So what's wrong with TOAW from a CM'ers standpoint?

[Redwolf]

While waiting for on-topic games...

I saw a lot of TOAW bashing in the recent doctrine/theory discussions.

Anybody would like to share better approaches to higher-level games? Any games on the operational level you like better?

[Guest Mike]

I find little or nothing wrong with ToaW, except for players abilities to use it!

[Redwolf]

Yeah, the dialog boxes suck. Or did you mean something else

[JasonC]

I find nothing even remotely sensible about TOAW. I have found it largely unplayable and quite unrealistic.

They seem to have had the brainstorm that they would count all the beans behind the scenes to "relieve" the player of any concern with the actual determining factors in combat power, then expect it to be "accurate" because beans are counted minutely. It isn't, it is hopelessly broken and the incentives it sets up frankly silly.

Attacks are best conducted as mega overstack affairs at about 10 to 1 odds. No, the concentration penalties do not remotely forbid this.

Supply is so uncontrollable, the only way to actually manage it is to deploy some units so far from the action they won't have occasion to draw it. Artillery supply in particular is frankly unbelievable. Air winds up getting used in the first few attacks per turn, then air units have reduced supply state etc.

None of the combined arms relationships of the real weapons are seriously present. Instead, units are bags of diversified combat power, supply and quality dependent to be sure, but not equipment dependent in any serious way.

The much older V4Victory series was better in every way. It also had a few drawbacks in supply depiction, but nothing compared to TOAW, and subject to realistic forms of player control. A few house rules were all it really needed (e.g. not to allow wholesale reassignment of artillery to some HQs then starved of supply, allowing too many combat units to be oversupplied. Also attacks below a threshold forbidden to avoid gaming the fatigue system).

If the operating thesis is that logistic thruput is the real generator of combat power, you can't abstract it and take it entirely out of players' hands and expect a livable game to result. A strategy game has to leave the major determining levers in players' hands.

The contact-withdrawal behaviors are also silly, far too draconian and "tar baby" esque. A Victory Lost gets the right effect much more cleanly by simply having a 2 MP movement penalty for leaving *or entering* a ZOC (and makes them cumulative).

The user interface is so painful that playing a game is a chore, not fun. And that is nothing compared to designing a scenario - much worse.

There is a decided tendency to giantism. When you have abstract units as bags of subelements, what sense does it make to allowed huge stacks of the things, and then attack from every adjacent hex? A cleaner design would have units at most one a hex in typical situations, and better still leave "luft" between them.

Tiller's campaigns on the other hand have the grand tactical giantism problem - they seem to think a game is more interesting and grognardesque if there are 1500 counters per giant scrolling off the screen map. This simply makes them one unplayable and two unrealistic in the coordination the player has over that giant command span.

There are vastly better designs in modern board wargames, and even better ones in many much older board wargames. Instead of learning the art of game design from advances made in past games, what TOAW has done is thrown all of that away to lean on the bean counting of the computer, and then done that badly.

You want a list of the sorts of design innovations a computer operational game should have? It starts with a list of the right board (and computer) wargames to learn from.

To the green fields beyond - mobile vs. static combat, types of artillery fire, simple breakdown rules, exploitation soft ZOCs.

Battle for Stalingrad (John Hill) - attritionist combat, huge terrain effects, continual initiative movement interrupted by enemy reactions

Napoleon at Bay - command initiative system, operational points limits to reflect logistic and command constraints, movement attrition directly tied to stacking (big stack equals slow and straggling), simple but effective bridge destruction and repair and its effects on supply.

Terrible Swift Sword - step losses, simple morale states, command spans, higher level morale through brigade combat effectiveness system based on cumulative losses to a formation, firepower based attritionist combat, classified weapons.

V4V - different supply states for higher HQs, player allocation of budgeted supply among HQs, strong unit quality and combined arms effects, simple replacement procedures, simple fatigue and disruption modeling that penalizes overuse of a few strong units without rotation

A Victory Lost - multiple moves by HQ, in unknown order, initiative system. Soft ZOCs, dramatic movement differences across unit type and quality, river supply effects, unkillable HQs that displace when overrun.

Vietnam (VG) - pursuit based multi-round combat (you *want* to retreat - standing stock still gets you clobbered - can you even retreat fast enough to break contact etc), defensive reaction moves prior to combat resolution, cumulative attrition, political state tracks and influence on them.

Paths of Glory - threshold-attritionist combat, op point activation system, strategic movement, card events and "hand management"

The point is not to have every conceivable bell or whistle, but it is to do some serious game design, pick the aspect or relationship you consider crucial to the type of combat or affair you wish to model, and model it accurately with a few simple systems that put the key variables and decisions about them in the player's hands.

Not hidden under a computational hood, broken and unfixable.

[Cuirassier]

Interesting discussion. I was thinking about buying TOAW III, but if its unrealistic and unplayable, then I won't.

How is Korsun Pocket in comparison to these games? I was thinking of getting it.

JasonC,

How do I get A Victory Lost for computer? You talked about the game on a different thread and it sounds quite well done.

[sross]

I tried hard to get into the large Panzer Campaigns scenarios, but it is just too big. Having to control every little unit individually is terrible. The scope of the operations was good, now if they would implement a command system like HTTR and COTA, it would be a hell of a game system.

With TOAW, it is a great game concept, but as JasonC states, broken in some important ways. The lack of true combined arms simulation is very frustrating. The supply system always irked me as well. No control over the replacement system was a big pain also. Even though you received 50 PzIV's, you couldn't choose to rebuild a depleted unit to full strength, instead they shotgunned them around to all the other units. No player input into the logistical end is a major flaw of the game.

[markshot]

JasonC,

Greetings! I read here a lot more than I post. Yours often tend to provide some of my favorite reading, since I tend to learn a lot.

I realize that the topic of this thread is TOAW, but since you have broadened it a bit to discuss other models of ground combat simulation (although all appear to be turn/hex oriented), I did want to ask you opinion of another system. Assuming you know the system, if you can find the time, I would be interested to have you respond in some detail as to what you think works (simulates reality) properly and what does not.

The system in this case is the Airborne Assault engine. Currently, three games have been produced off this code base: RDOA (originally published by Battlefront), HTTR (Matrix), and COTA (Matrix) with a BFTB (Matrix) currently under development. Admittedly, the model of game play is very different: no turns, no hexes, hiearchical chain of command with delegation to AI subordinates, and order delays. Myself, I am unfamiliar with all these other games which are being discussed, since RDOA was my second ground combat game after CMBO.

In the interest of full disclosure: I have been a beta tester with Panther Games since RDOA in 2001. However, I am not looking to debate or score points or advocate this series. Mainly, I would just care to know if you have tried these games and if so what is your impression of how accurately they model the major issues of operational combat. So, any opinion you offer positive or negative will satisfy my query.

Thanks for your time.

[Redwolf]

Tiller games (Panzer Campaigns, foo campains) are all nonsense, IMHO.

I liked the "Decisive Battles" series (Korsun Pockets etc.), but couldn't really get into them, TOAW "felt" better.

I forgot many TOAW details, last time I revisited it it seemed OK. My thought of going back into TOAW is part of the reason for this question.

[PFMM]

While we are at it I will put my plug in for SSG's ancient War in Russia.

Army level units with division resolution with regimental replacements,and weekly turn based. It has two levels of command (High and District) with slots for up to eight human or computer players.

At the High level (OKH,STAVKA) you set activity levels and the send reinforcments to each district, with the West front a concern.

At the District level (North,Center and South) you have to manage each army and their manuever elements (composed of divisions), allocating air and support as required.

Partisans, supply, fatigue, administration, rail lines, forifications, lend lease, weather, and even Hitler and Stalin (who you do not control) are covered.

The entire campaign is playable in an afternoon, with smaller scenarios available. If you don't like anything it comes woth an editor.Best of all it's free as SSG released it to the public domain.

Caveat, it's old and will require an emeulator.

[JasonC]

Korsun pocket is marginally worse than the V4Victory series. A bit more playable than the giantism-infected Tiller Panzer campaigns IMO.

As for A Victory Lost, it is a board wargame rather than a computer game. But VASSAL is a system for playing board wargames by mail using a computer assistant, basically. One of the players (at least) is supposed to own the physical game, and you need to know the rules. If you google VASSAL you will find it easily. The Aide De Camp series of "gameboxes" are another way board games have been "ported" to the computer. I have one for Battle for Stalingrad (as well as the physical game).

I'd be perfectly willing to play by email either, with anyone interested. But understand, these aren't programs you can play solo vs. an AI, there is no other side unless you have another human, and somebody involved should know and own the game.

Paths of Glory can be played that way too. But there is also a java engine that executes orders and updates the map automatically. Some find that a bit clumsy (especially inability to take back a mistaken move, fiddly interface etc), but it is fine for me.

As for the airborne assault games, I have not played them, other than trying the demo of the original briefly. You should understand that I deeply dislike real time strategy games in general. I find the concept a contradiction in terms. I have no problem with plotted simultaneously move or "we go" as in CM, and like it even. I have no problem with real time for things like flight sims where reaction is clearly most of the point.

The only real time strategy game I can stand is the Total War series. And for me it works only because the battles it is meant to depict had serious control issues - as in, you only get to make about 3 real decisions per. The strategic level has problems, but is the real source of interest. On the other hand I like all kinds of turn based games, board computer and traditional (chess, go e.g.). Strategy to me is about carefully matched wits, and Total War just has most of its wit applied at the mustering and operational move stage, not the tactical resolution.

When I tried the airborne assault demo, I felt like all the problems in TOAW were probably there plus another, that I had precious little control over anything my forces attempted. Perhaps because I did not invest any time, I also saw no real controls conveying critical variables to the player. It felt like some movie director forcing me to sit through his version of a sim. I'd rather have root canal that let somebody else sit in my chair during a strategy game, so I did not put up with this for long.

What has blasted wargame design on computers is the desparate desire to avoid the actual effort of game design, in return for engaging in a bit more engineering, programming, or art work. It emphatically does not work. The best computer wargames are ports of successful board wargame systems. The very best add features the computer makes possible or playable, but still do game design - and also include highly flexible editors and scenario creation mechanisms to ensure replay value, and to correct inevitable balance errors.

But the standard of game design in the board community remains vastly higher than in the computer community. CM and Tac Ops stand out in the latter - and Tac Ops has a quite dated and finicky interface. In the former, in contrast, you can find entire new genres on 5 year time scales with very high replay value, and the design standard continually rises. The level of historical work and accuracy also puts 99 out of 100 of the computer offerings to shame.

Board game designers are game designers who know their profession and consider it one. Most computer games are comic books.

[markshot]

JasonC,

Thanks for your thoughts.

In addition to your comments about board game (of which I have no knowledge) designers being real game designers, I do feel that additional CPU cycles and polygons has not necessarily improved PC gaming over the years. There were more than adequate CPU cycles for good game play design 12 years ago, but not for much else. So, a larger portion of projects went into game play. These days less goes into game play and more goes into the sizzle factors. {Of course, based on your point, it might be that the early PC game designers were in fact simply prior board game designers.}

Possibly a bit OT ... but I have recently been playing a PC port of a highly acclaimed board game, 1830. Focuses mainly on railroad stock trading - but is nothing like any other RR games out there. It's one of the best PC strategy games I have ever played; solid AI, no holes, and very meticuously balanced. If you are interested: 1830 Board Game by Avalon Hill (Francis Tresham, Bruce Shelly) in 1989; 1830 PC Port by Simtex (Bruce Shelly, Russ Williams) in 1993.

References

en.wikipedia.org/wiki/18XX

en.wikipedia.org/wiki/1830_(board_game)

[ March 14, 2007, 07:44 PM: Message edited by: markshot ]

[Redwolf]

CM is also an example of putting computer cycles to proven good use. It uses them for better LOS computations in a 3D space in a game where that makes a real differences.

The CPU cycles put into many hex-like games are trying to replace simpler mechanisms, and the result is not always clearly better. The problem is that a carefully chosen set of simple rules can be better understood and brought into balance, and the more complicated rules in the computer versions can get carried away and come up with unwanted effects.

Airborne Assault does put CPU cycles to new and good use, too, even though the result is quite a clearly an improvement as it is for CM.

[PFMM]

I should be working...

The problem of computer game design is one I know painfully well. First a bit of history.

Back in the sixties and seventies WARgame design was mainly concerned with presenting the matter over a table top in a manner unaquainted people could comprehend. They were simple. With the eighties and more familiarity in the market another force became more evident, and for about a decade games became increasingly complicated and convaluted (ASL is one of those survivors). Eventually it was relised that this was doing more harm than good, which brings us to the current level of innovation we see.

The eighties also saw the beginnings of the personnal computer, and many board game designers fantasised (me too!) about how their games would have a wider audience as the computer would track every detail and enforce the rules. All it needed was an interface.

They were wrong. The issue is both one of resolution and meaningfull procedure. And it's a mistake I also made.

For example if you were to make a game of, say, 18th century naval warfare with say fifteen minute turns you would be able to provide meaningfull results (based on historicall outcomes) to your players, and they would be able to respond in kind on the results. However, if you model every single cannonball in realtime (for which there is no meaningfull historical outcome), you will produce rubbish with gibberesh as a response.

It's a trap that many are prone too, ,as we have games of whole wars (Ancient, Napoleonic, WW2,etc.) stretching decades modelled on individuals. One systematic error anywhere results in nonsense. If it were a religion, Gary Grisby would be it's Pope.

When computer wargame designers learn from the previous errors of their boardgaming cousins and pick appropriate scales for their subject matter, maybe we will see better gaming.

JasonC, I'd take you up on the offer, but my gaming budget is controlled by my better half these days. I am currently playing Fire in the Sky by the same designer of A Victory Lost. I'd recommend it.

[JasonC]

PFMM - by "gaming budget" I assume you mean time. Because VASSAL is free. It is a worthwhile point to notice, however - the truth of the matter is time limits gamers more than anything else, which puts a premium on playability.

Yes there was a tendency to monster games and overly complicated rules in the later SPI genres. Innovation did not happen in the biggest projects, but instead in minor games, frequently overlooked outside the designer community and a modest number of hard core grogs.

Occasionally an innovation was so powerful and successful it created an entire genre (e.g. Terrible Swift Sword went on to spawn a whole series of US civil war and Napoleonic game systems, be ported to the computer as the Battleground series, then complexified and messed up as the present Campaign series of those, etc).

But a lot of them just got "left", because they were overshadowed by first, the late SPI giants and second, the first solid computer wargames.

An example is the attritionist each side fires combat system of Battle for Stalingrad, which languished in obscurity until rediscovered by the Paths of Glory style of game from GMT.

If you look at why Combat Mission works, it is not primarily because the computer is exploited for massive CPU stuff. Yes that is used in a couple of ways, notably the general accuracy of the AT fire modeling. But as always, it is not possible to do that sort of thing without leaving outcome bugs and points of controversy, then stressed by players etc.

But CM mostly works because of game design level innovations. "We go" is a very successful adaptation of the plotted movement system used in board wargames like Sniper, Wooden Ships & Iron Men, and the air combat games like Dauntless. Waypoint movement is a computer-enabled innovation in that, previously seen in Tac Ops and in space games like Stars.

CM also exploits the computer to give true double blind play, a simple point but one board games find essentially impossible. The spotting rules have gamable aspects and "borg sighting" is a constant source of comment on accuracy, but those are irrelevant - as a pure game design issue, units first unseen and then uncertain, losable again, is a decided improvement over ASL "concealment" counters in a way only the computer makes possible.

The CM morale and fatigue systems mimic types seen in V4V and some of the grand tacticals, as well as ASL, but significantly ramified it. No Tobruk like literalism stood in the way of handing out arbitrary classes of quality level - something seen already in board games as early as Wooden Ships & Iron Men.

Command delay was a CM innovation and cleanly implemented. It leaves players more control than is probably realistic, but a strategy game can't take them out of the driver's seat and work, so they get it about right.

Other places they tried to use engineering literalism or simplicity left holes. The arty model is rough, for example. Big shells do too much. Air strafing is too powerful. If one only wanted to poke holes in realism issues one could do so all day. But it is irrelevant to the basic success of the design. We go plotted moves of morale sensitive infantry and realistic AT modeled armor captures the fun and strategically interesting aspects of tactical combined arms. Toss in an editor that lets you replay any kind of battle and you have a lasting hit.

The point is the few key game design aspects got right make all of the difference. The same will be true for a truly successful series of operational scale wargames on the computer. Chrome and eye candy will follow easily and are not the point.

But so far, no one has done it.

[markshot]

JasonC,

I'm back!

So, you are looking for the perfect operational game and you find WEGO to be the right balance of eliminating turn oriented artifacts while still allowing you the ability to fully intellectualize the contest.

Well, let me ask you your impressions of the AGE engine (WEGO and operational) used to power Birth of America and an upcomming game on the American Civil War. I realize that some might say that AGE is beyond operational and actually strategic, but for me I would say that the engine is less than strategic (compared to the attempts by Paradox), since it is mainly focused on depicting warfare and not nation management.

Phil Thibaut the lead designer on these projects does, in fact, have his first entry into game design in board games. You'll see some of that legacy in some of the game concepts such as hit points/strength versus PC games using personel and casualty figures.

Full disclosure: Surprise, surprise ... I am a member of the beta team for this developer too. Once again, not looking to score points here, but just to get feedback and impressions. However, I promise you, this is the last query and the only other organization which I test for. {Well, I did involve with MMG for a while, but let's not go there.}

Thanks again for your time.

[Guest Mike]

Jason is a realism Grog - he wants everything on the computer to behave exactly as it did in "real life".

I like to play games. As long as they have a reasonable correlation to "real life" then that's good enough for me.

TOAW works.

It creates "realistic" results for historical scenarios viea "realistic" intermediate postions.

Hence all of Jasons objections are specious IMO.

His comments that it is "unplayable" are a joke - I am playing several games PBEM at the moment and have no trouble with it at all. Indeed I play it much mroe than any of het CM series.

Certainly it has limitations - it is ONLY a computer game - let's be real.

[PFMM]

No Stalin's Organist. You are wrong.

The crux of this argument is the integral design versus differential design.

The integral design is a process of averages. One examines a whole process, movement of a unit say, and it's likeley outcomes, ie it goes here or here or here. It's accurate, but the actual process of how and why it occurs is a mystery, with the net result that the movement of one unit is much the same as another unit.

The differrential design displays the individual parts that make up that process and tell you why it ended up where it did and how it did it. The problem here is that any error in the modelling of the parts will result in the whole being inaccurate and what the player sees off.

If you do research for a differential design you will quickly discover that the historical data is not there or is incomplete. So one compromises.

The other issue of a differential design is it requires a lot of processes and is complicated and convaluted. Even on a computer even if you had all the information, you're not going to be able to do them all. So one compromises.

The final measure of these compromises is how well does the simulation stack up to the expected outcomes, what insights does it give to why and how things happened, and does it both entertain and challenge.

The games we are highlighting here make these compromises in a nice way. They grant us interesting, challenging decisions, show us how they are done and why and are simple enough to able to do in a reasonable length of time. The others become a chore in comparison and are as illuminating as an accountants spreadsheet.

That is what we are discussing.

[Guest Mike]

Pardon me for beign a consumer and voting with my money then.

your post is meaningless to me - not only is it meaningless I am actuaslly not interested in trying to discover what you are trying to say.

You guys may wish to spout about computer design philosophy, but to me the only thing that matters is "Does it work?"

ToaW works.

[Reckall]

Regarding TOAWIII, I should ask to JasonC which is the last version that he tried, since the game went a long way in these nine years - and it is still going.

Anyway, the purpose of the game is to represent the problems that a commander could face at operational level: what's the strength/proficiency of my divisions, how the road/rail net is helping/hindering my advance, what about the air power factor etc. In this it surely works beautifully, and it is a great game.

I understand how, by involving too many factors in the combat calculations, you risk to increase the end margin of error. So, if your data about the impact of a certain weapon in combat are slightly off, and your proficiency model is slightly off too, the end result is exponentially off.

However, it is also worth mentioning how the above factors *which in TOAW can be manipulated* also grant flexibility to the scenario designer, who can tweak them (and many others, like special events, pre-determined variations in supply over time so to simulate contingent factors, attrition and MANY others) so that the end result reflects his view of the battle

(BTW: of course one expects that the designer will conduce some tests on his scenarios - which is what happens: no one in the TOAW community puts together some units and wild calls and labels it "a scenario"; this is why, for example, the very intensively tested "Campaign for North Africa" serie by Bob Cross simulates so well the ebb and flow of the operations in that theater).

It is also worth noticing how, in the wargaming world many whines about how "something is not realistic" often boil down to "this thing doesn't portray MY vision of what is realistic". For every NATO/WP "expert" sure that "whoever thinks that the war would have lasted more than a week is an idiot" you will find a NATO/WP "expert" quoting number and facts re: "any cretin can see how it would have developed into a months-long attrition war".

[H.G. Welles wrote a very funny tale on this phenomena, called "A Moth" ( http://whitewolf.newcastle.edu.au/words/authors/W/WellsHerbertGeorge/prose/stolenbacillus/moth.html ) Just read the first part and you will understand what I mean ]

So, fine: this is what a flexible tool is for. If you do believe that the Rumanian Army at Stalingrad was exceptionally unlucky you can tweak things in TOAW (and other games) so that the rout result is at one end of the bell curve. If you instead think that the result was just what expected, you can put it at the center of the bell curve in your scenario - and so on. This is one of the things that TOAW allows you to do.

Re: Panzer Campaigns, I have some of the titles, and I have still to see someone pointing a gun to my head and forcing me to play the ENOOOOOOOORMOUS full campaigns. They are there for those that really want them, but each game usually is broken down in dozens of small to medium scenarios. I just finished playing the German Breakthroug at Vyazma in Moskow '41 (small scenario) agains a friend: we finished it in an evening and we had a blast. Other scenarios, like Salerno, Crete and Anzio, are small enough to allow you to play the full campaign in a reasonable time - so even that can be done. And Normandy '44 is basically about the various sub-operations: the rules themselves state that the full campaign is more a tool for scenario designers.

However, everyone is entitled to his opinions, of course, but I would suggest to ask around for a compendium of them before making your call. I could point you to people able to "explain" how CM basically ruined forever the reputation of serious wargames

[ August 15, 2007, 03:34 PM: Message edited by: Reckall ]

[PFMM]

I am fine with people having their own tastes and preferences, and they can play what they like without any fear of criticism from me.

My point is that in my profession to rig a series of procedures to produce an expected result is unethical.

If one has to tweak the fatigue rate to a little quicker than in reality, or increase the supply consumption compared to what it was in order to reproduce some other historical outcome I'd say your design is wrong. And what is worse you are giving people the wrong reasons for the original outcome. At worse you are lying. Just becuase it's hidden does not excuse it, in fact it makes the offence worse.

If this was not an issue we would all be playing chess, go, checkers, etc.

[Ralph Trickey]

Originally posted by PFMM:

No Stalin's Organist. You are wrong.

The crux of this argument is the integral design versus differential design.

Do you have any references for those terms? I've never heard of them, and google doesn't bring up anything like that.

What you're talking about sounds more like what in a QA environment would be called White Box Testing and Black Box Testing.

In essence White box means that everything is known and there are no unknowns. That means that you can do complete testing. Black box means the opposite, you can set the inputs and test the outputs, but you don't know what actually happens inside.

As applied to games, I have no idea of what you're talking about, that's why I asked if you had some references.

Thanks

[ March 15, 2007, 07:20 PM: Message edited by: Ralph Trickey ]

[Ralph Trickey]

Originally posted by PFMM:

My point is that in my profession to rig a series of procedures to produce an expected result is unethical.

If one has to tweak the fatigue rate to a little quicker than in reality, or increase the supply consumption compared to what it was in order to reproduce some other historical outcome I'd say your design is wrong. And what is worse you are giving people the wrong reasons for the original outcome. At worse you are lying. Just becuase it's hidden does not excuse it, in fact it makes the offence worse.

If this was not an issue we would all be playing chess, go, checkers, etc.

OK, I was going to stay out of this, but I have to take exception to this post. BTW, out of curiosity, what is your profession?

TOAW was designed to cover roughly from 1900 to the near future. It does best at certain scales and time periods, and there are some settings that make life VERY challenging for the designer.

TOAW does a resonable job of covering that period, although, of course, compromises are made as you get further away from the 'sweet spot,' this is arguably the scale that the Korean war scenario was designed to cover.

Instead of locking everything down and trying to cover everything from within the engine silently, it allows the designer to override some behaviors depending on the operation being covered. I wouldn't call that 'cheating' any more than shifting my car into 2nd gear instead of cruise when going down a hill is 'cheating.' It's adjusting the system for variables that can't be anticipated by the system.

No engine can take into account all the possible political and economic factors that affect the combat, that's why TOAW allows the scenario designer to adjust those factors.

Thanks,

Ralph

[PFMM]

Ralph Trickey,

The terms are defined in the post.

As to your second post I've never owned TOAW III, have not refered to it ever, and till now known what it was. I would advise you to ask people who own and use it for comments.

[JasonC]

Markshot - I do not think we go is any magic bullet, and I doubt it would be helpful for an operational scale game. In fact I expect it would harm it.

I don't quite see why this is so hard to grok, but evidently it is. Game design is a process in which an intelligent person with a clear thesis about the key aspects of some strategy game subject, models those key aspects, and puts control over them in the hands of the players. Leaving the outcome to their respective wits and getting his own tush off their table - but having selected the key variables and parameterized their control.

Leave out any idea of simulating everything. It is impossible and would leave no strategy anyway. Strategy requires playable control.

We go, I explained, was a successful adaptation of plotted movement systems. What sort of board wargame used plotted movement systems? Tactical air combat, because anticipation of enemy moves was key in that.

Individual infantry combat with firearms. Why? Because without it, one side or the other is a "phasing player" shooting another guy standing there, and how did the opportunity for them both to see each other arise anyway? Etc. In other words, the unexpected collision of two intentions is the usual way men who can each kill the other easily, wind up in each other's LOS.

Now, tank combat is a natural extension of this. When in addition you have battle resolution within a plotted minute out of player hands, you get a workable strategy game instead of a broken clickfest shooter. Imagine somebody instead reasoned "the key thing when a Panther and a T-34/85 face each other is the actual shooting, so players have to control the exact mil along which the gun is laid and the precise instant of firing or the game will have no point". They'd be making an arcade game, not a strategy game.

A tactical scale in which things happen by haphazard collision, or in which head game mutual anticipation is critical, is the underlying motivation for a we-go or plotted system. Now, there is no sign whatever that such things are critical variables in operational campaigns.

Similarly, pushing a real time engine on top of an operational strategy game is pointless. It is an engineer's fetish or an attempt to differentiate a product as having a new bell or whistle that hasn't been used before. At its best it might be part of a thesis about Boyd decision loops or something, but it would still be highly dubious as a means.

What is every operational narrative full of, instead? Initiative, roles of attacker and defender, a clear sense that someone is driving events and someone is being driven. That does not mean "losing" - the initiative can be turned into a fiasco quite readily. But at the operational scale, possession of initiative is a military reality, and not an abstraction imposed by a game system.

So you should have one player moving at a time, turn based.

Well, old board wargames made everything symmetrical for no particular reason. Maybe its seemed fair, or to involve fewer mechanics or something. But all the best games threw this bit of brainless design out the window, one way or another.

What systems have been used to model initiative in a turned based game?

Command rating systems like OSGs Napoleon At Bay. What was this? Well, each army (several on the Allied side) has some limited number of administrative points to allocate, representing logistical and command capacity. A force issued one of these can move normally in its own phase for such things. But there are not remotely enough of them to go around, and a force that spends them before the enemy loses its flexibility and becomes vulnerable.

Any commander can instead try to exercise his own initiative. This means a blindingly simple mechanism - roll 1d6. If it is less than or equal to the commander's rating, the force can move. Otherwise, it sits stock still for that phase. This initiative mechanism is also allowed for "forced marches", basically during the enemy's turn.

Now give a typical separated major general an initiative of all of 1. Give a modest corps commander a 2. Give Napoleon a 5 and Blucher a 4 and you get the idea. The lumbering Austrians, unsure they actually want to unseat the husband of their princess to restore a British toady, are led by a 2 - but have some administrative points. You can move a big army with few points if they stay together - but that will lead to march attrition (from high stacking) and slower moves.

Suddenly initiative and maneuver mean things. There is also great risk and variance realistically introduced. It was enough to practically *be* the game system. Other mechanics were distinctly secondary.

That is adapted to pre-modern operations. Take a second case, Battle for Stalingrad, adapted to battalion sized units in an overall clash between modern armies.

The Germans always move first. No symmetry. They can move all they want, as many units as they want. They can attack when they want. Each time they attack, there is a modest chance of triggering a Russian reaction. Mechanically this is done by chit draw, so a string of non-reactions increases the chance of a reaction, while several rapid reactions decrease it.

This can happen as many times as is happens, in a single game turn. The turn continues as long as the German player wants to move units. When he has moved them all or no longer wants to move any more, the Russian gets one last reaction phase.

Now here is the biggest kicker - each Russian reaction can only move 1d6 stacks, plus any arriving reinforcements for that phase. The last can move 2d6 stacks, and any arriving reinforcements left for the turn as a whole. They can also move units close to their army HQ, every phase.

German initiative, written into stone. With only flurries, usually local and urgent, of Russian response. The Russians are therefore forced to deploy in ways that promise decent defense no matter what the Germans try or how long it is before the Russians will get to move the local forces again.

Well that is just unfair! Who cares, it is accurate. Now the balance - combat is violently attritionist, terrain gives huge bonuses to defenders, and defenders usually get to fire first. Abstract, what does it mean in practice? It means the Russians can be nearly certain a battalion deployed in heavy city will take at least one German battalion with it, if it is attacked. And the Germans do not have more battalions.

Now the Germans can dance and dazzle to their hearts content. But the Russians will only die if their terrain adjusted price has been paid in blood, in full. So the Germans have to pick carefully who they can *afford* to kill. Oh and to remotely be able to afford killing enough, they have to pay attention to all available combined arms bonuses (armor hurt less by arty, engineers helping to negate defensive terrain, etc)

Utterly brilliant design. No we-go needed.

Or take a larger scale assymetric depiction of initiative in A Victory Lost. Each turn, both sides have an allowance of HQ activations, which they pick from a pool of available HQs. The Russians pick their overall pool at start, without the Germans knowing it. Each turn, they pick 3 out of that pool as the active armies for that turn. Plus a Stavka chit that activates every Russian HQ.

The Germans get a varying number of HQs, starting at a pitful 3, and rising as high as 6 for 2 turns in February. (Turns are weeks). They can pick any HQ on the map. They also get Manstein chits after a while, one then two total. These can be used to activate any HQ, chosen at the moment it is drawn instead of at the start of the turn. (Flexibility). But they have no chit that activates all HQs.

Who has the initiative? Well, you draw out the chits one at a time and that side moves any unit in command range of that army. And attacks. If they move a unit into command range of another HQ drawn next, fine, move them again. Fight again. It is perfectly normal for the Russian spearheads to move 2-3 times per week, while the rest of the army moves once. The Germans, well some don't even get to move - axis minors, often enough. Others may get 4 panzer divisions in their sector and move them 3 times, or bounce them from one part of the line to another, playing "fire brigade".

Best part, quite often the same side will get 2 moves in a row, sometimes on the same sector of the map. Make a hole, run through the hole. Very fluid situations result, compared to any symmetric move system.

The card driven games instead give players many options among which operational moves is only one. So they can force the op tempo and move and fight repeatedly - but only at the cost of letting other useful tasks languish. Which is more nearly right for a strategic rather than operational scale.

Another system I've played has players actually bid for initiative, by declaring how many attacks they will conduct if they win the bidding. Bid high and you can "drive", but you have to smash into the enemy the stated number of times. With sometimes expensive attritionist combat.

Then there are deployments and roles. TOAW actually uses some of these, so do the Panzer Campaigns, and V4V had some of it for artillery. Nam from VG had reaction moves before combat. What all these can involve is local adaptation to the falling hammer. A general support artillery battery fires in support of the threatened point. A unit set to defend by fighting withdrawal backpeddles instead of standing still to be assaulted. A reserve formation reaction moves to a threatened subunit of its own formation. These can all be automatic rather than player input. (Battle for Stalingrad had "instant counterattacks" if the enemy moves up to units but doesn't immediately attack them - with terrain ignored on both sides, meeting engagement style).

With these sorts of options for handling initiative and the flow of play, why on earth are we stuck with real time or we go or symmetric phased turns? Real time is an arcade approach and pointless for operational scale. We go is tactically golden, operationally crippled.

To keep playability high, avoid the face to face solution of vast numbers of interleaved small movements. Instead let activations create a clear phasing side who can move and fight at will (in any sequence desired, not all fights after moves), and let the other one make small SOP local reactions but nothing more.

OK, that handles the time mechanics. Then there are the battle resolution mechanics. There are three clear things to avoid that past games have suffered from. One, mechanics so opaque the players feel completely outside the loop. Two, rewarding only overstacking and 10 to 1 gang ups. Three, unplayable, byzantine procedures.

If the player cannot predict the likely outcomes within some known risk parameters, then he can't make strategy and the system is busted.

If all it takes to kill enemy without loss if making a big fist, there is a single unrealistic optimum imposed by the game system that will falsify every method used.

If it takes half an hour to resolve a combat, nobody will play out a campaign.

There is room for design within all of that, and some depends on scale. At the lower end, it is critical to get combined arms effects right, because the point of sub-division tasking was and is to get such things to break favorably to oneself and unfavorably to the enemy. If levers for doing this are not present or powerful, the game is busted.

Virtually every odds based resolution system fails on the second point. Sometimes it won't matter for the success of the overall game, but usually the game would be vastly better with a better combat resolution system.

There are seriously only 2 - odds based and retreat focused, and attritionist aka firepower resolution. The former is biased toward allowing units to fight repeatedly without material change in their abilities. It is maneuverist in tendency and what it encourages. It typically results in surround kills being the only serious way to destroy enemy units - with huge overstack attacks a back up or breakthrough creator.

Attritionist firepower, on the other hand, works by step loss principles, or gets nastier still and removes entire units at every "shot". The latter is more effective at the smaller scale. The principle is that every combat leaves both sides materially weaker.

Which naturally puts emphasis on whatever replacement, reinforcement, or logistical processes model reconstitution of weakening forces. It also emphasizes any available method of hurting the enemy assymmetrically, without the usual cost from exposing similar maneuver forces to effective enemy fire.

Suppose we are trying to design a game that will get company to division command about right. We need combined arms modeling, we decide only attritionist resolution is proper for the subject matter. Next we have to design the actual flow of actions in the combat resolution. And should not hide them all under a computational hood, to be guessed at but not manipulated by the players.

The point is emphatically not to figure out what should typically happen and put that in the center of a bell curve of outcomes, as one poster put it. It is instead to hand the levers that determine what happens to the players, and get them out of the designers grubbing little scripting and movie directing hands. It is to be a strategy game, not your take on likelies.

If the players manage to pit infantry against tanks in open ground, it should be a slaughter, not an odds column. Outliers right and dependent on the players, that is the goal. Knight takes queen. If that can't happen, there is nothing for players to manipulate, they might as well call an actuary and collect interest on the battle outcome.

Computers promised to make some of the more intricate combat resolution procedures of board wargames actually playable - if not made too obscure. What are examples?

OK, first the attacker's artillery has its effects, which suppress various defending units. Next the defenders AT fire occurs, boosted by terrain benefits. It isn't odds but threshold (plus chance) based - is this AT weapon effective vs. that type of armor, if there is any. Then let the attacker AT fire occur, and as a special thing, allow attacking tanks (and SP guns natch) only, to affect non-armor defenders at this stage, too. Next combined defender artillery and any remaining maneuver unit soft fire against the attackers. Defending maneuver units should get massive boosts to their firepower from terrain or improved positions at this stage. Remove all resulting losses. Last, attacker's remaining soft firepower occurs (from infantry and support weapons or guns etc). Give the defenders terrain benefits here, but allow them to be reduced by various forms of combined arms (e.g. each paired armor and infantry, or combat engineers, reduces terrain bonus this amount etc).

Battle for Stalingrad had a trivial condition on taking ground - your fire has to destroy all the defenders, or they keep it. Depending on scale and typical terrain, one might instead "weigh" the remaining live armor and infantry vs. remaining defenders, and give the location to the attackers if the (terrain adjusted, perhaps) odds are long enough.

All of which can happen in a blur of CPU processing, as long as how it happens is clear to the players and they know how to manipulate it. E.g. if I need to avoid strong gun based AT, I need sufficient arty that fires before it does, to suppress the defending ATGs, say.

Now the sustainment and logistic modeling bit. Here the enemy is abstraction and loss of control, on the other hand, and forcing players to count beans, on the other hand. The solution is to give realistic higher level controls to players that meaningfully impact how his forces regain strength lost through combat. Not hidden, not automated. But not a spreadsheet fest.

Lots of board wargames got this pretty close to right. You can have supply units and expend them. You can pick HQs whose subordinates will have full supply or attack supply, vs defensive supply. You can assign replacement steps to specific units if they are rare enough. You can time reinforcements in return for VP effects or later costs of some other sort. You can have units left out of battle recover disorder or fatigue, to given an incentive to rotate them. Etc.

These are the elements of a successful operational game design. For any given campaign, there will be other specials that influenced the outcome - "chrome" or events etc.

All of it has to be rapidly learnable and playable. The players have to know their decisions are in the driver's seat, and must understand clearly what will cause what.

No, you can't tell them to just do realistic things and learn what typically happens by watching and seeing - they need to be able to calculate ahead of time and manipulate all the variables you have chosen to model.

[BloodyBucket]

Wow. I play TOAW, have a good time and get plausible results....and I think JasonC gets just as much pleasure from pondering the nitnoids of what consim theory should be.

What might make a pure and righteous consim game is a fun topic for debate and thought. I would venture to guess that what makes a good sim that sells well is somewhat tricky to nail down, or, as they say, weaklings and children would do it.

I find TOAW quite playable and realistic enough since I view it as a game, not a tool or the Gospel of Combat. I don't think of it as the zenith of game design, never to be bettered.

Now, a game with the features that JasonC advocates would be great...or dull as dry toast, depending on how it was presented, how the interface worked, etc.

I think there is room in the gaming world for a lot of different levels of abstraction, detail and seriousness. I dread the day when the issue of how many attack factors can dance on the head of a pin is resolved once and for all...I think the debate is entertaining.