🧙 The Art of War – Top 100 Alliances Data Review and updates

Hey All! I am the War Score Data Tracker for Crew-Saders and have posted some updates in various threads – mostly centered around explaining the war matching system using Data to wayward souls that just don’t understand 1. It felt as if perhaps there needed to be one place that tied it all together and post updated rankings periodically for those interested.

So to begin, when I did a search of the forum to see if there was already a thread that puts the current war matching methodology and discussion with data all in one place, there are so many threads started, merged, closed by members that just don’t understand the system 2. So, it made sense to me to start a new thread to hopefully summarize it better (not buried in a 1000 post string) using data. Just the way my mind works so maybe it will help others as well.

War Matching is based solely on your alliance war score and the number of members opted in for war, not the total alliance score. SGs matching algorithm attempts to match the “closest” war score alliances with the same number of members. Does it happen that there is a “mismatch” to the number of members fighting against each other? Yes, as there may not always be the exact number of alliances in each of the right tiers. But it really is the exception and not the rule. “War block” will also factor here – meaning you cannot face the same alliance within X wars (currently set at 5) 3.

WAR SCORE FACTORS

Your alliance war score is based on 4:

  • Top 30 heroes of each member opted in for war including Emblems
  • From those 30, more weight given to “best” 5 heroes in each roster
    • This was done to avoid members from fielding a “sub-par” defense in order to draw a more favorable match-up
  • Highest Troop of each color per member
  • Player count of the Alliance “opted in” to participate in the war
  • War history - AKA War Penalties (will be explained later as this is where some alliances have learned to “exploit” the system or most of the “unfair” matchmaking occurs)

Your alliance war score IS NOT based on:

  • The Titan score of the alliance (what * Titan the alliance is hitting)
  • The Trophy score of the alliance (the individual “cups” of the members)
    • “Cup Dropping” does not factor into alliance war matching
  • The number of members the alliance has (the formula ignores members “opted out” for war)
  • The length of time an individual member has been in an alliance
    • Mention this as there seems to be a misalignment of many that “new” members mean that the alliance is getting big fish just to skew matching – it doesn’t. Many reasons members leave and come back – even in between matching and start of war
  • The level of the individual members of the opponent vs. the matched alliance
    • Some people just “farm” faster than others – does not equate to the power of their roster
  • The TP of the individual teams displayed either for Raid or even on the war field
    • Prevents individuals from purposefully fielding a weak team to get a more favorable match-up
    • Displayed TP does not show the strength or depth of a member’s roster. A super strong defense may mean that the rest of the bench was “neglected” – war is two-sided… offense and defense.

ALLIANCE WAR HISTORY COMPONENT – AKA WAR PENALTIES 5

Using the first three bullets of the alliance war score factors as a “base” Alliance War Score, a new alliance’s war score will start here as there is no “history” to factor into the score. Every time an alliance wins a war, they will “earn” a penalty which is simply an addition to their base war score in order to match them with other alliances in the same skill range. Win = +1 penalty. Loss = -1 penalty. An Alliance can earn up to 20 war penalties before they are considered to be at “War Cap” i.e. War score will no longer increase based on War wins. The actual math behind how much a war score will go up or down with each win/loss is not known, but via the data tracked over the last few months of the top 100+ alliances, a full 30 member alliance with high heroes and troops will increase/decrease 2000-2500 points.

How do you know what “penalty” your alliance is at? Track it like I do. Unless you have been tracking since the formation of the alliance, there is no way to truly know what penalty you are at unless and until you hit the Cap and WIN = 20 penalties. The war score won’t change then.

How do you know what “penalty” your opponent is at? You can’t unless you track the war history of ALL the alliances you may potentially match. This is impossible outside of the Top 100 where war scores are visible. Which brings me to… the data! But before I get there… why do we care about penalties? This is where much of the match-making frustration is coming from. Consider the below example:

  • My Alliance is at 20 Penalties (at Cap) matched against 20 Penalty Opponent (at Cap)
    • Base War Scores are fairly even across both Alliances (our “peer”)
  • My Alliance is at 19 Penalties matched against 20 Penalty Opponent (at Cap)
    • Our Base War Score is higher than the Opponent (we are “stronger” – fighting down)
  • My Alliance is at 20 Penalties (at Cap) matched against 19 Penalty Opponent
    • Our Base War Score is lower than the Opponent (they are “stronger” – fighting up)

That’s fine for alliances that have been “established” and have a war history for consideration in the matching. However, when a new alliance forms, they have zero penalties. They “start” in the middle of the ladder. They may be “matched” to an alliance with a much lower base war score but have made it up to 15 penalties. So even though the war scores appear to be an evenly matched war, the strength and depth of the newly formed alliance is much greater than the alliance happily grinding through the system 6 7.

Chili Transparant #MustHaveGottenLucky!

21 Likes

TOP 100 ALLIANCE DATA ANALYSIS

So… data! My tracking is done with the top 100 alliances because that’s all I can see. It’s not intended to be “elitist.” But understanding the TOP also helps to explain for the masses and see how the changes SG makes affects the community as a whole. @Zombies and I have been tracking war scores for the better part of a year now with a 2 month break when SG increased the War Penalties Cap from 10 to 20 on November 29th, 2019 1. Picked it up again once the first alliances reached CAP in the middle of January 2020.

**Please vote for this idea so that this manual process can be “easier!”

War scores are tracked for alliances with a full 30-man roster so it compares apples to apples. Side benefit – shows alliances that are chronically not at 30-man roster (purposefully?) :thinking:

I track any alliance that has hit the top 100 that has a war score over 128K full (30 members). Right now I’m tracking 126 alliances so there has to be a cutoff somewhere.

If an alliance is below 30 for 5 wars they come off my list until they are full and over 128K again. And, of course, frequent name changes can throw me off… :stuck_out_tongue_closed_eyes:

This is by no means intended to be a “perfect” system as it is still very manual on a dynamic process. War scores and alliances are captured at a “point in time” and not reassessed throughout the war break period between war end and the next matchmaking unless there is an outlier that I need to revisit. As an example, I capture the score at 30 members. Two situations may occur:

  1. The alliance loses a member after I record the score and actually end up warring at 29 members. I don’t notice and may record a “Cap Win” even though they didn’t war at 30 (A win at 29 is NOT a CAP win ethically)
  2. The alliance is at 30 but has a member opted out - will appear as a loss to me sometimes if the alliance actually won their previous war. Their score went up ~2500 for win but goes down ~5000 for opted out player so appears to go down ~2500 = loss. I can sometimes “catch” this within next battle or two if score deviates more than expected with a win/loss

Plan is to update this post on a monthly basis with updated rankings for those interested.

Legend

Top 50 Alliances by # CAP Wins since May 1st:

Top 50 Alliances by CAP score (highest score they have achieved to this point):

Top 50 Alliances by last 10 war average:

Enjoy!
Chili Transparant #MustHaveGottenLucky!

13 Likes

Great summary!

Thank you.

3 Likes

Thanks for all your work on this. Quite interesting to have a true war power rankings. We will get you one day @Aquaginera_7DD :panda_face:

5 Likes

Thank you for all the hard work that went into the making of it, and the hard work of maintaining it. :slight_smile:

3 Likes

(Deleted weird post)

3 Likes

Cookie Settings