Rarity

[Sergei]

This is an old gripe, but here I go anyway (figures made up):

Suppose there are 10 thousand T-26's and only 2,500 T-34's, but these T-26's are split equally between five submodels (2,000 each). This makes T-26 five times more common than T-34, but in CM T-34 has lower rarity than T-26 because rarity is calculated per submodel.

This causes that the rarity value has nothing to do with the actual rarity or commonness of a vehicle.

[Le Tondu]

Easy solution: Turn rarity off.

No more problem. In the end, it is your gaming experience that matters.

[Sergei]

Yeah, but then there's no rarity. Which isn't a bad thing, you get more varied quickies by not limiting yourself just to rarity. But what I was wondering is how to correct this problem.

Maybe some day there will be a "crappiness" pricing option: all prices are raised exponentially so that a 50 point tank would be 2500 and a 200 point tank 40000 points, or 25 and 400 to remove unnecessary 0's. This way you could buy 16 pieces of **** instead of an Überpanzer as earlier you could only buy 4 POS's. It might make for some more interesting battles, dont'cha think?

[JasonC]

I've suggested in the past rating items of a similar nature and having rariety work by what portion of the distribution that item comes from. I think this would solve the problem you've noticed. While a few of the judgments involved are subjective, the main outlines certainly are not.

What do I mean by grouping and distributions? Well, consider long 75mm StuGs in late 1942. As the most common production item in the assault gun category at that date was indeed a long 75mm StuG, these get low rariety. But actually they represent the top couple of percent of German AFVs at that date, by gun and armor ability. The low number implies CM players can have one in every late 1942 battle. As a fact, defenders against the Stalingrad counterattack e.g. faced T-34s in Pz 38s. Quality as good as a long StuGs was rare - various lesser types were common.

So what one does is rank the available AFV types, the available arty types, the available towed gun types - at any given date. There were 3500 German AFVs, to this date 45,000 Russian towed guns had been produced, there were n Russian artillery regiments, divisional or independent. OK. The median quality AFV, gun, or regiment had equipment X. X or worse are low rariety, aka "common", aka just pay what it is worth in combat terms and it is yours.

But only the top quarter had equipment as good as Y. So Y is 40% rariety. Only a tenth had equipment as good as Z. So Z is 80% rariety. The cherry picker's special that only 3% of the actual force could hope to see costs 150% rariety.

I think this would do what QB players actually want a rariety system to do - which is encourage use of vanilla equipment and discourage cherry picking. Right now the rariety system does not do that. The tiny premiums for thick StuGs soon after introduction, or Tigers by 1943, or T-34s and KVs in late 1941, just don't remotely balance the enourmously greater ability the highest quality armor confers. As a result, people tend to conduct fights using a given item a year earlier than the average actual participant saw it.

[Other Means]

as usual JasonC sums it all up.

this is IMHO exactly what rarity *should* do. it should give you a force mix representative of the average unit on the ostfront. yes, it should be possible to get all the top of the range equipment in scenarios etc but you should get what they had for QB's.

also, if the AI made sensible force choices i think a lot more people would leave it to choose them.

maybe a bugger to impliment but i think if you've set up a QB of, say, area south, 02/03, it should give a force mix representative of what would be there. i know it's a hellish amount of data to collect but it would be fun.

[Vanir Ausf B]

Hopefull BFC at some point will abandon the 'rarity by price manipulation' model they are using and go to a binary system where units are either available for purchase or not. It probably won't happen for CMAK, but I have hope for CMX2.

[JasonC]

Here is an example of the sort of figures you'd expect to see with my "position in a distribution" rarity system. I did the German AFVs available at the start of the war, as an example.

sIG-33 - +150

Pz IV F - +100

Pz IV E - +80

Pz III H - +65

StuG B - +50

Pz Jgr I - +40

Flamm II - +40

Pz IV D - +30

Pz IV C - +30

Pz III G - +20

Pz 38 E - +10

Pz III F - +0

Pz 38 A - +0

Pz 35 - +0

Pz II F - +0

Pz II C - +0

Pz I - +0

Somua - +30

Hotchkiss - +20

Renault - +20

Right now, the only effect of the existing rarity numbers is to increase the cost of sIG, Flamm II, and Pz Jgr I types. Modestly, by enough to force you to take only 1-2 if you take any. The best types available, the III H and IV E, which represent the top 10% of German AFVs, are presently available at +10 rarity, which has no effect on unit choices. In QBs, there is no reason anyone would take a IV C or a III G, let alone a Pz-38 A.

With my distribution position revisions, there is a stiff point penalty for cherry picking the latest, thickest types. You can take full sized German models with good armament for modest +20-30 rarity - but without the thick armor. You can take the improved 38 E, with decent armor but a so so gun, for +10. The types with 37mm and 20mm main armament and thin armor are all available at rated strength cost - the whole bottom half of the type distribution. The non-standard, captured French types are rarer.