Shuffle war strategy

Subject
Shuffle war strategy

Author’s intent

I am not advocating the Shuffle war strategy.

But knowledge is power.

And knowing what is happening with your opponent is half the battle to defeating your opponent.

If this turns out to be an opponent’s strategy, screenshots sent to customer service is your current remedy.

Edit

Switch after each war chest not after each win as previously discussed.

Shuffle war strategy

It would be more effective to lose ten wars ( twenty now) so both alliances are at maximum loss modifier, then alternate wars chests

Example

Start

Alliance A 20x losses
Alliance B 20x losses

Alliance A, 5x wins

Open war chest in Alliance A.

Alliance A 15x losses
Alliance B 20x losses

switch

Alliance B, 5x wins

Open war chest in Alliance B.

Alliance A 20x losses
Alliance B 15x losses

switch

Alliance A, 5x wins

Open war chest in Alliance A.

Alliance A 15x losses
Alliance B 20x losses

Switch

War Chest

Since war chest points carry over, it would make more sense to just alternate wins and losses.

It would just look odd since 6 war chest points does not go evenly into 25 until 6x war chests are opened.

SGG is very secretive about information, but their core principle of merciless RNG means the war chest rewards should be as good, or better, than opening one chest

Setup

While getting Alliance A and B down to 20x losses, you would be fighting in Alliance C.

Or deliberately losing, you still get the same 2x ascension item roll per week for losing all your wars and winning all your wars.

War Rewards

But if everything works out, once Alliance A and B are at 20x losses, your account would win every war and get winners ingredients, battle items, etc.

Difficultly

Now that the method has been explained to me, it is easier than coordinating centers.

And cheap at 100 gems.

Sucks for their opponents, but the Devs built the math around their system instead of their system around the math.

Line / Discord

A Shuffle war strategy would require a 100% participation in Line / Discord and communication between teammates.

But that is already necessary for a good war.

Leader Anchor

Alliance A and B would each have a leader that has individually opted out of war.

The leader account would most likely be an alt account.

Closing the loophole

That would be interesting.

I just think they should stick with Elo’s math

(Warscore Manipulation (not matchmaking) - #51 by Gryphonknight)

But until shuffle war strategy becomes wide spread, the Devs are unlikely to do anything about it. They have higher priority fires to fight.

12 Likes

I would like to re-iterate that I believe that this sort of exploit should not be in the game as the artificial modification of an alliance’s war score this way puts their opponents at considerable unfair disadvantage by facing a team far stronger than it should do.

Disclaimer: Alliance swapping is not the only explanation for why there may be a disparity in on-field strength between two alliances in war

However I do want to caveat that given the number of complaints the forum sees about a mismatch in the strength of the defences between two sides in war that this exploit is NOT the only reason you can get this disparity between defence teams. You can get this as a result of your opponents’ rosters are heavily imbalanced towards their war defence with their later attacking teams being much weaker than yours and you can get this as the result of your alliance’s superior tactics and communication means you can beat much stronger alliances with mediocre strategy and participation rates. In both these it turns out that despite having the weaker war defences you actually have the advantage over your opponent so don’t assume you’re at a disadvantage just from defences. Indeed if you want your alliance to do well in wars in the more legitimate way then you need to get used to and develop strategies for beating stronger opponents as the performance modifier will present you with stronger and stronger opponents with each win.

Titan loot losses vs more frequent War Chests and wins then and potentially after the change to past 20 wars in performance modifier

Up until now whilst alliance swapping strategies were a great way of exploiting the war matchmaking system to have a good win record it wasn’t really worth doing from a loot point of view as whilst yes you were getting twice as much war rewards and opening war chests twice as frequently – it doesn’t really make up for the loot lost from high level Titans in the temporary period after each switch as whilst one alliance was losing repeatedly the Titans are also likely escaping and so when players switch to that alliance there is a certain number of days required to battle the Titans back up to a reasonable level. (Now I think there is potentially one method to mitigate the Titan issue a bit which would require a lot of timing and co-ordination but I’ve not tested it – nor do I plan to as I disagree with war score resetting in general – so shall refrain from posting an untested theory). But really it wasn’t the wise course of action in terms of loot but in terms of egos that wanted the great war record cheaply.

However the news a couple days ago that war performance score has changed from the last 10 wars to the last 20 wars significantly reduces this downside somewhat as whilst it will take roughly the same amount of time to battle the Titan up to a decent star level the alliance will have longer to enjoy the Titan at the higher level before needing to switch so will spend a lower proportion of their play time each month against low level Titans. This may be enough that it may actually now be worth it in terms of loot as well now – only time will tell us if that’s the case.

An alternative strategy for casual and mid-tier alliances – doesn’t increase the amount of loot you’d get normally but adapting to the algorithm and carving out a USP from that may help with recruitment and therefore hit bigger titans

As a parting point I think there is something a lot of the middling and casual alliances are missing with regards to how the war score works that they could use to give their alliances better USPs for recruitment. Generally speaking the war matchmaking algorithm is designed so as to find the strength of opponents that you need to be presented with to get your alliance into a position to maintain a 50% win-rate against. What do I mean by this?

Well I don’t mean that all alliances will have a lifetime 50% win rate, that is nonsense. To illustrate let’s imagine an alliance starting out:

Alliance 1: Let’s say that the good gameplay skills and superior alliance wide tactics means that this alliance can expect to beat an average alliance (average in terms of skill and tactics) with a roster that is 20% stronger than Alliance 1. When Alliance 1 is formed they will have a rush of wins until they reach +8 wins. Now at +8 Alliance 1’s war score is the strength of their roster +20% so they are now at the score where facing other alliances of the same score (assuming the other alliance hasn’t artificially manipulated their score) at which both alliances have about an equal chance of winning the fight. Now if Alliance 1 wins then their next war is going to be against an opponent that they’re less likely to win against, if they lose against them Alliance 1 is back at +8 again though they may have some good fortune and beat the team they were expected to lose to but then they will have two more wars against alliances that surpass what their skills give them an even chance of winning so the short streak of 2 wins may then be followed by 2 losses to get back to +8. Etc etc, once an alliance is at the performance modifier that reflects the quality of their gameplay skills and war strategy that they will oscillate around that level performance modifier with an on-going 50% win rate until either they hit the cap of +20 or -20 wins or they improve their skills and war strategy to improve their effectiveness.

So whilst those alliances that can maintain a good win streak at +20 wins or continue to lose at -20 wins will buck 50% win-rate oscillation trend the vast majority of alliances will be subject to it and I think there are many out there who are subconsciously aware of this trend whenever they’re expecting a recent war streak to suddenly go into reverse.

But there is an important realisation to be had here and that is that unless your alliance is willing to be competitive and organised enough to be able to maintain a good win-record past +20 wins (which would translate to a roughly +50% modification to the war score based solely on your roster strength) then the war matching algorithm is design to make your alliance oscillate between wins and losses to maintain an on-going 50% win-rate. Even if you all suddenly decide to use the daftest possible strategy after a period in which it takes your war performance modifier to drop down to the new equilibrium between wins and losses you will eventually begin oscillating around a new 50% win-ratio albeit perhaps focused around a performance of -4 wins rather than +8 wins.

How to learn to stop worrying and love the wars: Now given that the algorithm will mean for most casual and middle-tier alliances that they will all eventually settle at a point where they’ll have an on-going 50% win rate it is possible to adapt to that fact and use that to create a USP for your alliance to recruit players. What do I mean by that? Well you could do like everyone else and take every war 100% seriously but you’re still going to lose 50% of them despite the effort and potentially stress. You can also take the approach of your alliance trying their best on 50% of the wars and on the other 50% where it doesn’t matter if you win you could have novelty wars (perhaps using a different colour tank than I your serious wars, or a competition to see who can score the most points attacking with 3* heroes only etc) or use it as an opportunity for training or trialling different war strategies etc etc. Let’s take these as examples you could have as distinctive war policies for your alliance:

  1. Mid-week wars fought competitively with Holy tanks, during weekend wars we rotate with other tank colours for fun.
  2. Weekend wars fought competitively, mid-week wars we do training and experiment with war strategies.
  3. Busy during the week but want full war chest participation? We fight weekend wars competitively but let the mid-week wars slide.
  4. Hate the Field-Aid wars but don’t want to opt-out? We let those ones go but fight the others competitively.
  5. Full participation in competitive mid-week wars but optional novelty games at the weekend wars.

Now with these you’ll still wind up with a win-ratio and open war chests at the same rate as those taking every single war seriously but you’ll have more fun doing so and may recruit more players to help you tackle bigger Titans, even in the case of those who are attracted to your alliance as they have less time to play during either the weekend or midweek war as whilst they may be too busy to be able to be online to attack at a specific time in a war strategy they likely can still squeeze in titan hits that day – I mean everyone needs to go to the loo at some point – and morale is likely to be higher than it would losing 50% of the wars you’re taking seriously.

Just food for thought – having a distinctive USP like one of these may help you fill out your causal or mid-tier alliance for better Titan loot without impacting on your war loot materially.

PS. Just a quick heads up I may be slow to respond to any replies on this due to both a backlog of work that’s built up and providing palliative care for one of my dogs not expect to make it to the new year so please be patient if seeking a response.

1 Like

Titan loot

My retirement alliance is currently fighting 2* Titans.

For 1.75 years, my main alliance was fighting 5* Titans.

So Shuffle war strategy is valid for casual players, and players missing key 4* / 5* titan heroes.

Really crappy for their opponents, but valid.

Matchmaking

No.

I very strongly disagree.

Current matchmaking is an abomination.

It forces alliances that do not fit its narrow model to quit wars or individual users who do not fit to quit.

SGG ( the studio) only cares that the matchmaking decreases the number of blow outs of really high wins.

Players, and alliances, quitting war fits their model better than using actual Elo math and years of real world matchmaking experience with solving the insolvable 1:1 matchmaking problem.

The alliance based win/ loss rating shows a poor understanding of the basic assumptions, and math model, used by Elo to derive his ground breaking solution.

Not Elo math

No.

See above.

I hate with a burning passion level based, power based, one size fits all bot scaling and matchmaking.

Elo set out to solve a very real problem, and succeeded.

Elo’s #1 assumption

Elo’s #1 assumption was the players matchmaking number was unknown.

SGG is creating an artificial initial seed number then treating it as true.

Something Elo would NEVER do.

The math does not change just because your situation does not fit the math!

Real world solutions

Because Elo’s #1 assumption was the players matchmaking number was unknown you change the your situation to fit the math!

Elo assumes initial seed numbers are flawed and uses season resets, sliding K-Factors, qualifier rounds, and other methods to find the real matchmaking numbers.

Games, and organizations, have been using Elo’s math for decades.

This is all known!

Not perfect, but doable.

Separate matchmaking numbers

Elo assumes drastic condition changes require a separate matchmaking number ( looking at you Field Aid and Wednesday wars ).

Hard

But understanding the base assumptions, and models, for Elo’s math is tough.

So people can make mistakes.

People can also fix mistakes.

SGG hire an Elo expert for 12 months. Then consult with them, or their replacement, anytime Empires requires matchmaking.

1 Like

I don’t really know what I’ve written that has riled you enough to post this but I shall just point out that I have at no point posted a judgement on whether I believe the war matchmaking system is good or not (and as such I don’t know why you assume that I think it is), all I have set out to do is explain how it works and propose ways in which some alliances can adapt to how it works.

For the record I don’t believe SGG have any intention to change any of the matchmaking in any avenue of the game right now and as such have no intention to waste my time banging my head against that particular door (but even if I did I’d prioritise Raid Tournaments as the underlying logic there is the most broken in the game) and therefore the optimal strategy is to adapt to how things are rather than make oneself miserable arguing for something that is not gonna happen.

1 Like

@Gryphonknight some monthes ago, I sent a mail to SG explaining this strategy and I loved their answer : "It is not possible to cheat . You can find various opponents strategy unfair but you can always do the same thing yourself ".
My translation : “Anything that is not formally forbidden is allowed . Don’t complain if your opponent is more cunning than you”. Trying to achieve unjust advantage is a part of the game.
We will nevertheless continue playing fair. Hopefully, it only happened to us twice.

4 Likes

For what it’s worth, this exploit was already pointed out and explained over a year ago

Nothing was done about it then, doubt it is now, but probly good to have it brought to attention

I think there are much easier more beneficial exploits being abused but probly deserve their own topics

Think the “war swapping” situation fixes itself as alliances either grow in strength or performance and start climbing the ladder of more competitive alliances that are smart enough not to sacrifice their daily loots for such a small gain

War Shuffle lol.

If 30 Idiots want to exit a Alliance to reform elsewhere and have the possibility to win 10 Wars (or 20 now) in a Row. They lose there ability to Chain X titans (Example:14’s).

They bleed chance of mats like it’s going out of style (Titans > Wars).

Literally the dumbest “strat” I have read on the internetz

Would take less energy to work on Start, teach players on teach, work on Hero’s, Troops & Tiling for higher probability and just winning “extra wins” at your teams “War Cap”

1 Like

That’s what I’ve said about players going into an alliance to “spy”. It’s not a very nice thing to do but everyone can do it, so technically it’s not c-heating.

1 Like

That being so easily dismissed is because without going into an alliance’s line group, notes, etc

There isn’t much to learn from in game tactics such as when a team hits or doesnt hir or what color tanks they run

At least in my opinion

Not much to be gained from “war shuffle” either as zombie pointed out above

3 Likes

They do not. They change an amount of people, the alliance stays at it is. So they get their titan AND a lower war score.
I know some smaller alliances rebuild the whole alliance, but the greater ones seem to switch between several alliances.

Which are the “greater” ones?

If they’re switchin between alliances, odds are they’re still dropping titans if switchin stronger players with newer players unless their playin full time alts but even then odds of more than 5 to 10 people having an alt each to keep up with titans over 10* is slim unless they’ve already been in the game for a long time or spent a ton of cash which in either case they could be stringing bigger titans and manipulating war chests by just dropping from 30 to 29 members or 30 to 28,27, etc

I don’t see any way around it being a losing situation. Possibly a select few are killing titans over 10* and doin the alliance swap trick succesfully without dropping titans but i would expect it to be a very low %, low enough to be non issue for most of their opponents has they only face them once in a great while, sometimes once period

Exploit? Yup. Beneficial exploit? Not really. Chance of some diehards more clever than others to make it most beneficial? .0001% probly but not clever enough to accomplish the same goal in much easier ways

I meant “fuller” ones. I’ve seen it one time, there were at least 2 alliances, the players had a strength from 3,7 k to > 4k. I think it’s easer to switch between alliances than to create them new for 30-60 (or even more) players.

All 60 were 3700 to 4k?

That was, what I remembered, but that wasn’t correct. The war participants had that TS. The alliances are (atm) made up from lvl 67 down to lvl 24. Some of the higher players certainly never switched, some of them did. The lvl 67 joined one day before. Switches are 21-24 days, 8 days, 3 days, 1 day.
I just have this one example, we’re not high enough to see “manipulation” things often (this was the alliance with the spy and the shuffle of the war board, they take it serious, so I think the shuffling of members is also war related).

With that kind of gap between player levels, the switches still hurt them titan wise than having all the strongs in 1 team, stringing 14s, and dropping from 30/30 to 29/30 every 4th war or so

Wrong target

Sorry if you read the above as a personal attack.

It was intended as strong condemnation of the current matchmaking algorithm.

It was aimed at the Devs who might read this topic.

Math

This is true of raiding with 1* to 4* Heroes since raiding trophies use Elo’s math for matchmaking.

In my opinion, and other opinions, not using Elo’s math for war matchmaking means alliances will not self adjust to a 50% win / 50% loss record.

Fixed seeds and bounded adjusting

So SGG generates a fixed seed using the roster of the players involved.

The fixed seed is adjusted for 10 wins to 10 losses ( temporary increased to 20 wins to 20 losses )

Wins greater than 10 ( temporary 20 ) are ignored.

Losses greater than 10 ( temporary 20 ) are ignored.

K-Factor

If Individual war energies / war battles used Elo’s rating then the brilliance of Elo’s K-Factor would shine.

This would be similar to individual chess game wins/ loss/ ties.

The K-Factor is rendered null because war matchmaking always has same war score.

This is similar to giving an entire chess club, of thirty members, one matchmaking rating, and adjusting it once per month.

1 Like

Titans

In 2+ years, Gryphonkit ( my wife ) and I have never fought a 10*, or higher. titan.

It would be interesting to know what percentage of players are killing what titans.

The Perfect is the enemy of the good

Having thought about this, I think you are correct. My recent anger over 1* / 2* tournaments seems to have brought up all my unresolved issues with Empires.

It is the same as trophy dropping. There are advantages, and disadvantages, to trophy dropping. Just like there are advantages, and disadvantages, to not trophy dropping. Each player decides if they want to trophy drop.

Hopefully the Devs will increase it to 104 wins and 104 losses.

As far as 30/30 matchmaking goes, i do feel it’s rather fair outside of 1 particular exploit but i only know that in the top end.

The war swapping i suspect goes beyond top 100 and i can’t track that.

Just like i can’t say matchmaking is fair outside of the top 100 cause i wouldnt know. I don’t have matches to compare, scores to compare, alliances to track up n down etc

So it may just be top end fair

If it’s unfair after the top 100 then improvements could be made, if it’s similar fairness aa in the top then i see no reason to make major adjustments to how alliances are matched but instead alter the way war chest points are accumulated to deter exploits(similar to what they did with raid arenas to deter cup dropping)

I would love to see this.