9 Weeks of War Data - Analysis

Discussion in 'Wars' started by admin, Aug 8, 2013.

  1. I agree with everything hatf1eld and and would like to emphsize on the advantage ko system, that was the best change in these test wars and it would be a damn shame if you decide to leave it out of next season.
     
  2. I totally agree with : ER-Rudementary , Well put .
    So your basing everything off of plunder ? Hmm... Not CS/BFA/BFE , Just plunder :lol: . Now were back to almost season 1 standards , I thought the random ko times were to stop the bots and auto clickers. It almost seems that these test are going backwards , as it does you see less and less clans sign-up . I think there are some great ideas on this post as random ko times and other factors. Seriously basing studies off plunder is just ..yeah.. What are the main factors in match ups? Plunder... Hopefully you will continue to test before the season and get us back on the right track , Seems were back to square 1.
    Ty for the ee wars i do like and seems a lot of others do as well. I would like to see a graph on CS/BFA/BFE in the last two weeks .
    Cheers :?
     
  3. KaW Admin,

    I was warring and admittedly did not see the results of war 1. However, I do note that one war was a rematch of two clans Elysium vs. Warlor. With the set KO times the differential was even larger (with the winning clan winning by even more than last time).

    Admittedly this is anectodal and no conclusions should be drawn from one war (and perhaps a significantly different roster lineup was in either/both clans), but if it is indicative of the wars overall, then the it would seem to lend some credence to the people that say the Random/Advantage KO system merely masked an unfavorable matchup.

    Will you be sharing the data people have requested (CS, BFE, BFA) in order for us to make matchup formula suggestions?

    Cheers,

    Kage.
     
  4. honestly, if you reveal the matchup we can adjust what we need to to keep from getting outmatched.

    in prep for war, its like ok we got all our boats ready, oh man, they are coming by land, lets just kill ourselves and maybe next war will be by boat. if we knew how much of what was counted by what weight we could get ourselves matched better.


    unless you all just throw darts to give match.
     
  5. When is kaw going to address the gh issue?
     
  6. @kaw_admin,

    I'm afraid the match up imbalances are not narrowed with the new algorithm and there are already many credible reason posted above. However I personally think that there are two parts that Dev must look into seriously if you are serious into looking at effective solution to reduce the match up equality. they are both related to BFA!!!!

    Addressing the below creates a more balance warring-field and truly test the war strategies of every war clan. with this implementation, the key determination of war result is no longer skewed towards a measure of pocket dept or your kaw age.

    1. %BFA contribution
    My understanding is that it currently stands at 2%, simply put, if you are HLBC and have another 50 HLBC allies you are twice as strong. This 2% contribution is the cause of biggest imbalance in KAW and benefits mainly a)the cash burner, b) early starters, c) multiple accounts holders.

    The 2% had resulted in increasing bfa gap each time new buildings are introduced, also the most hated bit -- bullies. I would suggest the reduction of %bfa (to 0.2% instead or even 0.1%). It is understood that if this is implemented, the rich cash burner is going to protest like mad hence i am suggesting that this reduction is temporarily and applies only during EE wars ONLY.

    but in the long term, a reduction in %bfa contribution would certainly encourage PvP.. who won't want to hit back if you know you have at least 50% chance of winning? I am sure ally traders would like this reduction because being able to hit back (with wins) against aggressor after buying allies is a pleasing thing to do!!! No?

    2. %hit ratio (DTS/DTW)
    This logic is not known (at least to me) but it is obviously not working well. 90% hit ratio does not make any sense when 80% of the time you get only defeats.To ensure that team match up is more balance, i would recommend that if you do not have >30% chance of winning an attack, the opponent is DTS for you and you are DTW to the opponent. this would avoid one sided bashing and forum would get lesser Dev bashing too!!! :D

    I can understand why Dev is arriving to the current match up algorithm -- it reduces no match unless odd-man out -- but it does not reduce the match up imbalances. the direction Dev is heading by studying every possible areas that affects war outcome (bfa/bfe/stats) is great but the %weightage for the respective area seems largely wrong!!!

    For your consideration!!!!
     
  7. So far, so good. A nice alternative would be to keep the "blitz" style wars as an available option. No complaints from what ive seen the last 2 weeks.
     
  8. I think DTW isn't good idea ... better solution is maintain ability to hit each other ppl who are now in dtw ...
    player1 too big and player2 too small,
    then better than refuse hits is get down their plunder to very low level
    small will get bad plunder and big too ... or mdl KO plunder .... war idea maintained and ppl able still unload if haven't target
     
  9. *small KO plunder
     
  10. Well, the results of war 1-3 are in. Being halfway this weekend here are some thoughts on the results so far.

    The results from war 1:
    - an average plunder difference of 116%
    - 80% of the wars had over 25% difference
    - 30% had more than 100% difference
    - and 25% had even more than 200% difference

    War 2:
    - an average plunder difference of 71%
    - 65% of the wars had over 25% difference
    - 20% had more than 100% difference
    - 10% had even more than 200% difference

    War 3:
    - an average plunder difference of 106%
    - 68.75% of the wars had over 25% difference
    - 18.75% had more than 100% difference
    - 12.5% had even more than 200% difference

    If a difference of max 25% is considered a 'good match', then only 20% of the clans in war 1 had a good matchup, 35% in war 2 and 31.25% in war 3. Over more than half had a 'bad matchup', some had a really bad one (more than 200% difference).

    If this weekend really is about testing whether the hit ratio sytem gave 'closer matches' (in plunder) or if it was just the random KO system covering up the bad matchups, this data surely suggests the latter.

    And it seems that the worst matchups happened in the top tier clans warring, which has to do with the fact that the BFA diferences are too big there. Yes you can hit 100% of roster but entire clan fails 100% on their top 3. Good matchup? I think not.

    So far this proves all the points made: this hit ratio matchup system isn't working and the bigger the clans warring, the worse it turns out. The random ko system just covered up real bad matchups under the hit ratio system. The matchups seem even worse than in season 1.

    Solutions?
    1. Stop with the hit ratio algorithm; use season 1 algorithm and allow more 'no matches' again.
    2. Go with the idea the person above me posted: diminish the effect of BFA in EE wars to 0.2% and test with that.

    And be more transparent on how matchups work out with regards to difference in cs, BFA and BFE.

    Just my thoughts so far. Let's see what wars 4, 5 and 6 bring.
     
  11. And something I was wondering about - if one clan has 26 HLBC and the other 29 HLBC warriors and all can hit eachother, would that make for a 100% perfect match? I'd assume so since 100% can hit eachother. And there have been clans matched 3 down (not talking about ghansels, they can be a pain but the impact of 3 more ghansels is much smaller).

    26 vs 29 HLBC is more than a 10% difference in strength on its own and 15 full bars more in a 2 hour war (1 full bar at start, 2 xtals and 2 regens in 2 hours time for 3 people). 15 full bars more is over 15B difference plus a clan can completely overpower the other clan. I would not call that a 100% match.
     
  12. Good stats. Just one comment I would like to make, BFE is very important. To get mith to gain BFE is a hard task, please consider increasing aqua payout to allow 'newer' players a chance to acquire better BFE / equipment. I understand the 'it's a business' and need to make money from XSTAL sales. But do you really get much income from xstal to aqua transfer?? My guess is better aqua payout won't cost you a lot of income, I don't know anyone who swaps XSTALS to aqua.... Just a thought keep up the good work 
     
  13. Devs-


    You need to match based on Combined stat again. We get matches where all the enemy clan is 5m, maybe more stats bigger than each player.
     
  14. @Devs, Thank you for your efforts in continuing to improve our war experience.
    I can only speak on the level at which my clan fights and the addition of DTS/DTW did little to improve matches. In fact, we ended up in more unfair matches.
    This system allows for the exploit of stacking 29 lb into a war, which ruins the war experience and is against the spirit of what EE was created for. I remember the kaw admin post saying they wished to create clan loyalty and unity. Instead this creates an arms race to see who can gather in all the biggest lb into one clan.
    Plz reinstate the no-match system. Id be happy to rotate my biggest guys to war to prevent a no match and allow for fairer wars. We aren't interested in beating up smaller clans anymore then we wish our 3 lb guys to get beat up by 29 of them.
    At the level in which we war BFE does nothing. You can have a Lv 10 sword but you will not win an attack on a top 50 lb with it.
    Matching by Hit Ratio also doesn't work to produce fair matches. Top 50 lb can hit any hlbc account, that same account can not win an attack in return. So simply being able to hit the button doesn't allow for a proper war experience or environment when all you do is fail. It will simply cause less clans to war.
    The 1hr EE wars were simply a plunder and hide war. It produces the same environment as a regular system war in which you sit in pin. Plz leave that war system within the system wars that are already established and let EE wars be what it was created for. I saw 2 many wars in which one team totally dominated the war with actions and KO only to lose the war. That seems ridiculous to me. Keep EE wars in the 2hr format that allows hits exchanged and strategy to allow come backs. Then keep the system wars. That allows both types of wars available for all who wish to war.
    Again thank you for your continued efforts to improve our game.
     
  15. BRING BACK 1 HOUR WARS Ty :ugeek:
     
  16. Bump as a reminder for season 2
     
  17. Bring back 1 hour wars