Five Tribes Five Tribes

Forums

Recherche
Forums » Memoir '44 - English » Issues with probability - example Patrol Cars
Montrer: Messages du jour 
  
AuteurSujet
Tempest_Au
Junior Member
Major

Messages: 12
Enregistré(e) en :
September 2011
Issues with probability - example Patrol Cars Thu, 08 September 2011 11:05
Hi there,

I've got a few issues with the relative probabilities of hitting things in Memoir. Granted, it's not meant to be a completely accurate simulation, but these anomolies are irritating.

Firstly, let's look at the base set - and the relative probabilities of killing off assorted units.

Infantry - 4 tokens per unit, and each attack die rolled against them results in a token killed 3/6 of the time.
Therefore, it takes an average of 8 attack rolls to kill off an infantry unit. This is our "baseline" of comparison.

Tanks - 3 tokens per unit, and each attack die rolled against them results in a token killed 2/6 of the time.
Therefore, it takes an average of 9 attack rolls to kill off a tank unit (12 if there are 4 tanks in the unit). Looks like tanks get a bit of a raw deal, but given their mobility and attack range, it evens out.

Artillery - 2 tokens per unit, and each attack die rolled against them results in a token killed 1/6 of the time.
Therefore, it takes an average of 12 attack rolls to kill off an artillery unit.... Even given their range, this seems a little excessive - but it's not completely unreasonable.

Now lets look at something I spotted whilst deciding what expansion sets to get - The Long Range Patrol Car. Bear in mind these are jeeps - unarmored vehicles.

The base chance of hitting one is the same as hitting an infantry - 3/6. However, any hits are re-rolled, and only confirmed on a grenade - making the actual odds of scoring a hit a whopping 1 in 12! So, even if there is only 1 car token in the unit, it will take on average 12 rolls to kill it - making it harder to kill than a tank unit!!

All is quiet in the Western Front, until 3 Panzers come crashing through the hedgerows immediately in front of an allied infantry unit, waving their 88mm canons and coaxial machine guns menacingly. The infantry curse, believing that their commander will send them in head-to-head against the giant steel beasts, but instead he climbs into his jeep. "Don't worry men, I'll take care of these" he says as he proceeds to demolish the tanks with his service revolver

This is nuts, really nuts. Surely there must be an error?

If the object was to make patrol cars "just a bit harder to kill than infantry", then the rule should be changed to "A Patrol Car unit is treated as Infantry when being targeted in battle. The attacker scores 1 hit for each Infantry dice symbol rolled against the Patrol Car unit. Any dice with a Grenade symbol are re-rolled, with an Infantry dice symbol or Grenade symbol indicating a hit, and all other symbols on the re-roll ignored."
      
clexton27
Senior Member
Historien accompli

Pages Perso
Messages: 3089
Enregistré(e) en :
February 2007
Re:Issues with probability - example Patrol Cars Thu, 08 September 2011 11:40
I think the basic idea was to recreate the "hit and run" effect of a fast patrol car attack, not exactly replicate the effect of using a jeep versus an infantry or armor. This game is not meant to be an truly accurate representation of battlefield forces.

You can use the patrol car units as intended or create your own "house rules" on them. Many in the forum have done so with many of the units and have their own "house rules" for many different types of units. Use the SEARCH to do the topic "house rules" and I am sure you will find a plethora of tried and true rules.
      
Tempest_Au
Junior Member
Major

Messages: 12
Enregistré(e) en :
September 2011
Re:Issues with probability - example Patrol Cars Thu, 08 September 2011 20:47
stevens wrote

I think the basic idea was to recreate the "hit and run" effect of a fast patrol car attack, not exactly replicate the effect of using a jeep versus an infantry or armor.


Strange that the designers would choose to create an "ultra jeep" to accomplish this, when they could just as easily chosen to allow Patrol Car units to perform a second move immediately after attacking (in the same manner as a "behind enemy lines" card).

stevens wrote

This game is not meant to be an truly accurate representation of battlefield forces.


True, but all the same the game doesn't have cyborg laser-troops either - because such a unit would be completely unrealistic.

stevens wrote

You can use the patrol car units as intended or create your own "house rules" on them. Many in the forum have done so with many of the units and have their own "house rules" for many different types of units. Use the SEARCH to do the topic "house rules" and I am sure you will find a plethora of tried and true rules.


ROFL. If I didn't live in a house full of "rules lawyers", this would be the ideal solution. Sadly, there is only 1 way to get a rule change implemented in this house without causing the entire game to be relegated to the "games that can't be played pending the change being signed by everyone in triplicate, sent in, sent back, queried, lost, found, subjected to public inquiry, lost again, and finally buried in soft peat for three months and recycled as firelighters" - and that is to find something "official" somewhere, duly referenced, and attach it to the game rules as an addendum.
      
Nygaard
Senior Member
Général de brigade

Pages Perso
Messages: 1014
Enregistré(e) en :
May 2006
Re:Issues with probability - example Patrol Cars Thu, 08 September 2011 21:15
I don't like the patrol car rules either - so I avoid scenarios that include them - or make scenario specific rules if I want to use the nice little minis... Smile
      
Fred the Obscure
Senior Member
Lieutenant

Pages Perso
Messages: 170
Enregistré(e) en :
April 2006
Re:Issues with probability - example Patrol Cars Fri, 09 September 2011 09:25
Greetings Tempest_Au,

Maybe you need to feed your rule lawyers to the Bugblatter Beast of Traal, that would solve most issues. Very Happy

Fred the Obscure,
Durban, South Africa
      
Tempest_Au
Junior Member
Major

Messages: 12
Enregistré(e) en :
September 2011
Re:Issues with probability - example Patrol Cars Fri, 09 September 2011 16:04
Fred the Obscure wrote:

Maybe you need to feed your rule lawyers to the Bugblatter Beast of Traal, that would solve most issues. Very Happy


If you are not going to contribute something at least semi-constructive to this thread Fred, I'll read you some of my poetry...

      
clexton27
Senior Member
Historien accompli

Pages Perso
Messages: 3089
Enregistré(e) en :
February 2007
Re:Issues with probability - example Patrol Cars Fri, 09 September 2011 17:20
Quote:

All is quiet in the Western Front, until 3 Panzers come crashing through the hedgerows immediately in front of an allied infantry unit, waving their 88mm canons and coaxial machine guns menacingly. The infantry curse, believing that their commander will send them in head-to-head against the giant steel beasts, but instead he climbs into his jeep. "Don't worry men, I'll take care of these" he says as he proceeds to demolish the tanks with his service revolver

This is nuts, really nuts. Surely there must be an error?

If the object was to make patrol cars "just a bit harder to kill than infantry", then the rule should be changed to "A Patrol Car unit is treated as Infantry when being targeted in battle. The attacker scores 1 hit for each Infantry dice symbol rolled against the Patrol Car unit. Any dice with a Grenade symbol are re-rolled, with an Infantry dice symbol or Grenade symbol indicating a hit, and all other symbols on the re-roll ignored."


Consider this, if that one patrol car attacks an armor unit he will, if adjacent, get to toss three dice, 2/6 chance of hitting with each of 3 dice. More than likely, he will not destroy a full strength armor unit. However, that armada of 3 panzers attacking back will get usually at a minimum if they position correctly, at least 9 dice and possibly up to 18 dice with overruns. I will take a chance of 9-18 dice versus 3 dice any day. So although the unit has a difficult probability of getting hit, it would not do well to be ganged up against in a no retreat situation. In one of the scenarios that came with the Patrol Cars (Jeeps) -RAID ON BARCE- they were not able to enter towns to hide and so moving attacking and running (rather than hiding) was the order of the day. This of course is the attempt to simulate the effects of a Raid (what the author intended).

I know many people don't always see the LOGIC in the game or unit rules, but personally, I am not much of a tweeker and just enjoy the game as it is. I know you will find a satisfactory way to enjoy the game as well or else you will find some other game to play. So I wish you well.

[Mis à jour le: Fri, 09 September 2011 17:21]

      
50th
Senior Member
Spécialiste des blindés

Pages Perso
Messages: 1412
Enregistré(e) en :
October 2006
Re:Issues with probability - example Patrol Cars Fri, 09 September 2011 19:29
I use the house rule that when you re-roll the dice after the hit, you destroy a figure on an infantry symbol and grenade, rather than just a grenade. That way they are easier to destroy than Tiger tanks.

      
Nygaard
Senior Member
Général de brigade

Pages Perso
Messages: 1014
Enregistré(e) en :
May 2006
Re:Issues with probability - example Patrol Cars Fri, 09 September 2011 23:22
50th wrote on Fri, 09 September 2011 19:29

I use the house rule that when you re-roll the dice after the hit, you destroy a figure on an infantry symbol and grenade, rather than just a grenade. That way they are easier to destroy than Tiger tanks.




That would actually make a lot more sense...
      
RBorg
Game Designer
Cadet

Pages Perso
Messages: 238
Enregistré(e) en :
December 2003
Re:Issues with probability - example Patrol Cars Sat, 10 September 2011 04:56
Show down
Take a tank unit with 3 figures, and pit it against a the Patrol Car unit with 1 figure and see which wins. I believe the odds are in the armor unit's favor, but I'm no stat man more of the against the odds guy Embarassed

Thanks for an interesting post.
Richard Borg
      
Tempest_Au
Junior Member
Major

Messages: 12
Enregistré(e) en :
September 2011
Re:Issues with probability - example Patrol Cars Sat, 10 September 2011 05:09
stevens wrote on Sat, 10 September 2011 01:20

Quote:

All is quiet in the Western Front, until 3 Panzers come crashing through the hedgerows immediately in front of an allied infantry unit, waving their 88mm canons and coaxial machine guns menacingly. The infantry curse, believing that their commander will send them in head-to-head against the giant steel beasts, but instead he climbs into his jeep. "Don't worry men, I'll take care of these" he says as he proceeds to demolish the tanks with his service revolver


Consider this, if that one patrol car attacks an armor unit he will, if adjacent, get to toss three dice, 2/6 chance of hitting with each of 3 dice. More than likely, he will not destroy a full strength armor unit. However, that armada of 3 panzers attacking back will get usually at a minimum if they position correctly, at least 9 dice and possibly up to 18 dice with overruns. I will take a chance of 9-18 dice versus 3 dice any day. So although the unit has a difficult probability of getting hit, it would not do well to be ganged up against in a no retreat situation. In one of the scenarios that came with the Patrol Cars (Jeeps) -RAID ON BARCE- they were not able to enter towns to hide and so moving attacking and running (rather than hiding) was the order of the day. This of course is the attempt to simulate the effects of a Raid (what the author intended).

I know many people don't always see the LOGIC in the game or unit rules, but personally, I am not much of a tweeker and just enjoy the game as it is. I know you will find a satisfactory way to enjoy the game as well or else you will find some other game to play. So I wish you well.


You are basing your argument on an incorrect premise. You are assuming that there are 3 full tank units, while I am expressing that there is 1 tank unit, with 3 tokens (little tank models). More than likely I have failed to express myself clearly, so I will try again.

For the purpose of this example, we will assume that all battles are occuring on flat terrain with unlimited line of sight (i.e. no bonuses/penalties to rolls). I will call this an "empty" environment for brevity.

We will also assume that combat starts with one of the units moving up to the other (i.e. close assault).

Back to my example, yes, the patrol car (a 1 token unit) gets to throw three dice - each has a 2 in 6 chance of scoring a hit against a tank. So, by the laws of probability, the most likely outcome is that one of the 3 tank tokens will die. There is also a 50% chance that the tank unit will be forced to retreat due to a flag being rolled, however as the patrol care does not have overrun capabilities, this won't affect the overall outcomes in an "empty" environment.

Now let's look at the tanks. They also get to roll 3 dice (irrespective as to whether there are 1, 2 or 3 tokens in the unit) - each with only a 1 in 12 chance of hitting (and of course destroying) the patrol car. By the laws of probability, there is only a 50% chance that the tank unit will kill the patrol car. Of course, the tanks also get a 50% chance of pushing the patrol car back with a flag, and hence obtaining a free overrun attack. Net result - 75% chance of actually hitting (and hence killing) the patrol car in an empty environment. (Of course, inevitably this will end up somewhat less due to terrain/line of sight issues occasionally preventing the second overrun from being able to be utilised against the patrol car)

So, while it is probable that the tank unit with 3 tokens will defeat a patrol car unit with 1 token, it still stands that the tanks will lose 1-2 tokens in the battle.
      
Tempest_Au
Junior Member
Major

Messages: 12
Enregistré(e) en :
September 2011
Re:Issues with probability - example Patrol Cars Sat, 10 September 2011 05:25
RBorg wrote on Sat, 10 September 2011 12:56

Show down
Take a tank unit with 3 figures, and pit it against a the Patrol Car unit with 1 figure and see which wins. I believe the odds are in the armor unit's favor, but I'm no stat man more of the against the odds guy Embarassed

Thanks for an interesting post.
Richard Borg


Hi Richard,

Yes, the tanks would usually win (in fact the odds are roughly 5-2 in their favour). The point I was trying to make was that the patrol cars are proportionately overpowered compared to other units (even if you look at the tokens/unit as being representative of "hit points" as opposed to separate individual units).

I don't have every single expansion / scenario for Memoir '44 yet (give me a couple of months, I actually love the game for it's combination of strategy and relative simplicity), and hence you would be in a better position to answer this than I, but if any scenario were to place more than 1 patrol car token in a unit (or require the destruction of more than one patrol car to earn a victory medal), they certainly would be excessively overpowered.

All that aside, please allow me to thank you for a wonderful game. Although my comments (now and in future) may appear at first glance to be "complaints", please understand that they should be viewed as observations/suggestions to consider for possible improvement. I'm a "crusty old fart" and tend to have difficulties expressing myself in a dispassionate manner.
      
Achtung Panzer
Senior Member
Leutnant

Pages Perso
Messages: 1081
Enregistré(e) en :
December 2007
Re:Issues with probability - example Patrol Cars Sat, 10 September 2011 10:33
50th wrote on Fri, 09 September 2011 18:29

I use the house rule that when you re-roll the dice after the hit, you destroy a figure on an infantry symbol and grenade, rather than just a grenade. That way they are easier to destroy than Tiger tanks.


Should they be? We're comparing a large but very tough easy-to-hit target with a very fast, agile thin-skinned vehicle. Apples and pears?
      
sam1812
Senior Member
Brigadier General

Pages Perso
Messages: 2292
Enregistré(e) en :
August 2006
Re:Issues with probability - example Patrol Cars Sat, 10 September 2011 15:04
Sandwiches and flies.
      
clexton27
Senior Member
Historien accompli

Pages Perso
Messages: 3089
Enregistré(e) en :
February 2007
Re:Issues with probability - example Patrol Cars Sat, 10 September 2011 18:31
Quote:

Sandwiches and flies.


Great example. I think I would rather eat the sandwich.
Laughing

[Mis à jour le: Sat, 10 September 2011 18:31]

      
Tempest_Au
Junior Member
Major

Messages: 12
Enregistré(e) en :
September 2011
Re:Issues with probability - example Patrol Cars Sun, 11 September 2011 04:26
Achtung Panzer wrote on Sat, 10 September 2011 18:33

50th wrote on Fri, 09 September 2011 18:29

I use the house rule that when you re-roll the dice after the hit, you destroy a figure on an infantry symbol and grenade, rather than just a grenade. That way they are easier to destroy than Tiger tanks.


Should they be? We're comparing a large but very tough easy-to-hit target with a very fast, agile thin-skinned vehicle. Apples and pears?



In some ways you are right, but all you have done effectively is prove my point.

I'll even use the Panzer IV Ausf A (the very first model) as an example. We have a tank that is armed with a Kampfwagenkanone 37 L/24 (KwK 37 L/24) 75 mm (2.95 in) main gun (which would probably use HE shells in this case), and 2 7.92 mm (0.31 in) MG 34 machine guns - one mounted in the turret, one in the front hull. At it's weakest point (the rear of the tank) the armor was still 10mm rolled plate steel - more than adequate to stop anything short of a 20mm AP shell (and even that would need to be a 90degree hit). I can't put my finger on the exact figures at the moment, but from memory the Ausf A had a turret rotation of aproximately 20 degrees a second.

Our Patrol car - the Willys MB/Ford GPW "Jeep", is unarmed, and unarmoured. In order to cause damage to a tank, one of the vehicle occupants will have to have some form of anti-tank weapon. The problem with this is that the ranged infantry AT weapons available at the time were not at all suitable for firing on the move - Recoilless rifles were simply too big and cumbersome to fire from a jeep, and a bazooka (not the most accurate weapon at the best of times) would be not only a difficult shot but also risk torching the jeep with the backblast. This leaves our heroic Jeep Assault Squad with the only feasible option - satchel charge/AT grenade lobbed from a rapidly moving vehicle at point-blank range.

Now, I can honestly say I've never tried to hit a moving jeep from a tank, but assuming the jeep didn't suddenly appear 100 meters away driving full tilt toward the rear of the tank, my money would be on the net result being a car full of dead GI's.

As for jeeps being targetted by infanry, again, I've never tried to shoot a moving car. I've regularly killed kangaroos and emus going full tilt at over 200 yards with my .308 from a crouch position, and they are a little smaller than a jeep...
      
RBorg
Game Designer
Cadet

Pages Perso
Messages: 238
Enregistré(e) en :
December 2003
Re:Issues with probability - example Patrol Cars Sun, 11 September 2011 04:41
I guess, that my love of the old TV show RAT PATROL has influenced patrol cars.

As I said interesting discussion. I look forward to more. Now if I could only read French, I bet in his Memoire 44 book Alexis Beuve would provide us with some great factual stats on the subject. Anyone have his outstanding work?

Richard Borg
      
Tempest_Au
Junior Member
Major

Messages: 12
Enregistré(e) en :
September 2011
Re:Issues with probability - example Patrol Cars Sun, 11 September 2011 09:24
RBorg wrote on Sun, 11 September 2011 12:41

I guess, that my love of the old TV show RAT PATROL has influenced patrol cars.


I love "Hogan's Heroes" too, even if it's not anything remotely like what life in a German POW camp was like. Razz

RBorg wrote on Sun, 11 September 2011 12:41

As I said interesting discussion. I look forward to more. Now if I could only read French, I bet in his Memoire 44 book Alexis Beuve would provide us with some great factual stats on the subject. Anyone have his outstanding work?


Sadly, I don't read French either, and I don't know of any recorded accounts of jeeps going up against tanks. This is possibly because anyone who tried it either died, or if they somehow survived were given a psychiatric discharge. Laughing

I've got no issue with Patrol Cars being set up in a manner that facilitates "hit and run" tactics (even if this isn't all that realistic either, but this isn't supposed to be a truly accurate simulation), possibly through the use of a "behind enemy lines" type of move. I just can't see why the damn things need to be tougher (or at least "significantly tougher") than infantry.

Infantry can go to ground, take cover behind trees, even use comparably small terrain imperfections like drainage ditches (etc) to very effectively make it exceedingly difficult to shoot them. The guy in the jeep has 2 options: 1) Find a large object (like a house) to take cover behind, or 2) Put his foot flat to the floor, weave around like a maniac, and pray that the enemy are idiots and don't think to lead the target.
      
50th
Senior Member
Spécialiste des blindés

Pages Perso
Messages: 1412
Enregistré(e) en :
October 2006
Re:Issues with probability - example Patrol Cars Sun, 11 September 2011 15:34
RBorg wrote on Sat, 10 September 2011 21:41

I guess, that my love of the old TV show RAT PATROL has influenced patrol cars.

As I said interesting discussion. I look forward to more. Now if I could only read French, I bet in his Memoire 44 book Alexis Beuve would provide us with some great factual stats on the subject. Anyone have his outstanding work?

Richard Borg



I also loved Rat Patrol! Being 55 now, I remember it well, plus I now have the first season on DVD. Yes, the Germans never could hit those jeeps.
Does anybody know how many seasons Rat Patrol lasted?


      
Tempest_Au
Junior Member
Major

Messages: 12
Enregistré(e) en :
September 2011
Re:Issues with probability - example Patrol Cars Sun, 11 September 2011 18:31
50th wrote on Sun, 11 September 2011 23:34

I also loved Rat Patrol! Being 55 now, I remember it well, plus I now have the first season on DVD. Yes, the Germans never could hit those jeeps.
Does anybody know how many seasons Rat Patrol lasted?



2 seasons were made and shown in the US (to the best of my knowledge anyway).

A few episodes were shown in Australia and Britain, but it was quickly cancelled 'cause we all got annoyed with the fact that "Yanks" were implying they had anything to do with us "Rats" defeating Rommel (Yes, I am aware that the Yanks did join in for the final push to kick Rommel out of Africa, but prior to that it was all ANZACs and Brittish), the complete disregard for historical accuracy, and a non-Australian wearing a slouch hat (think of it as being as almost as insulting to us Aussies as someone using the "stars and stripes" as toilet paper would be to an American).

[Mis à jour le: Wed, 14 September 2011 03:18]

      
JJAZ
Senior Member
Lieutenant Colonel

Pages Perso
Messages: 623
Enregistré(e) en :
May 2008
Re:Issues with probability - example Patrol Cars Mon, 12 September 2011 17:35
RBorg wrote on Sun, 11 September 2011 04:41

I guess, that my love of the old TV show RAT PATROL has influenced patrol cars.

As I said interesting discussion. I look forward to more. Now if I could only read French, I bet in his Memoire 44 book Alexis Beuve would provide us with some great factual stats on the subject. Anyone have his outstanding work?

Richard Borg



If the odds would be facts, we all would win the lottery.
If all the facts would be true, then their would be no lies.
If there would be no lies, there would be no lotterys.
      
Tempest_Au
Junior Member
Major

Messages: 12
Enregistré(e) en :
September 2011
Re:Issues with probability - example Patrol Cars Tue, 13 September 2011 02:44
JJAZ wrote on Tue, 13 September 2011 01:35


If the odds would be facts, we all would win the lottery.
If all the facts would be true, then their would be no lies.
If there would be no lies, there would be no lotterys.



This is confusing.

Statistical odds are factual (presuming the mathmatics is correct). Random chance is random only in small samples.

A fact, by definition, is always true - otherwise it wouldn't be a fact. People may misrepresent data as fact when it in fact is not a fact, which can happen quite often. To explain:
Fact: Gravity on Earth exhibits an attractive force equal to 9.8 Meters per Second at sea level (accurate to 1 decimal place).
Fact: The front glasis plate of a Panzer IV will prevent penetration by small arms fire.
Misrepresented fact based on inaccurate assumptions: The Jeep was so fast and agile that it was harder to kill than a Panzer IV.
Absolute bulldung / lie: I know everything there is to know about WWII combat, and am the only authority that should be listened to.

And as for lotteries, I have a policy of telling the truth at all times - and I still put the occasional ticket into the lottery (whilst knowing full well that my chances of winning the jackpot are less than 1 in a million).

[Mis à jour le: Wed, 14 September 2011 03:17]

      
rasmussen81
DoW Content Provider
Designer's Oath

Pages Perso
Messages: 7161
Enregistré(e) en :
July 2007
Re:Issues with probability - example Patrol Cars Tue, 13 September 2011 04:00
Tempest_Au wrote on Tue, 13 September 2011 04:44


A fact, by definition, is always true - otherwise it wouldn't be a fact. People may misrepresent data as fact when it in fact is not a fact, which can happen quite often. To explain:
Fact: Gravity on Earth exhibits an attractive force equal to 9.8 Meters per Second at sea level (accurate to 1 decimal place).
Fact: The front glasis plate of a Panzer IV will prevent penetration by small arms fire.
Misrepresented fact based on inaccurate assumptions: The Jeep was so fast and agile that it was harder to kill than a Panzer IV.
Absolute bull / lie: I know everything there is to know about WWII combat, and am the only authority that should be listened to.

And as for lotteries, I have a policy of telling the truth at all times - and I still put the occasional ticket into the lottery (whilst knowing full well that my chances of winning the jackpot are less than 1 in a million).


Like Richard, I think this is an interesting discussion. I would respectfully request, however, that you avoid any kind of swearing on the forums to keep it completely family-friendly. I have students who sometimes read posts and I'm always happy to report to parents that this is a clean environment for the kids to be on. Let's keep it that way! Smile

Keep the good times rolling!! Very Happy

[Mis à jour le: Wed, 14 September 2011 03:51]

      
Tempest_Au
Junior Member
Major

Messages: 12
Enregistré(e) en :
September 2011
Re:Issues with probability - example Patrol Cars Wed, 14 September 2011 03:16
rasmussen81 wrote on Tue, 13 September 2011 12:00

I would respectfully request, however, that you avoid any kind of swearing on the forums to keep it completely family-friendly. I have students who sometimes read posts and I'm always happy to report to parents that this is a clean environment for the kids to be on. Let's keep it that way! Smile

Keep the good times rolling!! Very Happy


Sorry mate Embarassed

I'll try and remember that this forum is multi-national. While I'd never drop "F Bombs" etc, I sometimes forget that words like "Bulldung" are deemed by some to be swearing. Rolling Eyes

I'll edit out the offensive word in my post, if you could do the same with the quote ('cause I can't edit it from your post) Razz

P.S. I've taken the liberty of editing a couple of "questionable" words in earlier posts, but there are some "borderline" words common in the Australian idiom that I am completely unsure of. Please feel free to act as my "moral compass" and let me know if I cross the line again.

Tact and "Political Correctness" were never my strong suits. Sad

[Mis à jour le: Wed, 14 September 2011 03:27]

      
Tempest_Au
Junior Member
Major

Messages: 12
Enregistré(e) en :
September 2011
Re:Issues with probability - example Patrol Cars Wed, 14 September 2011 03:24
Le message n'a pas de contenu

[Mis à jour le: Wed, 14 September 2011 03:25]

      
rasmussen81
DoW Content Provider
Designer's Oath

Pages Perso
Messages: 7161
Enregistré(e) en :
July 2007
Re:Issues with probability - example Patrol Cars Wed, 14 September 2011 03:55
No worries about the other posts...and thanks for your understanding! One of the wonderful things about this community (although there are many wonderful things) is that everyone is so nice and everyone tries to be so polite. You certainly fit right in here in that regard and I look forward to more of your interesting comments and ideas!!

Thanks again and have a great day. Cool
      
red_zebra
Senior Member
Lieutenant

Pages Perso
Messages: 327
Enregistré(e) en :
February 2005
Re:Issues with probability - example Patrol Cars Wed, 14 September 2011 20:32
Patrol cars where not only Jeeps, there where plenty armoured cars also.
      
g1ul10
Junior Member
Maggiore

Pages Perso
Messages: 27
Enregistré(e) en :
April 2013
Re:Issues with probability - example Patrol Cars Sat, 18 January 2014 22:35
In the past days I started to explore the scenarios that I could possibly play with my just bought EP expansion. One possibility was to play with Jeeps but when I read the rules (Troop - 19 Long Range Patrol Cars) I was a bit puzzled. Searching the forum I found this tread. Since the discussion here (and also elsewhere) focused on the "relative" difficulty of hitting patrol cars I tried a simulation experiment. The point of the experiment is to discover what is the probability of a Patrol Cars (PC) unit to survive the confrontation with an enemy unit.

Patrol Cars vs. regular Tanks

First, consider the case in which a Patrol Cars unit confronts a tank unit in a plain (no obstacles) battlefield. One possible situation is depicted below

https://lh4.googleusercontent.com/-Jv97gRMDHGI/UtFhdSpxzBI/AAAAAAAAEnQ/KV7-nMPnrrI/s288/IMG_20140111_155642.jpg

I assumed that both units, at each round, attack each other in Close Combat. Given their mobility this is always possible, irrespective of the number of flags previously rolled. If the objective is to destroy the enemy before the enemy destroys you, the attack in Close Combat makes sense. Indeed in this way Patrol cars attack with 3 dice and Tanks gain the possibility of an Armor Overrun. I summarize below the results, which of course depend on the number of figures composing the units and on which side moves first.

Number of       Number of     First          % Victory of
PC figs         Tank Figs     Moving Unit    Patrol Cars

1                   3         Patrol Cars      51%
2                   3         Patrol Cars      81%
3                   3         Patrol Cars      94%
1                   3         Tanks            36%
2                   3         Tanks            70%
3                   3         Tanks            89%


Patrol Cars vs. Elite Tanks

Now consider the case in which the Tank unit opposing the Patrol Cars is an elite unit with four figures. A possible situation is depicted below

https://lh4.googleusercontent.com/-Z1imQ2WeHJM/UtFhd6yDagI/AAAAAAAAEmk/sHUEIFelN6c/s288/IMG_20140111_155407.jpg

The rule of engagement are as before: both units always attack in Close Combat. In this case the probability of victory of the Patrol Cars unit are as follows

Number of       Number of     First          % Victory of
PC figs         Tank figs     Moving Unit    Patrol Cars

1                   4         Patrol Cars      36%
2                   4         Patrol Cars      70%
3                   4         Patrol Cars      89%
1                   4         Tanks            27%
2                   4         Tanks            59%
3                   4         Tanks            82%


Conclusion and Outlook

My experiment confirms the initial estimate by Tempest_Au. A Patrol Cars unit with three figures is extremely effective against armored opponents. It can easily dispose off an Elite Tanks German unit. Also with one single figure, the PC Unit has a daunting offensive potential: I would avoid to leave a Tanks unit in its range without adequate support. What is still more impressive is the defensive capability of PCs. If you are considering to send your Elite Panthers unit to destroy that single isolated Jeep, think twice! You have a one in four probability that your Panthers are destroyed instead. Moreover, note that the presence of obstacles, like villages or woods, will give a definite advantage to the PC unit. I already started to explore how PC units behave against infantry. It seems they are also very effective in dislodging Panzergrenadier units from well defended positions (read village+sandbags).

Technical details

The percentage were computed simulating 10.000 independent fights with GNU Octave, an open source program for scientific computing (https://www.gnu.org/software/octave/). Source code is available upon request.

      
JFKoski
Senior Member
Historien accompli

Pages Perso
Messages: 603
Enregistré(e) en :
October 2005
Re:Issues with probability - example Patrol Cars Sun, 19 January 2014 03:46
I don't know of any scenario that uses more than 1 figure for a Patrol Car unit.

For those interested I posted my 2nd odds table in 2012 under my user pages to include Patrol Cars (and Tiger Tanks).

I kept my spreadsheet where I computed each die roll, but didn't go up to 6 dice (two 3-dice attacks against a patrol car). With 3 dice I got a 21% chance of a kill against a lone-figure patrol car. Extrapolating into two attacks, .21 + .21 * (1-.21) = .37 or 37%.

In Oath of Kufra there is one Italian Patrol Car. One PC is a nuissance. Will the Italians advance hoping it's invulnerable or keep it in reserve? Will the French shoot it in hopes of a kill that will neutralize the Italians' mobility? Or will they pursue a different strategy?

In Rommel's Right Hook (classified), the Germans have 2 patrol cars. Working together they could get behind a unit and block it's retreat for higher chance of hits. But it's a desert tank battle, so the tanks will more than likely go at each other first.

In Battle of Nezuet Ghirba (classified), the British have 3 Patrol Cars. These can run up and force the sandbagged artillery to close assault them, or go for the supply trucks or infantry.

In Raid on Barce (classified), all 6 British units are Patrol Cars. So there's no alternative but to attack them and hope for 4 kills. If the Italians miss in close, they get a free Battle Back shot! Medics is a dud for British, and we're waiting on a ruling whether Dig-in and BEL could order one unit to move and battle normally (as they could a tank or artillery in Knightsbridge since there's no infantry there) or whether those cards are duds too.

So far, I don't see a Patrol Car vs Patrol Car scenario.

If I had a knack for scenario design, I'd do one with 2 patrol cars in the same section, so they could do the runaround and block the retreat of infantry in the open.

[Mis à jour le: Sun, 19 January 2014 03:51]

      
Quit2
Senior Member
Historien accompli

Pages Perso
Messages: 1201
Enregistré(e) en :
July 2007
Re:Issues with probability - example Patrol Cars Sun, 19 January 2014 13:22
A 3 figure tank unit does not represent a real unit with 3 tanks. Just as a 4 figure infantry unit doesn't represent a unit of only 4 men. I think very few infantry units were made up of only 4 men.

As such, a one figure Patrol car is not just one jeep.

The number of figures in a unit represents how much damage that type of unit could take until it lost its combat effectiveness. The average hit rate reflects the easiness with which one could inflict damage to the unit that would result in getting closer to combat ineffectiveness of the unit.

That being said, I don't know if it should be more difficult to inflict critical damage to a jeep unit than to a tank unit.
A few ways one could explain this:
- If you damage 2-3 tanks of a unit on a road through dense woods, they would serve as road blocks limiting the combat effectiveness of the other tanks in the unit, whereas a jeep would still be able to manoeuvre in between broken jeeps on a road.
- To kill the effectiveness of a jeep, one should either break the jeep or kill all it's passengers, since just shooting the driver would have them replace the driver by one of the passengers. And with several infantrymen in a jeep shooting, taking out a few of those guns does not eliminate the ability to strike. Shooting 3 men in a jeep with 6 men still allows that jeep unit to do a lot of harm. With a tank, shooting the officer in the turret would already decrease the effectiveness of the tank.

Still, while this explains a bit, it doesn't seem to be satisfactory.

You could always use an house rule with those figures. I have once used them like this:
Both half track minis and jeep minis were used as a SWA: one figure to be added to an infantry unit:
- Half track: would serve as a permanent sand bag that moves along with the unit: it may ignore one flag, and if there is no terrain battle dice reduction, the unit is attacked with one less die.
- Patrol car: The unit may move up to 3 hexes and battle and move one more hex after all battle of that turn.
      
50th
Senior Member
Spécialiste des blindés

Pages Perso
Messages: 1412
Enregistré(e) en :
October 2006
Re:Issues with probability - example Patrol Cars Sun, 19 January 2014 15:43
Tempest_Au wrote on Thu, 08 September 2011 13:47

stevens wrote

I think the basic idea was to recreate the "hit and run" effect of a fast patrol car attack, not exactly replicate the effect of using a jeep versus an infantry or armor.


Strange that the designers would choose to create an "ultra jeep" to accomplish this, when they could just as easily chosen to allow Patrol Car units to perform a second move immediately after attacking (in the same manner as a "behind enemy lines" card).

stevens wrote

This game is not meant to be an truly accurate representation of battlefield forces.


True, but all the same the game doesn't have cyborg laser-troops either - because such a unit would be completely unrealistic.

stevens wrote

You can use the patrol car units as intended or create your own "house rules" on them. Many in the forum have done so with many of the units and have their own "house rules" for many different types of units. Use the SEARCH to do the topic "house rules" and I am sure you will find a plethora of tried and true rules.


ROFL. If I didn't live in a house full of "rules lawyers", this would be the ideal solution. Sadly, there is only 1 way to get a rule change implemented in this house without causing the entire game to be relegated to the "games that can't be played pending the change being signed by everyone in triplicate, sent in, sent back, queried, lost, found, subjected to public inquiry, lost again, and finally buried in soft peat for three months and recycled as firelighters" - and that is to find something "official" somewhere, duly referenced, and attach it to the game rules as an addendum.


This is exactly why I have a house rule that when you re-roll hits, you kill a PC on a grenade or an infantry symbol.
      
Sgt Storm
Senior Member
Lieutenant

Pages Perso
Messages: 904
Enregistré(e) en :
December 2006
Re:Issues with probability - example Patrol Cars Mon, 20 January 2014 07:20
Quit2 makes the most important point in this discussion. The patrol car single figure unit is an abstraction. It represents a mobile infantry force that is 50% more resilient in battle than a normal infantry unit. I think of it as a mechanized infantry unit with specialized troops, commandos for example.

This seems perfectly acceptable to me.

Now, as regards the Rat Patrol TV show. I have fond memories of watching that show as a child. However, I recently saw a few episodes and was sorely disappointed by what I thought was very poor acting, stock characters, and stale story lines.

On the other hand, the show Combat! I remember being excellent and I've seen a few recently and they were pretty good. But they don't have patrol cars, so not terribly relevant to this thread.
      
Sgt Storm
Senior Member
Lieutenant

Pages Perso
Messages: 904
Enregistré(e) en :
December 2006
Re:Issues with probability - example Patrol Cars Mon, 20 January 2014 07:28
One other point: basing the comparison of PC and Armor on the odds of winning a close assault attack does not take into account one of the most powerful aspects of Armor unit's attack. The ability to move 3 spaces and fire 3 die from a distance makes a difference.

Statistics and theory are great until you actually play a number of scenarios. I played all the official scenarios with PCs and don't recall them being overly powerful in practice.
      
g1ul10
Junior Member
Maggiore

Pages Perso
Messages: 27
Enregistré(e) en :
April 2013
Re:Issues with probability - example Patrol Cars Mon, 20 January 2014 23:02
Sgt Storm wrote on Mon, 20 January 2014 07:28

One other point: basing the comparison of PC and Armor on the odds of winning a close assault attack does not take into account one of the most powerful aspects of Armor unit's attack. The ability to move 3 spaces and fire 3 die from a distance makes a difference.

Statistics and theory are great until you actually play a number of scenarios. I played all the official scenarios with PCs and don't recall them being overly powerful in practice.


You surely know better then me. In order to explain the logic of my far from perfect experiment, let me notice that the ability to move three spaces and fire 3 die, while extremely effective against slow units like infantry, is not so much of a deal against units capable of moving FOUR spaces and fire. There is virtually no way an Armor unit can avoid to be engaged in Close Combat (CC) by a "mechanized infantry unit with specialized troops", if the latter so desires. Apart of course taking cover behind allied units. Moreover if the Tanks decide to engage the PC from afar, their survival probability worsen because they give up the ability to perform an Armor Overrun in their first turn. And in any case, in the next turn they have to sustain a CC attack by the PC. But as you rightly point out, all these considerations are just "theory" which is likely to evaporate fast in the heat of the battle Wink G.
      
Sgt Storm
Senior Member
Lieutenant

Pages Perso
Messages: 904
Enregistré(e) en :
December 2006
Re:Issues with probability - example Patrol Cars Tue, 21 January 2014 07:45
g1ul10 wrote on Mon, 20 January 2014 17:02

Sgt Storm wrote on Mon, 20 January 2014 07:28

One other point: basing the comparison of PC and Armor on the odds of winning a close assault attack does not take into account one of the most powerful aspects of Armor unit's attack. The ability to move 3 spaces and fire 3 die from a distance makes a difference.

Statistics and theory are great until you actually play a number of scenarios. I played all the official scenarios with PCs and don't recall them being overly powerful in practice.


You surely know better then me. In order to explain the logic of my far from perfect experiment, let me notice that the ability to move three spaces and fire 3 die, while extremely effective against slow units like infantry, is not so much of a deal against units capable of moving FOUR spaces and fire. There is virtually no way an Armor unit can avoid to be engaged in Close Combat (CC) by a "mechanized infantry unit with specialized troops", if the latter so desires. Apart of course taking cover behind allied units. Moreover if the Tanks decide to engage the PC from afar, their survival probability worsen because they give up the ability to perform an Armor Overrun in their first turn. And in any case, in the next turn they have to sustain a CC attack by the PC. But as you rightly point out, all these considerations are just "theory" which is likely to evaporate fast in the heat of the battle Wink G.


Very good points. I agree if you look at these units logically the implementation may be wanting. But on the other hand, the official scenarios that employ them didn't strike me as odd, so maybe its how you design scenarios with them.

At times I think of these special units, such as Patrol Cars, Tiger tanks, Half-Track, Trucks etc. as promotional items and not "serious" units. As such I don't give them all that much consideration and there really aren't that many official scenarios with them, so I'm not much bothered by these supposed shortcomings. I think they all could be implemented differently and maybe they would be better, maybe not.

But I've learned over the years its not worth thinking too deeply about it. Best to just enjoy what I have. The game's not perfect for sure, and I've griped about these types of things in the past, but its still at the very top of my list of light war games for sure.
      
g1ul10
Junior Member
Maggiore

Pages Perso
Messages: 27
Enregistré(e) en :
April 2013
Re:Issues with probability - example Patrol Cars Mon, 27 January 2014 01:21
I would like to conclude my little simulation experiment about Patrol Cars (PC) units investigating their effectiveness against infantry. Of course an infantry unit caught in the open is no match for a PC unit so I will consider a more elaborate situation.

Patrol Cars vs. Entrenched Elite Infantry

The situation is represented in the picture below

https://lh5.googleusercontent.com/-5vmpBJ6KP-4/UtFhcQHuTjI/AAAAAAAAEmU/YFs_iYZ36Uk/s288/IMG_20140111_155556.jpg

I assumed both units always attack each other in Close Combat. The PC unit attacks first. Its goal is to dislodge and eliminate the elite infantry in the village. I did the experiment with elite infantry because I wanted to ignore the possibility that, in following rounds, the infantry unit were unable to engage the PC in close combat. The fact that the Elite unit can move two spaces and fire implies that the possibility above can be neglected, for any practical purpose.

I summarize below the results, which of course depend on the number of figures composing the PC units.

Number of   % Victory of
PC figs      Patrol Cars
1             46%
2             80%
3             94%


Summary and Conclusion

PC units with more than one figure not only posses dreadful anti-tank capabilities, they are also more effective that a combat engineers unit in attacking prepared defensive positions. Even the single figure unit has almost one in two probability of conquering the village and eliminating opposing resistance. Given that the PC unit, because of its high mobility, can avoid a round of enemy fire before engaging the infantry in CC, I have the impression that, in this specific mission, the single figure PC unit is still better then an engineers unit, at least if appropriate cover along the approaching pathways is not available.

By way of conclusion, let me express a general caveat about these exercises. Their purpose is just to illustrate, mainly to me but also to all my interested friends, how this type of unit, which I never played before, compare with more standard pieces of the game. While I tend to agree with those who pointed out the extravagance of its uncommon resilience to enemy fire, I do not want to imply that it represents, by any mean, a flaw in the game. I think everybody will agree that, in a game, there are no right nor wrong rules, but just rules that are more or less effective in achieving a goal. It was already clarified that this unit and its rules were designed to add a certain degree of "cinematic" flavour to the game. I read my experiments to confirm that this design goal was achieved. On the other hand, my experiments suggest that scenario designers who need a set of rules able to convey the tactical flavour of the use of patrol cars (i.e. mechanized infantry units with light weapons used for patrolling, exploration and reconnaissance) in historical WWII situations should look elsewhere. In fact, taking inspiration form Sgt Storm post, I would say that official rules for PC are more reminiscent of an assault unit, a sort of "fast-moving armored elite combat engineers commandos with man portable anti-tank weapons", whose mere existence in the military doctrine of the beginning of the past century I would dare to consider unlikely (but whose role in modern warfare cannot be overstated). Needless to say (but you never known in forums) this is just my very humble opinion and it has nothing to do with the truth Wink

Technical details

See my previous post in this thread.

[Mis à jour le: Thu, 30 January 2014 22:28]

      
rasmussen81
DoW Content Provider
Designer's Oath

Pages Perso
Messages: 7161
Enregistré(e) en :
July 2007
Re:Issues with probability - example Patrol Cars Mon, 27 January 2014 03:31
Interesting numbers. I would point out (again, since someone else already said this) that there are no Official Scenarios where the PC units have more than one figure...
      
Sgt Storm
Senior Member
Lieutenant

Pages Perso
Messages: 904
Enregistré(e) en :
December 2006
Re:Issues with probability - example Patrol Cars Mon, 27 January 2014 03:48
Also, are your numbers reversed because its not clear why a 3 PC unit would have lower chance of victory than a single PC unit. Though maybe I don't understand your table.

I say this only to alleviate any possible confusion, not to prolong this thread. Also, its clear to me that PC units should in practice have a single figure like the Tigers.
      
Jeronimon
Senior Member
Brigadier

Pages Perso
Messages: 1121
Enregistré(e) en :
November 2007
Re:Issues with probability - example Patrol Cars Mon, 27 January 2014 10:18
I have no qualms prolonging this thread. Wink

I also spotted what Sgt Storm pointed out, the percentages make no sense this way:

g1ul10 wrote on Mon, 27 January 2014 01:21


Number of   % Victory of
PC figs      Patrol Cars

1             94%
2             80%
3             46%




Should we simply reverse the figures up down, or is this the percentage the winning percentage for the infantry in the town?


Adding to that: we played THE BASTOGNE CORRIDOR - BREAKLORD! by Burdie Smith recently and the tow patrol car unit was destroyed pretty quickly. (I know, that is just one example (n=1) Smile , but it is an example of an actual battle where a two patrol car unit did us less then we would have expected from the stats here. Smile )

[Mis à jour le: Mon, 27 January 2014 10:23]

      
g1ul10
Junior Member
Maggiore

Pages Perso
Messages: 27
Enregistré(e) en :
April 2013
Re:Issues with probability - example Patrol Cars Thu, 30 January 2014 22:36
Yes of course! Numbers should be reversed. I edited the post. Sorry for the mistake and thank you for pointing it out.
      
    
Sujet précédent:Air Check - Infantry/Armor on a Wagon
Sujet suivant:Strategy Guide
Aller au forum: