Comprehensive WitP ENG unit testing...

[Apollo11]

Hi all,

I finally managed to conduct my comprehensive ENG unit tests in WitP (I was hampered with but that in small scenarios you have to have large large number of LCUs in order WitP not to freeze after consecutive turn executions)!


Leo's WitP ENG unit testing


Testing finds

#1
According to WitP manual 1 ENG vehicle has value of 5 ENG squads.


#2
EXP and commander of ENG unit do not have any effect on construction speed.


#3
Fatigue itself didn't slow down construction speed (75% fatigued unit was as fast as 0% fatigued unit) but it produced disruption (because tired unit had to work) and disruption slowed construction!

For example if disruption was 25% then it, roughly, slowed construction time by 25% (i.e. unit was only 75% effective)!


#4
Presence or not presence of HQ (of any kind) does not have any effect on construction speed.


#5
ENG unit construction capabilities add up (in other words 2 same ENG units would build 2x faster and 4 same ENG units would build 4x faster).


#6
Construction cost in time of Airbase or Port object is same (i.e. it doesn't matter if you build one or another - if you build both your construction speed is 50%).


#7
When constructing Airbase and/or Port object (if their intended build values are below or equal to SPSs) the cost in time is as follows:

Size 0->1 = each ENG squad constructs 0.002500 (1/400) of object
Size 1->2 = each ENG squad constructs 0.001667 (1/600) of object
Size 2->3 = each ENG squad constructs 0.001250 (1/800) of object
Size 3->4 = each ENG squad constructs 0.001000 (1/1000) of object
Size 4->5 = each ENG squad constructs 0.000833 (1/1200) of object
Size 5->6 = each ENG squad constructs 0.000714 (1/1400) of object
Size 6->7 = each ENG squad constructs 0.000625 (1/1600) of object
Size 7->8 = each ENG squad constructs 0.000555 (1/1800) of object
Size 8->9 = each ENG squad constructs 0.000500 (1/2000) of object

NOTE: I am 100% sure in values above!


#7
When constructing Airbase and/or Port object (and if their intended build values are above SPSs) I can't calculate the cost in time.

It is because there appear to be "diminishing returns"!!!

In other words you can put as many ENG squads you wish but construction speed can't be accelerated further after some number of MAX ENG squads (this is 100% different that it was when you build below SPS limitations)!

For the time being I can't determine/calculate/estimate those modifiers but it appears that MAX ENG squads allowed number is determined by overall size of SPS value!

For example if you have beach (i.e. "dot") HEX with SPS (0 - 0) and attempt to construct base there all ENG squads above 200 don't count at all (i.e. equivalent of 200 ENG squads will build that place at same speed as equivalent of 1000 ENG squads)!!!

Perhaps developers can help here - I am kindly asking for their help in this matter!



Discussion

IMHO the construction speed appears to be too fast even if just single big ENG unit is building at one place.

Having several big ENG units at one place throws things out of whack 100%.


For example single US EAB unit has:

30 ENG squads
20 ENG Vehicles (20 x 5 = 100 ENG squads equivalents)

Or combined 130 ENG squads equivalents.

This means that such EAB ENG unit can construct imaginary HEX that starts with 0 - 0 (i.e. it is "dot") but which has SPS (9 - 9) to 9 Airbase and 9 Port in:

Size 0->1 : 1 / (1/400 x 130) x 2 = 6 days for 1 Airbase - 1 Port
Size 1->2 : 1 / (1/600x 130) x 2 = 9 days for 2 Airbase - 2 Port
Size 2->3 : 1 / (1/800x 130) x 2 = 12 days for 3 Airbase - 3 Port
Size 3->4 : 1 / (1/1000x 130) x 2 = 15 days for 4 Airbase - 4 Port
Size 4->5 : 1 / (1/1200x 130) x 2 = 18 days for 5 Airbase - 5 Port
Size 5->6 : 1 / (1/1400x 130) x 2 = 21 days for 6 Airbase - 6 Port
Size 6->7 : 1 / (1/1600x 130) x 2 = 24 days for 7 Airbase - 7 Port
Size 7->8 : 1 / (1/1800x 130) x 2 = 27 days for 8 Airbase - 8 Port
Size 8->9 : 1 / (1/2000x 130) x 2 = 30 days for 9 Airbase - 9 Port

Or combined 6 + 9 + 12 + 15 + 18 + 21 + 24 + 27 + 30 = 162 days (give or take few).

In just 5 months the biggest possible base that can be build (9 - 9) is constructed.

Is this OK?

IMHO no.

It is accelerated several times.

Now imagine 5 EAB units working together. Since they all add up their ENG squads they would have equivalent of 650 ENG squads (650 = 5 x 130) and this would mean that from 0 - 0 the "dot" can be build to 9 - 9 (in our example) in just 30+ days (or just single month)!!!



Conclusion

IMHO the developers (Matrix/2By3) should consider two things:

#1
Slow down the construction speed overall.

#2
Find a way to introduce "diminishing returns" when massing ENG units in one HEX (when building below SPS - apparently something already exists in code when building above SPS).


IMHO the base construction was one of the mayor obstacles in historic War in the Pacific and the whole war was, in fact, waged around bases that had to be constructed slowly because nothing could have been build "overnight"!

Therefore those two measures I suggested above would slow down base construction to acceptable and historic values!


BTW, I myself and several other (most prominent of them is, of course, "Nikademus") brought this up before but this time I added empirical data (from testing) into equation...



What do you think gentleman?

Matrix/2By3?



Leo "Apollo11"

[Herrbear]

Thank you for all your work in doing this. I found this very interesting. You have brought up some valid points for discussion.

[dtravel]

Devil's Advocate points:

1) What is the largest SPS of an undeveloped (dot) base?

2) How often do people actually have that many engineers (5 EAB or equivalents) all working in one base?



Otherwise, good info to know. Thank you.

[Montbrun]

Very nice data. The one restricting factor that I don't see, though, is supply levels.

Brad

[Apollo11]

Hi all,

ORIGINAL: Brad Hunter

Very nice data. The one restricting factor that I don't see, though, is supply levels.

Brad

Supply is, unfortunately, almost irrelevant... [:(]


In one of my tests (using WitP editor and my Template empty scenario) I created the following setup:

There was "dot" HEX with SPS 6 - 6.

I placed 50000 supplies there and one single EAB.

Both Airbase and Ports were to be constructed.

In just short 15 days the both Airbase and Port was 2 - 2 and spoilage was eliminated from equation.

From original 50000 supply there was left 35000 but it was more than enough to build that test HEX to MAX possible value because from that point onwards the supply demand was very small (to my huge surprise)...

BTW, after months of work the lone EAB was still 100% effective (1 squad here and there was disabling and enabling during that time but at the end all squads were 100%). Also fatigue was only 5% after months of hard work (and no rest).


Leo "Apollo11!

[Apollo11]

Hi all,

ORIGINAL: dtravel

Devil's Advocate points:

1) What is the largest SPS of an undeveloped (dot) base?

2) How often do people actually have that many engineers (5 EAB or equivalents) all working in one base?

Otherwise, good info to know. Thank you.

I am not worried at all about such extreme cases I described in my test (it was hypothetical HEX after all) - what I am worried are the possible B-29 bases that can be build almost "overnight".

Those HEXes are "middle" developed by Japanese but once in US hands the B-29 base can be established using ENG massing (or "hording" if you wish) in just few short days.

As for ENG massing - I am 100% sure it is tactics that many players use and will most certainly use (US player for their B-29 base for example).


Leo "Apollo11"

[Apollo11]

Hi all,

ORIGINAL: Herrbear

Thank you for all your work in doing this. I found this very interesting. You have brought up some valid points for discussion.

Thanks!

BTW, this is my "pet project" for long long time (I first brought this up almost 18 monts ago but those threads were lost when hacker attacked Matrix forum)...


Leo "Apollo11"

[mogami]

Hi, Thanks Leo you clearly have spent a lot of time working all this out.

The important data that I gather is 200 engineer sqd (or equal in vichecles is all a player needs to send to a base over SPS.

It does not really matter about base over SPS because a 0-0 can only be built to a size 3-3 and that is a very minor base. As long as making a major base over SPS takes more time. (turning a 6 into a 9) (and no impact from more then 200 engineer Sqd)

As far as fatigue if the unit is in a non malaria hex and has that great surplus of supply it will not get tired. You can't tell a unit to work overtime. The engineers do not work around the clock max efforts are when you place many engineer units in same hex and now we know more then 200 is a waste once a base reaches its normal max size)

It is disruption not fatigue that impacts combat. The effect of high fatigue is a unit will suffer increased disruption if exposed to attack or takes part in combat. Extra supply use will descrease fatigue and that will help lower the disruption. (parts become disabled from disruption not fatigue. )

What numbers would you expect to see compared to what you found in your testing?

A size 1 airfield is a length of field that has been smoothed out and objects removed
A size 2 airfield is the same only longer
A size 3 is the same only now support buildings exist
A size 4 has been given a hard surface
(also every increase includes adding additional strips)
A size 5 has long hard strips and many support buildings and aircraft parking/arming/hangers

At size 5 the number of aircraft reaches 250. 250 aviation points is the most ever required becasue the airfields larger then size 5 (6,7,8,9,10) have machine/repair shops that increase the effectivness of the aviation support troops. (There appear to be something along those lines with engineer troops maxing out effect at 200.) having 1000 aviation support points also provides no more effect then having 250 other then making it hard for the enemy to reduce the number of support troops below where they are at max effect)

[Mr.Frag]

You also might want to toss in the duration to take a size 3 (4) to a size 7 (4) airbase considering that *is* your complaint afterall since the B-29's will be flying out of Saipan, Tinian and Guam.

As you well know, the build time of a 0 to 9 base is not relevant to any discussions as there *are* no size 9 bases that start as a size 0. Don't exaggerate reality to improve your case, just report the facts. Let everyone draw their own conclusions based on those facts. It's far more helpful. [;)]

Your baseline numbers are great but the overbuild numbers are far more important as those are the ones that take till the cows come home.

[mogami]

Hi, Japanese players don't worry about how long it takes for the Allies to build Saipan, Tinian and Guam (at least they shouldn't) because part of the defense of this area is Japan's having already built them to max size and stocking them with max airgroups and aviation troops. If the Allies capture the airfields they will already be built (but hopefully with massive damage when Allies get them)

If Saipan, Tinain, and Guam were captured at size 0(4) 0(4) it would not be important how fast Allied engineers could build them to size 7 unless at the date of capture the Allies had all their B-29 groups built and ready to move there. If the Allies capture these bases before they have B-29 groups then build time is not important. If they capture them with B-29 unbuilt then it is a short delay. If they capture them built to max with B-29 then so what (are you seeing the important issue is not how long it take Allies to build these bases but if they capture them at all)

I don't think we can compute to the day how long it should take X number engineers to build any facility.
Leo did you notice any effect on speed based on the terrian in the hex? I would think it faster to build from 0 to 3 in clear compared to building from 0 to 3 in mountain hex.

[Apollo11]

Hi all,

ORIGINAL: Mogami

Hi, Thanks Leo you clearly have spent a lot of time working all this out.

The important data that I gather is 200 engineer sqd (or equal in vichecles is all a player needs to send to a base over SPS.

It does not really matter about base over SPS because a 0-0 can only be built to a size 3-3 and that is a very minor base. As long as making a major base over SPS takes more time. (turning a 6 into a 9) (and no impact from more then 200 engineer Sqd)

As far as fatigue if the unit is in a non malaria hex and has that great surplus of supply it will not get tired. You can't tell a unit to work overtime. The engineers do not work around the clock max efforts are when you place many engineer units in same hex and now we know more then 200 is a waste once a base reaches its normal max size)

It is disruption not fatigue that impacts combat. The effect of high fatigue is a unit will suffer increased disruption if exposed to attack or takes part in combat. Extra supply use will descrease fatigue and that will help lower the disruption. (parts become disabled from disruption not fatigue. )

What numbers would you expect to see compared to what you found in your testing?

A size 1 airfield is a length of field that has been smoothed out and objects removed
A size 2 airfield is the same only longer
A size 3 is the same only now support buildings exist
A size 4 has been given a hard surface
(also every increase includes adding additional strips)
A size 5 has long hard strips and many support buildings and aircraft parking/arming/hangers

At size 5 the number of aircraft reaches 250. 250 aviation points is the most ever required becasue the airfields larger then size 5 (6,7,8,9,10) have machine/repair shops that increase the effectivness of the aviation support troops. (There appear to be something along those lines with engineer troops maxing out effect at 200.) having 1000 aviation support points also provides no more effect then having 250 other then making it hard for the enemy to reduce the number of support troops below where they are at max effect)

Thanks Russell!


What I will now test is Tinian scenario.

Tinian is SPS 1 (Port) - 4 (Airbase) and has Japanese buildup 1 - 3.

Historically there were two USN engineer brigades (equaling 8 to 12 Seabee battalions).

They build the Tinian up to MAX in 1 year (12 months: Aug 1944 - Aug 1945).

Since MAX in our WitP game terms means that the Tinian was 4 - 7 I will now test how long it takes to do that.

I will use 10 Seabee battalions for this test.

BTW, 1 Seabee in WitP is 30 ENG squads + 30 ENG vehicles equaling 180 ENG squads altogether.

10 Battalions thus are 1800 ENG squads.



Leo "Apollo11"

[Mr.Frag]

Japan's having already built them to max size and stocking them with max airgroups and aviation troops.

Thats your way ... I'm a scorched earth type ... give nothing of value [;)]

[mogami]

Hi Leo thats a good test to conduct however you do understand the actual building of Tinian by the USA used 15,000 Engineer troops. Did not use the existing Japanese airfields and had B-29's operating less then 4 months after Tinian was captured. (Not bomb groups but there were B-29 flying missions from Tinian before the airfields even had hard surfaces! The 6th Bomb Group moved there for bombing operations in Jan 1945.
So you might say the Seebees did in fact build a size 7 airfield from scratch. They had a size 4 in less then 4 months (B-29 in WITP can fly from a size 4 airfield but only use extended loads and only fly normal ranges)

The airfields the Japanese had built and had used were smaller then those the USA had open in under 4 months. The USA did not use any of the Japanese runways but did use a number of existing buildings.

[Apollo11]

Hi all,

ORIGINAL: Mr.Frag

You also might want to toss in the duration to take a size 3 (4) to a size 7 (4) airbase considering that *is* your complaint afterall since the B-29's will be flying out of Saipan, Tinian and Guam.

Raymond this is exactly what I am doing right now (please look at my answer to Russell)!

As you well know, the build time of a 0 to 9 base is not relevant to any discussions as there *are* no size 9 bases that start as a size 0. Don't exaggerate reality to improve your case, just report the facts. Let everyone draw their own conclusions based on those facts. It's far more helpful. [;)]

True.... but those values are also nice to know (i.e. how long it takes to build 0 - 0 "dot" with SPS 9 - 9 into 9 - 9) wouldn't you say... [;)]

Your baseline numbers are great but the overbuild numbers are far more important as those are the ones that take till the cows come home.

I was unable to determine the "diminishing returns" (if they are that at all) formulas so far.

When base with SPS 0 - 0 was a "dot" (0 - 0) and I expanded it using equivalent of 100 ENG squads I got same time when using 1000. The limiting ENG number was, apparently 200 ENG squads equivalents.

Also the build times were very long and I am (as yet) still unable to calculate building speed.

My best guess so far is that each ENG squad equivalent builds 1/8000 in this case.

But when that same HEX is already at 1 - 1 things start to complicate and now upgrading that to 2 - 2 proved to be very strange because now 100 ENG squad equivalents seemed to be limit (i.e. 100 ENG squad equivalents at build were same as 1000 ENG squad equivalents).

Must test this some more...


Leo "Apollo11"

[Apollo11]

Hi all,

ORIGINAL: Mogami

Hi Leo thats a good test to conduct however you do understand the actual building of Tinian by the USA used 15,000 Engineer troops. Did not use the existing Japanese airfields and had B-29's operating less then 4 months after Tinian was captured. (Not bomb groups but there were B-29 flying missions from Tinian before the airfields even had hard surfaces! The 6th Bomb Group moved there for bombing operations in Jan 1945.
So you might say the Seebees did in fact build a size 7 airfield from scratch. They had a size 4 in less then 4 months (B-29 in WITP can fly from a size 4 airfield but only use extended loads and only fly normal ranges)

The airfields the Japanese had built and had used were smaller then those the USA had open in under 4 months. The USA did not use any of the Japanese runways but did use a number of existing buildings.

Russell RGR that!


As I initially wrote I will post all landmark times and I think 10 SeaBee WitP units would suffice

Since each WitP SeaBee is 30 ENG squads + 30 ENG vehicles so this equals 180 ENG squad equivalents x10 = 1800 ENG squad equivalents I think we can agree that this is 15000 people you mention in your message.


Tinian is SPS 1 (Port) - 4 (Airbase) and has Japanese buildup 1 - 3.

I will build it from there to 4 - 7 (i.e. MAX that it can reach).


Leo "Apollo11"

[Mr.Frag]

Raymond this is exactly what I am doing right now (please look at my answer to Russell)!

I was posting my reply while you & he replied. Phone rang and distracted me [;)]

An SPS 0 is a special location ... think it is a 10x normal speed.

sps 1 past should take 2x
sps 2 past should take 4x
sps 3 past should take 8x

Something like that ...

[mogami]

Hi, And to compare what the USA actually did run a second test after editing the airfield back to 0 (4)

[Mr.Frag]

And while you are at it, test with a malaria location like PM and see what the burnout rate is on them poor shovel pushers [:D]

[Apollo11]

Hi all,

Like I promised here is Tinian test scenario.


Unfortunately I think I have bad news and it seems that base construction is extremely accelerated...


Tinian is SPS 1 (Port) - 4 (Airbase) and has Japanese buildup 1 - 3.

I build it from there to 4 - 7 (i.e. MAX that it can reach).


I used 10 SeaBee battalions.

Since 1 SeaBee in WitP is 30 ENG squads + 30 ENG vehicles equaling 180 ENG squads equivalents altogether it means that 10 Battalions thus are 1800 ENG squads equivalents.


I build both Port and Airbase in parallel.


Port size:

0 turns = Port 1
4 turns = Port 2
13 turns = Port 3
24 turns = Port 4 (MAX reached - expansion stopped)


Airbase size:

0 turns = Airbase 3
3 turns = Airbase 4
10 turns = Airbase 5
25 turns = Airbase 6
34 turns = Airbase 7 (MAX reached - expansion stopped)


As you can see it is _ENORMOUS_ acceleration...


Leo "Apollo11"

[Apollo11]

Hi all,

ORIGINAL: Mr.Frag

And while you are at it, test with a malaria location like PM and see what the burnout rate is on them poor shovel pushers [:D]

Raymond, with construction speeds like we currently have in WitP malaria is just too slow... if you stack enough ENG units you can do "overnight wonders"...


Leo "Apollo11"