V-E Day games ... and granularity

Monday 8 May was of course V-E Day. Pierre kindly came over and we marked the occasion with two games with very different scales and settings but some striking similarities.

 

Yep, that's me in the middle, surrounded by some of my many BBB groupies.

First up was some WWII air action, the "Thunderbolts over Normandy" scenario from the Check Your Six! rulebook. I'm pleased to note that CY6 was recently voted "the best WWII air-combat ruleset" on TMP. I took on Pierre's four P-47s - a superior plane - with four Bf-109s with a couple of better pilots. Thanks in part to my better pilots and in part to Pierre being a relative CY6 novice, I soon slotted in two of my 109s right on the tail of his flight leader. I pumped lead into the P-47 but the tough bird mostly shrugged it off. He did take a critical hit which chewed up his elevator so he couldn't change altitude for the rest of the game - but chunks of debris flew straight through my best pilot's prop, taking his plane down.

After that it became a furball. I got a lot of shots in, but my undergunned 109s kept bouncing off the Jugs. Pierre didn't get nearly as many good shots, but I managed to fly my green pilot right in front of his damaged P-47 that was cruising round in circles and it blew me away. My two survivors narrowly escaped off the board, out of ammo, with their tails between their legs, while the USAAF boys all returned safely to celebrate in an English pub and make inflated claims about their kills.

Then we swapped the light blue cloth of the Normandy skies for a dark blue one for the Pacific. We played a simple "kill 'em all" naval encounter battle, using the Victory at Sea rules. Pierre had made some pretty islands for our squadrons to dance around. It was my turn to be the novice, commanding a USN Cleveland-class cruiser and four destroyers against Pierre's equivalent Japanese squadron. Pierre had no luck at all, as I won the initiative on absolutely every one of the seven or eight turns we played - on a 50/50 roll each time - and then his torpedo strike with the legendary Japanese Long Lance pretty much bounced off my cruiser, while my single destroyer's torp salvo sent his cruiser to the bottom on turn 2. I then mopped up three of his destroyers as well, for the loss of just one of mine.

Thus our afternoon had a pleasing symmetry to it, each of us winning one ultimately one-sided game each; and with the Allies beating the Axis in both, as is only right on V-E Day.

The CY6 and VAS rulesets are for very different types of warfare. CY6 pits a handful of individual pilots against each other in brief air battles of a couple of minutes, while VAS can bring together fleets of mighty capital ships and their many thousands of crew to pummel each other for hours. But what they have in common is quick, clean and simple game mechanisms which generate new and interesting decisions every turn and let players fight battles to conclusion in a sensible time with none of the brain pain that more complex games can inflict.

Which leads me to some more general musings about granularity, i.e., the level of representation in games - representation of combat units, of terrain, of tactical factors etc. Many of the wargame rulesets that disappoint me do so because they are too detailed and complex for the combats they want to simulate: the granularity is wrong.

Appropriate granularity in terms of number of units is related to number of players and how many discrete elements our brains can comfortably cope with: about 8 to 12 at most, I believe, and preferably a few less. This figure is perhaps not concidentally roughly the number that real military commanders tend to deal with when thinking "two levels down". A corps commander is interested in his to or three divisions and their two or three component brigades or regiments, and perhaps a specialist attachment or two; a company commander cares about where his platoons are and their constituent sections and HQs, etc. Maybe this is why chess with 16 pieces per player is way too difficult for some of us?

With air and naval games like CY6 and VAS, you might think the scale of representation of the combat units themselves is pretty much fixed. Unless you want to go to operational level games and start dealing in terms of entire squadrons of ships or planes as your units, a plane is a plane, and a ship is a ship, right? So we give each player half a dozen or so ships and we're happy. But ships and planes have a lot of rivets, and there are players out there who want to count them all and for them all to count. Hence rulesets that require you to keep track of many damage points, perhaps allocated to different parts of the unit; rulesets that require you to keep track of how many shots a plane has fired; too much detail, too complex, too granular. CY6 by contrast takes the view that air warfare is swift, unpredictable, and catastrophic. Thus if a plane is hit it is simply either unhurt, damaged (engine or airframe), or destroyed; with just enough possibility of critical hits to make the game realistic and interesting without being unplayable.

Similarly with 'terrain' (of which there isn't usually much in air or naval games). Traditional air combat wargames have tried to slice the airspace up into 50 or 60 or more altitude levels, and to capture fine differences of manoeuvrability with multiple turning templates of different radii. This can make the decision set confronting a player so big as to be incomprehensible and meaningless. Certainly that has been my own experience. But CY6 boils it down to just 6 tactical altitude levels (or maybe 12 in some scenarios), and uses a hexmat so that the hexes structure maneuver options. Now the decision set becomes limited enough to be able to make an intelligent choice, while still large enough to reflect planes' performance and offer interesting tactical decisions - the granularity is right.

Granularity applies to time as well. What time interval should a game turn represent? Since for me a game is about making decisions, each game turn needs to be long enough for something significant to happen every turn and new decision points to be reached. I've suffered through too many overly detailed games where each turn is a tiny fraction of time, units inch forward in tiny increments, we grind through the same round of inconsequential firing as the last couple of turns, nothing really changes, and then we do it all over again, ad tedium ... A fast game's a good game. A bit like the number of units, whatever period of time the combat as a whole is supposed to cover, the magic number of turns for a game seems to be about 8-12. Each turn is like a chapter in a book, where each chapter moves the story along significantly.

Of course, just as one can have too much granularity, it is also possible to have too little. Some years ago I read an academic research paper on modelling the survivability of combat helicopters on the battlefield. I wish I could find it again - if any reader can track down the reference for me I will be very grateful! The gist of it was that previous simulations had modelled the terrain in 10x10m cubes, and had found that helicopters were very vulnerable and quickly got shot down. Advances in computing power meant the authors of the paper had been able to model the terrain in 3x3m cubes. At that more granular scale, lots more trees and bushes and gentle folds in the ground were represented; and in the new simulation, helicopters survived much longer and were much more effective. Since attack helicopters such as the Lynx or Cobra are about 4m high, obviously they were able to benefit much more from cover which the earlier simulation had omitted. The influence of this 20th century helicopter research made itself felt in my rules for 19th century land warfare, Bloody Big Battles!. In BBB, the basic infantry unit has a frontage of at least 3", so in creating terrain maps for the scenarios, my rule of thumb was that if a wood or hill would be at least 3" across, it should be represented. This in turn fed into all the ranges and movement distances and zones of control, which are all set in 3" increments. The granularity is consistent.

Thinking of the standard 6'x4' wargame tabletop, another dimension of that space, apart from the granularity of the terrain represented on it, is how it relates to time: specifically, how far units can move in a turn. In BBB the basic infantry move is 12". Discount that somewhat for crossing difficult terrain, and for the chance that a unit may fail to move for command and control reasons, not to mention enemy action, and it could easily take a unit 6 or 8 turns to cross that 4' width. This means that in an 8-12 turn game a player needing his unit to cross the table will have some flexibility about, e.g., when to push on and whether to wait for preparatory fire to pave the way, while still being under time pressure that makes such a decision an important one.

The short moral of my overlong and rambling story? Perhaps simply that in designing games we should start by gearing them to giving each player around 8 units to worry about; and to games being completed in about 8 turns; and to that game length being in proportion to the table size. Keep the granularity consistent between troops, time and terrain.