BREAKING NEWS : Auto Assault Intensity Levels Aren't Sticky.

[yeti_c]

I recently ran a test game with DiM to see if the Auto Assault Intensity levels were sticky.

By Sticky I mean to see if they get stuck reading the same row from the Intensity Level file time after time.

During which we played a build game on AOM2 where we stacked huge armies and then DiM loaded one place with 500 armies - and I autoassaulted them.

I then logged the intensity levels for these assaults into a spreadsheet.

I then ran analysis on the levels to see if any rows matched the one before it.

In 23 different assaults of 500 deaths... the intensity levels were the same 1 time.

I therefore conclude - that the auto assault intensity levels aren't sticky.

http://www.fileden.com/files/2008/1/19/ ... estxls.zip

C.

[wicked]

Cool. I use AA almost exclusively and works fine for me. People who are anti-AA are like those old fogies who are anti-computer/technology... they just fear it because they don't know how to use it.

[Soloman]

Cool. I use AA almost exclusively and works fine for me. People who are anti-AA are like those old fogies who are anti-computer/technology... they just fear it because they don't know how to use it.

Lol I do also and while sometimes it does hurt other times it is great, I look forward to win we can set limits on the AA that way I feel there will be far less whining about it(sadly there will always be some whiners though).

[mibi]

In laymans please.

[Soloman]

In laymans please.

the same number combinations from random.org are not just recycled continuously on autoattack as some believe them to be(Thus a lot of people argument to never use auto attack) and they are all random lines of numbers same as regular attack...

[wicked]

In laymans please.

dice fair. ooga booga.

[jiminski]

I recently ran a test game with DiM to see if the Auto Attack dice were sticky.

By Sticky I mean to see if they get stuck reading the same row from the dice file time after time.

During which we played a build game on AOM2 where we stacked huge armies and then DiM loaded one place with 500 armies - and I autoattacked them.

I then logged the dice rolls for these attacks into a spreadsheet.

I then ran analysis on the dice to see if any rows matched the one before it.

In 23 different attacks of 500 deaths... the dice were the same 1 time.

I therefore conclude - that the auto attack dice aren't sticky.

http://www.fileden.com/files/2008/1/19/ ... estxls.zip

C.

bloody hell you two are dedicated to the cause! cheers chaps.. with all of that fairly useful data, are you looking at trends in probability?

It would also be interesting, in light of somes perception that auto creates a false; less random result due to the grouping of the number list, *(perhaps explaining the vast swings in positive or negative results for the attacker) to see how the overall pattern compares with 500 consecutive, non-auto attack rolls.

heheh and don't tell me it will be the same that is just not scientific!


*if you are thinking "Jm really does not understand how the numbers are taken and translated to rolls!" you would have a 100% probability of correctitude!

[wicked]

yeti, did you see any "streaks" in the rolls?

[yeti_c]

yeti, did you see any "streaks" in the rolls?

I wasn't looking for them - however DiM is looking into processing the XLS further.

So no doubt he will come through with some useful stats soon...

Any other wizards with Excel feel free to look at the numbers too.

C.

[yeti_c]

to see how the overall pattern compares with 500 consecutive, non-auto attack rolls.

First up - It's not 500 rolls - Each attack killed 500 armies.

Second up - I'm not about to repeat the process pressing attack one at a time!

C.

[jiminski]

to see how the overall pattern compares with 500 consecutive, non-auto attack rolls.

First up - It's not 500 rolls - Each attack killed 500 armies.

Second up - I'm not about to repeat the process pressing attack one at a time!

C.

First up - I am pulling your leg. (the thought of you and DiM pressing the button 10,000 times only to prove Xtratabasco right after-all, tickled my chestnuts for some reason!)

Secondly - Even though you are not willing to follow this experiment to its natural conclusion *wink*; well done and cheers.

[owenshooter]

In laymans please.

the same number combinations from random.org are not just recycled continuously on autoattack as some believe them to be(Thus a lot of people argument to never use auto attack) and they are all random lines of numbers same as regular attack...

that isn't why i don't use auto-attack. i don't use auto-attack for strategic reasons. if i lose too many after a few attacks, i find it better to have a larger number to start off of on my next turn, than to be left with a 3 facing a 7 on a strategic border or a bonus. has nothing to do with think auto is the boogeyman or evil, i think it is just bad in certain situations, it is that simple. unless the game is decided i have absolutely no use for auto-attack. is that so difficult to understand?-0

[Sentinel XIV]

Cool. I use AA almost exclusively and works fine for me.

Same here. I mean, their twelve step program worked wonders to get me off the sauce. I don't know what I would do without them.

Oh, and I also use auto-attack quite a bit, unless it is crucial that I don't whittle my armies down too much. That method only screws me part of the time.

[Soloman]

In laymans please.

the same number combinations from random.org are not just recycled continuously on autoattack as some believe them to be(Thus a lot of people argument to never use auto attack) and they are all random lines of numbers same as regular attack...

that isn't why i don't use auto-attack. i don't use auto-attack for strategic reasons. if i lose too many after a few attacks, i find it better to have a larger number to start off of on my next turn, than to be left with a 3 facing a 7 on a strategic border or a bonus. has nothing to do with think auto is the boogeyman or evil, i think it is just bad in certain situations, it is that simple. unless the game is decided i have absolutely no use for auto-attack. is that so difficult to understand?-0

LOL I said a lot of people not all people, and as I said you primary strategic reason will hopefully be taken care of win the make adjustable auto attack or whatever they are going to do with autoattack changes on to do list...

[DiM]

i'm currently compiling the results to see if there are any visible streaks in the rolls.

i have decided that anything bellow 4 rolls with same outcome is normal so i'm just looking for streaks of 4 or more of the same result in a row.

for example rolling double wins 6 times in a row.

so far i have done the first set of attacks (454 attacks) and found this:

so i found 12 streaks where 4 results repeated (6 double wins, 6 draws and 0 double losses)
and 2 streaks where 7 results repeated (7 draws and 7 double losses)

[t-o-m]

from what i understood...ouch!

i didnt really get very much of the 'sticky' but ill trust you two, yeti and DiM,
i really hate those streaks though, i just want them to be fairer from day to day, or game to game - because on some games it seems that i can win a 20v40 which i have done in the past, but in others i cant win a 15v2!

[DiM]

one more stat for today cause i have to go to work:

yeti killed a total of 11500 of my armies and in this process he:

1. made 23 auto attacks vs a stack of 500 of my defenders
2. lost a total of 9782
3. on average he lost 425,3
4. at most he lost 498 taking out 500
5. best result was when he lost just 340 to kill 500

[jiminski]

one more stat for today cause i have to go to work:

yeti killed a total of 11500 of my armies and in this process he:

1. made 23 auto attacks vs a stack of 500 of my defenders
2. lost a total of 9782
3. on average he lost 425,3
4. at most he lost 498 taking out 500
5. best result was when he lost just 340 to kill 500

yes that is genuinely very interesting!

It goes towards proving that the 'streakiness' is at least evened out and mostly works to the attackers benefit over a large enough control group.

I could be very wrong here but I think what most people get peeved about are the 10 to 50 range stacks' attacks (very much off the top of my head)
You will, through studying the large data base, be able to see patterns of win/loss peek and trough for the auto.
But i do think it needs comparison to a group of consecutive, individual attacks, in order to be able to allay peoples doubts that the auto disproportionately skews 'randomness'.

I am in a large build up game now...taking the odd 50 here and there would be light relief, so i can give you 'some' data for that if you are interested... and tell me how to pass it to you.

[detlef]

one more stat for today cause i have to go to work:

yeti killed a total of 11500 of my armies and in this process he:

1. made 23 auto attacks vs a stack of 500 of my defenders
2. lost a total of 9782
3. on average he lost 425,3
4. at most he lost 498 taking out 500
5. best result was when he lost just 340 to kill 500

yes that is genuinely very interesting!

It goes towards proving that the 'streakiness' is at least evened out and mostly works to the attackers benefit over a large enough control group.

I could be very wrong here but I think what most people get peeved about are the 10 to 50 range stacks' attacks (very much off the top of my head)
You will, through studying the large data base, be able to see patterns of win/loss peek and trough for the auto.
But i do think it needs comparison to a group of consecutive, individual attacks, in order to be able to allay peoples doubts that the auto disproportionately skews 'randomness'.

I am in a large build up game now...taking the odd 50 here and there would be light relief, so i can give you 'some' data for that if you are interested... and tell me how to pass it to you.

Well, this isn't hugely surprising. Assuming the various odds calculators are accurate, every time I plug in a very large attack (like at least 50 vs an army no greater) the odds are overwhelmingly in favor of the attack. Hell, 50 attacking 50 shows a 73% chance of success.

[edbeard]

In laymans please.

the same number combinations from random.org are not just recycled continuously on autoattack as some believe them to be(Thus a lot of people argument to never use auto attack) and they are all random lines of numbers same as regular attack...

that isn't why i don't use auto-attack. i don't use auto-attack for strategic reasons. if i lose too many after a few attacks, i find it better to have a larger number to start off of on my next turn, than to be left with a 3 facing a 7 on a strategic border or a bonus. has nothing to do with think auto is the boogeyman or evil, i think it is just bad in certain situations, it is that simple. unless the game is decided i have absolutely no use for auto-attack. is that so difficult to understand?-0

see I like auto attack because I think it cuts down on the variance of my attacks. Yes, obviously if I'm auto-attacking, I'll have more overall losses (and wins) because I'm using all my armies, but if I don't use the auto-attack and just peck (use regular attack), I'll miss out on those occasions where I win because I had a good streak of winning dice.

essentially, more times I roll 3 dice vs 2 dice = more likely I'll be close to the true probabilities. If I only peck and roll dice until I get to an 'acceptable point' then I believe the variability of my rolls is much higher.


Owen is right though. There are times where pecking away and stopping is the right thing to do. I just don't think this happens as often as Owen seems to think. If I'm going to attack somewhere, most of the time I'm doing it to take over that territory. Auto-attack away!

[lancehoch]

that isn't why i don't use auto-attack. i don't use auto-attack for strategic reasons. if i lose too many after a few attacks, i find it better to have a larger number to start off of on my next turn, than to be left with a 3 facing a 7 on a strategic border or a bonus. has nothing to do with think auto is the boogeyman or evil, i think it is just bad in certain situations, it is that simple. unless the game is decided i have absolutely no use for auto-attack. is that so difficult to understand?-0

Owen, if someone were to make a script that was an auto attack, but you could put in a set number of losses for that attack or minimum army size for that attack would you use it?

[yeti_c]

that isn't why i don't use auto-attack. i don't use auto-attack for strategic reasons. if i lose too many after a few attacks, i find it better to have a larger number to start off of on my next turn, than to be left with a 3 facing a 7 on a strategic border or a bonus. has nothing to do with think auto is the boogeyman or evil, i think it is just bad in certain situations, it is that simple. unless the game is decided i have absolutely no use for auto-attack. is that so difficult to understand?-0

Owen, if someone were to make a script that was an auto attack, but you could put in a set number of losses for that attack or minimum army size for that attack would you use it?

FWIW -> this is possible in the CM script.

C.

[Timminz]

that isn't why i don't use auto-attack. i don't use auto-attack for strategic reasons. if i lose too many after a few attacks, i find it better to have a larger number to start off of on my next turn, than to be left with a 3 facing a 7 on a strategic border or a bonus. has nothing to do with think auto is the boogeyman or evil, i think it is just bad in certain situations, it is that simple. unless the game is decided i have absolutely no use for auto-attack. is that so difficult to understand?-0

Owen, if someone were to make a script that was an auto attack, but you could put in a set number of losses for that attack or minimum army size for that attack would you use it?

FWIW -> this is possible in the CM script.

C.

So, the answer is "No. owen wouldn't use it."

[owenshooter]

Owen, if someone were to make a script that was an auto attack, but you could put in a set number of losses for that attack or minimum army size for that attack would you use it?

it is possible, and it does exist, and i don't use it. CM and BoB are not my friends, as i have voiced on more than a few occasions... this is a casual gaming site, i prefer to use my mind than a script. i appreciate all that yeti does, and understand all the various needs/uses/applications for the scripts, i just choose not to use them. kind of like how i choose not to play standard or freestyle. soooo, to sum it all up in one word... NOPE.-0

[Thezzaruz]

Well, this isn't hugely surprising. Assuming the various odds calculators are accurate, every time I plug in a very large attack (like at least 50 vs an army no greater) the odds are overwhelmingly in favor of the attack. Hell, 50 attacking 50 shows a 73% chance of success.

Yea they are accurate. In a situation where attacker and defender have the same number of armies on their territs it swings in favor of the attacker at 11 armies IIRC. When doing large battles that extra attacking dice really makes a difference.