Low War Participation Explains Shadows Above LossesClan Elder Grant- University of My Parent's House's Basement
Introduction: Clash of Clan wars are won and lost based on a simple numerical metric: stars won during battle. Win more stars than the enemy clan, and you win. Win less stars than the enemy clan, and you lose. The end goal of any war strategy should therefore seek to maximize the number of stars won. End of story, no exceptions.
The ensuing debate on how to achieve this goal must account for a variety of factors inherent to the game. For example, army size, troop upgrades, and barracks levels all influence the strength of a given invading army. At the clan level, decisions on which bases to attack first, or second, if at all, and in what order, influence the effectiveness of those armies in their respective engagements. How then, can all of these variables be reconciled to maximize the number of stars won?
One solution to this question rests on a useful formula for determining the number of stars won: S = nE (S = number of stars won, n = number of attacks used, E = effectiveness). This equation can be rewritten as E = S/n, revealing that E is simply the average number of stars won per attack. (E.G., if 80 stars are won during a war using 30 attacks, E=80/30 = 2.67; an average of 2.67 stars won per attack).
The beauty of this formula lies with the term E, because E can account for all the nuanced factors that plague discussions of strategy discussed above. Everything from army size, upgrades, and attack order are implicitly accounted for by the average number of stars won per attack. For example, if a clan experiences a sudden reduction in army size (hypothetically of course), this would lead to a concomitant reduction in the number of expected stars won. Similarly, if clan members misuse attacks and attack out of order (or attack bases they cannot score stars on), this would also lead to some amount of reduction in the expected number of stars won per attack. Therefore, S = nE should be considered a formidable solution for simplifying the nuanced calculations involved in war plans, as E serves as a relatively accurate catch-all for otherwise unknown variables.
The subsequent study investigates whether S = nE can adequately model and explain Shadows Above's war results by examining historical data regarding stars won, number of attacks used (a proxy for participation), and effectiveness.
Methods: The past two wars were analyzed for stars won, number of attacks used, and effectiveness. In one of these wars, Shadows Above won, and in the other, they lost. Unfortunately, more data could not be accessed, as those records have since expired from the Clash of Clans database.
Data and Analysis:
S = nE
| War 1
| War 2
|
| Shadows Above Enemy Clan | Shadows Above Enemy Clan
|
n (number of attacks used)
| 50 46 | 45 55 |
S (number of stars won)
| 84 48 | 80 83 |
E (average stars won per attack)
| 1.68 1.04
| 1.78 1.51
|
Winner | Shadows Above
| Enemy Clan
|
In addition to the data presented above, I have been also keeping track (to myself) the results of our past three losses. In each of these losses, our clan effectiveness exceeded that of the enemy clan, but we nevertheless lost the war. The realization of this trend, coupled with the data presented above, leads me to the following conclusions.
Discussion
Since winning a war is directly related to both n and E, a high n can compensate for a low E, and vice versa. Considering Shadows Above's data, our effectiveness (average number of stars won per attack), is consistently higher than that of our enemies,
even when we lose. This trend has been consistent for the clan's most recent battles, including 3 losses. Also remember that our effectiveness accounts for all of the nuances of battle plans. Therefore, we can conclude within reason, that our current war structure (how we determine who attacks who, and in what order, with what forces and with what reinforcements, etc), is superior to that of our enemies, and if this alone decided the victor, we should win every time. To repeat, our effectiveness has never been lower than that of our enemies, even if we lose the war. When we attack, we do well. We are always more effective than our opponents. So...why do we still lose wars?
We have established that Shadows Above's E values are consistently higher than that of our opponents, which indicts the other variable, n, in the war losses of late. The only mathematical way to decrease S given an increasing E is a weak n-value (low war participation). This assertion is supported by the fact that we are still losing wars despite E values that exceed those of our opponents. If our n-values were always at least equal to or greater than the enemy clan's and we maintain our current E values (which I think is reasonable), then we would always win. The only variable preventing this reality is a comparatively low war participation rate.
The implications of these analysis suggest that kicking decisions after wars should be based not off number of stars won, but on number of attacks made. Perhaps we should focus our efforts not on the dizzying analysis about what forces to attack with and where, (our effectiveness shows we're already good at that), but rather, that we need to focus on putting people in wars that actually fight. Volunteering to fight in a war, and then failing to attack, dramatically lowers the n-value, which forces everybody else in the war to carry higher E-values to compensate. This is difficult, and increases the odds of losing wars. Fortunately, the remedy for this is easy- don't volunteer for a war you can't fight in, for whatever reason. And if you don't use your attacks, you risk getting kicked.
The limitations in this analysis lie within the limited access to data. However, I encourage you to keep track of all of our wins,
and losses. I suspect future data will continue to support the claim that low participation is costing Shadows Above war wins, unless we decide to do something about it.