Gunnery Scoring and seconds to kill

[mapman]

The gunnery score is a combination of hit percentage and average seconds to hit.

I am working on a moving gunnery scr similar to Dagger's Tank Shoot on the Run, but one that does not need perfection to get any score at all. :shock:

I noticed in the scoring conditions that a (number of my choice) can be selected to multiply by an (time of my selection MINUS the average time to hit) and this brought up a few questions:

1) When does the clock start ticking on the time to hit? When the TC spots the enemy? When the enemy appears (in the tank range) or when it spots me (in the scr I am building)?

2) How much weight is given to the time to hit versus accuracy in the Tank Range?

3) If I see a tank and miss it twice and then hit it on the third time, is my time to hit based upon the total elapsed time from when the clock first started ticking until my third round hit the target or from the time my third round was loaded to the time I hit the target?

4) Based upon your experiences, what does everybody think would be the right mix for a moving gunnery scoring condition and should it include % of enemy units destroyed?

Right now I have 35 enemy units on my range and 40 sabots in my tank.

Last question.

How many of you would be interested in a moving gunnery course for a platoon and fully manned tank competition between the three or four tanks real time?

That is....have multiple targets over a course (impotent) and a human TC and a human Gunner and all three or four tanks driving the course and shooting in a race to see who kills the most and makes the finish line first?

Kind of a tank sporting event? SB Olympics etc. :D

By the way, the new Roadcourse map is great!

~ 8)

[Lone*star49]

The gunnery score is a combination of hit percentage and average seconds to hit.

I am working on a moving gunnery scr similar to Dagger's Tank Shoot on the Run, but one that does not need perfection to get any score at all. :shock:

I noticed in the scoring conditions that a (number of my choice) can be selected to multiply by an (time of my selection MINUS the average time to hit) and this brought up a few questions:

1) When does the clock start ticking on the time to hit? When the TC spots the enemy? When the enemy appears (in the tank range) or when it spots me (in the scr I am building)?
.........

:arrow: I believe it's when the TC makes a target call, the time starts ticking..
.................................................. ......

2) How much weight is given to the time to hit versus accuracy in the Tank Range?
.........

:arrow: Hmmm, time is based on kills, so if you have to fire a second round to make kill, that eats up the time, and accuracy is based on both total targets in course vs how many of them you killed, and say you kill 8 out of 10 targets, then accuracy would be 80%.. but if one of the kills took 2 shots, I think the accuracy would be around 75%
.................................................. ......

3) If I see a tank and miss it twice and then hit it on the third time, is my time to hit based upon the total elapsed time from when the clock first started ticking until my third round hit the target or from the time my third round was loaded to the time I hit the target?
.........

:arrow: total time it takes for that kill
.................................................. ......

4) Based upon your experiences, what does everybody think would be the right mix for a moving gunnery scoring condition and should it include % of enemy units destroyed?
..........

:arrow: I like the idea of say 35 targets, and 40 rounds of sabot, but I also like the idea of say 6 - 8 Heat Rounds for close target kills, and the idea of saving the Sabot for the harder targets ahead, while having to remember to switch the round indexing.. IMO kinda adds to the challenge.
.................................................. ......

Right now I have 35 enemy units on my range and 40 sabots in my tank.

Last question.

How many of you would be interested in a moving gunnery course for a platoon and fully manned tank competition between the three or four tanks real time?

That is....have multiple targets over a course (impotent) and a human TC and a human Gunner and all three or four tanks driving the course and shooting in a race to see who kills the most and makes the finish line first?

Kind of a tank sporting event? SB Olympics etc. :D
.........

:arrow: Sounds like a great challenge and contest to be had IMO
.................................................. .......................................

By the way, the new Roadcourse map is great!

~ 8)

...


LS :helmet:

[TaNkB0y]

1) When does the clock start ticking on the time to hit? When the TC spots the enemy? When the enemy appears (in the tank range) or when it spots me (in the scr I am building)?

If you want it to mimic the US FM 17.12 Tank Tables, The clock should start when the target is "fully exposed". So either when it begins moving or when it moves out from behind what ever cover you've hidden it in.

2) How much weight is given to the time to hit versus accuracy in the Tank Range?

Its all time to hit. Each engagement has a specified number of rounds permitted but its usually not a issue ( ie: 3 rds for a single tank engagement) and I don't know how you would track that with SB anyway.

3) If I see a tank and miss it twice and then hit it on the third time, is my time to hit based upon the total elapsed time from when the clock first started ticking until my third round hit the target or from the time my third round was loaded to the time I hit the target?

From the time the target becomes active or fully exposed to when you kill it. In real life, sometimes you get time taken off for obscuration (dust kicked up by the muzzle blast) or for breach malfunctions, but since those don't happen in SB they aren't an issue.

4) Based upon your experiences, what does everybody think would be the right mix for a moving gunnery scoring condition and should it include % of enemy units destroyed?
Right now I have 35 enemy units on my range and 40 sabots in my tank.

The more the merrier! You could replicate an actual Table VIII but I think it would be pretty short and boring in SB. I think 50/50 stationary to moving is good. It would break up the monotony of plinking tanks. Also having PCs and troops (and requiring the proper ammo) will make it more interesting.

How many of you would be interested in a moving gunnery course for a platoon and fully manned tank competition between the three or four tanks real time?

Thats Tank Table XII. Its the bomb!

That is....have multiple targets over a course (impotent) and a human TC and a human Gunner and all three or four tanks driving the course and shooting in a race to see who kills the most and makes the finish line first?

Hmmm sounds like fun, but it will be very tempting for the slow poke to bust his faster compeditors in the grill doors. :shock:

Kind of a tank sporting event? SB Olympics etc. :D

Thats what gunnery is!

By the way, the new Roadcourse map is great!

Thanks! :D

James

[eageleye]

Sounds great and as far as busting fellow competitors that would be automatic no go

[Lone*star49]

Sounds great and as far as busting fellow competitors that would be automatic no go
...

Yep, any Frags would be immediate grounds for disqualification of that persons a/o team run.. And may bring one up on charges of hostile actions against a friendly and a review in front of the SB Board of Inquiry, which may result in quilty party facing Battledog in the Walls of Woe Courtyard.. :shock:

:arrow: You have been warned..


LS :3starSK:


PS.. BD loves the WWC ~ hehehe

[mapman]

2) How much weight is given to the time to hit versus accuracy in the Tank Range?
.........

Hmmm, time is based on kills, so if you have to fire a second round to make kill, that eats up the time, and accuracy is based on both total targets in course vs how many of them you killed, and say you kill 8 out of 10 targets, then accuracy would be 80%.. but if one of the kills took 2 shots, I think the accuracy would be around 75%

Thanks for the feedback!

Maybe Al or Ssnake can clarify this for me. If I hit 8 out of 10 targets, that is 80% accuracy. Right?

If I kill 6 of those 10, that is 60% kill ratio.

If I hit the target in 2 seconds, but don't kill it, does that count in toward my average time to hit? (No...answering my own question...it is average time to kill!)

In the score screen (that I can't upload the screenshot of for some reason), it gives the following choices:

+ __ x hit percentage

+ __ x ( __ seconds - average kill time)

+ __ x fraction of destroyed enemy forces

etc.

Now in the Tank Range screen it only lists Accuracy % and average time to kill. If Accuracy % was just kills, we would all get 100% on the tank range.

I am wondering how Al set up the tank range scoring. For instance, if I used the above formula as follows:

+50 x hit percentage
+2 x (30 seconds - average kill time)

and I hit 90% with an average of 10 seconds to kill then

+ 50 x 90% = 45
+ 2 x 30 - 10 = 20

giving me a score of 65

So...how is this weighted in the tank range?

In a regular mission, we can not give the tank unlimited ammo. So while the fraction on destroyed enemy forces is not a factor on the Tank Range, it would be on a gunnery practice made in a regular mission editor.

So, if I added + 10 x fraction of destroyed enemy forces to the above and I killed all 35 units with my 40 sabots my score would be:

10 x 100% = 10 + 65 = 75 (out of a possible 120)

So...with all this in mind, I ask all of you statistical wizards out there what would be the best scoring combo formula for a series of gunnery scrs (that might also include some with just GAS in emergency mode)?

[Lone*star49]

...

:( math.... help!!

But the total kills = %accuracy

And the time it took for each kill, depends on how fast it took per kill.. thus 4.1 sec..

I have done the same @ 100% accuracy and had 9sec's 12 secs.. etc.. cause of misses and 2 shot kills..

So time would be key in contest as to winner../winners.

** But some could get 100% and have high times.. 12 sec's for example..

** And some could get 90% of targets with a faster time.. not by much..but 9.8 secs.. etc.

:arrow: So Winner mush have best Accuracy and time to be da King of the Hill...lol

That's all the math I know.. hehehe



LS :? (math).. eeek!

[smogover]

G'day Mapman,

What do you think of the following formula?

Score = ( Accuracy % * Time taken to kill ) * 100, ie.

Accuracy % = 80
Time taken to kill = 8
Score = 1000

Accuracy % = 100
Time taken to kill = 6
Score = 1667

I seem to remember something like it being suggested by someone in a previous thread for SB2 gunnery ratings (sorry, can't remember who :oops: ).

I think that if this formula is applied by using the percentage of shots that hit the target as the 'Accuracy %', & the total elapsed time from when the target appears till it is killed as the 'Time taken to kill', it should be a fair indication of how useful the persons gunnery skills would be in the real world.

By the way, totally agree with your comments about Dagger's scenario. Although it is great practice, that thing is too damb hard for me :(

Smog

[NEpi]

as the only "declared" statistician here i know of, i'll have to take the challange (although that's not what you need for such a question) :)

the problem with that calculation is that all the factors are interdependent:
low accuracy yields high kill times
low accuracy yields lower kill percentage

still, accuracy and kill time is not fully-dependent, while accuracy and kill percentage is, so i think using kill% is fault, unless you want to use it as a measure of accuracy.

now what's more important, accuracy or kill time? as i assume stronger effect of accuracy on time rather than the opposite (missing and having to wait for the loader is much more time consuming than taking the time to aim better in the first place), i suggest a stronger response to accuracy than to time.

Score=Accuracy*(kill-time/good-time)

accuracy is %*100
kill-time is average kill time per target
good-time is a good-enough average kill time. i suggest 5sec.
if you're a superb gunner, hitting 100% at less than 5sec on average, you're going to be higher than 100 score. if not, your score will be in the range of 0-100.

to adjust to manual GAS range, just change the parameters accordingly (change good-time, and add to accuracy):

GASScore = (accuracy+30)/(kill-time/GAS-good-time).

to find out what is the appropriate GAS accuracy and time bonus, you should test a good gunner (put not the freaky-freaky gun maniacs roaming this board ;) ), let him train for some time in these conditions, and then find out what's his accuracy and time score, and correct the formula to make him have the same score as the regular score.

i mean, have the GAS score fit the regular score empirically.

you can calculate that fit for any set of conditions, and test it on others as well. save the original data, so you could refit it later again.

don't forget to stop measuring time to hit when the player ran out of ammo. just finish the range at that point.

[Ssnake]

The clock for killing time starts ticking the very moment a line of sight between the target and your tank is established. Whether you or a virtual crew member actually recognize the target (or the target recognizes you) is irrelevant.

Kill percentage and accuracy are not identical. Accuracy means the fraction of rounds that actually hit the target (in relation to all rounds fired), irrespective of the damage they create. You can be 100% accurate with a rifle and still not score a single kill against a tank.

As far as the gunnery range is concerned, there is a slight bias towards accuracy. It hurts your score more to miss a single shot than to be a second late. We want to encourage you to aim precisely. However, this built-in scoring formula of the gunnery range applies only to this specific scenario.

[mapman]

Thanks Ssnake!!

[mapman]

Another question:

While setting up the red units for the moving gunnery I notice that there is an option of each unit to have Display on or off. I can not find anything in the manual on this. Can someone explain please?

[Ssnake]

I think that for computer controlled units there was the option to hide them in the map view from being displayed to friendly units unless a LOS was established. Do you mean that? Or are you referring to the option to hide checkpoints from the player's view?

[mapman]

I mis-spoke.

The option is for the Red units only (computer or player owned) and it is for the route options only.

In the drop down menu, right under Smoke Generator it says:

Set Display > On or Off.

[NEpi]

of course accuracy is not exactly kill-ratio, but both are tied together, as long as the ammo is good enough to have high kill/hit rate.

so how can such a formula be formed into a "regular" sce?

[mapman]

Ssnake Wrote:

I think that for computer controlled units there was the option to hide them in the map view from being displayed to friendly units unless a LOS was established. Do you mean that? Or are you referring to the option to hide checkpoints from the player's view?

mapman wrote:

The option is for the Red units only (computer or player owned) and it is for the route options only.

In the drop down menu, right under Smoke Generator it says:

Set Display > On or Off.

Does that help Ssnake?

[mapman]

Note of interest:

In the regular scenario scoring it give you average seconds to kill.

In the scoring screen for the tank range it states "Average Seconds to Hit."

[Lone*star49]

Note of interest:

In the regular scenario scoring it give you average seconds to kill.

In the scoring screen for the tank range it states "Average Seconds to Hit."
...

Thus, when the TC says "Fire" the clock starts ticking, so imagine if you will, on the tank range, when the first of ten targets appears, and you have (as gunner) your sights on the first target before the TC says "Fire" and fire exactly at the same time he say's it, you get a say.. 2 sec time for that kill, now, if you (as gunner) have the rest of the 9 remaining targets in your view before TC spots and gives order to fire, and you kill all 9 before he says anything.. your score time will remain as 2 sec's..

:?:

LS

[Ssnake]

Mapman, I'm sorry but is it really important for you, and impossible to find out by yourself?
I guess I COULD fire it up and have a look, but it's going to cost me time that I'd rather spends elsewhere. I'm going to do it if you tell me that it's important, though.

[mapman]

Not a problem Ssnake....I am sure somebody out there who is not involved in SB2 production knows the answer. I was responding to your request for clarification.

Thanks,

mm

[Ssnake]

...yes, I know. I always provoke more work for me. :o
Sorry.

[mapman]

Ok...so this thread has been feeding my thirst for Gunnery Scoring Knowledge resulting in this:

http://www.steelbeasts.com/modules.php?op=modload&name=PNphpBB2&file=viewtopic&t=2140

Let me know what you think on that thread please!