PLEASE PLEASE PLEASE Fix the odds predictor

koala404 · Post by **koala404** » Sat Jan 21, 2012 7:01 pm

Guys,

I love the game. I really do. And yet, there is just one glaringly big problem.

The combat damage predictor. I mean, seriously, why don't you just show us some random numbers. I actually did the math, and my random number generator was only 5% worse than your calculated odds.

What I really, REALLY, object to is the fact that it is WILDLY optimistic always. While I understand some random nature is good, and some unpredictability is good, what i TOTALLY OBJECT TO is the fact that it just is COMPLETE RUBBISH.

May I respectfully suggest that you either:
1. Make it actually predict the AVERAGE result (not some wild-ass best case in fantasy land)
2. Allow us to TURN THE DAMN STUPID THING OFF

As you might see from this post, I am a little bit peeved (read: thoroughly disgusted at your rediculous design decision) and my lack of mental balance by way of anger as a result is making me look for other publishers of wargames that provide me a more stress-free game that involves good decisions based on sensible information (not necessarily foolproof, just sensible would do me) rather than leading me up some garden path.

Annoyed, agrieved and really put out,

A Player of Panzer Corps/General/etc since day one.

VPaulus · Post by **VPaulus** » Sat Jan 21, 2012 7:10 pm

Sorry but this was already debated ad nauseam.
Most people seem to like it, others don't.
So it doesn't need a fix.
However, exactly for people like you, who want to play with it disable, there's a cheat code called Chess.
viewtopic.php?t=25990
This will makes all combats play exactly as they were predicted (unless rugged defense happens).

El_Condoro · Post by **El_Condoro** » Sat Jan 21, 2012 9:02 pm

Welcome to the forums Koala404 (not another Aussie?)

As VPaulus said, this has been debated a lot and the developers have stated they won't be changing it. There are a lot of guys who agree with you and a lot that like wild swings of luck. Personally, I'm in the former camp but the up-shot is the random factor as it is will continue.

As an example, last night in an MP turn my infantry attacked a tank in a mountain hex without artillery support - a 3-3 prediction turned into 2-8 result! No matter if I get a good luck turn later - this amount of swing should not occur IMO. Anyway, we have to accept that it is part of PzC and, in fairness, these sorts of things are extremely rare.

Dreadgod · Post by **Dreadgod** » Sun Jan 22, 2012 5:24 pm

I know the issue is closed, but I have to admit that I LIKE the unpredictability of the combat predictions. Like the old cliche goes, no plan survives contact with the enemy. At the scale we're talking about with PzC, a lot of unexpected results should be the norm.

The true test of a wargamer is when you're hit with these situations. Are you able to recover from a crushing roll of the dice against you? Are you able to capitalize on those rare tantalizing moments when an unexpected breakthrough is laid out in front of you?

Ryben · Post by **Ryben** » Sun Jan 22, 2012 10:09 pm

The true test of the wargamer is having the guts to bear how enemy AFV´s armed with MGs beat your panzers time after time.

I am aware of the randomness but i swear the original campaign was more balanced. I´m playing DLC 40 and sometimes i would like to throw the game across the window. Odds results here and there are fun but watching those pesky Dingos & Matildas attacking my Panzers III&IV every turn....it get on my nerves. Try to do that with a Panzer I and you get crushed. Not fair.

vasilak · Post by **vasilak** » Mon Jan 23, 2012 12:23 pm

Since this issue seems to concern a lot of people (me included, although not so much) I am wondering if there could be an option specifying whether the deviation from the predicted values is smaller or larger. If this does not involve big changes with regards to development then perhaps it could be considered...

airbornemongo101 · Post by **airbornemongo101** » Mon Jan 23, 2012 1:16 pm

Ryben wrote:The true test of the wargamer is having the guts to bear how enemy AFV´s armed with MGs beat your panzers time after time.

Veyr,,,very very good,,,awesome

Amaranthus · Post by **Amaranthus** » Mon Jan 23, 2012 1:34 pm

vasiliak, I think you're asking for something like I suggested here: viewtopic.php?p=294763#294763

That is, averaging of the combat dice.

vasilak · Post by **vasilak** » Mon Jan 23, 2012 1:42 pm

Amaranthus wrote:vasiliak, I think you're asking for something like I suggested here: viewtopic.php?p=294763#294763

That is, averaging of the combat dice.

Well actually no. What I am saying is to have 2 different ways of evaluating combat results. The first (the current one) allows for a bigger deviation from what is predicted (in other words allowing more randomness), and the second one allowing for smaller deviation, so that extremely unexpected results occur only very rarely. The second option could also be provided as a cheat if necessary. However, as I said I am not sure how easy this is to develop.

Amaranthus · Post by **Amaranthus** » Mon Jan 23, 2012 1:46 pm

What I suggested would achieve what you want. Use 1 combat dice, and you get the same result as the game currently yields. If you took that average of 2 rolls, then you'll get unexpected results slightly less commonly. Average 4 dice and you'll get outlier results considerably less often (but it is still possible).

You could make it a user option by specifying the number of combat dice, with the default at 1 (which is how PzC currently plays).

robc04_1 · Post by **robc04_1** » Mon Jan 23, 2012 1:47 pm

Any game that uses random numbers is going to have times where there is significant deviation from the norm. For example in a computer baseball game, if a batter has an expected .333 batting average in the simulation engine, they only will hit between .313. and .353 68% of the time in 600 at bats if random numbers are used to generate the results. That leaves room for a lot of unexpected results. It is just the nature of random numbers.

vasilak · Post by **vasilak** » Mon Jan 23, 2012 1:51 pm

Amaranthus wrote:What I suggested would achieve what you want. Use 1 combat dice, and you get the same result as the game currently yields. If you took that average of 2 rolls, then you'll get unexpected results slightly less commonly. Average 4 dice and you'll get outlier results considerably less often (but it is still possible).

You could make it a user option by specifying the number of combat dice, with the default at 1 (which is how PzC currently plays).

Well in that case yes, this is more or less what I am saying

So since this was mentioned already and the issue is closed I guess it won't be considered, right?

Amaranthus · Post by **Amaranthus** » Mon Jan 23, 2012 1:57 pm

I'm not sure - what was discussed previously was whether the randomness should be in the game or not, and the near-universal response was "yes". However, the exact model of implementation of the randomness was never resolved.

I suggested the an averaging of combat dice for 3 reasons:

1. It would be simple to implement, probably just a few lines of code.
2. Default could be 1 dice, which would be exactly the same as in the current PzC (i.e. default is no change).
3. It has been implemented in Commander: Europe at War - Grand Strategy already, and seems to work really well there.

I only posted this idea regarding the use of this mechanic for PzC a few days ago, so I'm not sure if the Lordz guys would consider it for an upcoming patch.

bebro · Post by **bebro** » Mon Jan 23, 2012 2:40 pm

I didn't participate in those other RNG/prediction debates, mainly because I got the game relatively late last year.

My 2 cents:

I find the results sometimes pretty outrageous indeed, but can live with losses. Still, I'd probably like a somewhat toned down randomness as optional. Maybe like others said in three steps between no random(chess)/medium/high(current), if that's possible.

Personally what I don't like with the current situation is not so much unexpected high losses sometimes, but rather the opposite: I remember quite some cases where either AI or my troops where caught in situations that would absolutely justify heavy losses (fighters attacking unescorted bombers, inf/art attacked in trucks etc), but thanks to randomness (iow: luck) took only minor or no damage at all. I expect to take significant losses myself in such situations, and likewise expect the AI to receive some.

As most others, I would not want to remove randomness completely. So the situations described above should not be totally impossible, but rather exception. I didn't run extensive tests, but at least my impression is they aren't rare.

Post by **IainMcNeil** » Mon Jan 23, 2012 4:29 pm

The predictor is not broken - it gives a true estimate with the info you have available if you understand what it is showing. It is an average.

It does not show you the spread of possible results with the chance of each - it only shows the average if you tried the fight 1000 times. A 1-1 result could end 10-0 or 0-10, but they are very unlikely. The info you really want would be much more complex to show and need some sort of probability graph and is not something we plan to add!

vasilak · Post by **vasilak** » Mon Jan 23, 2012 6:25 pm

iainmcneil wrote:The predictor is not broken - it gives a true estimate with the info you have available if you understand what it is showing. It is an average.

It does not show you the spread of possible results with the chance of each - it only shows the average if you tried the fight 1000 times. A 1-1 result could end 10-0 or 0-10, but they are very unlikely. The info you really want would be much more complex to show and need some sort of probability graph and is not something we plan to add!

I believe the following adjustment is very simple and would probably require very little coding.
Let's consider 2 options for calculating losses in battle:
1) As it works at the moment.
2) First calculate the values as in 1). Then filter the resulting values to limit the deviation from the average values predicted before the actual battle took place.
For example, assume a 2x3 prediction exists for the upcoming battle. "Rolling the dice" gives a 0x5 (as calculated by the current method).
It is easy now to define a function for calculating the final value for each side taking also into account the "distance" of the dice results (0x5) from the predicted values (2x3). Assume for example that this function is defined as follows: X = x1 + (x2-x1)*50%, where X is the final value, x1 is the predicted value, and x2 the value calculated by 1). So in this example if we were to apply the above function we would have a result of 1x4 instead of 0x5. This is merely ensuring that the dice results are 50% closer to the predicted values.
Obviously a more realistic method could be used but in this case it is easy to see that it would probably require just an extra calculation to filter/restrict the results as they are now calculated by the game.

charonjr · Post by **charonjr** » Mon Jan 23, 2012 10:13 pm

Or much more simple, just cut-off/ignore the extreme values - e.g. everything above +/- 3 wil only taken into account up to 3.

Longasc · Post by **Longasc** » Tue Jan 24, 2012 12:33 am

Holy cow, player from day one and NOT used to random results.

Yeah, it was already debated ad nauseam.

How about not throwing a tantrum when things go wrong and planning with the possibility of bad results in mind?

Post by **IainMcNeil** » Tue Jan 24, 2012 10:40 am

We dont want to cut off the extreme results - this is part of the core design

It is working as designed and showing an accurate prediction of the current design. We just don't agree it needs changing - sorry!

airbornemongo101 · Post by **airbornemongo101** » Tue Jan 24, 2012 10:52 am

Longasc wrote:Holy cow, player from day one and NOT used to random results.

Yeah, it was already debated ad nauseam.

How about not throwing a tantrum when things go wrong and planning with the possibility of bad results in mind?

That's my solution

,,as a matter of fact my computer almost went to jump school

yesterday . I came real close to throwing it out the window

I wonder what the P.L.F would be for a computer

.......