is this: suppose your weapon has A points in reload, B points in shield damage, C points in armor damage, D points in ionization, E points in impact, and F points in recoil.

It is then optimal to have A = B+C+D+E+F. This is because, each of the things on the righthand side gives some benefit per shot (how much is up to the designer to decide) and the thing on the left multiplies how many shots are fired in a given time interval. Assuming the designer has allocated points optimally between B, C, D, E, and F for their purposes, then the rate of benefit is proportional to the product A(B+C+D+E+F), which for a given total number of points is maximized when A = B+C+D+E+F. It is another case where the benefit is proportional to the square of the points allocated.

Thus if you invest heavily in range, you benefit a lot by the squared return-on-investment, but your damage rate is cramped by a different square-law. On the other hand if you invest heavily in damage, you benefit a lot on that square-law, but have severely reduced range. On the third hand, if you try to make some sort of hybrid, with medium range and medium damage, the result should be much weaker than either specialized style. So if I fix the range issue, I will also have to fix the reload issue, otherwise there will be only one local extremum in the design space. I’d like to fix both.

I don't really have time to respond to each point I want to respond to, but I'll try to make some good points none the less.

As for the contest itself, contrats Eugene. I did an analysis of the recoil ships too and came to the conclusion they'd have problems against more heavily armored opponents due to some shots running out of steam if there wasn't sufficient range. I also expected more entries and for those entries to put a damper on the recoil ships making it far into the playoffs or whatever determined the winner. Where as my ship had enough fire power (probably a bit too much) to destroy any possible ship if all the weapons hit.

As for the next contest, a 2000 point cap is nice, to change things up. I'd caution against adding too many things. Asteroids are no good as kev mentioned, they're highly dependent upon the observer. Jamming might work, or interference.

I would probably half the costs of shield/armor regen as it is now it's pretty worthless, even with double regen, it's still mostly worthless, just a little less so.

Keep armor/shield costs at 1 each, should consider this a baseline.

Possibly make ammo proportional to the damage done, so 1/10th or 1/20th per damage.

As for proximity and blast radius, just keep it simple, proximity = blast at all times, or blast is always 0. Cost is proximity^2.

As for homing weapons being too common, allow point defense (turrets only). Maybe disallow firing at enemy ships with point defense weapons, or maybe not, might want t investigate that.

You'll want to also allow the durability field with some cost eqn otherwise there won't be any homing weapons.

QUOTE (Qaanol @ May 4 2010, 06:46 PM) <{POST_SNAPBACK}>

QUOTE (Eugene Chin @ May 3 2010, 04:25 PM) <{POST_SNAPBACK}>

Another alternative would be to fix the weapon speed for each given guidance type. That way, the players only pay for weapon lifespan, the point investment becomes proportional to the final range, and you could set the values so unguided shots were faster than guided weapons.

That is a worthy idea.

Bad idea, fixing max speed makes it really easy to know just how fast your ship needs to go to evade the weapons.

Also if you change speed+lifespan so that the total range is linear, you'll end up with really really fast weapons that get there almost instantaneously but have a really short lifespan. Better to keep them decoupled and open to all values. If 2x per point of each doesn't work, try squaring it and dividing by 10 or something.

I would keep it so firing at the opponent when both start in system is reasonably feasible, while when they jump in from other ends they should not be able to fire on each other.

Finally, I'd suggest not letting the next final submission day be 2 months away. Hammer out what the new point costs/rules are going to be over the next week or two and say final designs are due by the end of may. Also design your own ship and call him the final boss or something, and see how it does against all the submissions, you'll have some fun with that I think.

Cham

Thanks for putting this together, Qaanol. Very fun. I wish I submitted my idea now, because I think it would have fared well against the selection. There's always next time.

I am also in favor of shortening the deadline to the end of May or so. Whether you give people two weeks or two months, people will always work at the beginning of the contest, or right before the deadline.

Any word on the next round?

I’ve been a little busy with real life. Graduating from college, moving, and some other things. The most likely scenario is I will get bored one weekend and crank out the revised point values for the 2k contest. Probably won’t be until next month though. Although, if you feel like writing up a point system, then I can look it over and maybe be ready sooner.