Computer experience bonus vs force size handicap

[Robert Olesen]

In ROQC I have been using the force size handicap to model difficulty levels for the player. I'd like to make use of the computer experience bonus as well, but I'm sure as to how these two compare with each other. I haven't used the computer experience bonus myself and I don't have the time to play a lot of battles to check it out.

I'm hoping that someone here has accumulated some experience on this and is willing to share it.

For example, is +100% equivalent to +50% and experience +2, +75%/+1, +50%/+1 or something else?

I'd love to have a rule of thumb, e.g. +25% matches +1 experience bonus - but that may not be possible.

[Joachim]

Originally posted by Robert Olesen:

In ROQC I have been using the force size handicap to model difficulty levels for the player. I'd like to make use of the computer experience bonus as well, but I'm sure as to how these two compare with each other. I haven't used the computer experience bonus myself and I don't have the time to play a lot of battles to check it out.

I'm hoping that someone here has accumulated some experience on this and is willing to share it.

For example, is +100% equivalent to +50% and experience +2, +75%/+1, +50%/+1 or something else?

I'd love to have a rule of thumb, e.g. +25% matches +1 experience bonus - but that may not be possible.

Hi Robert,

IMHO all unit prices increase about the same percentage when going up from reg to vet - but rounding makes up for a huge error when calculating that factor. So you need to look up the price of a very expensive unit (KT? Huge arty? Turn rarity off to get rid off another potential error source) - the rounding error is smallest there.

Notice the cost as conscript, green, reg, vet, crack and elite.

Look at the percentage increase (or the factors green/conscript, reg/conscript, vet/reg etc.).

If the factor is the same regardless of experience, you have got your rule of thumb - that factor gives away the percentage for going up one exp level.

Gut says the factor is not the same for every exp level, but it does not differ very much - at least until veteran to crack.

Even if it does differ very much, you could still get your rule of thumb but it has to include the "base" exp of the AI force.

Hope tht helps

(Don't have the game available right now...)

Gruß

Joachim

[atiff]

Close, but not quite....

What Rob means is, if you have a +100% battle, how does that compare to a battle with a lower +X% and an AI Experience handicap (ie, +1, +2 or +3, defined on the last screen before going into the battle)

[Robert Olesen]

Scarheard, that's an obvious way of loking at it that I haven't thought of. I'll do a few computations.

But there's also the aspect of player perspective, as atiff points out.

For example, if you were up against a +100% enemy force, and you were to trade that with +50% and +x computer experience bonus, which value of x would you feel gives you the same challenge?

The point value does not necessarily reflect the difficulty. For example, weakened troops cost 90% of fit troops, and unfit costs 80%. That does not in my experience reflect the handicap of being unfit or weakened - the cost reduction should be higher to accurately relect the handicap, at least when you're attacking.

[ September 23, 2003, 06:35 AM: Message edited by: Robert Olesen ]

[Joachim]

[Edited 'cause this post is much better]

Ok, now I know what you mean.

The question I was answering is:

What is the point value of a given force if it has a +1 exp modifier.

The answer to that is difficult. Consider the following prices (no rarity, conscript to elite) ) of a

PzIIIj__________________85____92___100___116___133 ___150

214mm Rocket spotter___502___539___574___659___748 ___840

As you can see, there is no exact rate at which the prices are increasing. The tank goes up 50% from regular to elite. 574*150% = 861 <> 840!

As a guestimate for the factors I'd say

8% from conscript to green,

8% from green to reguler

15% from regular to vet

14% from vet to crack and

13% from crack to elite.

If you pay 1000 pts for an all regular force, you'd have to spend 1150 pts for the same as veteran.

But a +1 exp bonus for a 1000 pts green force is worth a mere 80 points!

The question you are asking is:

Does a 10000 pts regular force with a +1 exp bonus (which is - according to my calcs a 1150 pts vet force) fight as if it is worth 11500 points or is there some additional effect.

Consider a force of 100 regular PzIIIj tanks at a cost of 100pts each. With a +1 exp bonus, the same force would cost 11600 points (ignore the 10 pts rounding error...) if bought as vet without a bonus.

Do those 100 veteran PzIIIj figth as well as a force of 116 regulars, which you would get with a 16% pts bonus (lets assume for simplicity a 16% bonus is in the game).

Or do the rather fight like 125 regulars, IE a 25% points bonus.

If they fight like 125 regulars, they should cost as much as 125 regulars! IE the price increase from reg to vet should be 25% instead of 16%

With this example, you can see that your question (as I understand it) is the same as "is the price increase of a unit worth what you get for". My guess is that BFC carefully selected the price increases, so roughly it should be ok.

...and this discussion leads directly to the quality vs quantity thread.

Albeit with a small twist:

The question is not if you prefer to fight with quality or quantity, but if you prefer to have a quantity AI or a quality AI.

And I say I'd need to know terrain, mission and my own quality and force type to answer that..

Gruß

Joachim

[ September 23, 2003, 12:03 PM: Message edited by: Scarhead ]

[Robert Olesen]

OK, here are some numbers based on a sample of units (four German, four Soviet, June 42, different purchase categories)

Each experience level adds ca 14% to the cost (that's probably programmed to be 15). This gives the following comparisons, approcimately:

Conscript: 100

Green: 114

Regular: 128

Veteran: 149

Crack: 168

Elite: 190

These are approximations, and the cost increase seems to be less for artillery than for infantry.

So, the rule of thumb is:

- Going up one level corresponds to ca. 15% cost increase.

- Going up two levels correspond to ca. 30% cost increase.

- Going up three levels correspond to ca. 50% cost increase.

This applies to all level jumps, not just from Conscript.

Now bear in mind that a change from +50% to +100% handicap is a 33% relative increase in the total enemy force size. So, according to the data above +100% should be equivalent to +50% with +2 computer experience bonus. Do you think that is a fair assessment of the actual difficulty difference? Would you rather face 2000 points of Conscripts than 1500 of Regular troops? Or 2000 points of Regular troops instead of 1500 points of Crack troops?

Similar calculations can ofc. be made. From +25% to +50% is a 20% relative increase, indicating +1 experience bonus. +50% to +75% is a 17% relative increase, indicating +1

[ September 23, 2003, 03:11 PM: Message edited by: Robert Olesen ]

[atiff]

That all seems pretty reasonable to me, Rob. Looks like you have found your answer.

[Joachim]

Originally posted by Robert Olesen:

Would you rather face 2000 points of Conscripts than 1500 of Regular troops? Or 2000 points of Regular troops instead of 1500 points of Crack troops?

In a non-ladder play I don't care what I face. I'm not playing to win, I'm playing to fare as good as possible. So the enemy size does not matter

As my tank numbers above suggest, tanks increase less for the first two exp. levels. In a pure armor battle, it would be harder to play 2000 pts conscripts than 1500 regulars (if BFC knew what they did when doing the pricing).

Now for a Infantry or CA game, if I have a crack force, I'd prefer to fight the regulars, as I fear my crack units run out of ammo before killing all conscripts. If I have greens or conscripts, I'd prefer to fight the conscripts.

If I'm solely looking for a given number of kills for one of my units to increase experience of that unit (as BCR does) - it is easier to score kills on conscripts, so I'd tend to fight lower experience.

Gruß

Joachim

[Robert Olesen]

So, Scarhead, the short version of your answer is: It depends. That's true, ofc.

But I don't agree that BFC necessarily got the cost vs. difficulty comparison right. They had a lot of numbers to input, and I find - personally, that is - that the fitness cost modifiers don't match the difficulty the fitness changes introduce. And besides, the experience bonus is a separate handicap, it is probably not meant to be compared to a cost change.

Also, I noticed when doing my own cost samples that artillery and some armor units had a lower cost increase progression than infantry.

And I don't think the amount of experience you can accumulate vs. the various forces is a useful comparison factor. I'm looking for a comparison of the difficulty you have in achieving your objective or, to put it simpler, to win the battle according to the CMBB victory criterion.

I'll go with the rule of thumb I mentioned above to start with. It can't be all wrong.

[Joachim]

Originally posted by Robert Olesen:

So, Scarhead, the short version of your answer is: It depends. That's true, ofc.

But I don't agree that BFC necessarily got the cost vs. difficulty comparison right. They had a lot of numbers to input, and I find - personally, that is - that the fitness cost modifiers don't match the difficulty the fitness changes introduce. And besides, the experience bonus is a separate handicap, it is probably not meant to be compared to a cost change.

Also, I noticed when doing my own cost samples that artillery and some armor units had a lower cost increase progression than infantry.

And I don't think the amount of experience you can accumulate vs. the various forces is a useful comparison factor. I'm looking for a comparison of the difficulty you have in achieving your objective or, to put it simpler, to win the battle according to the CMBB victory criterion.

I'll go with the rule of thumb I mentioned above to start with. It can't be all wrong.

Short answer: It ain't that bad. It uses the numbers of BFC. It is even based on an obvious way of looking at it. (BTW: For any Southern German like me, "not bad" is not worse than "good". Not at all!)

The rule is entirely based on numbers, not on personal preferences whether someone prefers to fight with vets or greens. This way it is disputeable, but in the end you can alyways say it is based on the numbers of BFC and those should be blamed for setting them as they are.

Yet I would not play it vs some gamey bastard who sets the force types after the rule is negotiated. But I bet that's not what it is designed for.

I had even considered to put something like that into BCR for the large battles - I'd rather fight 17000 points regular than 17000 points conscript. This is a hardware restriction, not because I prefer fighting regulars :mad:

On the "experience gathering"

All my research in maths was based on global optimization over time. Single results in one period were less important than the whole. My guess is that keeping my force in good shape now will pay off in future battles. Strange habit. But it creeps into my thinking so often. Yet it is not a necessity for any campaign to factor that in, as this is just a personal preference. But somebody asked for personal preferences, so I said them. And as these are very subjective views and you want an objective view, I added the explanation for my prefs.

Gruß

Joachim

[Robert Olesen]

Originally posted by Scarhead:

On the "experience gathering"

All my research in maths was based on global optimization over time. Single results in one period were less important than the whole. My guess is that keeping my force in good shape now will pay off in future battles. Strange habit. But it creeps into my thinking so often. Yet it is not a necessity for any campaign to factor that in, as this is just a personal preference. But somebody asked for personal preferences, so I said them. And as these are very subjective views and you want an objective view, I added the explanation for my prefs.

Scarhead, that's not a problem. I obviously want to factor the experience effect into ROQC, but in this particular case I was concerned with an AI strength comparison for a single battle, not in particular how this affects the player's ability to accumulate core force experience. It obviously may affect that, but that's incidental unless the effect is unbalancing.

Thanks.

Where I come from (Western Jutland), "Not bad" is actually equivalent to "very good" in most other places.

[JAT]

Robert,

Just thought of something when I got a +1 Comp. Bonus on my next battle. When is the Bonus applied? Is it prior to the computer picking it's forces, or is it applied after the computer picks its forces? Makes a big difference. I was always assuming that it was applied before the computer picked. Do you or anyone else know?

jt

[Robert Olesen]

Before you pick your forces. At the screen where you pick side. That's actually after the computer has purchased the forces it is allowed to buy.

Bear in mind that the experience label on the enemy units doesn't change when you increase the computer experience bonus, only the behaviour.

[Joachim]

Originally posted by Robert Olesen:

Pleased to hear that we encircled those Northern Germans who don't understand the true meaning

Gruß

Joachim