Advanced Stats - Skeleton Army
Authors:
Last updated: 2026-09-05
Page change log:
- Created: 2026-09-05
Page contents:
Skeleton Army has been present in the game since it was originally released, back in 2016. This Epic card had been mostly relegated to gimmicky bait decks, but the addition of its evolution in October of 2025 made it meta again.
In this blog we’ll dive into some unusual stats and mechanics of this card. Due to the huge Skeleton swarm it spawns, it gives us an unusually detailed view of how pathfinding works in Clash Royale, and how has changed over time.
Card Basics
- Skeleton Army deploys 15 skeletons
- In the game files, it is known as "Skeleton Horde"
Leaderboard Stats
Skeleton Army is not very popular card on the top of the leaderboard. Among the top 1,000 players of every season, the amount of uses is usually around 10, which is 1% of the total. The usage peak was reached in June of last year, with 16 players.
The release of the evolution (October 2025) multiplied the usage of the card. During the first 2 seasons it was modest, but in December it reached a peak of 160 users, which is 16% usage.
Another interesting way to look at leaderboard stats is the average rank. Unpopular cards can have great results as hidden gems, sometimes they are quite hard to master; but this doesn’t seem to be the case with Skarmy. The normal form of the card has a rank around 600, worse than the expected average of 500.
The evolution has slightly better results, but still below average.
This rank chart also displays the best rank obtained each season, labelled with stars (★). In recent months, this rank has remained high for Evo Skarmy, but Normal Skarmy isn't in a good moment.
During the first 4 seasons after its release, Evo Skarmy ranked #1 in 3 of them; recent seasons aren't as good, but always between #10 and #80.
Split Tech
In this section we’ll explore how different deployments of the Skeleton Army can affect the damage output to opponent towers.
Although the scenario of an undefended Skarmy reaching towers is unrealistic, we think it’s a very interesting way to look at how the game engine affects troop interactions. Small differences are usually unnoticeable during the chaos of a battle, so this experiment is a very controlled sandbox that allows us to see unusual details.
Methodology
Skarmy was placed on each tile, and amount of damage when undefended vs Princess tower was tracked.
Data Collection
These interactions were tested at 2 different points in time:
- November 2025, which had Clash Royale version 12
- September 2026, which had Clash Royale version 16
Both versions had very similar results, but we'll also explore a few differences.
Main Results
The damage output caused by each tile was plotted in this heatmap.
In the Skarmy Damage Heatmap, Dark Green is lowest damage (0 DMG), Dark Red is highest damage (2,511). The data was collected at tournament standard levels, where each heat deals 81 damage, so 2,511 damage is equivalent to 31 hits.
Key Insights
What causes these weird discrepancies in damage? Some of the main factors we have identified are:
- True Blue / True Red (TBTR) variations
- Arena Side difference (Left-Right)
- River / Bridge difference
- Edge / River / Building dispersion
- Skarmy Clumping
- Skeleton Lagging
- Sniped by other tower
It’s worth keeping in mind that the Clash Royale game engine operates in a deterministic way. Once a troop has been deployed, it behaves automatically according to fixed rules. So all the results we’ll share today are completely reproducible and just dependent on card deployments.
What is True Blue / True Red (TBTR)?
This term refers to the inconsistencies in gameplay that depend on which side of the arena you are in. Card deployments aren’t completely symmetrical.
Each battle has a True Blue and a True Red player. Regardless of which point of view you have of the battle, the game engine always sees the True Blue as the battle host, and True Red as the guest.
You can find out more details of this phenomenon and how to identify whether you are True Red or True Blue in this YouTube video.
The main reason for TBTR discrepancies is that being TB or TR affects float values of deployment coordinates. Let’s look at a few specific examples:
| Player | Tile | X | Y | Relative |
|---|---|---|---|---|
| Blue | 1 | 16,500 | 14,500 | Top, Left |
| Blue | 2 | 1,499 | 14,500 | Top, Right |
| Red | 3 | 16,500 | 17,499 | Top, Right |
| Red | 4 | 1,499 | 17,499 | Top, Left |
| Blue | 5 | 16,500 | 17,500 | After Bridge, Left |
| Blue | 6 | 1,499 | 17,500 | After Bridge, Right |
In this diagram you can see that tile #3 and tile #5 seem the same. But depending on which player uses them, the troops are deployed in slightly different coordinates.
There are also clear differences if we look at relative placements:
- The True Blue player, when using their Top Left tile, they are deploying in ( 16,500 , 14,500)
- The True Red player, when using their Top Left tile, they are deploying in (1,499 , 17,499)
What should be simply mirrored positions have very different coordinates in the main frame of reference, which is located at the bottom right of the arena for True Blue. This is what likely creates all kinds of inconsistencies, as these asymmetric coordinates are used for troop interactions and pathfinding.
Main Factor details
Arena Side difference: Some interactions in CR can only occur on certain sides of the map (left or right), and some interactions can only occur on certain quadrants of the arena. This is slightly distinct but also intertwined with TBTR.
River / Bridge difference: This phenomenon is almost completely unknown. Unfortunately I couldn’t find any perfect examples to display this, but basically there are some very rare interactions that are only possible on certain edges of the river/bridge.
Edge / River / Building dispersion: When you place Skarmy in the middle of the arena, there are no ‘walls’ or things in the way, so it will spawn normally. But when you place it near the river, edge of the arena, or crown towers, it will cause the skeletons to be clumped/pushed and will spawn them in a weird configuration.
Skarmy Clumping: For the most damage possible on a tower, you want as many Skeletons clumped as tightly as possible. The further back you place the Skarmy, the more dispersed they will get - so less damage will be done.
Skeleton Lagging: There are a few Skarmy placements, where as the Skeletons approach the bridge, some skeletons will walk backwards, pushing other Skeletons. I do not have a great explanation for why this happens. I do believe sometimes this is caused by a troop being pushed into the bridge and ending up on the other side of something they wouldn’t normally pathfind to, so they backtrack for a second to ‘reset’ itself.
Sniped by other tower: If any Skeletons attacking the Princess tower are sniped by the other Princess tower, it will lead to a large loss in total damage.
Labelled Damage
Arena Average: 1,407 HP (~17 hits)
True Blue Average: 1,402 HP
True Red Average: 1,412 HP
Difference: 10 HP
Row / Column Damage Aggregates
Skarmy Damage Heatmap, with the average damage done to a Princess tower, per column and row.
Quadrant Damage Aggregates
Quadrant 1 Average: 1,432 HP
Quadrant 2 Average: 1,392 HP
Quadrant 3 Average: 1,371 HP
Quadrant 4 Average: 1,432 HP
If the game engine mirrored perfectly, all quadrants would be the same, but the most important are Q1=Q3 and Q2=Q4, to preserve side fairness.
Left-Right Symmetry
As we just showed, your 2 lanes of the arena are not equal. How big are these differences? Do they concentrate in specific tiles?
Here all the values are mirrored around the vertical axis of the arena, but with opposite signs. The left side has significantly more hits overall for both players, although note that it would be the right side if you are playing as True Red.
Blue-Red Symmetry
The previous section could be read as "which lane you should pick", but something you can't pick is whether you are True Blue or True Red. How balanced are the outcomes of these two options? Is one significantly better than the other?
To find out, instead of using reflection symmetry we'll use inversion symmetry.
Overall, the differences between Left-Right are much bigger than the differences due to Blue-Red. This means that you stand to benefit much more from identifying if you are TB or TR and playing the correct lane.
Army Lane Split Counts
This heatmap shows the number of Skeletons that go in the same lane as the tile they were placed in. For example, tile K19 has a ‘13’, which means when Skeleton Army is placed on K19, 13 Skeletons will travel down the right lane, and 2 will travel down the left lane.
Notice the numbers highlighted Red at L1 & L32. These are the only tiles where there is a discrepancy in the number of Skeletons traveling down the lane, when you do a vertical reflection over the river.
Interesting facts
There is no tile on the map where you can place Skarmy unsupported, and it will get damage on 2 Princess towers!
The tiles where Skarmy does the highest damage to Princess towers, are because the Skeletons are tightly clumped, but most importantly, are not sniped by the other Princess tower.
Skeleton Army (as well as other cards with multiple troops in their deployment) used to have their Spawn Order dependent on True Red True Blue, but after I made a video highlighting this issue, it was fixed (2025-03-31)!
The Skarmy Skeletons have names: Larry, Harry, Gerry, Terry, Mary, Barry, Cherry, Karrie, Faerie, Scary, Sherry, Gary, McClary, Huckleberry, and Ledoot.
But when Skarmy is evolved, the Skeleton General is Gerry. So who joins the Skarmy in place of Gerry when he equips his lance, shield, and moves to the
backmiddle of the group 🤔
Old vs New
The analysis we have seen until now was based on data from November of 2025. In recent months, there have been minor changes to the game engine that affected some of these results, so we decided to run our tests again and compare the two versions. Has TBTR symmetry been resolved? Let's find out!
Basic Comparison
The simplest comparison we can make is to compare the amount of hits from each tile in both versions of the game, here are our results:
Positive numbers (green), indicate tiles in which the damage has increased; negative numbers (red) indicate reductions.
We found these results to be very peculiar, and wish we had footage of both versions to compare them in more detail. Almost all of the changes happen in 2 columns of tiles (D & G), while everything else remains (almost) unchanged.
Left-Right Symmetry
Just as we compare the total hits of the old and new engine, we can also compare the symmetry they have. Have the pathfinding changes reduced differences? Or made them worse?
As we saw earlier, the hits didn't change for most of the tiles of the arena. Unfortunately, for those that did, most of these tiles became even more asymmetrical.
Blue-Red Symmetry
Now let's look at the same comparison for inversion symmetry:
The differences between Blue-Red also got a bit worse, but less than LR asymmetry.
Deployment Order
Skeleton Army deployment order is also a fairly interesting and complex topic. We won't cover it in this blog, but you can go down that rabbit hole with this video, which was also created by Milo the great.
Frame from "Reverse Engineering Skeleton Armys Spawn Order", by Milo the great
Raw Data
This section simply exists to facilitate access to our results, for anyone looking to produce other visualizations or explore the data in more detail. If you decide to use it, please mention the source clearly.
Hits per tile - Old (v12)
,,,,,,10,4,0,0,5,12,,,,,,
17,14,16,16,17,13,7,,,,,7,15,13,14,16,16,14
9,12,21,25,25,13,13,,,,,9,16,22,13,18,13,7
9,17,18,20,20,13,13,,,,,11,16,17,19,15,14,8
10,11,15,19,19,17,17,,,,,15,19,19,20,18,10,14
14,13,,,,18,18,6,0,4,9,17,18,,,,14,12
13,19,,,,23,23,6,1,6,10,23,21,,,,15,16
16,18,,,,25,25,10,1,7,13,24,21,,,,20,15
22,23,17,20,20,25,25,11,1,6,11,18,22,22,21,27,18,16
22,21,25,24,24,23,23,9,2,6,9,25,22,23,26,28,22,22
24,25,22,24,24,25,25,8,2,7,13,19,23,27,25,29,23,19
27,21,24,26,26,22,22,10,2,7,15,19,25,27,27,30,23,24
19,27,31,27,27,25,25,11,2,7,15,20,23,25,28,28,26,25
24,23,20,19,19,24,24,11,2,6,14,23,20,24,23,31,25,24
,25,23,26,26,21,21,9,2,7,15,20,20,25,24,26,22,
,,,,,,,,,,,,,,,,,
,,,,,,,,,,,,,,,,,
,20,28,31,31,23,23,8,2,6,14,20,21,25,31,23,26,
21,20,25,23,23,21,21,11,2,6,15,17,19,25,23,29,24,23
20,21,30,28,28,26,26,11,2,7,16,20,22,23,27,29,26,27
21,24,27,25,25,21,21,8,1,7,18,25,23,22,24,30,23,19
18,24,22,24,24,23,23,7,2,7,13,25,22,24,22,29,21,19
17,22,25,24,24,23,23,10,2,6,12,22,24,24,26,26,20,22
18,23,26,21,21,22,22,9,1,6,12,23,23,20,27,24,17,21
22,15,,,,25,25,11,1,7,12,24,22,,,,22,19
18,22,,,,16,16,6,0,7,13,18,18,,,,25,18
13,15,,,,18,18,8,0,4,9,13,14,,,,14,15
10,13,13,20,20,16,16,,,,,14,18,21,18,14,12,14
11,12,15,16,16,16,16,,,,,15,15,19,18,17,12,9
7,13,19,19,19,13,13,,,,,10,12,20,18,19,12,9
15,15,17,21,21,12,12,,,,,11,11,14,14,15,8,15
,,,,,,6,0,0,0,3,10,,,,,,Hits per tile - New (v16)
,,,,,,10,4,0,0,5,12,,,,,,
17,14,16,16,17,13,7,,,,,7,15,13,14,16,16,14
9,12,21,15,25,13,6,,,,,9,16,22,13,18,13,7
9,17,18,15,20,13,9,,,,,11,16,17,19,15,14,8
10,11,15,20,19,17,10,,,,,15,19,19,20,18,10,14
14,13,,,,18,10,6,0,4,9,17,18,,,,14,12
13,19,,,,23,16,6,1,6,10,23,21,,,,16,15
16,18,,,,25,16,10,1,7,13,24,21,,,,20,15
22,23,17,25,20,25,15,11,1,6,11,18,22,22,21,27,18,16
22,21,25,24,24,23,18,9,2,6,9,25,22,23,26,28,22,22
24,25,22,25,24,25,15,8,2,7,13,25,23,27,25,29,23,19
27,21,24,28,26,22,17,10,2,7,15,19,25,27,27,30,23,24
19,27,31,31,27,25,16,11,2,7,15,20,23,25,28,28,26,25
24,23,20,23,19,24,18,11,2,6,14,23,20,24,23,31,25,24
,25,23,29,26,21,19,9,2,7,15,20,20,25,24,26,22,
,,,,,,,,,,,,,,,,,
,,,,,,,,,,,,,,,,,
,20,28,25,31,23,21,8,2,6,14,20,21,25,31,23,26,
21,20,25,24,23,21,17,11,2,6,15,17,19,25,23,29,24,23
20,21,30,29,28,26,15,11,2,7,16,20,22,23,27,29,26,27
21,24,27,21,25,21,19,8,1,7,18,25,23,22,24,30,23,19
18,24,22,25,24,23,16,7,2,7,13,25,22,24,22,29,21,19
17,22,25,18,24,23,18,10,2,6,12,22,24,24,26,26,20,22
18,23,26,29,21,22,16,9,1,6,12,23,23,20,27,24,17,21
22,15,,,,25,16,11,1,7,12,24,22,,,,22,19
18,22,,,,16,14,6,0,7,13,18,18,,,,25,18
13,15,,,,18,10,8,0,4,9,13,14,,,,14,15
10,13,13,20,20,16,13,,,,,14,18,21,18,14,12,14
11,12,15,16,16,16,9,,,,,15,15,19,18,17,12,9
7,13,19,15,19,13,8,,,,,10,12,20,18,19,12,9
15,15,17,13,21,12,6,,,,,11,11,14,14,15,8,15
,,,,,,6,0,0,0,3,10,,,,,,Feedback
Did you know that TBTR could matter this much during battles? Are there other deep dives that you’d like to see for other cards?