This is a long post. You have been warned.

I’ve had to put the M-Engine play testing on hold since the start of the school year, and I’ll be too busy to resume testing it at least until summer break, but at this point I’m very satisfied with how well it works, its resiliency, and its flexibility.

It’s that last aspect of the M-E I’ve been focusing on recently.   We all know that there are players out there who want to play and complete games in one sitting.  Now there are some who may disagree with me, but this hasn’t really worked.  Imo, even play with more than 2 or 3 players on a ten-minute timer is not reasonable.  Consider that players in a 3-player game are able to take their turns every 4 minutes on average (this seems reasonable to me), that’s one round of turns every 12 minutes or 5 turns an hour = a 20 turn game lasting 4 hours.  In reality, players entering these games should probably allocate 5 hours just in case a game goes longer than average, and for games with four or more players, you’d better be prepared to sit at the ‘puter all day.

In theory, a SimulGear game should fair better.  With 3 people on the same timer averaging 4 minutes a turn, the same game will last 1/3 the time.  Unfortunately (at least according to the cognoscenti) SG is the kind of game where players prefer to take their time organizing their moves (i.e. along the lines of 20 to 30 minutes).

If we want to address the issue of play in one sitting, we need to acknowledge that the choices we have right now are inadequate.

M-E has the Simul-play advantage where a full round of “turns” lasts the same amount of time no matter how many players there are, but because it has the feel of a regular game, play moves quickly.  This in and of itself should be reason enough to consider its use.  I’m thinking that an M-E game like the ones described above should run between 1-2 hours, with many running closer to only 1 hour. 

But because M-E is able to work at the time frame of the throw of the dice, I’ve been thinking about tailoring it to work best for players who want as real-time an experience as possible given the constraints of the Flash-player (that apparently prefers to be refreshed).   They don’t want to move, wait 5 minutes, move, wait 7 minutes, move, etc..  Rather, they’d like to be making mid-turn decisions based on the outcomes of battles even as other players are doing the same, with these “rounds” lasting 1-2 minutes or even less.

If you’ve read this far, hopefully you’re one of those people who are interested in how this might work. Those of you who are familiar with the M-E know it’s too complicated to explain in a single post; you need to go to the web-site (see my signature below). What follows is a brief description of some mods that would let M-E fit the criteria above. 

Terms:

A ‘turn’ is composed of play from a point where players receive a card to the next point at which they are eligible to receive a card.

The length of time where all players are allowed to place a single list of orders is a ‘decision-point’. 

The term ‘rounds’ has a special meaning in M-E. It represents the simultaneous throwing of sets of attacking and defending dice in all battlefields.

The turn-timer may need to be renamed a decision-point timer, because a full turn (at the end of which players receive cards) will often require a number of decision-points.

The Concept:

I envision something like a 90 second decision-point timer being adequate for most boards. 

During each decision-point all players are allowed to place a single order for each territory they own.  These are Attack, Fortify, and AttackOrFortify.

Players who chose to attack can attack using the standard order form.  I.e., they can chose to attack with 3,2,1 or with more than 3, in which case the engine will continue attacking with 3 until victory is achieved or all allocated armies are depleted.

During any decision-point a player may elect to place no orders (pass).  If all players do this, the turn is ended, cards are allocated, and the next turn begins.  If any player places an order during a decision-point (even if all other players have passed), those orders are executed and another decision-point begins.  Players who have passed on previous decision-points may always place orders in subsequent decision-points.  This enables so-called blitzing, but also potentially enables defenders to counter mid-turn. It also gives players the option of attacking with 3, then attacking with 3 in the next decision-point, and so on - (though this may not be the best strategy). The turn ends only when ALL players pass.  If a player has not submitted an order by the end of the DP timer, it is considered a pass.  (Variations on this might have a player booted after a certain number of consecutive passes.)

Once all players submit their orders, the next decision-point begins.  The length of time between decision points can easily be just a few seconds if players are active.

One critical piece of keeping the flow of the game going ..and also keeping the time of the entire game as short as possible is how the Flash-player might be designed to refresh every 10 or 15 seconds after a player submits orders.

There are a number of kinks to be worked out such as decision-points for army placement, numbers of armies that advance, but these are simply mechanical rules that need to be agreed upon.

That’s it for now.  Any interest?

I will weight in. 

I like the ideas and agreed with this back when you first introduced it. 
I think the engine is fine without the decision points, but they would certainly add a more interesting element to things (do nothing during first decision point to see what everyone else is doing).  Unfortunately, this element could kill the engine (each player makes one attack per decision point to make sure the round is not over and see what others are doing, continue this each round and you are close to turn based).  I, of course, do not think everyone will do this so it is probably alright, but rounds can certainly be long if this occurs.

All that aside, I do like the basic engine and think that it has merit.  However, I would like to see some other things come about on wargear prior to introducing completely new game-play styles.

Good luck with another semester.

The problem I see is that this is really a fundamental change for both the Flash Player and the game engine in that it is moving away from turn based to true realtime play. This is a major design change as the site architecture is completely based around discete turns taking place rather than a continous system. For example right now there is no real problem if there is a network outage for a few minutes - this is even fine during 'realtime' play as the turn timer is 10 minutes.

Fundamentally the current system is designed around turn based play so any engine changes need to fit into this model. With your system as described the clients would all need to be in constant communication and monitored by the server to ensure they are connected much like a World of Warcraft or similar multiplayer game server.

I think I understand what you mean, but I don’t think I’m talking about is true real-time.   I'm essentially talking about a 90 second turn-timer, which admittedly does seem pretty tight, but I anticipate that at any decision-point, players will only want to make 1 or 2 attacks, especially at the beginning of the game.   This should only take a half a minute or so for most players, and hey, if all you can come up with in 90 seconds is 5 moves even though you wanted to make 8, that can be part of the game.  The timer part is critical to be sure, and I think I understand some of the limitations of the Flash player and refresh feature, but I also have questions about it.

If the perceived problem is that players would be expected to refresh the screen too often, why can’t there be some kind of poller on the FP (or even in another window) that alerts players to the status of timer and tells them when all players have submitted their moves? Couldn’t this be made to work even though the screen hasn’t been refreshed? This feature would make even more sense if it were activated for each player only AFTER they submit their move.

A different way to approach the problem: Because they are optional, players could all but ignore the turn-timer as they pertain to mid-turn decision-points.  They could just let them expire every 90 seconds with no penalty (the game would last longer, but with an average of 2-3 decision points per game, this probably wouldn’t be so bad). On the other hand, to speed up play, a check box on the Flash player could be programmed to “auto-pass all remaining DPs until the next turn”, and players could always check in and “un-click” this button during any DP.  Another idea is to make the timer for the first DP longer (like 3 minutes). Maybe all succeeding DPs could be shorter by 45 seconds until it hit something like 25 seconds, because at that point most players would just be blitzing with a large stack.

Another consideration is a time-out.  Players might get something like two 5-minute time-outs for bathroom breaks, etc.

Here’s a bizarre idea for a quick game timer if the above ideas aren’t tenable.  What if the timer cycled on the minute (or on even numbered minutes) to the second, regardless of whether or not all players submit their turns early? Players would have to make sure that their computer clocks were synched ..or maybe not.  This might work because all players (or all of the Flash Players) would know exactly when to refresh the board.  I’m just throwing ideas out here.

I starting to feel strongly that if Tom is willing to take the time to tackle a third option for the site, the game should probably be thought about in terms of addressing the wishes of the “Lightning Game” crowd.  The Basic M-Engine should do this pretty well, but I was just thinking that the inclusion of decision-points would add a real-time feel to the game, and players might prefer a gaming experience where they are more or less continually active.  This might be very popular if the average game can be kept down to an hour or so.

Of course, as Tom points out, the limitations of the Flash player could be the deciding factor where this is concerned.

M57 wrote:

In theory, a SimulGear game should fair better.  With 3 people on the same timer averaging 4 minutes a turn, the same game will last 1/3 the time.  Unfortunately (at least according to the cognoscenti) SG is the kind of game where players prefer to take their time organizing their moves (i.e. along the lines of 20 to 30 minutes).

I'm failing to see how another gameplay type would be any different.

Consider that players in a 3-player game are able to take their turns every 4 minutes on average (this seems reasonable to me)...

Why wouldn't this be at least the same case for your gameplay?

Yertle wrote:

M57 wrote:

In theory, a SimulGear game should fair better.  With 3 people on the same timer averaging 4 minutes a turn, the same game will last 1/3 the time.  Unfortunately (at least according to the cognoscenti) SG is the kind of game where players prefer to take their time organizing their moves (i.e. along the lines of 20 to 30 minutes).

I'm failing to see how another gameplay type would be any different.

Correct me if I'm wrong, but I recall people stating that they enjoyed sitting over a cup of coffee when working out a single turn playing BAO over on WF ..creating long strings of orders and phantom orders and agonizing over just what order to put them.  To really optimize things it might take as much as a 1/2 an hour.   Contrast this with the time it takes to take the average turn in a regular game.  I suspect most of us can analyze a position and carry out all of our attacks in a minute or two, maybe with the occasional 5 - 10 minute turn.

Let's say that the average time it takes to place a set of orders playing SG is 15 minutes, and WG orders take about 3 minutes to place (perhaps this is what you are disagreeing about). The difference is that in a three player game, WG players could theoretically complete a complete round of turns in 9 minutes, while SG players would take 15 to complete a round of turns (because all of the players orders are placed and adjudicated simultaneously).  If the amount of time need to place an individual M-E turn is about the same as a WG turn, but all turns run simultaneously like SG games, a complete round of turns in M-E could average as little as 3 minutes over the course of a game, theoretically speaking.

Consider that players in a 3-player game are able to take their turns every 4 minutes on average (this seems reasonable to me)...

Why wouldn't this be at least the same case for your gameplay?

Because the amount of options you have at any decision points are quite limited.  If I start a game with 10 territories, on most boards, I'll place my three armies and order an attack from 1 of those territories.  4 or 5 'turns' into the same game, I might order 2 or 3 attacks and a fortify.  Remember, there's no blitzing at any one DP.  You need to use multiple DPs to blitz.  I'm thinking the amount of attacks/fortifies that you might choose to make at any DP should only take about a minute or so to figure out and place.  I was thinking 90 seconds might be adequate but only play-testing will tell.  2 or 3 minutes might be better, any longer and the use of DPs doesn't make as much sense for the reason Alpha stated ..I.e., players choosing to attack with 3 at every decision point would make the games too long, though I don't know that this would be good strategy in most situations.

Here's a solution to the Alpha scenario where a player places only small attacks at every DP:

Limit the number of DPs to something like 4 or 5.  After the 5th DP, the next turn commences and cards and bonus armies are allocated.   Players who attack too slowly and fail to take a territory won't get a card.

Limiting DPs per turn would put a damper on full-out blitzing, but on the other hand it could actually speed things up because cards and bonuses will be cashed more often.

It follows that limiting the number of DPs per turn should enable longer timer lengths to be used. 

Essentially isn't your gameplay really a cross between SG and TurnBased?  SG in regards that everyone places orders at the same time and slightly TurnBased based in how the orders are executed.  Also can't there be more "loop"/scenario based orders with your gameplay (ie A-->B then B-->C if A-->B was successful), which ultimately gives even more options as to how an order set can be created (where as SG is limited to attacks of only adjacent territories), soooo that would seem that there is potential for an even larger turn time than SG.

I'm still not completely against another type of gameplay/order resolution, but I'm far from sold as to your system significantly decreasing turn time in comparison to SG.

Yertle wrote:

Essentially isn't your gameplay really a cross between SG and TurnBased?  SG in regards that everyone places orders at the same time and slightly TurnBased based in how the orders are executed.  Also can't there be more "loop"/scenario based orders with your gameplay (ie A-->B then B-->C if A-->B was successful), which ultimately gives even more options as to how an order set can be created (where as SG is limited to attacks of only adjacent territories), soooo that would seem that there is potential for an even larger turn time than SG.

Yes, as I mentioned ME is theoretically no faster than SG, but the reality of SG play is that the process of arranging order stacks and the amount of time it takes to strategize a turn makes "quick" SG play unfeasible.  And yes, as you know from helping to play test ME, "event dependent" and strings of orders can be placed in ME, and if these are allowed, it could arguably slow down play for similar reasons.

My re-think of ME for "pseudo-live" play involves a major reduction of the number of choices you have at any point in time. There is no blitzing, at least not in the sense that you can place a secondary (blitzing) order at any decision-point.  Each territory can receive a maximum of one order.  If you want to blitz, you simply order an additional attack at each future DP.   If mid-turn fortifies or army-placement are allowed, defenders could actually counter your blitzing maneuvers (but I'm getting ahead of myself).

Take a look at any game you play on most boards and I'll bet you make attacks from only 2 or 3 originating territories on most turns.  For instance, take a look at your test game.  If you back out all of the blitzing and event-dependent orders (because you wouldn't be able to place them in LiveGear), on most every turn, you would only be placing somewhere between 1 to 3 orders, even mid-game.

In fact, I suspect that at most DPs, players will only want to place 1 or 2 orders max and see what unfolds before placing additional orders from different territories at future DPs.  The only thing that I can think of that could potentially slow down things would be players who write what would amount to phantom fortify orders.  Actually, come to think of it, this wouldn't be possible because for instance, if I attack from Western US to Mexico and win, a fortify ordering Eastern US armies to Mexico would have to be placed at the next decision point (i.e. it would take an additional micro-turn because the territory needs to be owned in order for a fortify order to be executed), and of course, this is all mute if mid-turn fortifies are not allowed.

What I'm attempting to do here is create what amounts to a regular WarGear game (playable on most any standard board with standard dice and dice modifiers) where all players are simultaneously playing, and where the number of decisions players make at any given point in the game are small enough that "down-time" or "wait-time" is dramatically reduced.  If play moves quickly and players are constantly engaged, they won't mind hitting the refresh button (if the Flash Player can't be programmed to do it automatically) because they'll know that they can probably play.  There's nothing as frustrating as playing a lightning game and hitting the refresh button every 30 or so seconds (ten times in a row) hoping that it's your turn.

Of course, LiveGear has the additional advantage of no player going first, so even two-player games on "normal" boards will be "fair".

Hmm.. LiveGear.  I'm liking that name.

tom wrote:

Fundamentally the current system is designed around turn based play so any engine changes need to fit into this model. With your system as described the clients would all need to be in constant communication and monitored by the server to ensure they are connected much like a World of Warcraft or similar multiplayer game server.

If Growl is what I think it is (a turn-alert system), and given that what I'm talking about is not RT, just speedy play (like on a 60 or 90 second timer), I think it could work.  It's not like a two, or even ten second lag would ruin things.

you'll see the growl when you membership is about to expire.
or tom could post something on a growl to everyone saying 'hey this is a growl and this is what you'll see it for' :^)

I just uninstalled growl because a non-current version was apparently installed on my mac when I installed drop-box.  Growl apparently doesn't like it when their software is installed without the users knowledge.   Are we talking about the same growl?