Re: [NetEpic ML] [v5.0] Analysis - Help!

From: Daryl Lonnon <dlonnon_at_...>
Date: Thu, 16 May 2002 11:27:52 -0600 (MDT)

Peter Ramos wrote:
> Hi!
>
> Hm. IMO there's one important dependency, namely Move-CAF dependency. A
> good CAF is less useful if the unit in question is slow. IMO high CAF
> should be worth more if a unit is also fast. How difficult this is to
> accomplish I don't know.
>
> That's a very good point. You could combine move and CAF since they are
> co-dependent.
>
> Peter

I'm going to ramble for a long long time (you've been warned),

This same basic discussion came up on the GZG list about a month back
(except it was in regards to DSII). I invite you to join the
yahoo gzg list and look in their archives if you want to wade through
long posts (and arguments) on point systems.

In case you're wonder, I don't have a degree
in mathematics (I do have a minor), but I did think statistics was
one of the funner classes I had in college. So I'm warped that way.

Now DSII has a much more difficult task than NetEpic. It allows you
to generate vehicles with any stats you want (an example of this for
netepic would be allowing you to field ubertanks (mv: 50cm, Save: 1+,
Dice: 8, TSM: -5) and claiming to have a point system that would
roughly balance it against a "balanced" opponents force).

What the ending concensus was, is that everything effects everything else.

Defense allows you to survive longer and use your mobility and offensive
powers more (ie further into the game). Mobility allows you to get
in position to use your offensive abilities better and enhances your
defenses by allowing you to avoid potentially dangerous fire/situations.

So the formula is
Offense * Defense * Mobility = point cost

This is NOT the linear point system that netepic uses today.

Now the big difficult task is what values should be given to Offense,
Defense, and Mobility. They too aren't straight linear equations either.
They are also VERY MUCH dependent on the force and terrain they
are facing. In example, having the pop-up ability (a mobility advantage)
is useless on a wide open field; having a 1+ save against a force
that has no TSM less than 0 is priceless (and there attacks are
correspondently worthless); having a 6+ save against a force with
all their weapons having a TSM of -1 is worth no more than not having
a save at all.

So first you have to make assumptions (bad assumptions completely
invalidate your point system). How many dice at which TSM levels
do you expect to get targeted with. How many targets with what
saves will you have targetable with your weapons. How much
will terrain constrain your mobility (and effect the above targeting
and targetability).

In example: on the 1st turn my "average" opponents force will
throw:
1% of their dice with a TSM -5,
4% of their dice with a TSM -4,
7% of their dice with a TSM -3,
13% of their dice with a TSM -2,
25% of their dice with a TSM -1,
50% of their dice with a TSM 0

From this table I can calculate the statistics (and weight them
appropriately) for each armor value level. Using the statistics
I can get a good guesstimate on how much the Save adds to the Defensive
Value of this unit.

Using the above example:
A save of -4+ would be weighted 0 (ie invincible).
A save of -3+ would be weighted 1/6.
A save of -2+ would be weighted at 1.
<cut some values>
A save of 3+ would be (50% * 1/3) + 25% * 1/2 + 13 * 2/3 + 7 * 5/6 + 5 =
17 + 13 + 8 + 6 + 5 = 49
a save of 4+, would be (50% * 1/2) + (25% * 2/3) + (13% * 5/6) + 7% + 4% + 1% =
25 + 17 + 11 + 7 + 4 + 1 = 65
A save of 5+ would be (50% * 2/3) + 25% * 5/6 + 25% = 33 + 21 + 25 = 79
A save of 6+ would be (50% * 5/6) + 50% = 42 + 50 = 92
No save would be weighted 100.

To get the Defensive value (assuming Save is the only component to your
defensive value (CAF being another possiblity)), you'd take the
weight over the inverse of 100. i.e. Defense = 100/weight.

Now you do the same for your offensive system (assign the usefulness
value) and weight them appropriately to get the offensive value (keeping
in mind that stuff like troop transport are offensive capabilities
(since troops tend to be offensive in nature)).

Do the same for mobility characteristics.

Now mulitply them through.

If you haven't realized by now, all of the above is rather painful.
Pain notwithstanding, it's a much more mathmatically sound point system
model than a linear one (i.e. it's much more capable of dealing with
corner cases (like the ubertank example above), than a linear point
system).

But for netepic, I'd say it's pure overkill (mainly because there is
no "design your vehicles" element to the game). The concepts are
important to understand ... since they help with understanding
why purely linear point systems are broken and why (and HOW) you need
tweek points systems that result from them to give a really
good point system.

I'll stop rambling now,
Daryl
Received on Thu May 16 2002 - 17:27:52 UTC