Norm Koger's responses to older questions

[Don Maddox]

Since we have no idea how long the old Talonsoft forum migh be around I've decided to post some of Norm's responses to older questions. These might come in handy at some point.



Hmmm...

...equipment failure, especially in tanks and more specifically the German tanks (WWII) resulted in serious losses.

The game does have a small movement attrition effect which does not simulate the effects of weather or terrain, or as you say, mud!. Here's the line that calculates the breakdown rate:

breakdown=(100*(8+rnd(0,4))*the_cost)/(max*(units[the_unit].proficiency+100));

(A couple of "defined constants" have been replaced here by their numerical equivalents to improve clarity.)

"the_cost" is the movement point cost of the move. This calculation is performed for each hex moved.

"max" is the unit's original movement allowance for the turn.

So the breakdown rate is dependent upon the cost of the move relative to the unit's original allowance, and the unit's proficiency. Several things to note here: The cost of a move is dependent upon terrain, and includes the effect of mud, snow, etc. So the rate does increase when things get muddy or snowy. Units that are in bad shape to begin the turn will have a lower initial movement allowance, and will experience higher breakdown rates for any particular move. High proficiency units will lose less of their equipment - a kind of "first order" attempt to model likely improved recovery capabilities associated with higher quality units.

Finally, routed and retreated units suffer twice the breakdown rate in any given move. This is in addition to the losses they suffer as a result of the initial disengagement and retreat.

So a patch that included setting a forces attrition %age would be very welcome.

Not a bad idea. It's on the tinker list. :cool:

[Don Maddox]

Brian Rundgren -- Have you given any consideration to matching the combat simulation of TOAW to a strategic/resource management engine?

I have. Actually, the "strategic/resource management engine" is the simpler side of what would have to be done. There is also the issue of handling force evolution over longer periods - and that one is trickier. The current system can deal with limited equipment transitions (as in the Korea 50-51 scenario) and some cases of formation reorganization. But for a really good "world war" style capability, it would also have to be able to deal with: arbitrary formation creation and swapping, unit renaming and re-iconing and general reorganization... Entirely aside of interface questions (which are substantial), the PO development work would be pretty darned hairy.

It's _MUCH_ easier to do this kind of thing in a single scenario game than in something like TOAW. You don't have to worry about serious, relatively easy to use editor capability or a lot of unforeseen combinations. I may very well tackle the European (Britain-Urals, Norway-Africa) or Pacific theaters of WWII at some point in the future. But these games won't have nearly the editability that TOAW has.

[Don Maddox]

By the way, this tinker list. We now know at least 1 subject. Any chance of revealing others?


It's a huge list, easily enough to keep me busy for another year or so. I'm not sure it would do any good to make it public though, as there is very little chance that anyone could figure out a way to keep a roof over my kids' heads during the process of implementing it. The game industry really isn't set up to support that kind of substantial post-release development. I wish it were. Tinkering is fun. The only way folks have seemed to manage to get this kind of thing to work for any length of time is to release well implemented single scenarios as standalone games using the same game engine with minor improvements from one "game" to the next. And these days even that model only really works by direct marketing.

There really are no plans for immediate sequels or reworks for TOAW. I'm fully committed for the next year or so.

On the other hand, one of my commitments is flexible. And I'm starting to think seriously about one of the BIG items on the TOAW tinker list for inclusion as part of another project. Remember TANKS? Might be interesting if a revisited interruptible real time version used the same equipment database as TOAW - CW, and could import / export battles and results...

[Don Maddox]

You definitely can prioritize in WGOTY/ACOW

Ralph Thunderpuppy -- I haven't checked to see if this feature WORKS, but if it does, it should be very useful. On the one hand, you can set certain specialized units to get few or no replacements (paratroops, any units that can't draw on the general pool of manpower). On the other hand, you can also simulate units going into battle before everyone's shown up: just start them with, say, 50% of their authorized equipment, then give the unit a very high replacement priority so that it will fill up quickly.

The feature is very heavily used in my Middle East 73 and Middle East 2000 scenarios. Check the replacement priorities in the editor.

On the other hand, you can also simulate units going into battle before everyone's shown up: just start them with, say, 50% of their authorized equipment, then give the unit a very high replacement priority so that it will fill up quickly.

That's something else that's done in Middle East 73, where I've also added a twist. There is an entire formation of Israeli units (the "Mobilization" formation) that never show up on the map. I use these units to add tailored brigade size "chunks" of equipment into the replacement pool at specific points in the scenario. They "appear" only to be immediately destroyed and their assets redistributed to originally badly understrength units using the event engine.

If two units are equally understrength, but one is in full supply and the other is at 1/4 supply do they both get the same number of replacements?

Yes, on average - assuming that neither is embarked. Units embarked on ships or trains are not eligible for replacements.

How about if one unit is 5% understrength and the other is 50%?

The replacement routine receives all requests for replacements from eligible units, then calculates a chance that any particular item of equipment will be replaced. For each item requested by each unit, that chance is modified by the individual requesting unit's replacement priority (assigned by the scenario designer) and tested. If the test passes, the item will be received. If it fails, the item will not be received. So eligible units with equal replacement priorities will tend to receive equal _proportions_ of their requests. Ex: If there are enough Shermans in the pool to fill 30% of replacement requests, all requesting eligible units will, on average, receive roughly 30% what they need to get back to 100% authorized strength in Shermans.

The unit checks start in a randomly determined location in the unit list on each turn. This keeps low volume items from being sucked up by the same units by virtue of their position in the unit data structure turn after turn.

Sometimes it seems as if there are replacements available in the pool even though there are still understrength units out there. Why is this?

The per-item request filling scheme, modified by unit replacement priorities, introduces a certain amount of inefficiency into distribution. Feels about right...

[Don Maddox]

How does the "minimize" "limit" and "ignore" statis works in an attack, is it for each unit or the entire group...

Loss tolerance sets the limit at which the unit will consider retreating from a defense or breaking off from an attack. It is a per-unit setting, although it is frequently set for an entire group of units.

All else being equal in a given battle, units will take fewer losses at lower loss tolerance settings. But in some cases a low loss tolerance can lead to individual units pulling out of combat and leaving others to face the music on the next round of combat. This can actually lead for higher overall losses because some units pulled out to minimize their own losses. Unless you're trying to do something clever, it's generally a good idea to use the "limit losses" setting for large groups. Trying to tailor loss tolerance to a situation works best with small numbers of units.

[Don Maddox]

...still trying to figure out why some units can't be broken down and other can...

There are two reasons why units that should be able to break down wouldn't be able to:

There are already as many units in the unit list as allowed by the game, and new units can't be created because there isn't room for them.

The formation to which the units belong is maxed out - no room left in the unit list. In some scenarios this also implies problems with basic scenario design, as the PO is optimized to handle smaller numbers of units in individual formations. Packing formations full of units is a good way to force the PO to concentrate in small areas. This is great for high density scenarios or earlier periods, but it has a tendency to give the PO tunnel vision (ignoring flanks, etc.) if there aren't enough units in other formations to cover. General rule of thumb for WWII and later: try to limit typical formations to no more than 9 "line" combat units with the remainder reserved for support. If you check my scenarios (Korea 50-51, Nomonhan 39, Middle East 73 and 2000) you will see that the formations are relatively small in most cases.

[Don Maddox]

Airborne tips

Do you have any tips to get the PO to deploy large numbers of airborne troops? I had some trouble in "Invasion Zealand" for TOAW II, and I know that the designer of a Taiwan scenario has troubles to. So far we have figured out, that it helps to:

1. Give the paras one single objective - or objective very close to each other.

2. Deploy the paras at an airfield from the beginning of the scenario.

3. To surround the airfields by nonplayable hexes.

Any thoughts?

Make sure the airborne units are supplied. It's been a while since I looked at the airborne override code, but I believe it's set up to be particularly aggressive with airborne units on turn 1 if they are supplied and can't reach their objective the "normal" way. In any case, you could use an event to temporarily set the airborne formation loss tolerance very high for a turn or so - then set it back down again the next turn if you don't want them acting like Klingons when they hit the ground. You might also try using the po_log function to look at the po logic. Finally, it might be worth looking at the way airborne ops are set up on the Normandy scenarios. The situation is similar (forcing the PO to jump into locations where the odds don't look particularly appealing).

[Don Maddox]

Tell me more about the unentrenching effect

If it's temporary and not described in the manual, you're about our only source of information. Is there any way of ascertaining how much it contributed to the success (if any) of an attack?

The data and algorithms are all pretty much hidden unless you run the game with the toaw_log command line parameter. If you do generate game logs, the suppression is noted in the log with a line like "zzz kg/min weight of long range fire suppress entrenchment level in location x,y by pp%."

Essentially, what happens is that the weight of incoming artillery is compared to the size of the target (a strip a few km deep that runs across the target location), and a suppression figure is calculated for that combat only. It can be as high as 100%, in which case the combat proceeds as though the entrenchment level in the location were zero. All artillery, aircraft, etc. in TOAW are rated internally for kilograms per minute on target (that's where anti-personnel strengths come from), and individual shell weight (which is used for entrenchment suppression). Rate of fire will give something like a 122 mm gun about the same anti-personnel capability as a 305mm gun. But for suppressing fortifications, the 305mm piece is more than 15 times as effective due to the massive difference in individual shell weights. If you are wondering what good all those big, low rate of fire (and low apparent strength) artillery units are useful for in some scenarios, just try committing them to normal attacks on heavily fortified locations.

[Don Maddox]

Hi Norm--

I was talking with Erik Rutins via another web forum, and he mentioned that he's working on an updated Korea 99 scenario where the 2nd Division's 3rd Brigade is modeled using the Army's new Initial Brigade Combat Team structure. Since I've taken some time to follow the proceedings of the Army experiment rather closely, I'm concerned that TOAW will be unable to accurately model such a force. Perhaps you have some recommendations?

Now, obviously, since the Army is really still in the early stages of developing what this force will look like, and various platforms needed to complete the force (Crusader, Commanche, Future Combat System, etc.) aren't even in production yet, it's impossible to model the force in its entirety in TOAW. The Inital BCT is planning on using the LAV series (or something similar), but I see that as just a way, for the most part, to facilitate experimentation until something better comes along (the example I used with Erik was Heinz Guderians cardboard panzers. I guess Patton's trucks labeled "TANK" also serve as a comparable example).

That does make it rather difficult.

...in a more conventional situation the BCT is expected to serve almost as a screening force, finding targets that its advanced firepower can kill from long range.

You get a saturation if you're not careful in scenario design, so you have to be careful. But you can partially model this kind of thing by substituting Reconnaissance Assault Teams and AT+ Teams for some of your Assault AT+ Squads. That could substantially improve the unit's recon capability, which would have a noticeable effect on its first round combat performance.

...My concern is that I've never seen a consistent ability in the game for certain platforms (airpower/artillery/helos) to inflict damage on armored, or even vehicular forces...

Actually, airpower and helos can be dramatically effective. I was running some tests a few weeks back where 6 ship A-10 flights were getting 40%+ kill rates per round against T-80's in isolated Russian style tank battalions. In many cases the tank battalions were evaporating after one strike. A lot depends upon scenario scale, the state of all units involved, and the extent to which air defenses disrupt the attacking air units.

You might notice a difference in specific scenarios with the changes to interdiction in v1.03. Under certain circumstances (basically, if the interdicting units aren't dog tired after the bad guys have moved) units will now be targets of "target of opportunity" attacks even if they stay put. You could do a Desert Shield scenario in TOAW now. It would be boring, but it would work.

On the other hand, you have a point about artillery. TOAW doesn't really deal with exotic artillery capabilities - in part because these haven't been extensively tested (the effects are theoretical), and also because I wasn't able to get much info on things like stockpiles of exotic rounds. Mass exotic artillery capabilities haven't really been with us very long, if they even exist now. I could easily be wrong, but my sense is that even toward the end of the time frame TOAW claims to model, the number of exotic artillery attacks you'd see in a campaign would be relatively small due to limits on exotic ammunition supply. It's like the case with some of the uber-at rounds the Russians theoretically have. They developed some pretty good stuff in the late '80's and early '90's, but did they ever get around to making it in quantity?

One thing I'd still like to do in TOAW is an "interdiction" order for artillery. Particularly in the 2.5 and 5km scales, this would really improve artillery modeling. And it could happen, depending upon the results of potential background deals that are beyond my control. As a side effect, naval combat and ship - shore interactions would improve a great deal. We'll just have to see whether the incentive pops up. If it does, modeling of speculative capabilities and equipment for the 2000-2020 period would see considerable attention.

...in the real world if I kill a division's trucks, its tanks will grind to a halt...

Well, that happens in TOAW too, depending upon scenario design - but it takes a while. If transport asset sharing represents a substantial part of your forces' resupply capability (something you will only really notice if you let your units stay in one place for a turn or two) you will see a real hit on supply when you lose a lot of trucks. The effect won't be limited to the losing unit because of the way transport asset sharing works. But the entire force will see an effect if enough trucks are shot up. Of course, that brings up another area that could be improved: modeling of losses to strategic transport capability - but that's another topic :)

[Don Maddox]

Depot simulation using existing capabilities

One thing I've noticed in most scenarios is that the designer will go ahead and place a supply point in the middle of the board, or in a particularly important spot like a city. Inevitably, a situation will pop up where the bad guys will get cornered in such a location, get surrounded, and then NEVER DIE because they've got supply there...

There are so many things in the scenario structure that are almost never used. One of them is the ability to pick up or place supply points by event.

A few months back I was working on a 1904-05 Russo-Japanese War scenario. (The scenario still languishes on my hard drive - no time.) Historically, the Russians stockpiled mountains of supplies in Port Arthur - a fortified location at the end of a rail line on the tip of a peninsula. They started out with minimal supplies there and built them up as quickly as possible through the rail line once the war started. In the end, they built a substantial supply dump that kept local defenders supplied for months - eventually running down to zero on key elements (mainly food). Impossible to model in TOAW? Not really.

Here's how I was going to treat that supply point: 1) The Russian supply point was placed at Port Arthur in the editor. 2) For every turn the Russians controlled certain key points along the rail line, the event engine variable would be incremented by a somewhat randomized number. The points were carefully chosen to catch almost any break in the rail line - particularly near or along likely lines of early Japanese advance. 3) When any of the key rail line points became Japanese controlled, the event engine variable would begin to be decremented rather than incremented. 4) When the event engine variable reached zero value, the supply point in Port Arthur would be removed by an event.

There is no reason other scenario designers can't do similar things. Granted, there is only one variable available so multiple depots would have to be treated more simply than the one in my example. But it would be _easy_ to remove supply points based on a random number check after certain other events occurred - most likely occupation events.

Map design and OOB creation are only part of TOAW scenario design. The single most powerful tool available to TOAW scenario designers is the event engine. It is the tool that has the greatest ability to make different scenarios feel different, rather than generic map and counter exercises.

I've considered adding depots, including tonnage based supply capability. One of the reason I haven't is that scenario designers haven't used most of the features already available to do similar things. This isn't an indictment of scenario designers. Far from it. I am very impressed by some of the scenario design work I've seen. But if the existing stuff is too difficult to use effectively, why should I expect that something even more complicated would be used enough to be worth development - particularly since any change in code, even one not actually used by anyone, has the chance to introduce new bugs?

[Don Maddox]

What is the formula that the WGOTY program uses to decide when a player's turn is over?

At the end of all combats the number of "rounds" expended is totaled. The turn ends if 1) a random number from 1 to the maximum number of remaining rounds equals 1, and 2) The current player's force proficiency is less than a random number from 1 to 100.

The number of rounds expended depends upon the mean expended movement of the attacking units (originally this was the max expended movement) before the attack as well as the number of rounds actually spent in the longest combat.

So your turn is more likely to end if you move a lot of units a high proportion of their movement allowance before committing them to attack. It is also more likely to end if your force proficiency is low.

What if a unit attacks successfully and advances, expending a great deal of its available movement. Does this increase the likelihood of the turn ending?

It won't have an effect on the determination for the current set of combats. Expended movement fractions are only checked at the beginning of combat to set the effective "start round". (They are, however, reset to a maximum value at the end of each set of combats, based on the effective start round and the max number of rounds expended in any particular combat.) On the other hand, the units in your example would have an indirect effect on the determination at the end of the next set of combats in your turn if you use those units to attack again.

[ALF]

Good thing Maddog, dig up some more! :)