Page 1 of 2
Chance of SU to be commited to a battle
Posted: Thu May 26, 2022 11:53 am
by Stamb
I noticed, playing as a Soviets, that if HQ did not move from a start of a turn then there is much bigger chance to get 4-5+ SU into a battle when attacking with deliberate attack.
And in the same time even if you move 1 hex then chances are very low and it is common to see 1-2 SU from a 6 that are assigned to a HQ.
I did not find anything about it in a manual.
Is there some hidden undocumented game mechanic?
Re: Chance of SU to be commited to a battle
Posted: Thu May 26, 2022 3:07 pm
by Stamb
found this:
HQ units that have moved during the current player-turn will see -1 deducted from
the initiative value when determining if an attached Support unit is committed to a
deliberate attack.
still looks like to me that is is much more than -1
as even the best leaders struggle to commit SU when HQ moves
Re: Chance of SU to be commited to a battle
Posted: Thu May 26, 2022 5:22 pm
by loki100
probability theory ...
A lot of Soviet leaders have low init scores, so -1 off say '5' is far more a hit than -1 off '7'
I'll leave you to do the actual maths
Re: Chance of SU to be commited to a battle
Posted: Thu May 26, 2022 8:29 pm
by Stamb
math is one thing
but this is my experience:
i have army with a leader which has initiative 5
HQ does not move, i make an attack and get 4+ SU
i have another army, and i have leader with initiative 6, like Tolbukhin
i move this HQ so initiative should be 5
and as you can see i get only 1 SU to a battle from this HQ
- damaged.png (760.51 KiB) Viewed 759 times
you may say that this is bad luck
but i noticed such behavior in many other situations
that is why i wrote that it looks like it is actually not -1 to initiative
Re: Chance of SU to be commited to a battle
Posted: Thu May 26, 2022 8:50 pm
by loki100
2 bits:
a) as before, single instances in a complex system tell us little, how many SUs were available? how many attacks were made? and then repeat around 10 days (to eliminate the more extreme variances)
b) 7-1 = 60% chance to pass, 5-1 = 40% chance to pass - quite a difference
Re: Chance of SU to be commited to a battle
Posted: Thu May 26, 2022 9:06 pm
by Stamb
a) as i wrote, this behavior is consistent, i just did not have screenshots of every battle
i had 6 arty units in both cases
b) not sure why do you use 7 for initiative
in previous example i described cases with Boldin (initiative 5)
- 1.png (285.64 KiB) Viewed 746 times
and Tolbukhin has 6 initiative, thus moving his HQ reduces it to 5, so his is the same as Boldin
- 2.png (261.7 KiB) Viewed 746 times
also this is /12 checks, as penalty for a Soviet armies is still active
so chances to commit arty should be even lower
Re: Chance of SU to be commited to a battle
Posted: Fri May 27, 2022 1:03 am
by DesertedFox
In addition, the chances
of Support Units being committed and the effectiveness of
artillery are also influenced by the number of CPP a unit
possesses.
For each support unit attempting to be
committed, the leader of that headquarters must pass an
initiative check. The support unit must then pass several
checks, with the checks becoming more difficult based both
on the number of support units already committed and the
total number of non-construction support units attached
to the headquarters unit. This means that Headquarters
units with large numbers of non-construction support
units will have more opportunities to commit support
units; however the overall probability of each support unit
being committed will be less than if the headquarters units
had fewer non-construction support units
Re: Chance of SU to be commited to a battle
Posted: Fri May 27, 2022 7:44 am
by loki100
Stamb wrote: Thu May 26, 2022 9:06 pm
...
b) not sure why do you use 7 for initiative
....
trying to explain probability to you - but if you don't grasp that then you will struggle (specifically just made up a number so that hopefully the variable impact of losing a single point became clear).
as ever, there are so many variables, in every instance, its only by repeating the same thing multiple times will you generate data that supports or disproves your point.
There are several psychological factors at play here including base rate neglect and when we notice something (attribution), but I'll spare you and others a quick recap of my PhD.
What you need is a HQ with say 5 SUs and a given leader. What % of those SUs turn up in each battle if (a) not moved or (b) moved. Does the variance between those 2 percentages match the impact of taking 1 off the relevant initiative score - and remember the lower the final score the more random the system actually gets (we're back to base rates here)
Re: Chance of SU to be commited to a battle
Posted: Fri May 27, 2022 7:52 am
by Stamb
Stamb wrote: Thu May 26, 2022 9:06 pm
...
b) not sure why do you use 7 for initiative
....
loki100 wrote: Fri May 27, 2022 7:44 am
trying to explain probability to you - but if you don't grasp that then you will struggle (specifically just made up a number so that hopefully the variable impact of losing a single point became clear)
i have no problems calculating probability...
but thanks anyway
and yeah, once i have free time
i will definitely make such test
Re: Chance of SU to be commited to a battle
Posted: Fri May 27, 2022 4:03 pm
by Stamb
here is test setup
19th army with Boldin which has 5 initiative + 6 arty SU
HQ did not move from the start of a turn
i will perform same attack 10 times
- HQ.png (5.33 MiB) Viewed 669 times
- 1.png (5.64 MiB) Viewed 669 times
- 2.png (5.65 MiB) Viewed 669 times
- 3.png (5.59 MiB) Viewed 669 times
Re: Chance of SU to be commited to a battle
Posted: Fri May 27, 2022 4:05 pm
by Stamb
- 8.png (5.58 MiB) Viewed 667 times
- 9.png (5.58 MiB) Viewed 667 times
- 10.png (5.58 MiB) Viewed 667 times
Re: Chance of SU to be commited to a battle
Posted: Fri May 27, 2022 4:07 pm
by Stamb
Now i move same HQ with same leader with same SU to 1 hex west of Smolensk
so there will be minor river as in a previous battles between HQ and units
and perform another 10 battles
- 1.png (5.57 MiB) Viewed 664 times
- 2.png (5.62 MiB) Viewed 664 times
- 3.png (5.59 MiB) Viewed 664 times
- 4.png (5.58 MiB) Viewed 664 times
Re: Chance of SU to be commited to a battle
Posted: Fri May 27, 2022 4:09 pm
by Stamb
- 9.png (5.58 MiB) Viewed 662 times
- 10.png (5.58 MiB) Viewed 662 times
Re: Chance of SU to be commited to a battle
Posted: Fri May 27, 2022 4:14 pm
by Stamb
from a manual:
HQ units that have moved during the current player-turn will see -1 deducted from
the initiative value when determining if an attached Support unit is committed to a
deliberate attack
i found another leader that has same stats as Boldin, except of an initiative which is 4 instead of 5
- HQ not moved -1 initiative.png (6.04 MiB) Viewed 659 times
perfect
if everything works as documented then we should see similar results as Boldin had when HQ was moved
next 10 battles are made with a HQ that did not move, just another leader to simulate -1 in initiative
- 1.png (5.63 MiB) Viewed 659 times
- 2.png (5.58 MiB) Viewed 659 times
- 3.png (5.59 MiB) Viewed 659 times
Re: Chance of SU to be commited to a battle
Posted: Fri May 27, 2022 4:16 pm
by Stamb
- 8.png (5.58 MiB) Viewed 656 times
- 9.png (5.58 MiB) Viewed 656 times
- 10.png (5.58 MiB) Viewed 656 times
Re: Chance of SU to be commited to a battle
Posted: Fri May 27, 2022 4:21 pm
by Stamb
and this table represents all of this battles
- total.png (19.83 KiB) Viewed 654 times
what do i see
1. if HQ does not move then chances to get SU into battle are the highest
2. if HQ moves (Boldin 5 initiative -1 for HQ move = 4 initiative) commits less arty
3. if HQ does not move but has leader with 4 initiative, thus same as in 2 then there are at least 1 more SU in comparison to a case when HQ moves but with Boldin (5-1= 4 initiative)
my verdict - as i wrote initially
undocumented mechanics or bugs
Re: Chance of SU to be commited to a battle
Posted: Fri May 27, 2022 4:24 pm
by Stamb