46 replies to this topic

[Phaesphoros]

Shameless *bump* because of update (BJ)

[Modo44]

Endo Steel and Ferro Fibrus change a lot, actually. They allow lighter 'mechs to carry much more than a simple comparison shows. You can often use all tonnage with some room to spare.

[Phaesphoros]

Modo44, on 31 May 2013 - 12:02 AM, said:

Endo Steel and Ferro Fibrus change a lot, actually. They allow lighter 'mechs to carry much more than a simple comparison shows. You can often use all tonnage with some room to spare.

Yes they do change a lot for an individual mech, but they don't change much in the comparison between different mechs if you apply these upgrades to every mech.

E.g. using an XL engine, AWS and CTF have the same pay load (- armor) at roughly 76 km/h no matter if they both use endo or both don't.

[Modo44]

I am talking about differences between sane builds of different 'mechs. You will not fit FF on most heavies or assaults, nor will you fit ES or use XL on most assaults. This gives the lighter 'mechs considerably more actual average payload than your data suggests.

[Phaesphoros]

*bumping* for updated charts

[Tragos]

Why shouldn't you use Endo on Assaults? It fits perfectly. Endo AND Ferro...well, that is tricky, but still possible for some cases. But Endo fits nearly every time and grants you 4-5 tons.

Edited by Tragos, 19 July 2013 - 02:55 AM.

[Reod Daie]

This is great. I thought that more mechs would have sharp diminishing returns. Thanks for putting this together.

[The Duke of Dirty]

How is this surprising? The heavier your 'mech, the more tons of guns you can carry. durrrr.

[Windsaw]

I think the chart I would like to see most is missing here:
STD Engine, but with Endo.
I mean, if I think about putting an STD engine in, then Endo Steel automatically becomes mandatory except for the most heavy builds.

[Phaesphoros]

Windsaw, on 22 July 2013 - 05:47 AM, said:

I think the chart I would like to see most is missing here:
STD Engine, but with Endo.
I mean, if I think about putting an STD engine in, then Endo Steel automatically becomes mandatory except for the most heavy builds.

I'd disagree that endo becomes mandatory when using a STD engine (most of the time I'm using a STD engine to get 2 more DHS). Of course there are cases this is useful, e.g. AC/20 in side torso.

Nevertheless, I added the graph. Voilà! Put some spoiler tags not to bloat the post too much.

Edit: Creating new graphs is neither a problem nor very time-consuming (any more), since I automated it to a high degree. 80 % armor? 2 % of max tonnage extra weight per 'Mech? Only 'Mechs heavier than 80 tons or faster than 40 km/h? No problem.

Edited by Phaesphoros, 22 July 2013 - 08:58 AM.

[Dracol]

Nice work. Been thinking about how to tweek some of my builds and this is a great reference.

Personally, I tend to have the same amount of armor on legs as i do the arms, on non-jumping heavies and assualts.

Any who, thanks for working the numbers. Greatly appreciated.

[TehSBGX]

DanNashe, on 21 March 2013 - 09:53 AM, said:

Cent: safe to xl at at least 90 kph.

I've Piloted Cents a lot and the XL engine thing with them is pretty debatable. If you're going 98 kph in a cn9-a, people are gonna aim for your legs usually, but getting hit in the side torso can happen too especially if they aim for your CT and your speed throws them off. So I kinda agree 90 kph and higher XLs can be worth it, but it's more so "slightly risky" instead of a flat safe.

[The Duke of Dirty]

These charts forget to include the impact of heat sink slots on larger engines. A 400 engine can house 6 dbl heatskins, the equivalent of 18 extra slots. This allows you to use many energy weapons (which are small and light) and include armor/skeleton upgrades on assault 'mechs that cannot usually afford the slots to use both upgrades.

If you could figure out how to include that, it would be real cool!
EDIT:
example http://mwo.smurfy-ne...a7f0423e78304d5

Edited by Pwnius, 23 July 2013 - 11:28 AM.

[Phaesphoros]

Pwnius, on 23 July 2013 - 11:14 AM, said:

These charts forget to include the impact of heat sink slots on larger engines. A 400 engine can house 6 dbl heatskins, the equivalent of 18 extra slots. This allows you to use many energy weapons (which are small and light) and include armor/skeleton upgrades on assault 'mechs that cannot usually afford the slots to use both upgrades.

If you could figure out how to include that, it would be real cool!
EDIT:
example http://mwo.smurfy-ne...a7f0423e78304d5

I wouldn't say they forget to include the extra heat sink slots for >= 300 rated engines. It is an important factor for hard points, but not so much for payload. Yes, it might allow you to use endo.

But these kind of data visualizations always need some restrictions / simplifications, otherwise it couldn't be a useful overview. E.g. including all upgrades (endo, FF) and engine types (STD, XL) in one graph would just result in a mess, even if that is useful for some kind of comparison.

I could make a graph that uses endo-steel structure IFF the engine rating is greater than some value (like XL350 for 4*3 = 12 DHS critical slots) if that'd be useful for you.

[The Duke of Dirty]

Phaesphoros, on 23 July 2013 - 12:02 PM, said:

I wouldn't say they forget to include the extra heat sink slots for >= 300 rated engines. It is an important factor for hard points, but not so much for payload. Yes, it might allow you to use endo.

But these kind of data visualizations always need some restrictions / simplifications, otherwise it couldn't be a useful overview. E.g. including all upgrades (endo, FF) and engine types (STD, XL) in one graph would just result in a mess, even if that is useful for some kind of comparison.

I could make a graph that uses endo-steel structure IFF the engine rating is greater than some value (like XL350 for 4*3 = 12 DHS critical slots) if that'd be useful for you.

Well, the weight value of a slot can be calculated. For a dragon with max armor, the armor/skeleton upgrades will save 5.66t (total). This makes each slot worth ~.2 tons.

Each heat sink slot is worth 3 slots of double heat sinks = .6 tons .

Simply adjust the larger engines by -.6 tons per heatsink slot.

However, the value of each slot will need to be calculated for each 'mech.

Edited by Pwnius, 23 July 2013 - 01:29 PM.

[Phaesphoros]

Pwnius, on 23 July 2013 - 01:27 PM, said:

Well, the weight value of a slot can be calculated. For a dragon with max armor, the armor/skeleton upgrades will save 5.66t (total). This makes each slot worth ~.2 tons. Each heat sink slot is worth 3 slots of double heat sinks = .6 tons . Simply adjust the larger engines by -.6 tons per heatsink slot. However, the value of each slot will need to be calculated for each 'mech.

Here you go:


However, I still think this metric is not useful for everybody. More critical slots doesn't necessarily mean you'd better put endo on it, for example if you need as much DHS as possible (where for most chassis >= 50 t, critical slots are the bigger problem).

[The Duke of Dirty]

BA11ER status: ACHIEVED.

Edited by Pwnius, 24 July 2013 - 11:35 AM.

[evilC]

A very interesting set of charts, thanks!
So it seems a general rule of thumb is that the 300 engine (STD or XL) is the most efficient?
In particular, the QKD and DRG with 300 engines seem to be about the best bang per buck possible.
Particularly heartening to see that for all my fave builds, they seem to sit exactly in the sweet spot for that chassis

Is there any chance this data could be made into a graph that can be examined more closely? Could your automation process be tweaked to create the graphs as a google charts page?

If you are interested in doing this but need assistance with the coding or whatever, I would be willing to help.

[Phaesphoros]

evilC, on 25 July 2013 - 08:48 AM, said:

A very interesting set of charts, thanks!
So it seems a general rule of thumb is that the 300 engine (STD or XL) is the most efficient?
In particular, the QKD and DRG with 300 engines seem to be about the best bang per buck possible.
Particularly heartening to see that for all my fave builds, they seem to sit exactly in the sweet spot for that chassis

Is there any chance this data could be made into a graph that can be examined more closely? Could your automation process be tweaked to create the graphs as a google charts page?

If you are interested in doing this but need assistance with the coding or whatever, I would be willing to help.

Hmm interesting idea. The program I'm using currently, gnuplot, can already create somewhat interactive graphs via SVG. In fact, I currently do export to SVG, adjust some things manually and finally export to png.

The problems with interactive graphs are:
1) I currently don't have any website to host them
2) can't probably insert them into the forums directly

Is there any way to host the interactive graphs on google? (I've lost track of their services long ago o.O)

From a technical point of view, I don't think it would be a problem to use google charts. I currently extract the data from the game files, process them, and pass the results to gnuplot. Could as well use google charts for the output part. It should even be possible to get the data off smurfy's API and have the whole thing online.

The one problem I don't know atm how to solve - aside from hosting - is that I currently adjust the placement of the labels manually. They're inserted and placed automatically, but my algorithm isn't sophisticated enough so I'll have to manually adjust them. Using a legend only isn't very useful due to the number of lines. I didn't find any feature in google graphs how to place labels (besides the annotation function), do you have any idea?

[evilC]

I would imagine if you approached one of the hosts of various MWO sites, they would embed your code in their site (Smurfy's mechbay would be a good match IMHO).

All you would need to do is have a bit of javascript that would generate the graph, and they could easily add it to their page.

Do not forget that html / js works on a local computer - you can build the code required to generate the page, then package it as a zip file and send to the webmaster who adds it to his site.

I would maybe look for a piece of gnuplot -> html software? I have no experience of gnuplot, but if it is a gnu app, chances are someone has written some kind of converter.

The label problem could be solved via a key, and also when you mouse over a line it could show the label as a tooltip.