Gnome76
01/30/05 12:17 AM
68.0.125.156
|
So I was thinking, BattleTech is primarily about 'Mechs, but vehicles are a big part of the game. 'Mechs have to be built in 5-ton increments, which limited the original engine chart to 5-ton increments. Also, everything had been rounded to .5 tons. So, unless there's already a published formula for it, here's the best I could come up with:
Engine Weight=((1.08569678^(Engine Rating*0.12)))
This doesn't fit the progression perfectly, with the medium range of engines coming out lighter (by up to 1.387 tons) and the smaller and larger engines generally being heavier (by up to 2.147 tons), but the average in differences came out to less than a gram, with which I am happy.
Questions?
Comments?
A more accurate formula?
|
Toontje
01/30/05 05:07 AM
84.24.165.226
|
Did a curve fit using a spreadsheet? you might try something that is (tonnage = a(rating^b)+C) instead of the current tonnage = a(rating^b).
IIRC lifting the starting point a bit results in a more flat curve, thus smaller engines become lighter, medium heavier, and large lighter.
Rather to blow up, then.
|
Gnome76
01/30/05 09:28 AM
68.0.125.156
|
Well, mine is actually W=a^(R*b), and looking at the syntax for your suggestion, it would just add weight across the entire range of engines. But I only did mess with this for about half an hour.
|
Gnome76
01/30/05 09:33 AM
68.0.125.156
|
Here's the .XLS file I used, if anyone wants to tweak with it.
Attachment (540 downloads)
|
Gangrene
01/30/05 08:43 PM
24.6.228.14
|
I messed with this a while ago because I was not happy with the limitations of the table format. I did a curve fit and came up with a 7th order polynomial thats pretty close. The greatest discrepancy is about 0.3 tons. If X is the engine rating, the tonnage can be computed by:
t = 3.4515E-16*X^7 - 3.634E-13*X^6 + 1.47E-10*X^5 - 2.5384E-08*X^4 + 1.418E-06*X^3 + 1.7231E-04*X^2 + 0.0113*X + 0.285
The order has to be so high because of the highly discontinuous regions of the table near the low end.
Gangrene
Edited by Gangrene (01/30/05 08:45 PM)
|
Nightward
01/31/05 12:03 AM
203.214.146.19
|
~.^
I hate you. Nobody should ever be that good at math. It should be illegal.
Yea, verily. Let it be known far and wide that Nightward loathes MW: DA. Indeed, it is with the BURNING ANIMUS OF A THOUSAND SUNS that he doth rage against it with.
|
davion76
01/31/05 12:13 PM
12.219.244.44
|
I suspect that he used the curve fit in excel. Its really quite powerful and easy to use.
|
Wraith
01/31/05 03:30 PM
129.101.55.124
|
Excel, at least 2003, only goes to 6th order polynomials, definitely other programs out there that can do more though, I have, off hand, at least two that do automatic curve fitting. Much more satisfying to do it by hand though 
-Wraith
|
Gangrene
02/01/05 12:54 AM
24.6.228.14
|
Something like that. I did this a while ago, but if my memory serves me correctly I used the polyfit function in Matlab. I might try a 2nd or 3rd order by hand, but a 7th is too much work. I did try the 6th order available in Excel, but testing the coefficients proved to yield poor results. I have heard that Excel might have some calculation problems with high precision numbers/higher order polynomials, so maybe thats the problem.
Anyway, another idea I have had was to find a function that fits the rating vs tonnage data over the region of the table that does not have repeated engine weights. Then classify the other portions of the table by different functions or fixed values. I don't really have the time to invest in this, though.
Gangrene
Edited by Gangrene (02/01/05 12:56 AM)
|
wartang
02/01/05 11:19 AM
192.235.2.30
|
so what about the jump from 395 rated i think its weight is 26.5 to the 400 rated engin wich is something like 52.5 I dont think it would fallow the curve
i love this game
Edited by wartang (02/01/05 11:21 AM)
|
Silenced_Sonix
02/01/05 12:55 PM
168.209.97.34
|
Try an expert system that generates the wieght based on which 'section' of the curve the engine rating falls in. Last time I checked, the curve would follow a certain pattern for a few units, and then change to another, different pattern. You could try something like that, although any more in-depth questions on the topic will not be answerable by myself.
Evolve or Die
|
Toontje
02/01/05 03:57 PM
84.24.165.226
|
All curves can be fit with polynomals, even random numbers on an x-axis. It's just a question of adding enough orders. The more the better the fit.
Rather to blow up, then.
|
Wraith
02/01/05 04:15 PM
129.101.55.124
|
Of course, the more orders the more likelihood that the curve beyond the bounds of the data will be horrible.
Maximum orders on a curve fit is equal to the number of terms you are trying to fit.
-Wraith
|
Gnome76
02/02/05 05:42 PM
68.0.125.156
|
No, Excel doesn't handle precision calculations very well.
Also I decided to make it include the Large engines from Max Tech, and that made me need to use a 9th order curve fit, since some of the smallest engines went into negative weights with a 7th order.
And so I present to you Gnome's Level 4 Engine Table, with weights for engines rated 1 to 500, listed in metric tons to three decimal points (in case you want to get down to kilograms), for standard, light, XL, XXL, compact/vehicle std, vehicle light, vehicle XL, vehicle XXL, and ICE.
...in another .XLS file...
|
Gnome76
02/03/05 03:48 PM
68.0.125.156
|
Okay, apparently my last post didn't take, so I'm going to try it again...
Added in the large engine (405-500) table, had to use a 9th order curve fit, now have engine ratings from 1 to 500 and level 3 engine types, and I'm calling this the Level 4 Engine Table.
Also, if the majority of people can actually see my first attempt at putting up the lvl 4 engines, let me know.
|
Gnome76
02/04/05 01:08 AM
68.0.125.156
|
Okay, now that Nic says stuff should be working again, let's try this a third time...
My "level 4" engine table, with engine ratings 1 through 500, with compact, light, XL, XXL, and vehicle L/XL/XXL, generated by a 9th order curve fit, weights displayed in metric tons to 3 decimal places (so, essentially, Kg's)...
This is what I'll be using along with other house rules for fractional accounting, such as gyros being 1% of the engine rating, armor being 625 kilograms per point (521 C FF, 558 IS FF), ammo being allotted per round instead of ton or half-ton, and speeds being based on Engine Rating / Unit Tonnage * 10.8 WMP, 16.2 RMP (for purposes of role-playing sessions, which is mostly what my group does anyway).
If you can't tell, I like complexity and micro-management.
Anyway, my third (and most likely final) attempt to attach my level 4 engine table...
--EDIT--
If you look, you'll see you need to change the Standard Engine Weight for a 427-rate from 92.xxx to 82.xxx .... all the other engine types go off of that one, so it should be the only necessary change.
Attachment (669 downloads)
Edited by Gnome76 (02/04/05 01:17 AM)
|
Gangrene
02/06/05 02:54 AM
24.6.228.14
|
Why make a table? Why don't you just stick with the algorithm? Are the people in your group not capable of using a calculator?
Gangrene
|
Gnome76
02/06/05 11:20 AM
68.0.125.156
|
My TI-89 is about 250 miles away.
|