Spire of Eternity Sqaud size test

[Dominionofgod]

Iv been doing a few tests lately in my two cities to compare a higher squad size vs lower squad size in combat. - Of course my ultimate first objective is to be able to do my 2500~ points tourney and all spire fighting.. and be able to fully retrain my army by the next weeks new spire/tourney. But it does bring me to that old question about squad sizes. Now i know in the last tourney over haul they changed everything around some and it was stated that squad size should no longer play a major role in this. But wheres the fun in taking them at there word...

Im also in an interesting position to test a few things. Iv been on hold in the tech tree while i use up my expiring buildings that iv won in the last several years. I wanted to use them all up before starting woodelves. Which gave me a lot of down time (good to work on AW tho). So with that down time iv now reached the point where both of my two cities are at the same point in the game. So i built there cities up the same and have nearly two identical cities... except for there squad sizes. So it was time to experiment on some things!

For this specific test i wanted to see the results from the spire.

-------------------------------

City set up:
My two cities, Dom and Necro for short.
Both are elves.
Both are at the exact same spot in the tech tree (unlcoked advanced scouts in woodelves and then stopped)
* Only tech tree difference is some optional squad size upgrades, obviously.
Both have a 700% relic boost form relics
Both have a 88% relic boost form the mountain halls AW
So Both have a total of 788% boost.
Both cities have the exact same AW's.
Both cities have the exact same AW levels
Both cities have placed the same number of regular and premium expansions
Both cities have the same military build.
Same lvl barracks/merc camp/training ground.
Both have lvl 10 fire phoenix and lvl 10 brown bear placed.
As stated, all the AW's are same level, so all the military aw's are the same level.
(Golden abyss, Mountain halls, Bulwark, Shrooms, Needles, Monastery, Prosperity tower)

Differences:
They have different boosted goods, thus different goods factories placed. (This shouldnt rly effect anything in the test. But different factories does mean slightly different working populations on each city. Which then in turn means different values on golden abyss and Mountain halls population bonus. Which again, should not effect anything. But im trying to be through.)

The bulwark and shrooms produce units every 3h based on squad size. Different squad sizes means different values of production. However for this test, no units were collected so its not a factor in the results.

So im not sure how much more alike i could make them. I think i have all the basis covered militarily wise.


*The notable difference is that my
Dom city has a squad size of 861
and Necro has a squad size of 726



The Test Parameters:

I ran a previous info gathering test to prove that the spire encounters for both of my cities have the exact same enemy formations and the exact same amount of enemies. As you can see in the 'First spire encounter' picture, they are the same. - Note that the goods requested for negotiation sometimes change between the two cities. But the enemies are always the same.

I wanted to compare the final results of how many troops were lost from each city after a full run of the spire. And since the enemies are the same for both cities, i can send in the same formation in both cities to get a direct comparison or who is doing better/worse/same per encounter.

I had a lot of time instants saved up so i was able to do the full spire all at once, in both cities. So no additional troops were collected at all during this test.

Both cities had there fire phoenix fed, as well as 1 Dwarven armor, 1 magnificent mage multiplier, and 1 enlightened light range building out and active.

I sent in the exact same troops in both cities for each of there matching encounters.

Should an encounter fail, i would then negotiate as to not mess the numbers up from an 'additional fight/encounter'. That way both cities had the exact same number of encounters.


Expectation:

Since the enemy formation and numbers are idential in both cities, i would expect the larger army from Dom to yeild better results/less losses. And the smaller army from Necro to take heavier losses and more defeats.



The Test Itself:


As the text in each picture says... this one shows my starting amount of squads

First encounter in both cities. Identical enemy formation and numbers.

Spires first floor boss. Sending in the same troops in both cities, but Necro and his smaller squad size actually had more surviving units then Dom and the larger squads.

Second floor boss. Kinda speaks for itself. Same enemies. Sent in the same troops. Necro with the smaller squads win, while Dom and the larger squads failed. (much to my dismay) And Dom failed quite spectacularly too. Most of the enemies are still alive.

Third floor, 6th encounter. Necro wins while Dom fails again. - Granted, Necro won by a small margin. And Dom failed by a small margin. - But a win is a win and a loss is a loss.

Third floor boss. - I messed this one up in terms of getting the picture for necro. I got excited after finishing Doms pic that i was nearly done.. and accidently clicked 'back to spire' button on necro before grabbing the pic. ><
However, its worth noting that i did not expect to win this fight given the enemy formation, but it seemed a waste to do all this in the spire and then not fight the last boss and get accurate results. So i fought him anyway.

Dom failed (as i assumed both would). But necro surprised me and actually won the whole thing. Not only that, he did it in incredible fashion. The vast majority of his troops survived. I was very very surprised by it. I really really really wish that i grabed that pic...

Test results:

There were two points of comparisons i wanted to make from the start.
1. Encounter vs encounter from one city to the other. And that result is fairly clear. All three encounters that Dom failed, Necro won.
2. The over all troop loss.

I compared the squads as well as converted them to actual troops lost a well. You can see the final numbers on the right side of the docupicture.
Over all, necro and his smaller squad size, lost less troops.

So...
Neco and his smaller squad size won all the encounters and lost less troops
Dom and his larger squad size was defeated 3 times, and lost more troops.

Now iv done everything i can think of to make every variable (other then SS) identical. If iv forgotten or over looked something, please tell me. I want to make these as accurate as i can.

I will note two things:
1. im not sure if the terrain is the same on both cities. I auto fighted the whole thing. - If the terrain is different, there is no way for me to control or compensate for that variable that i can think of. Other then to hope that it balanced itself out over the long run.

2. This one was brought to my attention. I did not think to check for this as i did this whole thing right after i woke up.
Was the squads i sent into combat in each city of the same size, or were they different?
That one ill admit could change the processing of the results some. But since i finished the spire to do this test, i do not have another encounter available to check this right now. I will have to wait until next weeks spire and compare that as i go up. So, I/we will have to wait a little while to get this answer.


--------

Fyi; Im not presenting this as some definitive proof type of thing or anything like that. It was merely a single test in a single area of combat (spire).
If iv forgotten something, over looked something, got something wrong.. etc.. please let me know.

For two cities so much the same, they had notable different results. And obviously, seeing if i can duplicate these results will take time..


Edit: Forgot to mention, i only used barracks troops for this test.

 

[Gladiola]

Unless I am reading the battle results wrong (which is entirely possible) Dom has fewer squads in the fights. Is that because some of the squads were wiped out in earlier waves? Or did that player start with fewer squads?

If larger squad sizes _on both sides_ lead to more total wipeouts of one or more squads (which seems possible for roughly the same reason tiny squad sizes lead to 100% survival), then that could lead to more defeats for someone with large squad sizes.

Very interesting, and makes me want to skip optional squad size researches for now.

 

[Dominionofgod]

Unless I am reading the battle results wrong (which is entirely possible) Dom has fewer squads in the fights. Is that because some of the squads were wiped out in earlier waves? Or did that player start with fewer squads?

The system doesnt show dead squads in 2nd or 3rd waves. That is what your seeing. So if i sent in 5 squares of units into battle, and on wave 1, one of those squares was annihilated.. then when you get the starting graphic for wave 2, it will only show the 4 'alive' squares of units and not the dead one(5th). - So when you see less then 5 squares, its because the rest were killed in earlier waves.

Very interesting, and makes me want to skip optional squad size researches for now.

I started necro a long while after Dom. So when i found out about the benefit of skipping optional squad size upgrades (before the tourney change) i started doing that on both cities. But necro was earlier on in the chapters then dom was. So he got to skip more of them. Now that they are the same in chapter progression, its interesting to compare them.

I also wanted to sort this out before using those troop instants that give squads. ... Turns out that was a bigger rabbit hole then i intended lol.

 

[Deleted User - 1178646]

Check manually if both battles use the same maps.
Because a single different obstacle on the map could lead to an entirely different outcome.

Unless all factors are the same (SS should not matter as difficulty ratio should be the same) you are comparing apples to pears.

 

[Gladiola]

Ratios are funny things. Half of 4 is 2, and half of 100 is 50. But 50 is 48 more than 2. It's a bit like Schroedinger's cat, which has a certain probability of being dead or not dead. On average, theoretically that should result in that ratio of live cats to dead cats being the same no matter how many cats (or troops) there are. However, if rounding is resolved in favor of the player and there is more rounding at smaller squad sizes, it seems possible that having smaller squad sizes could still lead to favorable outcomes.

Rather than an exact probability, it's more like something is approaching a limit. As the numbers get bigger and bigger, the limit of the average probability is achieved. Picture a staircase where one unit can survive, or two units, or three units, etc., in a rising curve. If the "uneven" units are rounded down (and therefore do not die) and the curve approaches the given probability of annihilation/defeat (the limit) more exactly with larger numbers, then having smaller sizes would be beneficial.

 

[Dominionofgod]

Check manually if both battles use the same maps.
Because a single different obstacle on the map could lead to an entirely different outcome.

have to wait until next spire to check that.

Unless all factors are the same (SS should not matter as difficulty ratio should be the same) you are comparing apples to pears.

Im doing everything i can to eliminate any variables. Specifically to make sure that its a direct comparison as it can be. Thats part of why i posted this test. I have to ensure that the test itself is accurate before jumping to final conclusions about the results.

I acknowledged that i need to check on the number of units being sent into combat on my side in relation to the two cities; since if that is different then it will effect the results.
And i acknowledged the terrain could change outcomes. Iv seen that before.
I also have to wait until a new spire starts to begin checking any of that since im finished for this week form doing this test.

So im trying to see if there is anything else i need to check/account for.

Edit: Also worth mentioning; this started a few weeks back as a different test. One designed as a broad, overall test. Taking both my cities that were militarily identical (aside from squad size).. run all tourney, run all spire, retain units all week. And check at the start of the next week which city was closer to being back to full unit strength. Naturally, at the end of that everyone complained because of variables. - Which is true, but that is the essence of the game.
So iv begun to try and break it down into smaller more controlled tests. The spire seemed a good place to start because i could do the whole thing at once, and i already new that the enemy units and formations were the same in both cities. So i begun there. And now im trying to elimiate as many variables as i can to get a more accurate result.

 

[Dominionofgod]

UPDATE

i went though and checked a few things in the spire.
The squad sizes i was sending in against the enemies were the same on both cities.
However the terrain was different for each city.

And according to the believed formula;
The activated obligatory techs are the same in both cities.
The Aw's and levels are the same
The expansions and primium expansions are the same.
the relic boost is the same.

So i have made everything i can the same.
The two things that i can not control are the terrain in combat, and the damage range of the units. Obviously, two massive factors of battle that can seriously alter the outcome of a fight.
There is nothing i can do about those two. The only thing possible would be to record the outcomes and look at the combined data over a longer period of time in hopes that those differences would even themselves out between the two cities.
As far as i can see atm, all that combined is as close to exact as a player could get.