Attack Times

[Monroe]

I've been watching what attack times allies choose to use and it's been an interested progression over the last couple of years. 4 years ago most attacks were launched 22:00 GMT, 2 years ago a lot of attacks were launched 3-4 GMT over the course of the last 4-5 rounds or so I've noticed it continue to shift forward, and now most attack seem to launch between 5-7 GMT, and I've seen some attacks launching as late as 12 GMT. While some of this shift is clearly because there is an advantage in landing last, I am beginning to wonder how long this slow shift will continue. Also I'm not sure the 'landing last is better' philosophy is entirely to blame for this shift. I wonder if some of the shift has to do with the changes to PA game demographics, or attack philosophies, or the simple change made to the game to allow prelaunches. I wonder if anyone else has noticed this slow shift, and if they have any better explanation as to why this is happening. I don't feel the shift is necessarily go or bad, but it'll be really funny (at least to me) on the day we've gone the whole 24 hour cycle are are back to launching attacks in the European evening.

[Zaejii]

heh. being in America, the attacks have gone at night from what i've seen. make a rule so i can sleep for a round w/o having to be up handling reporting for everyone since the Europeans are asleep! lol.

[JonnyBGood]

Four years ago most good alliances still started launching their first waves at 3am. Launching around 7am alliance def is likely to be tapped out. People are really just trying to find the optimal time to attack at but truth be told that's largely contingent on other factors and it's not going to be the case that the ideal time for everyone to launch at will be the same.

[Jester]

6 years ago, in round 4, Cray told me that CELL rarely launched before 6:55 CEST*.

In round 3, however, you had to be insane not to launch just before the game went down for burp tick, as you got a 20 minute tick when the game came back up.

On a whole, however, it's all about what you're trying to achieve. Plain gal raids are best launched late, so that people's fleets are out and unable to return in time. Alliance raids are best launched early**, so your own members don't have incoming already, and so that the targeted alliance loses initiative and spends their fleets defending.

If you're solo raiding, it's best to launch when you're awake, so you can recall/resend and reuse your fleet. Your fail rate will be higher when launching solo, but you can compensate for this with fakes and resending.

* Which was gametime back then.

** Initial launches early, probably good to have additional later launches/waves of course.

[Monroe]

Originally Posted by Jester

On a whole, however, it's all about what you're trying to achieve. Plain gal raids are best launched late, so that people's fleets are out and unable to return in time. Alliance raids are best launched early**, so your own members don't have incoming already, and so that the targeted alliance loses initiative and spends their fleets defending.

I'm going to disagree with you on Alliance raids, purely on the point that most top ally players attack with two fleets and defend with one regardless of whether they get incomming or not. I perfect example of why ally raids work just fine is CT, who afaik launches the latest attacks of the bunch atm, yet they seem to be doing just fine right now. So launching ally attacks late is not always a significant handicap, unless you want to infer they would be kicking butt this round if they launched earlier?

[Tietäjä]

Originally Posted by Monroe

I'm going to disagree with you on Alliance raids, purely on the point that most top ally players attack with two fleets and defend with one regardless of whether they get incomming or not.

In fact, a lot of top players recall their attacks should they get incomings and perform inter-galaxy defence maneuvers. It's really case-dependant, and definately not a "most".

[Monroe]

Originally Posted by Tietäjä

In fact, a lot of top players recall their attacks should they get incomings and perform inter-galaxy defence maneuvers. It's really case-dependant, and definately not a "most".

Well I calculate that in order to be in a t50 position in a non XP dominated round you need to launch on average somewhere between 1.5 and 1.75 attacker per day for the entire round. This is because unless you have some fore knowledge of who will and will not get defense it is tough to land more then 1 attack per night on average. Therefore in many cases (though certainly not all) a top player will need to keep attacking even if he gets incomming in order to get the XP and roids he needs to stay on top. That being said of course there are plenty of exceptions to the rule, but I feel my 1.5 to 1.75 number is about right. Therefore my original position of "most" (ie significantly over half) is an accurate description of the state of affairs.

[Rinoa]

Personally Ive been using the happy medium of 5am gmt launches for a good while now.

[Tietäjä]

Originally Posted by Monroe

Well I calculate that in order to be in a t50 position in a non XP dominated round you need to launch on average somewhere between 1.5 and 1.75 attacker per day for the entire round.

Care to share your calculations?

[Jester]

Originally Posted by Monroe

I'm going to disagree with you on Alliance raids, purely on the point that most top ally players attack with two fleets and defend with one regardless of whether they get incomming or not. I perfect example of why ally raids work just fine is CT, who afaik launches the latest attacks of the bunch atm, yet they seem to be doing just fine right now. So launching ally attacks late is not always a significant handicap, unless you want to infer they would be kicking butt this round if they launched earlier?

Comparing current alliances to 'optimal' alliance behavior is pretty ludicrous.

[Monroe]

Originally Posted by Jester

Comparing current alliances to 'optimal' alliance behavior is pretty ludicrous.

Why? Optimal behavior is by definition the behavior that leads to the best result. If an alliance is currently following a strategy that is leading them to a winning position, even if it is somewhat counter intuitive, why is it a bad comparison?

[Jester]

Originally Posted by Monroe

Why? Optimal behavior is by definition the behavior that leads to the best result. If an alliance is currently following a strategy that is leading them to a winning position, even if it is somewhat counter intuitive, why is it a bad comparison?

You are, of course, correct. But there's a clear difference between 'winning this' and 'theoretically optimal behavior which would have given the apriori maximized probability of winning the round'.

Originally Posted by Monroe

Well I calculate that in order to be in a t50 position in a non XP dominated round you need to launch on average somewhere between 1.5 and 1.75 attacker per day for the entire round.

I have no idea how you came to this number, but under the assumption that by "average" you mean that this is a statistical figure of some sort, rather than a necessary criteria, you are again (probably) correct.

However, you have again drawn an incorrect conclusion from this. Consider:

Planet A, B and C are in the same galaxy. They have 2k roids each. Their fleets are big, so they usually get through on about 30-40% of their attacks. Each night they attack two targets of 500 roids each. Their expected cap: 300-400 roids with 2 fleets.

Assuming a roid saved is worth the same as a roid gained, the expected value of single fleet covering a wave of incoming from these planets on eachother is 1k roids for 2 fleets.

It should be fairly obvious that these planets are better off 3-fleet defending when they have incoming.

Many planets will cancel attacks, organize ingal defense, and then launch new attacks in the afternoon when their fleets have been freed up. These attacks will be recallable early in the evening if they're defended, freeing up the fleets for ingal defense again.

Out of curiosity, have you ever held a top planet or been in a top galaxy?

[Zaejii]

well if the waves they purposefully put in front of them to draw def didn't crash on defense (thus dropping the target's value, and stealing a few roids in most cases) then by the time they landed they probably would be getting more score / xp / value out of their late attacks.

[Ultimate Newbie]

Wouldnt launching in the middle be best? Because then much of the defence has already been soaked up, but you tend to be first into the waves that actually land, thus the highest roid caps are directed to you, rather than later roid caps which are smaller?

[Makhil]

i've not landed an attack sent after 6 am, but landed a lot when sent between 3 and 4 am... the main problem is never ally def, but gal def.
I have the feeling it's the semi active planets which are the main threat : they wake up comfortably around eta5/6 and have their 3 fleet slots ready for you

[Tietäjä]

Originally Posted by Makhil

I have the feeling it's the semi active planets which are the main threat : they wake up comfortably around eta5/6 and have their 3 fleet slots ready for you

I wake up comfortably at 3-4am and I haven't launched an attack in two weeks or so!

[JonnyBGood]

Once you have roids it's more important to keep them than to cap more ones if you gain more score with the defence than the attack. You can also use your fleet to help others by forcing people to spend or sucking in defence etc. You can land attacks at any time really, if you've noticed a good bunch of elviz's roidcaps this round seem to come after midnight and before 6am.

[Tietäjä]

Out of curiosity, where are the other at least three top100 finishes in the PAX period? Can't find them.

[Monroe]

Originally Posted by Tietäjä

Out of curiosity, where are the other at least three top100 finishes in the PAX period? Can't find them.

I can't honestly tell you, I don't keep track of my finishes. But it seems to me it's completely irrelevant to the discussion of this thread anyway.

[Tietäjä]

You do realize, what an ass you make yourself appear like, when you exaggerate your record? It gets hard to tell when you're actually on facts and when you're just exaggerating something to make your point.

[Monroe]

Originally Posted by Tietäjä

You do realize, what an ass you make yourself appear like, when you exaggerate your record? It gets hard to tell when you're actually on facts and when you're just exaggerating something to make your point.

I have made no purposeful attempt to exaggerate anything, you tend to not to want to believe me no matter what I say, so I have no need to prove anything to you.

[Gerbie2]

Originally Posted by Tietäjä

You do realize, what an ass you make yourself appear like, when you exaggerate your record? It gets hard to tell when you're actually on facts and when you're just exaggerating something to make your point.

You do realize, what an ass you make yourself appear like, when you make remarks like that when you present no proof whatsoever someone is exaggerating something?

[Tietäjä]

Originally Posted by Gerbie2

You do realize, what an ass you make yourself appear like, when you make remarks like that when you present no proof whatsoever someone is exaggerating something?

You're joking, right? Oh. furball beat me to it.

[furball]

Originally Posted by Tietäjä

You do realize, what an ass you make yourself appear like, when you exaggerate your record? It gets hard to tell when you're actually on facts and when you're just exaggerating something to make your point.

Originally Posted by Monroe

I have made no purposeful attempt to exaggerate anything, you tend to not to want to believe me no matter what I say, so I have no need to prove anything to you.

Originally Posted by Gerbie2

You do realize, what an ass you make yourself appear like, when you make remarks like that when you present no proof whatsoever someone is exaggerating something?

Given that the PA Wiki has the top 100 planet ranks from every post-PaX round, exaggeration of records is fairly easy to spot /o\


EDIT: Jester, can you either unprotect the PA Wiki main page (spam ) or update it to add Rounds 20/21 to the 'Past Rounds' section?

[Jester]

Apologies in advance for this slightly off-topic post, but since this forum has a couple of PAWiki readers/editors, I figured it's better to answer here than in PM.

Originally Posted by furball

EDIT: Jester, can you either unprotect the PA Wiki main page (spam ) or update it to add Rounds 20/21 to the 'Past Rounds' section?

The main page is protected, but the pages that generate the actual content for the main page aren't. If you view source on the main page, it's explained in comments on that page.

[furball]

Originally Posted by Jester

Apologies in advance for this slightly off-topic post, but since this forum has a couple of PAWiki readers/editors, I figured it's better to answer here than in PM.

The main page is protected, but the pages that generate the actual content for the main page aren't. If you view source on the main page, it's explained in comments on that page.

Thanks, I've updated the wiki a bit.

Originally Posted by Monroe

Does anyone have anything relevant to add to the actual topic in this thread, or should we move this thread to GD so Keizari can continue his efforts to discredit my knowledge of PA?

You mean PD, surely. And there's a report post button especially for this sort of thing, so feel free to use it rather than complaining in the thread.

[Monroe]

Does anyone have anything relevant to add to the actual topic in this thread, or should we move this thread to GD so Keizari can continue his efforts to discredit my knowledge of PA?

[Jester]

Originally Posted by Monroe

Does anyone have anything relevant to add to the actual topic in this thread, or should we move this thread to GD so Keizari can continue his efforts to discredit my knowledge of PA?

I'm still waiting for your reply to my post.

[Monroe]

Originally Posted by Jester

I'm still waiting for your reply to my post.

I did respond, but you probably lost it in the useless clutter that is in this thread. I basically said that I wasn't asking about individual attack preferences but ally attack preferences and was wondering what your thoughts were on optimal ally attack times, rather then individual attack times.

[Jester]

Originally Posted by Monroe

I did respond, but you probably lost it in the useless clutter that is in this thread. I basically said that I wasn't asking about individual attack preferences but ally attack preferences and was wondering what your thoughts were on optimal ally attack times, rather then individual attack times.

Despite dropping names, I was not speaking to their individual habits at all.

[Monroe]

Originally Posted by Jester

Despite dropping names, I was not speaking to their individual habits at all.

Yes I understand that, that's not what I was driving at. The attack times for individual success is not necessarily the same as that of the attack times that are good for an ally as a whole. For example it may make sense for allies to launch somewhere between 4-6 GMT in order to have the greatest chance of success attacking, however for an individual player perhaps it's better to launch slightly later since most def will be soaked up in the ally attacks.

The original thought in the thread before it got all muddled in pointless bickering was that ally attack times are slowly shifting later and later. If this is true, is there really an 'optimal' attack time, or are attack times really arbitrary and mostly dependent on when an ally attacks relative to other allies and not any particular time of day?

[Colt]

Originally Posted by Monroe

Does anyone have anything relevant to add to the actual topic in this thread, or should we move this thread to GD so Keizari can continue his efforts to discredit my knowledge of PA?

He's not discrediting your knowledge of PA, he's asking for proof of your allegation that you've been "top100 atleast four times during PAX", yet haven't shown any proof of this....

On topic..

I always found the best times to attack were mid-late morning, 07:00 - 11:00. Although it all depends on your targets circumstances as to how succesful you are..

[lokken]

It depends on the target and how the target defends himself.

[AdmV0rl0n]

Originally Posted by Monroe

I've been watching what attack times allies choose to use and it's been an interested progression over the last couple of years. 4 years ago most attacks were launched 22:00 GMT, 2 years ago a lot of attacks were launched 3-4 GMT over the course of the last 4-5 rounds or so I've noticed it continue to shift forward, and now most attack seem to launch between 5-7 GMT, and I've seen some attacks launching as late as 12 GMT. While some of this shift is clearly because there is an advantage in landing last, I am beginning to wonder how long this slow shift will continue. Also I'm not sure the 'landing last is better' philosophy is entirely to blame for this shift. I wonder if some of the shift has to do with the changes to PA game demographics, or attack philosophies, or the simple change made to the game to allow prelaunches. I wonder if anyone else has noticed this slow shift, and if they have any better explanation as to why this is happening. I don't feel the shift is necessarily go or bad, but it'll be really funny (at least to me) on the day we've gone the whole 24 hour cycle are are back to launching attacks in the European evening.

I can't speak for the top players or alliances. But I will drop in my own opinions, just for the sake of discussion.

1. We've aimed for 1am/2am launches, and tried to do so en masse. We've only really been active in the last third of the round. Any semi organised grouping can go from nothing to some form of ranking by simple play.
The logic of the timeframe is that you hope to hit the magic ETA5 before people wake up. To maximise this, fleet weight should be CO/FI/FR

2. If you are small, there are lots of broken gals and inactives to hit.

3. We avoided the 'mega strike', massive rises in roids bring unwelcome visitors, you lose what you gained. Lots of people gain hundreds of roids, in flashy clever strikes, then have wolves at their door, and the rise in roids does not yet allow enough fleetbuilding to make solid use of roids in that way.

4. Our strikes boringly aimed to increase roid count *every* night, and secondary aim was make it a landing fleet every time, as far as possible. Flying fleets that get deffed wastes ETA, and leaks intelligence. Every hour should be gainfully used.

BS / CR ETA over 90 days = @98 Missions (x3 possible fleets =194).
FI/ CO ETA over 90 days = @154 Mission (x3 possible fleets =462)
FR/ DE ETA over 90 days = @120 missions (x3 possible fleets =360)
The race, and therefore alliance that has the right mixture, in pure, outright attacking terms has massive advantages over a whole round. (IMHO, if this remains as is now, with stats, TER is a very bad race and needs help next stat time... Being slow is one thing, being slow and having hundreds of less viable launches over 90 days is crippling)

By my rough reckoning, alliances should be using CO/FI for attacking as much as possible, in the hunt for roids, ships, xp, score. There is no difference in ships, fi/co have as many targets as any other ship type, this does not change. The only think that matters is successful landing through a whole round. Even if you don't, mixed C0/FI alliance fleets become the most available resource in an alliance fleets arsenal, short ETA's mean higher availability round the clock for both attack and defense.

5. Three fleet like crazy at every opportunity while small/medium, but keep to the smaller, more successful strikes that give a steady rise in roids. Especially if you keep fleet weight at FI/CO/FR/DE. You won't begin to register in other alliance eyes until you reach a certain size.

Downsides.
-----------

Our tactic is severed by late attacking committed by other alliances. By the time we wake, our ships are down to a few ticks to landing, you can be in position that your fleets are out of reach when those late attackers come.

Late attacking recovers/wins higher roids if you retal. (You pick up of others landings, they land first, giving you higher roid count).

Late attacks work against opponents already committed - leaving them a choice. Land your own attacks, but have the attackers take a higher share because they hit you after your landing, pull your attack and try and defend, even though half your ally will be asleep, or half way through an attack. Launching late against a committed alliance is a killer.

There is a downside though. If you run late, the risk is that galaxies in particular, wake up with more than ETA5 on INC and put up stern resistance. Any late raid on a galaxy therefore, should be more than the weight of the possible def - ie crushing. But even then, because the late raids land in daylight hours, the attackers get a very clear view of the play, and can always pull raids that get covered.

Ad

[Monroe]

AD, thanks for posting your comments, they are very insightful on how a smallish ally thinks about attack times. Fundamentally from an attack perspective it is better to get roided before you roid then the reverse because you loose less total roids. This has been the interesting part about CT's strategy this round is they have realized this and so chosen to pick an attack time that falls after pretty much every one else has launched. So what happens (if your watching the ally rankings) is they go way down in roids and usually loose a couple of spots in the ally ranks, then a couple of hours later jump back up at or near the top. If this strategy works (which is yet to be seen) I would expect other allies would for r23 push their attack times back to counter this effect. This is where some of my basis for the original thought in the thread comes from, because the next logical step would be for CT to push things back even further to stay launching last until we've completed the full 24 hour cycle. This leads me to suppose that there is no fixed optimal attack time, and that it is only relative attack times that matter. So as long as you attack a couple of hours after your opponents you will have the advantage regardless of what actual hour of the day that happens to be.

[AdmV0rl0n]

Originally Posted by Monroe

AD, thanks for posting your comments, they are very insightful on how a smallish ally thinks about attack times. Fundamentally from an attack perspective it is better to get roided before you roid then the reverse because you loose less total roids. This has been the interesting part about CT's strategy this round is they have realized this and so chosen to pick an attack time that falls after pretty much every one else has launched. So what happens (if your watching the ally rankings) is they go way down in roids and usually loose a couple of spots in the ally ranks, then a couple of hours later jump back up at or near the top. If this strategy works (which is yet to be seen) I would expect other allies would for r23 push their attack times back to counter this effect. This is where some of my basis for the original thought in the thread comes from, because the next logical step would be for CT to push things back even further to stay launching last until we've completed the full 24 hour cycle. This leads me to suppose that there is no fixed optimal attack time, and that it is only relative attack times that matter. So as long as you attack a couple of hours after your opponents you will have the advantage regardless of what actual hour of the day that happens to be.

The one caveat I see in all the above is that if it comes off, you gain. If you get it wrong, you've lost your roids, and you lose all that fleet time/ETA.

Also, I believe that the game has shifted. Main def comes from in gal, if you can't def your roids, don't bother, just go raiding, with 3 fleets (assuming you are smaller in all cases, Gal, Player, Ally), because there is only so much you can do when undersized in comparison to opponents.

ZU was hit by several waves this week (hence our dented rating in the ally charts), all by overmatched opponents. The only option has been to go attacking at every opportunity. Also, since we started late round, our fleet weight has moved heavier, and we have indeed lost ETAs and availability through heavier fleets (DE/CR/BS)

One last thing. I believe a 90 day round favours XAN significantly.

[qebab]

In response to the original post I would say that it depends a lot on what sort of attack it is. If it is a fullscale war, I would launch constantly to tire members and officers in the other alliance, and break their morale. In pia fleetcatches were often used to demolish large planets, which ruined morale further, especially if they were used on officers and these subsequently went inactive. Something I'd like to try once is to concentrate fire on a relatively smallish group of hostile players (Officers if possible) for 3-4 days, give them no pause and see if they could break.

Yet again, in pia, not pa, my galaxy was waved constantly for more than 45 hours. Pia has 30 minute ticks. It took me 3-4 days to recover from exhaustion and I lost the will to play shortly after. Morale and stamina is more important than roids in a war situation, and good alliances will take that into account. Morale can't be regained as easily as roids, and not everyone can ignore exhaustion and keep going when they're tired beyond belief.

If you aim to win a war, attacktimes matter less than volume of fleets launched. If you're simply after roidgains, you want to land after the attackers on your players land, and you want to launch when alliances are starting to run out of defense fleets.

[Tietäjä]

Originally Posted by qebab

In response to the original post I would say that it depends a lot on what sort of attack it is. If it is a fullscale war, I would launch constantly to tire members and officers in the other alliance, and break their morale.

I think the original post was assuming fullscale war, or at least that's the hint I've gotten from Conspiracy High Command posting on the alliance discussions side. Still, on a full scale war too, you probably want to have some landmark launch, at least when you start hitting!

[Monroe]

Originally Posted by qebab

If you aim to win a war, attacktimes matter less than volume of fleets launched. If you're simply after roidgains, you want to land after the attackers on your players land, and you want to launch when alliances are starting to run out of defense fleets.

I would tend to agree with this, which would tend to validate my initial thought (that optimal attack times are relative not fixed). In winning in the end it comes down more to the number of fleets launched then any other single factor. Obviously ally politics, timing of attacks, race distribution can all come into play in determining success, but in the end the key factor is almost always the will to keep fighting. One key point on this is I can't recall a single instance (someone can correct me if I'm wrong) of an ally being in the top spot for a significant period of time, then loosing it, and still coming back to win the round. It seems the player base (or at least those at the top of the food chain) has a strong tendency to simply give up and not push through the tough times in order to gain long term victory.

[furball]

I've split off the posts on Terran, the thread was becoming too unwieldy.