With the introduction of the Clan Wars API (thank you, Fizzer), collecting and analyzing data has become significantly easier and more efficient. This forum post focuses on Season 58 of Clan Wars while also touching on some historical data for context. Please note that while I’ve made every effort to ensure the accuracy of my data, I welcome feedback if any discrepancies are noticed.

Season Overview: Season 58 stands out as the most active season in at least the last 15 (based on available data). The season spanned 24 days and featured the following appearances and games for each of the 11 templates:

SEAD: Appearances: 60 | Games: 917

ComBomb: Appearances: 63 | Games: 395

SE1W: Appearances: 62 | Games: 603

S1v1: Appearances: 65 | Games: 369

CommLD: Appearances: 64 | Games: 212

MALD: Appearances: 64 | Games: 165

Volcano Island: Appearances: 24 | Games: 85

French Brawl: Appearances: 23 | Games: 78

Guiroma: Appearances: 22 | Games: 77

Strat Greece: Appearances: 29 | Games: 67

S2v2: Appearances: 41 | Games: 52

To illustrate these statistics, the first graph provides a dual y-axis view comparing the number of games played and total appearances for each template. Additionally, a second graph calculates the ratio of games per appearance by template.




Analysis: The data suggests that the templates can be categorized into two primary pools based on appearance frequency:

Pool 1 (Standard Templates): SEAD, SE1W, ComBomb, CommLD, MALD, S1v1

Pool 2 (Player-Suggested Templates): French Brawl, Guiroma, Strat Greece, Volcano Island

Key Observations:

Templates in Pool 1 appear more frequently than those in Pool 2, with S2v2 being a notable outlier in terms of both appearance frequency and games played per appearance.

The ratio of games played per appearance suggests that CommLD and MALD align more closely with Pool 2 templates rather than Pool 1, despite their higher appearance frequency.

Following that , another topic I want to comment upon is template frequency day after day. Plotting the number of appearances for each template from Day 1 to Day 24 of Season 58 results in these 11 figures:



Analysis of Pool 1 Templates
SEAD & SE1W:

SEAD appears to be the most popular template overall, yet it has surprising gaps in availability (e.g., appearing only twice or fewer for half of the season). This suggests that even the most favoured templates are not consistently distributed.

SE1W, despite being similar in design, shows higher day-to-day variance. This could point to a deliberate attempt to balance its availability or perhaps a natural outcome of randomness in scheduling.

S1v1 & ComBomb:

These templates maintain a more stable 2–4 appearances per day, but their occasional absence (e.g., S1v1 missing entirely on a day) -> lack of strict scheduling consistency.

The observed fluctuations may result in uneven opportunities for players who prefer these templates, depending on their daily availability.

Analysis of Pool 2 Templates

Expected vs. Actual Appearances:

The data suggests that Pool 2 templates (Volcano Island, French Brawl, Guiroma, Strat Greece) were expected to appear roughly once per day, yet this is not always the case. The inconsistency, with some templates skipping days entirely and reappearing multiple times later, can lead to frustration among players who favour these templates.

For instance, Strat Greece, which appeared 29 times, should theoretically have been available daily but seems to follow no consistent pattern.

Impact on Player Experience:

Given that players can participate only once per day, the inconsistent availability of Pool 2 templates makes it harder for those who prefer these templates to join games. If a player's preferred template doesn't appear on a specific day, they miss their opportunity entirely.

Unique Players and Timeslot Popularity



Season 58 saw 665 unique players participating from a total of 89 clans, with daily participation remaining relatively stable throughout the 24-day season. The number of unique players per day ranged from around 220 to 260, with a slight decline in activity as the season progressed. This pattern aligns with expected player behaviour:

Timeslot Popularity



Each day is divided into six timeslots, and the data shows clear patterns in player preferences:

Slot Rankings:
Most Popular Slots: Slot 1 and Slot 6 consistently attract the highest participation. Their timings likely align well with peak player activity across multiple time zones.
Least Popular Slots: Slot 2 and Slot 3 have the lowest participation. Especially Slot 2.

Trends Across the Season

A closer look at individual timeslots (Figure: Timeslot Popularity Over Time and Heatmap) reveals:

Slot 1 and Slot 6 consistently maintain higher participation, even on days when overall activity is lower.

Slot 2 and sometimes Slots 3/4 frequently dip below 30 unique players per slot.

Days 17 and 20 stand out with significantly lower participation in Slot 1. These spikes correlate directly with template selection. Analyzing the impact of templates on participation shows that template selection drives slot popularity. Templates like SEAD, SE1W, ComBomb, and S1v1 consistently attract more players. When these templates are available in a given timeslot, participation rises noticeably.

Day 17, Slot 1: Featured S2v2, CommLD, MALD, FB, and Volcano Island.

Day 20, Slot 1: Featured S1v1, S2v2, CommLD, MALD, and ComBomb. In both cases, the absence of small earth popular templates led to a significant drop in participation.

Edited 1/19/2025 07:34:05

Template Usage and Player Profiles

Player Participation in Templates - Histogram Overview

• Out of the 665 unique players, 68% (roughly two-thirds) participated in only 1–3 templates, emphasizing a preference for specialization.
  
• A minority of players engaged with a broader variety of templates, with only a small fraction playing more than 5 templates.
  

This raises two questions:

Why do most players limit their participation to just a few templates?

What impact does this specialization have on template scheduling and overall engagement?

Participation by Template

Players have been grouped into four categories based on their level of engagement. The results, solely for season 58, regarding how many games each player has played are on these links (images too big for the forum) (https://raw.githubusercontent.com/5siell/wz-random-code/refs/heads/main/clan%20wars%20data/10.png, https://raw.githubusercontent.com/5siell/wz-random-code/refs/heads/main/clan%20wars%20data/11.png)

• 1 Game: Players who only tried the template once.
  
• 2–5 Games: Players who played the template moderately.
  
• 6–10 Games: Players with significant involvement.
  
• >10 Games: Players highly invested in the template.
  

• Popular templates like SEAD and SE1W have a significant proportion of highly engaged players (14-30% playing 6+ games)
  
• Less Popular templates have a higher percentage of players that have tried it once with minimal engagement
  
• This suggests that the availability and quality of specific templates heavily influences retention and sustained participation
  

As the exact numbers can be found at (https://raw.githubusercontent.com/5siell/wz-random-code/refs/heads/main/clan%20wars%20data/12.jpg), a graph of the percentages of engagement are in this Figure


One-Trick Players (OT) vs. Dabblers (DB)

Using the scatterplot of Timeslot Appearances (Supply) vs. Distinct Games Played (Demand):

Low X, High Y (Under-Represented):

• Those templates see more games played relative to their appearances, indicating unmet demand and potential player frustration.
  

High X, Low Y (Over-Represented):

• Those templates appear frequently but may not see proportional engagement, suggesting over-saturation in scheduling.
  

This distribution highlights the need for better alignment between template scheduling and player preferences.

Historical Analysis of Template Supply vs Demand

Season-by-Season Observations (Excluding Season 55), from Season 50-57
Images at:

• https://raw.githubusercontent.com/5siell/wz-random-code/refs/heads/main/clan%20wars%20data/15.png
  
• https://raw.githubusercontent.com/5siell/wz-random-code/refs/heads/main/clan%20wars%20data/16.png
  
• https://raw.githubusercontent.com/5siell/wz-random-code/refs/heads/main/clan%20wars%20data/17.png
  
• https://raw.githubusercontent.com/5siell/wz-random-code/refs/heads/main/clan%20wars%20data/18.png
  
• https://raw.githubusercontent.com/5siell/wz-random-code/refs/heads/main/clan%20wars%20data/19.png
  
• https://raw.githubusercontent.com/5siell/wz-random-code/refs/heads/main/clan%20wars%20data/20.png
  
• https://raw.githubusercontent.com/5siell/wz-random-code/refs/heads/main/clan%20wars%20data/21.png
  

SEAD and SE1W Dominance:


• Across all seasons, SEAD and SE1W maintain exceptional popularity, with the highest ratios of games per appearance (G/A) consistently ranging between 10–15.
  
• Their overwhelming dominance makes meaningful comparisons with other templates difficult, justifying their exclusion in the combined analysis.
  
  Volcano Island and Strategic Greece:
  
• Introduced in Season 56, these templates start with low appearances and demand. However, early data suggests they may be absorbing some player interest from other strategic templates like S1v1, Guiroma, and French Brawl.
  

Combined Analysis Across Seasons 50-57

I have a mistake in an image here that I'll revisit later
---



Excluding SEAD and SE1W from the analysis because they are so popular that they make it impossible to draw results.



Key Observations:

• ComBomb leads G/A among standard templates at 4-5 games per appearance on average
  
• S1v1 follows closely at 3-4 games per appearance
  
• CommLD and MALD appear to have the same games per appearance as Guiroma and French Brawl
  
• S2v2 are so at the bottom I'll refrain from commenting till the end of the forum post
  

Since 7 seasons might not be the biggest sample for this analysis, here's the same graph but for all observations from season 40 to 57, excluding again 55.


Graph Explanation:

Timeslot appearances (X-Axis): Reflects how frequently a template is offered.

Distinct Games (Y-Axis): Measures how much demand a template generates whenever it is offered.

G/A Ratio: Combines supply and demand by calculating games per appearance of the template. High G/A ratios indicate the template generates more games per timeslot when it is available.

SEAD and SE1W excluded from this analysis as they are by far and without any doubt the most popular templates

Interpretation:
General Trends:

CommLD and MALD:

• G/A Ratio: 2–3 games per appearance.
  
• Despite frequent appearances (high supply), these templates generate relatively low engagement (low demand), which could result from:
  
  
  
  • Saturation: Over-representation reduces excitement to play
    
  • Specialised interest: These templates attract a niche group of dedicated players, while casual players may avoid them
    
  • Template Complexity: Factors like game length or skill requirements may deter broader participation
    
  
  
  French Brawl and Guiroma:
  
• G/A Ratio: 3-4 games per appearance.
  
• These templates appear half as frequently as CommLD and MALD but still manage to attract a comparable amount of games. This indicates:
  
• Broader appeal among casual players
  
• Potential under-representation, as they maintain stronger demand relative to their supply
  
  SEAD and SE1W:
  
• The templates have such high G/A ratios (6+ or even 10-15) that they skew the analysis.
  
  Strategic Greece and Volcano Island:
  
• Their performance will need to be monitored in the future as they were introduced just 2 seasons ago
  
  

Edited 1/19/2025 08:09:40

Before drawing arbitrary or blatantly wrong results, to compare better these templates, I look at the historical data from S41 onwards with the number of unique players per template as in


For the four templates the graphs of unique player engagement across seasons are the following images, while for the remaining templates for those interested the links are:

S2v2: https://raw.githubusercontent.com/5siell/wz-random-code/refs/heads/main/clan%20wars%20data/27.png

S1v1: https://raw.githubusercontent.com/5siell/wz-random-code/refs/heads/main/clan%20wars%20data/28.png

SE1W: https://raw.githubusercontent.com/5siell/wz-random-code/refs/heads/main/clan%20wars%20data/29.png

SEAD: https://raw.githubusercontent.com/5siell/wz-random-code/refs/heads/main/clan%20wars%20data/30.png

ComBomb: https://raw.githubusercontent.com/5siell/wz-random-code/refs/heads/main/clan%20wars%20data/33.png






Unique Players per Template:

CommLD and MALD:

• Attract 80-110 unique players per season, with a higher proportion of multi-game participants. These templates seem to appeal to a smaller but more specialised audience.
  
  French Brawl and Guiroma:
  
• Attract 70-100 unique players per season, but most participants play 1-5 games. This indicates more casual interest compared to CommLD and MALD
  
  

Preliminary Conclusions

Given the data,

• CommLD and MALD may be indeed over-represented if the goal is broader engagement as they disproportionately attract specialists. While they appear more frequently, the same players dominate their participation, skewing G/A downward.
  
• French Brawl and Guiroma, with half the appearances, achieve the similar levels of unique participation, indicating they might be under-represented in the current sustem and could benefit from more appearances.
  
• The introduction of Volcano Island and Strategic Greece (speculation) will result in a drop in participation in Strat templates, if we group the above two + S1v1 in the same category, as a player that is proficient in one, typically can play all of them.
  

Consensus

While the data does not statistically confirm anything, a recommendation to reduce the frequency of CommLD and MALD appearances to allow for greater variety and possibly increase player engagement across all templates is made taking into consideration:

• High appearance frequency: CommLD and MALD appear twice as often as French Brawl and Guiroma
  
• Similar unique player counts: Despite appearing more frequently, CommLD and MALD attract roughly the same number of unique players as the other 2.
  
• Lower G/A ratios: CommLD and MALD consistently have fewer games per appearance, indicating that they might not generate as much interest per slot.
  
  

This forum is my documentation for this Forum Request, as well as a place of discussion https://www.warzone.com/Forum/808955-clan-wars-template-balancing

My mind is foggy from all these so I'll revisit the forum request and I would edit it accordingly if something pops up I am mistaken about, or I never thought.

Regarding S2v2 lastly, something I didn't go through whatsoever in the posts above, S2v2 has consistently attracted a niche group of highly skilled players, while newer or less experienced players avoid it due to a perceived lack of competitiveness and frequent losses. This imbalance discourages broader participation and limits its overall engagement.

Personally? It's my favourite template from the whole selection. Objectively? Consider either revising its design to make it more approachable for a wider range of players (more banked time + less frequent as a template but still availability at least once per day. Not this uneven fluctuation where it doesn't appear for 2 days then suddenly is available 3 times in the same day) or reevaluating whether it fits within the goals of Clan Wars as a format.

Edited 1/19/2025 07:56:20

bump