Type of hero probability

[Ursulet]

Hi there,

I am in zolotonosha and i had a 88 gun at 294 kills, with a 6 kills fighter target. I shot and got a range hero. So i decided, for fun, to shoot at it 100 times and write down what heroes i 'd got.

A1 : 19%
A2: 6%
A3: 4%
D1: 24%
D2: 10%
D3: 6%
Spotting: 14%
Movement : 8%
Range: 9%

I thought u had an equal chance to get any hero, but its obviously not the case. While i think that spotting, movement, or range must have a 10% probability to draw, attack+3 or defense +3 must be around 5%.

All in all, 50% chance to get a defense/spotting hero on a AA gun.

[captainjack]

The pattern for heroes is equal (1 in 5) chance for A, D, Spot, Movr, Ini. However, you get Range* instead of Ini if in AA or Art mode when the hero is awarded.
Then, if A or D, 4 in 7 are ÷1, 2 in 7 are +2, and 1 in7 are +3.
So your spread looks reasonable for 100 tries, but isn't a perfect match because real life doesn't do stats until you do lots if runs - you'd probably be close to the official spread if you did 1,000 tries.
I've had one campaign where almost half hereoes were spotting, and not long after I had hardly any. Funny how I hate them when they come, but miss them when they're gone.
* Note that the Range instead of Ini might be if Range is 1 or more rather than specific to Art or AA.

[Ursulet]

The pattern for heroes is equal (1 in 5) chance for A, D, Spot, Movr, Ini. However, you get Range* instead of Ini if in AA or Art mode when the hero is awarded.
Then, if A or D, 4 in 7 are ÷1, 2 in 7 are +2, and 1 in7 are +3.
So your spread looks reasonable for 100 tries, but isn't a perfect match because real life doesn't do stats until you do lots if runs - you'd probably be close to the official spread if you did 1,000 tries.
I've had one campaign where almost half hereoes were spotting, and not long after I had hardly any. Funny how I hate them when they come, but miss them when they're gone.
* Note that the Range instead of Ini might be if Range is 1 or more rather than specific to Art or AA.

Indeed i was in AA mode and got no initiative hero. However when you say that The pattern for heroes is equal (1 in 5) chance for A, D, Spot, move and range/ini i am fairly suprised. I know i only did 100 tries, but i got 69% A or D heroes. Very far from your 40%.

[faos333]

this is an issue I have encountered also, the probability of heroes appearing, although I have not done a similar test, tends to be non typical.

[Ursulet]

Made another 100 tests with a fighter in poznan.

A1: 17%
A2: 4%
A3: 6%
D1: 21%
D2: 13%
D3: 4%
Move: 10%
Spotting: 12%
Initiative: 12%

So attack and defense heroes are 65% vs 69% in the first run.

[Ursulet]

Over 200 tests, chances you get an attack hero is 28%, 39% a defense hero, and circa 10% for each of the remaining types.

[PeteMitchell]

2 x 100 tests... wow!!!

[Ursulet]

Yeah i have too much time on my hands.

[goose_2]

God Bless you for this. It shows there is a huge chance you will get Attack or Defense hero, and if you are not careful most will be the disappointing +1 att or +1 Def. Ugh

Thanks for the perseverance on this one

[heinzrondorf]

Yeah i have too much time on my hands.

Great thread and work! Thank you very much for this initiative and taking the time to do this!

I agree with previous posters that the sample needs to be a bit bigger but if your results stand even with a bigger sample you may be on the verge of debunking what I and most people have always considered a fact - that heroes are distributed randomly. Very exciting .

A bit off topic but speaking of heroes… Even though those +A3 and +D3 heroes seem to be quite elusive you sometimes hit the Panzer Corps hero jackpot. On my current Rommel play through I got these two heroes for my Bf109E fighter in the space of one (in-game) day at Albert canal (GC40). Check the dates the heroes were awarded.



I will admit I cheated a bit in order to get the first, +A2, attack hero as I reloaded when I first got a spotting hero and then a move hero as the first hero, but the second hero, the +D3, was pure luck and came out of nowhere at only 300-400 something kills. What an awesome unit he is going to be as I am only about to start the Stonne scenario in GC40 and he has close to 500 kills already. With a bit more luck he might reach three heroes in GC41 or 42 but I might cut down on his deployment by then in order to bring Otto Kittel from Streets of Moscow up to speed.

Overstrengthened with 3 stars he is a beast and usually delivers 8-10 kills per shot against enemy aircraft and often 3-5 kills when strafing ground units. For a fighter this is almost (+A3 would of course be even better) the perfect hero combo in my opinion and to have it this early in the GC is awesome . I can't wait for the Dunkirk scenario when I get to match him up against the RAF fighters.

[Ursulet]

Yeah + 3 attack are quite rare on fighters. They are so cool to finish off enemy tanks or infanteries.

I once had a stuka with +3 A in la hague and next turn i got a second +3A. Like a mini rudel.

[Ursulet]

Made a last 100 runs with a panzer A in poznan.

Had inversed trends for attack and defense.

A1: 17
A2: 11
A3: 8
D1: 13
D2: 8
D3: 8
Ini: 14
Move: 10
Spot: 13

As you can see that run was luckier. Better d3/a3 heroes.

All in all after 300 runs
A1: 18%
A2: 7%
A3: 6%
D1: 19%
D2: 10%
D3: 5%
Ini/range:11%
Move: 10%
Spotting : 13%

Attack and defense are more balanced.

[Ursulet]

Made a last 100 runs with a panzer A in poznan.

Had inversed trends for attack and defense.

A1: 17
A2: 11
A3: 8
D1: 13
D2: 8
D3: 8
Ini: 14
Move: 10
Spot: 13

As you can see that run was luckier. Better d3/a3 heroes.

All in all after 300 runs
A1: 18%
A2: 7%
A3: 6%
D1: 19%
D2: 10%
D3: 5%
Ini/range:11%
Move: 10%
Spotting : 13%

Attack and defense are more balanced. I suspect that if i would do it 10000 times, there would be 33% chances to get an attack heroe, same for a defensive one, 34% to get one of the others.

[captainjack]

Now I'm back on PC and able to find and copy things:
viewtopic.php?f=121&t=101357&p=879600&h ... ro#p879600
Includes some links back to Rudankorps' explanations.

One of the reasons for clusters of heroes is that statistics is really a thing for lots of samples, rather than smaller runs. so during a campaign you will see a lot of variation that looks strange but would even out over 10 or 20 campaigns.

The other factor is that random number generators in computers aren't really random.

Your patterns from the sets of one hundred runs (great dedication to do that) seem to be more or less right but still a bit biased, so that's probably a system fault because the randomizer isn't quite random.

Talking of which, I was about to complain about getting spotting heroes on a Stug3 and a 105mm gun, but then I remembered the two spotting heroes on my cavalry unit and remembered that what's useful can very according to circumstances.

[Ursulet]

Now I'm back on PC and able to find and copy things:
viewtopic.php?f=121&t=101357&p=879600&h ... ro#p879600
Includes some links back to Rudankorps' explanations.

One of the reasons for clusters of heroes is that statistics is really a thing for lots of samples, rather than smaller runs. so during a campaign you will see a lot of variation that looks strange but would even out over 10 or 20 campaigns.

The other factor is that random number generators in computers aren't really random.

Your patterns from the sets of one hundred runs (great dedication to do that) seem to be more or less right but still a bit biased, so that's probably a system fault because the randomizer isn't quite random.

Talking of which, I was about to complain about getting spotting heroes on a Stug3 and a 105mm gun, but then I remembered the two spotting heroes on my cavalry unit and remembered that what's useful can very according to circumstances.

I read it but apart from the dice roll to determine if you get +1,2 or3, it doesnt not explain what chances you get for each type.

[captainjack]

I'm not on PC, so can't easily check references, but what I understand is that once you are awarded a hero, the type of hero is allocated randomly with an equal chance of each of the 5 types (attack, defence, initiative, move, spotting). That's the 1 in 5 chance of each type of hero. If ranged units are awarded initiative, they get range instead, so range isn't a separate hero but an alternate to initiative.
Then, if it's an attack or defence hero you get the +1 to +3 awarded in what should be randomly allocated with 4/2/1 in 7 chance respectively.
To save yourself the work of running the same event 100 times (or as well as, since it is very interesting to see what happens on your runs) you could load up different saved files and see what mix of heroes you got.
If there is a big difference, it may be the the randomiser is set at the time you get awarded the hero, so it has the same bias each time you run it. But since each hero event has it's own history influencing the randomiser, setting (number of units on map, number killed, prssting, hours run etc etc) it should even out over a whole load of hero awards. I hope that makes sense. If not, I'll try again tomorrow when on the PC when it's easier to type and think at the same time.
While I think of it, do others agree that initiative on US and Italian tac bombers is even worse than spotting on artillery?

[Ursulet]

Speaking of which, i am playing us corps right now. First time ever. Got 7 Heroes

A2 on the german rad recon
Spotting on arty
Spotting on genie
D3 on fighter
A3 on arty
A1 on rangers
D1 on wolverine AT.

[Ursulet]

And a spotting hero on my strat bomber. I have long given up reloading. I just dont bother.

[faos333]

Speaking of which, i am playing us corps right now. First time ever. Got 7 Heroes

A2 on the german rad recon
Spotting on arty
Spotting on genie
D3 on fighter
A3 on arty
A1 on rangers
D1 on wolverine AT

D3 on fighter and A3 on arty very well indeed,
spoting on arty bad

[Ursulet]

Speaking of which, i am playing us corps right now. First time ever. Got 7 Heroes

A2 on the german rad recon
Spotting on arty
Spotting on genie
D3 on fighter
A3 on arty
A1 on rangers
D1 on wolverine AT

D3 on fighter and A3 on arty very well indeed,
spoting on arty bad

Yeah i wont complain even though it would be better on a tank/AT. They are 3 *, backed up by arty and they keep getting attacked by the germans.