Why bother with T3 - numbercrunching

[601854]

After being told by my alliance I need lots of T1 troops, I decided to do some testing to see whether this is true. I did a total of 12 attacks against lvl 6 and lvl 7 pict camps. My results are as follows

2000 T1 against lvl 6 Picts: 1212 survived, resource lost: 394000
2000 T2 against lvl 6 Picts: 1713 survived, resource lost: 287000
2000 T2 against lvl 6 Picts: 1662 survived, resource lost: 338000
1000 T2 against lvl 6 Picts: 604 survived, resource lost: 396000
1000 T1 + 500 T2 against lvl 6 Picts: 282 T1 survived, resource lost: 359000
1000 T1 + 500 T2 against lvl 6 Picts: 231 T1 survived, resource lost 384500
800 T1, 300 T2, 150 T3 against lvl 6 Picts: 103 T1 survived, resource lost: 348500

2000 T1 against lvl 7 Picts: 677 survived, resource lost: 661500
2000 T2 against lvl 7 Picts: 1557 survived, resource lost: 443000
2000 T3 against lvl 7 Picts: 1775 survived, resource lost: 450000
1000 T2 against lvl 7 Picts: 281 survived, resource lost: 719000
500 T3 against lvl 7 Picts: 128 survived, resource lost: 744000

All in all, my resource losses in attacks seem quite stable. To avoid confusing, I ignored gains from the attacks. In the mono troop type attacks I attacked the lvl 6 camps with horses only, the lvl 7 camps with ground troops only. When I combined T1, T2 and T3 troops against the lvl 6 camp, I used T1 cavalry and T2/3 ground troops.

Conclusion
No matter what troop types I used or what setup I used, the resource losses seemed about equal. Even adding ground T2 troops together with lvl 1 horses against lvl 6 camps, did not change the results. To me, it would seem that T1 fights first until they are all dead, then T2, then T3. Considering that T2 troops yield 16 might per 1000 resources and T3 troops yield only 12 might per 1000 resources, I fail to see why anyone would use T3 troops. They take longer to train, yield less might and provide no considerable advantage over a double amount of T2 troops. Only advantage would be faster march times, but since they usually get send together with T2 troops, this advantage is usually lost.

[Unregistered]

After being told by my alliance I need lots of T1 troops, I decided to do some testing to see whether this is true. I did a total of 12 attacks against lvl 6 and lvl 7 pict camps. My results are as follows

2000 T1 against lvl 6 Picts: 1212 survived, resource lost: 394000
2000 T2 against lvl 6 Picts: 1713 survived, resource lost: 287000
2000 T2 against lvl 6 Picts: 1662 survived, resource lost: 338000
1000 T2 against lvl 6 Picts: 604 survived, resource lost: 396000
1000 T1 + 500 T2 against lvl 6 Picts: 282 T1 survived, resource lost: 359000
1000 T1 + 500 T2 against lvl 6 Picts: 231 T1 survived, resource lost 384500
800 T1, 300 T2, 150 T3 against lvl 6 Picts: 103 T1 survived, resource lost: 348500

2000 T1 against lvl 7 Picts: 677 survived, resource lost: 661500
2000 T2 against lvl 7 Picts: 1557 survived, resource lost: 443000
2000 T3 against lvl 7 Picts: 1775 survived, resource lost: 450000
1000 T2 against lvl 7 Picts: 281 survived, resource lost: 719000
500 T3 against lvl 7 Picts: 128 survived, resource lost: 744000

All in all, my resource losses in attacks seem quite stable. To avoid confusing, I ignored gains from the attacks. In the mono troop type attacks I attacked the lvl 6 camps with horses only, the lvl 7 camps with ground troops only. When I combined T1, T2 and T3 troops against the lvl 6 camp, I used T1 cavalry and T2/3 ground troops.

Conclusion
No matter what troop types I used or what setup I used, the resource losses seemed about equal. Even adding ground T2 troops together with lvl 1 horses against lvl 6 camps, did not change the results. To me, it would seem that T1 fights first until they are all dead, then T2, then T3. Considering that T2 troops yield 16 might per 1000 resources and T3 troops yield only 12 might per 1000 resources, I fail to see why anyone would use T3 troops. They take longer to train, yield less might and provide no considerable advantage over a double amount of T2 troops. Only advantage would be faster march times, but since they usually get send together with T2 troops, this advantage is usually lost.

Thx for testing Terrantal!

[Unregistered]

Good analysis. However might is a very poor indication of army strength. This is most evident in wall defenses where a level 1 trap is worth an astounding 18 might, but also in T3. They are twice as strong attacking, and twice as healthy, as T2 troops, despite only 33% more might (if you are only interested in the scoreboard, build nothing but T1 and wall defenses). As you advance resources won't be much of an issue, instead it is training time. It is easier to queue up 10 hours of T3 training before you go to sleep, or even 24 hours or more if you want to take a day off.

However the biggest advantage of T3 is in formations to take advantage of their attack while weaker armies provide a human shield. Use one type of T1 or T2 troops combined with T3 to take the losses initially. Therefore troops of one type i.e. miltiamen up front will take the initial attacks, EVEN IF the attacking armies are weak against ground. This can protect the T3 which attack first preferentially.

Another smaller advantage is in very large attacks, you can max out your troop limit for your rally point very quickly. So either you send multiple attacks, or send more T2/T3.

T3 used to have the advantage of less upkeep, which is now no longer an issue with 2 cities and no troop desertion. People would dismiss T1 when T3 finished training.

Try out some different formations and you'll see the results. Hope that helps!

[ocupunikojid]

it is extremely confusing that you send different amounts and listed the survivors. I do notice that every time you included t3 your resource loss was lowered. I believe that more testing is needed to show why t3 are worth is. Also while testing you should keep track of research healing potions, poison edge, and the level of the knight you are sending.

t1 have an attack of 1, t2 attack is 2, and the benefit of t3 is that their attack is 4.

try sending more then 150 t3 against lvl 6. sending only 150 their combined attack does not affect the battle enough.

t1 is a meat shield for t2 and t3 and I see no use to send the t2

So if you send the same troops 800 t1, 300 t2,150 t3 and record the losses then send 800 t1, 300 t2, and 2000 t3 and finally send 800 t1, zero t2 and 2000 t3 you should see a significant difference.

Here are my results with a level 4 knight and 9 poison edge 8 healing t1 lt cav, t2 cav t3 cats

first attack 0 t1 survived 285 t2 survived and 150 t3

second attack 162 t1 survived 300 t2 survived 2000 t3

third attack 162 t1 survived 2000 t3

[Unregistered]

it is extremely confusing that you send different amounts and listed the survivors. I do notice that every time you included t3 your resource loss was lowered. I believe that more testing is needed to show why t3 are worth is. Also while testing you should keep track of research healing potions, poison edge, and the level of the knight you are sending.

t1 have an attack of 1, t2 attack is 2, and the benefit of t3 is that their attack is 4.

try sending more then 150 t3 against lvl 6. sending only 150 their combined attack does not affect the battle enough.

t1 is a meat shield for t2 and t3 and I see no use to send the t2

So if you send the same troops 800 t1, 300 t2,150 t3 and record the losses then send 800 t1, 300 t2, and 2000 t3 and finally send 800 t1, zero t2 and 2000 t3 you should see a significant difference.

Here are my results with a level 4 knight and 9 poison edge 8 healing t1 lt cav, t2 cav t3 cats

first attack 0 t1 survived 285 t2 survived and 150 t3

second attack 162 t1 survived 300 t2 survived 2000 t3

third attack 162 t1 survived 2000 t3

Interesting test Farmer BOB...!

What about, if you use only t1 and t3 cav?
Or what about only t1 mil and t3 rams?
Thx

[ocupunikojid]

well I don't have heavy cav but I don't believe it would make a difference which t3 you use since they are not being hit but the original poster used cav so I tried to match his choice of t1 troops. MM should have lost less and archers should have lost more by a small percent.

[baangelstivanovy9060]

it is extremely confusing that you send different amounts and listed the survivors. I do notice that every time you included t3 your resource loss was lowered. I believe that more testing is needed to show why t3 are worth is. Also while testing you should keep track of research healing potions, poison edge, and the level of the knight you are sending.

t1 have an attack of 1, t2 attack is 2, and the benefit of t3 is that their attack is 4.

try sending more then 150 t3 against lvl 6. sending only 150 their combined attack does not affect the battle enough.

t1 is a meat shield for t2 and t3 and I see no use to send the t2

So if you send the same troops 800 t1, 300 t2,150 t3 and record the losses then send 800 t1, 300 t2, and 2000 t3 and finally send 800 t1, zero t2 and 2000 t3 you should see a significant difference.

Here are my results with a level 4 knight and 9 poison edge 8 healing t1 lt cav, t2 cav t3 cats

first attack 0 t1 survived 285 t2 survived and 150 t3

second attack 162 t1 survived 300 t2 survived 2000 t3

third attack 162 t1 survived 2000 t3

The problem with testing like this is that a bigger army always results in less overal losses. Compare fe my 2k T1 against lvl 7 camp, vs 2k T2 against the same camp.

The setup I used thereafter assumed the same attack power for every attack, which made it possible to compare resource losses.

[Unregistered]

it is extremely confusing that you send different amounts and listed the survivors. I do notice that every time you included t3 your resource loss was lowered. I believe that more testing is needed to show why t3 are worth is. Also while testing you should keep track of research healing potions, poison edge, and the level of the knight you are sending.

t1 have an attack of 1, t2 attack is 2, and the benefit of t3 is that their attack is 4.

try sending more then 150 t3 against lvl 6. sending only 150 their combined attack does not affect the battle enough.

t1 is a meat shield for t2 and t3 and I see no use to send the t2

So if you send the same troops 800 t1, 300 t2,150 t3 and record the losses then send 800 t1, 300 t2, and 2000 t3 and finally send 800 t1, zero t2 and 2000 t3 you should see a significant difference.

Here are my results with a level 4 knight and 9 poison edge 8 healing t1 lt cav, t2 cav t3 cats

first attack 0 t1 survived 285 t2 survived and 150 t3

second attack 162 t1 survived 300 t2 survived 2000 t3

third attack 162 t1 survived 2000 t3

Well

800 MM, 300 Sword,150 rams = 724 mm loss
800 MM, 300 Sword,2000 rams = 706 mm loss

Both with lvl 9 poisen, lvl 8 healing and lvl 3 knight
With lvl 113 knight, there is no difference!

I don´t think it´s worth it

[baamasmaarleyay5296]

Its the strongest army to send within the 100k rally point limit

I only build T1 and T3. different situations require different troops depending on strategy.


I'd also recommend u repeat ur tests with mixed tiers as they look wrong

[Unregistered]

So to sum up everything above and add a little more:
- Train T2 if you ONLY want max might, time is NOT an issue, BUT you ARE constrained by food prod rate
- Train T3 if you want max might &/or max atk/life, time is NOT an issue, and you are NOT constrained by food prod rate
- Train T3 if you're "active" (atk or get atkd a lot) AND you have time on your side to prepare for next assault
... If you want the strongest first wave of attacks against a peer opponent, T3 is the way to go
- Train T1 if your T3 forces are depleted and you need to build back power or might fast
... Works well for sustained fights when you need to build atk-life quickly
- Train T1 if you need fodder for farming or for attacking pict camps

Last time I checked (yesterday), at my current point of progress:
- T1 trains approx 115 might/min (fastest)
- T1 trains approx 29 atk-life/min (fastest)
- T2 trains approx 87 might/min
- T2 trains approx 11 atk-life/min
- T3 trains approx 52 might/min
- T3 trains approx 9 atk-life/min
- atk-life/upkeep is the same for all tiers (0.25)
- might/upkeep is highest with T2 @ 2.0 (v. 1.0 & 1.5 for T1,T3 respectively)

(disclaimer: these production stats are not going to be the same as yours, but the T1 v. T2 v. T3 ratios should be close ... I'll keep an eye on these to see if the relative values change much as I progress)

This also assumes that an attack value of * =1, and **=2, and ****=4.

Haven't looked at defenses yet, but I to. Would appreciate a second opinion on my analysis.

[Unregistered]

More on my post immediately above ...

After further review, it appears that the following relative might & atk-life training "velocities" are fairly constant no matter what "level" you are or how developed your city is.

Calculating relative training velocities using T1 values as the basis (unit value 1):
- Atk-Life (T3/T2, T2/T1, T1): 0.8, 0.375, 1
... Which says that T3 builds attack power 80% as fast as T2
... And T2 builds attack power 37.5% as fast as T1
... So, therefore, T3 builds attack power 30% as fast as T1
- Might (T3/T2, T2/T1, T1): 0.6, 0.75, 1
... Which says that T3 builds might 60% as fast as T2
... And T2 builds might 75% as fast as T1 - Atk-Life (T3/T2, T2/T1, T1): 0.8, 0.375, 1
... So, therefore, T3 builds might 45% as fast as T1
These relative training velocities were calculated from the basic/starting training rates AND are consistent with what I am experiencing now (earlier /min data, lvl 35) to within 0.1%.