AA Unit tests

[PoorOldSpike]

...Continue to do your researches and stats I always like Threads with Screenshots Pictures and not too cryptic Data..

Thanks mate, and you'll notice that my stunning tests are usually in direct response to people who've asked a question, and I'm only too glad to run tests to find out what the engine does so that the raw hard data I glean can be used as a solid base line foundation on which to build our tactics in whatever way we see fit, I wouldn't dream of telling people what style of play to adopt..
Hey grasshopper, what sage words of divine wisdom can you offer us to help us survive on the dynamic cyber battlefield?

[turn4441]

I'm sorry, but I don't get it. The 'tests' you glorify, while useful and I don't mean to disparage them, are simply 'controls' that take the basic attack and defense values from the equipment file (whether viewed in spreadsheet or whatever other mode) you vilify as maybe out of date or wrong and run them through the combat algorithm to get a result in vanilla circumstances with no experience or heroes included. How do we know your equipment file is correct? Yes, the results of combat against various other units are not in the equipment file (why would they be? It would have to show results for combat against every other unit in every circumstance), but if you know the combat algorithm, you can calculate them and get the same results as your tests (hence the utility of an equipment file, with unit attack/defense values, to start with when programming a game). Conversely, in the case of unit cost, it just means that the data you're looking for is available in a straight-forward manner (fixed cost) and you don't need to go to the in-game purchase screen and write down the values which never change and come straight from the exact same equipment file which you are also using (and saying may be wrong) to generate your tests and claim is old or out of date.

Bottom line, I don't understand why your tests are more legitimate than the equipment file when they use the attack/defense values from that very same equipment file to get the values that are used to generate the results of your test.

[PoorOldSpike]

...I don't understand why your tests are more legitimate than the equipment file when they use the attack/defense values from that very same equipment file to get the values that are used to generate the results of your test.

Ah, but the way I present it is much prettier...
Hurly confirmed it when he said- "I always like Threads with Screenshots Pictures and not too cryptic Data"
And my hands-on testing means we can see with our own eyes how the units perform.
I mean, if you join an online dating agency you can browse the womens dry statistics endlessly, but to really know how they perform you have to actually meet them and give them a thorough hands-on testing..

[hurly]

you only see and hear what you want to see

hurly also said

You sound like a car salesman telling me from the Booklet how low it is in Fuel Consumption and it has almost no unwanted emissions ........ under Test Conditions in a Closed Room in a Simulation of something No one is really sure about what it is, except that these Conditions never Exist


Hmmmmm thinking about it, I realize none of this Data and Arguments are relevant in Real Life Conditions, especially cause there are no new Cars delivered with Zero Mileage on the Instruments

and

I would rather listen to the Solid Advice the user Analog Gamer gave you

and regarding your cherrypicking quote

I always like Threads with Screenshots Pictures and not too cryptic Data, but one thing I learned here in my Panzer Corps Career. Always listen to the Advice of Experienced Players especially those who have seen it all more than once.

what do you think the word "but" stands for ?

so if you quote me, quote the whole passage, especially when the meaning is the complete opposite of your personal interpretation

while I said I like threads with pix and not too deep and complex data in general, I never said I find your researches very helpful. It's a bit like eating popcorn. You eat a lot of sweetened stuff with a lot of air inside.
But if you like Popcorn and don't mind the time you spend making it, go ahead. That's ok, just don't expect a very deep discussion about it

To say it crystal clear

I personally don't think your work is very helpful and effective as you get results from conditions that never happen in the game.

[PoorOldSpike]

..I personally don't think your work is very helpful and effective as you get results from conditions that never happen in the game.

Never say never!
Rather, take grasshoppers sage advice and "expect the unexpected"..
As a holy man myself, whenever I buy any game I like rigorously testing it to find out its hidden depths, then pass on that sublime knowledge to my disciples in sensational articles, tutorials and scriptures in assorted wargaming forums to help people get the best out of the game, and the feedback I get from them indicates I must be doing something right, for example-

Geordie-"As a player PoorOldSpike is top notch and his file return rate is amongst the best I have encountered..his tactics work"
RocketMan- "PoorOldSpike is great guy to learn tank tactics from with unsurpassed file return times (it is like he is an AI or something with him returning files any time of the night or day).."
MrBadExample- "PoorOldSpike is excellent player, good guy and gets turns out quickly! Highly reccomended .."
Enigma- "PoorOldSpike is great tactician, superb return rate and some really wonderful comments with them too. Well worth playing! Also a great teacher in helping you improve your own gameplay and a really good playtester.."
Fredrock- "PoorOldSpike is interesting Tactician, good player, really great Banter,got to give him a try.."
British Tommy- "PoorOldSpike is a very good tactical player, well worth playing against.."
Von Schwendemann- "POS is excellent player with an outstanding return rate. If you haven't played him you are missing out. Have fun.."
Neubaufahrzeug- "Played twice in mirror mode against POS and he is a fair, competent and fast player. CM is living with such guys - let´s go on.."
Koen- "PoorOldSpike I must admit that you're crazy and something's totally wrong with you but you're honest and straight to the point.."
Mad Russian- "To give PoorOldSpike his due, he does good CM support..Every single time POS and I have played each other he is a grand opponent. His return rate is better than ANYBODY's!!!"
Hirr Leto- "PoorOldSpike is deranged, crazy and full of fight. POS will bring it like no other... tactically sound player and yes, 'very' dangerous with armour.”
Lord Bane- "I have been playing the game for many years but have learned quite a bit about the game by reading your fantastic tactical posts."
The Coil- "PoorOldSpike is great when you want to understand the mechanics of the game"
Wardog- "PoorOldSpike you are a tough little f*cker to play against"
Gunter Severloh- "Spike is the best of us I would think for screenshots"

My unearthly knowledge of games also naturally helps me do well in PBEM and multiplayer, here's a 6-inch trophy I won for topping the Rugged Defense club ladder, it sits on a small table in my living room carefully positioned so that it's the first thing guests see when they arrive, and I can keep steering the conversation towards it during the course of the evening..

[AnalogGamer]

If I had the time, this the kind of thing I would do. Used to in Combat Mission. I also used pen and paper, as I never had a use for spreadsheets as a house painter. There are easy to use free ones out there now for those of us less digitally-inclined.

I would like to offer some advice and a request or two for further data, if you have the inclination.

Run an even number of at least 20 tests with each individual data point. 50 is better but life intrudes.

If you could, please run tests on each variation. 1-star/no over, 1-star/11 strength, and so on. No need to add heroes, just the raw stats on experienced and over-strength would be terrific.

The reason is I would like to see who scales up the best for the most effect, and where any "sweet spots" are for effect-v-cost.

Ammo, mobility, and a host of other factors go in to each unit choice. There is no "best" unit, only ones that are better or worse at certain tasks and situations. But more data is always a good thing.

Thanks for your efforts.

[stormbringer3]

I'd love some info about this also. There are a lot of experienced players here who could give opinions about the "sweet spots." For example, a lower cost unit like infantry might be worth going to 12-13 or more but a high cost tank not as high.
Thanks for any opinions.

[hurly]

said yadda yadda yadda

I let your words speak for yourself

to find out its hidden depths, then pass on that sublime knowledge to my disciples in sensational articles, tutorials and scriptures

My unearthly knowledge of games

to help people get the best out of the game

well something like ?

there's a standalone 'Moscow 41' scenario which I'm familiar with but I think AnalogGamer is talking about the one in the Campaign which has different units in it, but I'm not familiar with that one so I can't help him.

Why can't my flak unit fire at this P-51?
I'm sure there's a simple explanation but I'm racking my brain trying to think what it could be!
I've selected the flak and moved the cursor over the P-51, but if I now click to fire, it won't fire, it'll just move under the plane despite it having a firing range of 2, so what's going on?

is a little bit confusing

you may or you may not have unearthly knowledge of things

but I highly doubt your knowledge of Panzer Corps justifies your really really confident behaviour

There are some very good and fine Players here in the Forum that are true masters in the game and have proven it time and time again but none of them sports an attitude like you do

Maybe it's time to do your thing, which certainly has a place here if someone burns his time to do it and it is also appreciated by the community.
But a more back to earth approach in your tone would be appreciated even more.

[PoorOldSpike]

..If you could, please run tests on each variation. 1-star/no over, 1-star/11 strength, and so on..

I'm not quite sure what you mean exactly mate, so if you care to explain a bit more I'll get on it.
Meanwhile here are a few basic tests I just ran, I lined up 12x 1943 Wehrmacht infantry attackers against 12 x 1943 US Infantry.
I gave 6 of the Germans experience levels from 0 to 5, and gave the other 6 Germans strength sizes from 10 to 15.
The US are standard 0-experience/ 10-strength units.
I then fired each German separately and the results indicate that experience is worth more than unit size.
Example- the 15-strength German (0 exp) predicted results are 3/5, but the smaller 10-strength German (5 exp) prediction is 1/6.
Conclusion- maths is not my strong point but the test seems to suggest that a small highly-experienced unit has better combat power than a large-sized inexperienced mob, unless I've overlooked anything?

[PoorOldSpike]

..a more back to earth approach in your tone would be appreciated even more.

We holy men don't do "back to earth" mate..
Jesus said- "I know where I came from and where I am going, but you have no idea where I come from or where I am going....you are of this world, I am not of this world...though you do not believe me, believe the miracles...I'll tell you things hidden since the creation of the world" (John 8:14/ 8:23/10:38/Matt 13:35)

You'll have to send out a probe to find us..

STARGATE-THE MOVIE
[The probe continues to move up the ramp, then appears to be pulled through the Stargate.]
SHORE -"The thing has locked itself onto POS somewhere in the Kaliam galaxy."
DANIEL- "Where are we on that map?"
[Daniel moves to the far end of the star chart, then follows to where the trajectory guide is now sitting.]
CATHERINE- "That's right, Jackson. POS is on the other side of the known universe"

[turn4441]

If you review the combat logs (pre- or post-combat), you'll easily see why experience is much more important. Each level of experience (star) gives a unit a +1 modifier to each of its Initiative, Attack, and Defense values whereas having more strength simply gives the unit additional dice rolls and, therefore, a slightly better chance of getting a kill or suppression, or enables them to kill/suppress more of the defenders units should they get the proper rolls, for example, a 5-strength unit could only inflict 5 kills, even if it got all killing-value rolls whereas a 12-strength unit could inflict 12 if the enemy has that much strength.

Bottom line: Experience alters the combat odds (kill/miss) more in your favor as experience goes up, .e.g., a 0 exp unit might have a kill percentage in a certain combat of 50% whereas with a 501 experience level it might be 66%
Strength does not change the combat odds, it only gives you more chances to achieve those odds, e.g., a 10 strength unit would have 10 rolls at 50/50 each whereas a 15 strength unit would have 15 rolls at 50/50 each.

[AnalogGamer]

Thanks. Those kind of tests that you just showed are exactly how I ran my Combat Mission tests. Line up a bunch and test 'em one at a time.

What I mean by my earlier request re: differing exp/strength is to test a 1-star/10-strength AAA against a plane, then a 1-star/11-strength, and so on up to 5-star/15-strength.

What is the difference between a 5-star/10-strength and a 2-star/12-strength? And all the other strange permutations therein.

This could find where any falloff may be. Example - If a 14-strength vaporizes most targets, then a 15 may never be needed.

It could also tell me where my over-strength should NOT fall below for best effectiveness.

If you would like to personalize the first data set, then use my two faves... the 7/1 and towed 88mm. They are the only AAA I use anyway, unless your tests tell me otherwise... hehehe

Thanks again.


EDIT - Turn's above post just sank in. Thanks for the info. One increases chance of damage, and the other increases number of rolls per-combat. They made it extra fuzzy.

[captainjack]

The link below shows the percentage kill/suppress/miss for direct attack (tank vs tank) and indirect (artillery and strategic bombers).

http://slitherine.com/forum/viewtopic.p ... lit=combat

Assuming you know the experience bonus for the units you are considering (I think it is the file called experience - whatever, but take care not to fiddle with it if you open it), and the Rate of Fire (from the equipment file), you can get a very reliable comparison of effectiveness. Be warned that this can be quite time consuming. It is teh kind of analysis where a spreadsheet could be useful to test out several variations of experience and units to see whether it's easy (1* vs 1* is the same as 2* vs 2*) or complicated (4* vs 4* iis much better than 1* vs 1*. More complex variations come from the different way experience affects different types of units (eg AA attack +2/star, but fighter defence +1/star).

This link
http://slitherine.com/forum/viewtopic.p ... ge#p537053
is also very informative.

The key point is that when attacking, it's the actual difference in attack and defence that matters, not the relative difference - 5A vs 3D and 25A vs 23D are both +2 to the attacker.

[PoorOldSpike]

If you would like to personalize the first data set, then use my two faves... the 7/1 and towed 88mm. They are the only AAA I use anyway, unless your tests tell me otherwise... hehehe
Thanks again..

Okay I just conducted the following test to find out how experience level affects the 7/1, I'll do more tests later to analyse unit size effects, and also test the 88 Flak, a mystery surfaced which I can't explain so maybe you hotshots can have a crack at it?..

Top pic is the test setup, 6x 7/1's versus 6x B-25B's, all units on both sides are size 10, vanilla Pz Corps, Colonel level, no leaders.
The planes are all experience '0', but I gave the 7/1's experience levels 0/1/2/3/4/5 (note their stars in the screenshot).
I then cursored over the planes to bring up the crosshairs readout (including using ctrl-leftclick detailed prediction readout), taking care to stagger the attacks to avoid any 'mass attack' bonuses.
Note the '0' experience flak AND the '1' experience flak both surprisingly had the same 2 kills/2 suppressions readout!
Checking the detailed readout I saw that the '0' exp flak was Miss 78%/ Suppress 7%/ Kill 15%, but the figures for the '1' exp flak were Miss 74%/ Suppress 7%/ Kill 19%.
So, although '1' exp boosted the miss/sup/kill figures, the boost is so slight as not to boost the actual 2/2 readout, and we can therefore conclude that the '1' exp level unit is only very slightly better than the '0' unit for gameplay purposes.

Bottom two pics- these are just mostly for eye candy, the first is the '0' exp unit in action, and the second is the '5' exp unit at work. As we'd expect, the '5' makes mincemeat of the plane squadron, killing 6 of them and suppressing the remaining 4.
The mystery I mentioned is this- why does the '5' exp units predicted readout say "Air attack Experience bonus +10"?
It's a 5-star unit so where does the +10 come from?
And to further confuse the issue, why is its "Air Defense Experience bonus" only +3 ? Why not +5 to match its 5 stars?
(Same thing occured with the other units that had exp levels 1/2/3/4, their "Air Defense Experience" bonus was always a bit higher than the amount of stars they'd got, I dunno why)

Conclusion- just as we'd expect, experience boosts killing power, so the test is mainly of interest just to see basically how much hurt a 7/1 can inflict on the great iron birds on paper, for example a standard 10-strength 5-exp 7/1 can carve up an entire 10-plane squadron real bad with kills and suppressions, although of course during actual gameplay the results will no doubt fluctuate up or down as various factors come into play, plus Lady Luck will be getting in there too..

[dalfrede]

FYI

[PoorOldSpike]

Thanks Dalfrede, if there are already charts like that around that cover everything about PzCps I needn't bother making a fool of myself by doing any more tests..

[AnalogGamer]

Thanks for the work Spike!

The 7/1 info you provided is just the sort of thing I wanted. Not being much a data-miner of having the time to do so, every little bit helps.

Thanks to dalfrede as well for pointing out another resource.

[huckc]

FYI

Context and credit, rather than just the raw chart:

http://steamcommunity.com/sharedfiles/f ... =305486619

It saddens me how people can steal others' work while at the same time making it worse.

[PoorOldSpike]

Incidentally, all my tests, tutorial articles and screenshots etc are 100% my own work, I've been posting them in various game forums for the past 16 years and if anybody wants to copy and paste them I don't mind in the least, as I'm only too glad to help the communities.
All I ask is that if people are going to put them into a book they're writing, I hope they'll please cut me in for a share of the royalties..

[sorenthewild]

FYI

Context and credit, rather than just the raw chart:

http://steamcommunity.com/sharedfiles/f ... =305486619

It saddens me how people can steal others' work while at the same time making it worse.

How do you know that Delta 66 and dalfrede are not the same person?
Or that Delta 66 is the original author?