Successful attack percentage stat

[Dark_Angel]

Very simple really, I think there should be a stat in Personal Stats that shows the percentage of "Successful" attacks you've been on.

Now obviously, define successful, the stat could be made very simple or more complex.

For example if an attack ends up with a grab, it counts as a successful run. If you attack 100 times over the round, you land on 40, thats a 40% success rate.

OR - a successful attack could be defined as an attack that grabs land, but some sort of calculation could be made to determine whether the land grabbed will earn back the money it cost to grab - this idea probably giving a better indication of whether the attack was truly successful or not.

Something I'd quite like to see tbh.

 

[f0xx]

This is quite a good idea, but I am sure DS will not like the definition of successful

 

[pinpower]

What about if you attack and recall? Would that count against you...or would it only count if a "battle tick" occured?

Also, what about people who rarely attack for land. I am playing puppets at the moment and havnt attacked for land for nearly a week but ive sent out ~10 attacks a day to bribe...does that mean my "successful attack %" would be really low. When really my attacks have (apart from a couple ) been pretty successful.

Overall i think its a nice idea, but realistically would be too hard to work out. What if the attack was just to kill someone, you took some heavy losses but stole a tiny bit of land however did loads of damage to him, wiping him out. But land wasnt what you were after, you just REALLY wanted to kill that guy who'd been rushing/spying/piggying your alliance all round. Now he's dead. I'd say thats pretty successful

At first glance, I think there are too many variables for it to be a fair/true statistic.

 

[Dark_Angel]

Perhaps more accurately, the stat could be "Successful Grab Percentage Stat", aimed more at those who want to know how many of their attacks end up with a grab.

I personally attack exclusively for grabs, I rarely attack for any other reason. I can imagine such a stat being useless for routes who attack purely to bribe, but for others, it'd be worthwhile IMO.

 

[pinpower]

I dunno, i think alot more people attack for kills/bounty than you might think. Its not just bribers this wouldnt work for.

And for people who do attack for land mostly but occasionally send the odd bounty attack out...it would mean there "Successful Grab % Stat" was alot lower than it should be, and i know personally i wouldnt want a stat telling me my "land attacks" were unsuccessful when i know they have been just because of this.

I was thinking if you really wanted to add this it could only count attacks with geos/wheelies/thieves in for the stat, but sometimes people will send land stealers in a "kill mob" for bounty purposes as i believe a lower than eta 5 attack nets less bounty.

 

[Weeble]

Successful could be defined as % of troops returned at the end. If you send out an attack and only 14% of your original mob returns, that's probably not that successful, unless there was a large land grab, or what-have-you... whereas if you send out an attack and return with 120% original mob size/value/whatever (due to bribing, etc) then I'd say that's pretty successful.

Trouble with this kind of stat is, as you say, it's hard to define successful within this context. Great idea in theory but the number of factors people consider as to whether an attack is successful or not is huge, and therefore, very hard to accommodate!

 

[pinpower]

The % thing seems like a good idea but where would the limit be...as you often send flak for the purpose of dying! (in a round about way of course) So sometimes you'll loose 80% of your troops (flak) but your geos will get through and land!

I just think there's way to many factors that have to be considered for any sort of rule/formula to be coded in to work this out.

 

[f0xx]

I really like this idea, so I think we should try and find a way to make it work. Some people like to attack for land, others like to attack for bounty and then we have those who attack for bribes/converts. So we have 3 scenarios in which, each of those people will consider their attacks for successful.

Scenarion 1 - Attacking for land
An attack should be considered success if you steal an amount of land, which till the end of the round should make you profit i.e. you will make bigger profit from the land to cover the funds lost for the units to steal it.

Scenario 2 - Attacking for bounty
This one is considerably more simple. If the bounty + insurance funds are more than the funds you've lost while attacking, then the attack should be considered as successful.

Scenario 3 - Attacking for bribes/converts
This one is quite simple as well - if you make more money from bribes/converts + insurance than what you've lost while attacking, then the attack should be considered as successful.

There is also another scenario, in which people have all the three, or just 2 of those scenarions at once. Lets say I am a vamp player who is attacking a bigger solo RPG. I will get converts, I will get bounty and I will get land.

So SCENARIO 4 should be the only scenario possible i.e. if the bounty you make, bribes/converts, insurance and land you steal make you more money till the end of the round than what you've lost for the attack then the attack should be considered as successful.

If the land scenario seems a bit too hard to code, then a much more simple code can by implemented which will consider only wether you land or not. If you land, the attack can be considered as successful.

 

[timtadams]

I like this idea. And how about this for a very simple solution:

An attack can be defined as successful if:
- Your earnings (including insurance, bounty, bribes and converts) exceeds your losses (taking injuries into account as well)
- You land

First point satisfies bounty hunters/bribers
Second satisfies people going for land

Another point is what if multiple people are attacking one target?
Do you define successful as the results of the overall BR or is each person calculated individually?
For this i would say the battle report as a whole. Because if someone provides the doggies for LET flak, or the POMs (cause they cant do damage without hypnos) then it would be unfair to those players, whose participation was effective, but they still suffered loses without doing much damage.

[edit] you got in before me f0xx
And you had almost the same solutions as me, which is good.

But my last point is still good, in my opinion

 

[lavadog]

Maybe you should mesh successful grab and successful bounty hunt/bribe together in one equation. You could possibly do the mathematics so that, when you don't grab but kill/bribe a relatively big amount of troops relative to your score/target's score and lose relatively few, the kill factor outweighs the lack of a land grab. Vice versa for a small amount of kills and a big landgrab.

EDIT: or simply code something like:

if (landgrab > 0) {*some calculation on successfulness*}
else {*some calc on damage done/recieved/bribes/converts etc*}


Java ftw

 

[f0xx]

I like this idea. And how about this for a very simple solution:

An attack can be defined as successful if:
- Your earnings (including insurance, bounty, bribes and converts) exceeds your losses (taking injuries into account as well)
- You land

First point satisfies bounty hunters/bribers
Second satisfies people going for land

Another point is what if multiple people are attacking one target?
Do you define successful as the results of the overall BR or is each person calculated individually?
For this i would say the battle report as a whole. Because if someone provides the doggies for LET flak, or the POMs (cause they cant do damage without hypnos) then it would be unfair to those players, whose participation was effective, but they still suffered loses without doing much damage.

[edit] you got in before me f0xx
And you had almost the same solutions as me, which is good.

But my last point is still good, in my opinion

Yes, the last point is good. But I believe that when attacking as a whole, the only factor which would determine whether an attack is successful or note is if you have suffered more or less damage then the defending side. Although, I believe that such attacks shouldn't be involved in this stat at all.

PS. I was also surprised how similar our two posts were. Your post popped up 2 seconds after I posted mine too.

 

[timtadams]

I believe that when attacking as a whole, the only factor which would determine whether an attack is successful or note is if you have suffered more or less damage then the defending side. Although, I believe that such attacks shouldn't be involved in this stat at all.

I agree too, i did consider that also but wasnt sure. But now i think i would be in favor of it not being involved in the stat at all. Simply because it can start getting difficult to determine whether the attack was successful or not. Also with waves ect, some ticks could be a loss overall, but most are a success, ect. If the stat didnt factor landing then when some people land they might still take loses they were unsuccessful. There are also other situations as well

I hope you can understand were im coming from

 

[Dark_Angel]

Scenarion 1 - Attacking for land
An attack should be considered success if you steal an amount of land, which till the end of the round should make you profit i.e. you will make bigger profit from the land to cover the funds lost for the units to steal it.

That was my initial thought, tbh, and I think it makes the most sense. I can't see the calculation being overly difficult to code, it'd simply be a case of determining how much revenue the land will produce over X remaining days, and checking whether that amount is larger than the total amount lost over 3 ticks. If it is, the attack was a "True" success, if not, it wasn't. (Obviously if bounty/insurance could be incorporated this would give an even more accurate reading).

-

With regards to producing a Stat that incorporates this idea but doesn't play around with the word "Successful" too much, this word having different connotations depending on who you're talking to, perhaps this would be more appropriate:

"You have landed profitably on x% of your attacks"

Or something

 

[Polo]

I like this idea. Personally, I'd define a successful attack as one where you either bribe/gain bounty equal to or greater than your losses or if you steal 6.99%+ land (because wheels minimum grab is 7%). Calcing whether the land will replace your losses over the entire round is just stupid.

 

[Alcibiades]

I like this idea. Personally, I'd define a successful attack as one where you either bribe/gain bounty equal to or greater than your losses or if you steal 6.99%+ land (because wheels minimum grab is 7%). Calcing whether the land will replace your losses over the entire round is just stupid.

Good idea DA, i like Polo's version of it ^^

Plus you then avoid weird **** like Negative Eff for losing no flak and landing etc.

 

[DarkSider]

What about a fully customizable search instead ?
You can make your own stat you want to look for and give the rules you want to search for and order.
Like companies with the most attacks sent/landed, companies with most BR's in which BH > looses or BH + insurance > looses or the total bounty - looses or total bounty + insurance in attacks - looses or anything like that that gets through your mind. Or move onto allies and make land gained - lost in wars, gained from a specific alliance, indiv player gains against alliance etc.
It would require quite some effort from Azzer to add in database all that data (it can be done sometime between ticks so the ticker doesn't take longer) but it can be fun

 

[Weeble]

Why not just "Was this attack successful in your mind?" confirmation after every attack that doesn't *need* to be answered? That way you can choose whether it was successful or not (or ignore the stat completely if you don't care). If this is just a personal stat, then there's no need to find an automated solution as it's all subjective.

 

[f0xx]

I think the point is not to be just a personal stat.

Something like "most successful attacker", which will be visible to all.

 

[Tombi]

I think the point is not to be just a personal stat.

Something like "most successful attacker", which will be visible to all.

Might aswell hand it over to DS already then..

 

[f0xx]

I think the point is not to be just a personal stat.

Something like "most successful attacker", which will be visible to all.

Might aswell hand it over to DS already then..

Is that your only argument against the stat being visible?