We are once again looking for some awesome new moderators to help us out with keeping
the forum running and playing a crucial part in our Empire community! If
this is something you think might be of interest to you, HERE

Spy Fails

Let's talk about the elephant in the room and get this crap over with.  GGE if you want to fix something,  then do something about the spy fails.  The amount we get are ridiculous to say the least and no where near what the percentage you have them at happening.

I can understand failing constantly on a player that has several guard houses,  but to fail so much on npc castles is beyond stupid.  It doesnt even matter if they know we are coming,  they arent going to change.  

It's clear you want to "balance" the game by taking away our coins,  rubies,  and feathers as often as you can,  but at least be honest about it.  If i'm going to get a failed spy 40% of the time,  then just change it so it says so.  Otherwise fix your crap so it works properly.  This has to be my number one issue with this game.  Well,  other than everything else yall do to cheat and rob us.
CBee @ usa1


Comments

  • Lady Nym (US1)Lady Nym (US1) Posts: 390US1
    Hear! Hear!
  • Yeah, I managed to fail 7 times in a row on a 4% failure rate earlier this morning. -_-
  • bernhardt (US1)bernhardt (US1) Posts: 732US1
    If you read the spy% box, it says "The risk of getting caught," which due to a poor translation should read "Chance of getting caught."  It not a failure rate.  This is the GGE 'out' of it not being bs.  All that said, I've never spied a real player, failed and not gotten caught that I know of.  

    I just use the percentage as a rough guide as to how often I won't succeed relative to the base failure rate.
  • BM FujiwaraBM Fujiwara Posts: 575
    I've heard this complaint a lot, and I'm not particularly convinced. It feels like the human's ability to spot patterns noticing rare occurrences to me ... HOWEVER, it's nice and easy to test. So I'm going to run 40 spy runs at 60% fail, 50% fail, 40% fail, 30% fail, 20% fail, 10% fail and 5% fail and post the results here. If we get another 4 people doing it, we'll have a fairly decent amount of precision (each success/fail would move the results by half a percent, which should be enough clarity for them to be neatly in order).

    If anyone wants to take part, post away.
    fujiwara @ en-1
    fujiwara @ us-1

  • kookiekooks (US1)kookiekooks (US1) Posts: 252US1
    edited 01.08.2018
    Results thus far:

    at 10% failure rate = 3 fails out of 40 tries.  
    at 20% failure rate = 10 fails out of 40 tries
    at 40% failure rate + 14 fails out of 40 tries
    at 50% failure rate = 19 fails out of 40 tries

    working on 60% failure rate and will update when I complete that series.

    Warning this sample size is too small to draw any real conclusions
    Post edited by kookiekooks (US1) on
    kookiekooks @ usa 1
  • BM FujiwaraBM Fujiwara Posts: 575
    edited 01.08.2018
    Nice, thanks kookie, here are the findings of my runs:

    60% -- 22 fails out of 40, expected 24 fails
    50% -- 22 fails out of 40, expected 20 fails
    40% -- 16 fails out of 40, expected 16 fails
    30% -- 10 fails out of 40, expected 12 fails
    20% -- 6 fails out of 40, expected 8 fails
    10% -- 3 fails out of 40, expected 4 fails (inc kookie's results, 6 fails out of 80, expected 8 fails)
    5% -- 3 fails out of 40, expected 2 fails

    None of these are good enough sample sizes to be conclusive, but these results are seriously close. I'm rather suspicious that this one of those "there are pixels to click on the technicus upgrade button to improve your chance of getting success" that I've seen, which always amuse me greatly

    Results in more detail:
    https://prnt.sc/kdfyqk



    fujiwara @ en-1
    fujiwara @ us-1

  • I know failing to spy someone five times in a row makes for some awkward PMs
  • Herveus (AU1)Herveus (AU1) Posts: 10,861AU1
    I know failing to spy someone five times in a row makes for some awkward PMs
    Just say you gonna hit them makes it even better then lol

    You need or want help to become a better player shoot me a message in the forums or here I'll gladly help you out for free
  • Herveus (AU1)Herveus (AU1) Posts: 10,861AU1
    Nice, thanks kookie, here are the findings of my runs:

    60% -- 22 fails out of 40, expected 24 fails
    50% -- 22 fails out of 40, expected 20 fails
    40% -- 16 fails out of 40, expected 16 fails
    30% -- 10 fails out of 40, expected 12 fails
    20% -- 6 fails out of 40, expected 8 fails
    10% -- 3 fails out of 40, expected 4 fails (inc kookie's results, 6 fails out of 80, expected 8 fails)
    5% -- 3 fails out of 40, expected 2 fails

    None of these are good enough sample sizes to be conclusive, but these results are seriously close. I'm rather suspicious that this one of those "there are pixels to click on the technicus upgrade button to improve your chance of getting success" that I've seen, which always amuse me greatly

    Results in more detail:
    https://prnt.sc/kdfyqk



    Gonna do it as well once I am home 

    You need or want help to become a better player shoot me a message in the forums or here I'll gladly help you out for free
  • My results:  Complete

    at5% failure rate = 0 fails out of 40 tries
    at 10% failure rate = 3 fails out of 40 tries.  
    at 20% failure rate = 10 fails out of 40 tries
    at 30% failure rate = 13 fails out of 40 tries
    at 40% failure rate + 14 fails out of 40 tries
    at 50% failure rate = 19 fails out of 40 tries
    at 60%FAILURE rate = 23 fails out of 40 tries 
    kookiekooks @ usa 1
  • janina1 (US1)janina1 (US1) Posts: 49US1
    Most of the results are not similar to mine.

    However, my spy rate is only crap against bcs/fls.
  • kookiekooks (US1)kookiekooks (US1) Posts: 252US1
    edited 02.08.2018
    Let see some results to back up your statement.  Mine are against the current FL's.  except for the 5% failure.  before doing this exercise I would have assumed you to be correct, but my evidence states otherwise.  The GGE failure rate seems very close to what I and Fujowara have shown
    kookiekooks @ usa 1
  • CBee (US1)CBee (US1) Posts: 210US1
    i appreciate yall trying to help with these statistics but anything over 10% rate is not an issue because those are real players that i most like would hit blind anyway.  
    CBee @ usa1


  • kookiekooks (US1)kookiekooks (US1) Posts: 252US1
    edited 02.08.2018
    I agree with you CBee that most players will attack blind especially the higher level attacking and defending other high levels.  But Testing spy failure is easier than large scale testing of technicus odds and gem up-grade odds.  Yet if the figures provided for spying are correct then a spot check of the others will allow a better conclusion on those also.

    Plus you are the one that indicated you believed GGS percentages were inaccurate with the initial post.  
    kookiekooks @ usa 1
  • Results thus far in graph form

    https://i.imgur.com/I192t6B.png
    kookiekooks @ usa 1
  • Seems pretty reasonable.  The thing about people is that we never notice when it succeeds at the first time but we do when it fails for the 4th time.  So without testing(which the results above prove this), we can't determine that the spy rates are false.  In reality however they are almost perfect to what the game says.  If you put a bunch of marbles in a bag there is a possibility that you will never pick the marble you want(although a really low one)....same goes with this game.

    Hope this clears it up.

    200,000 MP
    Level 46
  • PJH_ (GB1)PJH_ (GB1) Posts: 318GB1
    I wish I could pick lottery numbers as well as my spies manage to hit that "rare" run of bad luck :)

    I console myself that by the 4th time trying to spy the same target that its the equivalent of them wearing a Hi-Viz jacket and carry a placard while drinking a "martini, shaken, not stirred"   .... move on 'this isn't the target you are looking for' :)
  • kookiekooks (US1)kookiekooks (US1) Posts: 252US1
    edited 23.10.2018
    deleted
    kookiekooks @ usa 1
Sign In to comment.