Jump to content


Real traverse speed calculation


  • Please log in to reply
13 replies to this topic

Dr_Oolen #1 Posted 08 May 2015 - 05:26 PM

    Lieutenant

  • Player
  • 23250 battles
  • 1,780
  • [IDEAL] IDEAL
  • Member since:
    03-13-2012

I've always been interested in how things actually work in this game, one of them being the effect of terrain resistances on the maneuvrability of tanks. Well, I measured a couple of things and figured out how terrain resistances affect traverse speed of tanks on different surfaces.

 

All the info/measurements can be found in a spreadsheet here:

https://docs.google.com/spreadsheets/d/1nLSD6E9tFdtbUeoWGyU-V6Rj0AyUnGXVCPG2vDMSgWs/edit?usp=sharing

 

I guess that should answer the hot topic about cromwells and their performance compared to each other. And you can calculate real traverses for any tank using the little calculator.

 

EDIT 10.5.2015 - RichardNixon went through the spreadsheet and found that i made an error in the spreadsheet which applied some coefficients twice, rekkin expected values for soft/medium terrain for some tanks. Fixed it and now everything is ok and well within measurement error.


Edited by troolenhardy, 10 May 2015 - 09:28 AM.


ClassicFrog #2 Posted 08 May 2015 - 06:34 PM

    Major

  • Player
  • 47145 battles
  • 2,998
  • Member since:
    04-05-2012

What is the real vs measured? As in, expected vs actual or in-garage vs actual?

 

Also, why not put the measurements into degrees per second, to compare the true value to in-garage value? I guess it would look be more informative. (frogs aren't the smartest amphibians out there)


Edited by ClassicFrog, 08 May 2015 - 07:28 PM.


Dr_Oolen #3 Posted 08 May 2015 - 06:55 PM

    Lieutenant

  • Player
  • 23250 battles
  • 1,780
  • [IDEAL] IDEAL
  • Member since:
    03-13-2012

View PostClassicFrog, on 08 May 2015 - 06:34 PM, said:

What is the real vs measured? As in, expected vs actual or in-garage vs actual?

 

Also, why not put the measurements into degrees per second, to compare the true value to in-garage value? I guess it would look be more informative.

 

Yes, the first 3 columns on the right of the tank names are expected values calculated with the equation. The measured ones, in game (with estimated ~5% error), are on the right. All those values are in degrees per second. The traverse that is shown in the game (but including bonus from clutch braking) is in the area on the very left, also in degrees per second.

ClassicFrog #4 Posted 08 May 2015 - 07:22 PM

    Major

  • Player
  • 47145 battles
  • 2,998
  • Member since:
    04-05-2012

View Posttroolenhardy, on 08 May 2015 - 05:55 PM, said:

All those values are in degrees per second. The traverse that is shown in the game (but including bonus from clutch braking) is in the area on the very left, also in degrees per second.

 

Yeah, my brain had a massive fart and I mixed soft and hard values. English is not my first, and I somehow read soft as hard and hard as soft. Doing that I then thought tanks can't turn faster on soft ground than they do on hard, so it must be seconds ... oblivious to the fact that 50 seconds full turn is totally ludicrous. Don't mind me :hiding:

Edited by ClassicFrog, 08 May 2015 - 07:24 PM.


RichardNixon #5 Posted 08 May 2015 - 08:40 PM

    Captain

  • Player
  • 4395 battles
  • 2,351
  • Member since:
    06-26-2013

I figured out the same formula here. Your version is missing the weight modifier, although it doesn't affect premiums:

http://forum.worldoftanks.eu/index.php?/topic/495217-tanking-w-science-all-about-that-fuel/page__pid__10654374#entry10654374

 

Not sure if you have any tanks with pivot in that list.

 



Dr_Oolen #6 Posted 08 May 2015 - 09:36 PM

    Lieutenant

  • Player
  • 23250 battles
  • 1,780
  • [IDEAL] IDEAL
  • Member since:
    03-13-2012

View PostRichardNixon, on 08 May 2015 - 08:40 PM, said:

I figured out the same formula here. Your version is missing the weight modifier, although it doesn't affect premiums:

http://forum.worldoftanks.eu/index.php?/topic/495217-tanking-w-science-all-about-that-fuel/page__pid__10654374#entry10654374

 

Not sure if you have any tanks with pivot in that list.

 

 

Nice, i didnt measure accurately enough to be able to figure out the weight factor, will put it into my spreadsheet and put your name down as a source. I wanted to do real hp/t (so basically acceleration as hp/t cant be measured in the game) and how terrain affects it aswell. Guess ill just wait for your results if youre already in the process of figuring it out, or i could help with the testing if you wanted.

 

I did look at pivot vs not pivot tanks but with my measurement errors i couldnt see any trend in measured vs calculated numbers that would suggest there is any difference.



RichardNixon #7 Posted 08 May 2015 - 09:59 PM

    Captain

  • Player
  • 4395 battles
  • 2,351
  • Member since:
    06-26-2013

View Posttroolenhardy, on 08 May 2015 - 09:36 PM, said:

 

Nice, i didnt measure accurately enough to be able to figure out the weight factor, will put it into my spreadsheet and put your name down as a source. I wanted to do real hp/t (so basically acceleration as hp/t cant be measured in the game) and how terrain affects it aswell. Guess ill just wait for your results if youre already in the process of figuring it out, or i could help with the testing if you wanted.

 

I did look at pivot vs not pivot tanks but with my measurement errors i couldnt see any trend in measured vs calculated numbers that would suggest there is any difference.

 

What measurement method did you use? I turn down WoT's graphics settings, use Fraps @ 60fps lock and Virtualdub for frame counting. You can do shooting mechanics fairly accurately with that, so it's plenty for measuring a 360 degree traverse.

 

I have some numbers on acceleration but I haven't managed to guess a halfway working formula yet. The acceleration vs time curves don't follow any simple rule: The first km/h take a while to accumulate, as though there's some kind of torque curve in there. Would be tough to reverse-engineer that, if it's true.

 



Spotterino_Aervorn #8 Posted 08 May 2015 - 10:04 PM

    Staff Sergeant

  • Player
  • 23634 battles
  • 439
  • [S4LT] S4LT
  • Member since:
    05-10-2012

Great stuff.

+1



Dr_Oolen #9 Posted 09 May 2015 - 11:41 AM

    Lieutenant

  • Player
  • 23250 battles
  • 1,780
  • [IDEAL] IDEAL
  • Member since:
    03-13-2012

View PostRichardNixon, on 08 May 2015 - 09:59 PM, said:

 

What measurement method did you use? I turn down WoT's graphics settings, use Fraps @ 60fps lock and Virtualdub for frame counting. You can do shooting mechanics fairly accurately with that, so it's plenty for measuring a 360 degree traverse.

 

I have some numbers on acceleration but I haven't managed to guess a halfway working formula yet. The acceleration vs time curves don't follow any simple rule: The first km/h take a while to accumulate, as though there's some kind of torque curve in there. Would be tough to reverse-engineer that, if it's true.

 

 

Edited the spreadsheet to include your findings.

 

I measured all tanks on lakeville, found some flat ground with the three ground types.

 

I lined up my tank with some feature i could recognise, started turning at some nice ingame time and turned, then when i reached the same spot again i looked at the time.

 

On soft ground i did 3 turns, on medium and hard ground i did 5 turns. This resulted in 30+ second intervals on soft ground (up to ~70) and 30+ turn times on hard/medium ground (up to ~70).

 

My estimate is that the original point where one starts to turn could be ~ 0.5 seconds off with the end point having higher error, perheaps around 1 second. Thats why i did 3/5 turns so that when averaged over the number of turns the error would be around half a second for one turn. Which for fast turning tanks on hard terrain makes the error +- 3° per second and for the slow tanks around 1°. Theoretically the soft ground measurements should be most accurate. But these are of course only guessed numbers (the orrors). Of course some lag could also have a noticeable effect.

 

When i incorporated your additions it did solve some tanks but introduced quite some discrepancies elsewhere. TBH i dont really care too much now to find out what the problem is or to remeasure stuff more accurately. Initially i just wanted to see what happens in general, only in regards to terrain resistances (so i didnt care at all about the effects of crew skills, pivot/non pivot, weights), which i found out and the results the initial equation produced were for my purpose good enough.


Edited by troolenhardy, 09 May 2015 - 11:43 AM.


Trent #10 Posted 09 May 2015 - 02:20 PM

    Captain

  • Player
  • 10516 battles
  • 2,433
  • Member since:
    07-20-2011

View PostRichardNixon, on 08 May 2015 - 08:59 PM, said:

 

 

Did you also specifically test how equipment's weight affects it? CartoonVillain in another thread reminded of something I've wonder for a long time, too: additional grousers reduce terrain resistance for medium and worst ground, but at the same time they weigh a lot.


Edited by Trent, 09 May 2015 - 02:22 PM.


RichardNixon #11 Posted 09 May 2015 - 04:13 PM

    Captain

  • Player
  • 4395 battles
  • 2,351
  • Member since:
    06-26-2013

View PostTrent, on 09 May 2015 - 02:20 PM, said:

 

Did you also specifically test how equipment's weight affects it? CartoonVillain in another thread reminded of something I've wonder for a long time, too: additional grousers reduce terrain resistance for medium and worst ground, but at the same time they weigh a lot.

 

I assumed that equipment weight counted the same as module weight, given that they add up in the garage. It's possible that there are bugs which cause some types of component not to be counted in-game, but I didn't test that far. Assuming that there isn't a bug, grousers would have minimal if any traverse benefit to a Pz IIIA on medium ground.

 

Edit: Checked the Pz IIIA + grousers case, comes out as the formula suggests, although maybe fractionally high. It's definitely taking the weight into account anyway.

 

 


Edited by RichardNixon, 09 May 2015 - 04:58 PM.


Trent #12 Posted 10 May 2015 - 08:41 PM

    Captain

  • Player
  • 10516 battles
  • 2,433
  • Member since:
    07-20-2011

View PostRichardNixon, on 09 May 2015 - 03:13 PM, said:

It's definitely taking the weight into account anyway.

 

Well, so it does hamper your performance on best ground - added weight, but no decreased terrain resistance.



Dexatroph #13 Posted 19 May 2018 - 12:49 PM

    Lieutenant Сolonel

  • Player
  • 46304 battles
  • 3,196
  • Member since:
    02-09-2013

I want to point out that the internal formula might have changed or have some other variable what is unknown.

 

Based on the spreadsheet (tanks.gg use the same formula) the KV-2 seems to have a better effective/real traverse speed than the KV-2R. 

https://tanks.gg/compare/kv-2?l=41211a&t=kv-2-r 

 

But I made some tests myself and in reallity the KV-2R turns faster (~20% better) than the regular KV-2 and what is calculated by the formula.

 



tmp #14 Posted 19 May 2018 - 01:42 PM

    Corporal

  • Player
  • 94800 battles
  • 175
  • [BADDY] BADDY
  • Member since:
    05-01-2011
One thing to remember is that it seems like the terrain resistance data on tanks.gg is not always correct.  I ran into this while trying to figure out the formula for terrain resistance - for example, T-34-85M should by numbers be significantly more sluggish than T-34-85, but in practice it isn't.  It was before it got a mobility buff, so I assume that they forgot to update the numbers.  Similarly, AMX 65 t has by my calculations terrain resistance of about 2.4 on _medium_ terrain rather than soft like it says on tanks.gg.  Granted, I may be wrong but it _is_ in practice more sluggish than tanks with comparable numbers.




1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users