PROJECT "RHUBARB" DEVELOPER DIARY

This sounds amazing.
Excellent work.

StiC

Looks fantastic. Thx for keeping us informed.

Map Building



Work begins in earnest in building the testing map for Project "Rhubarb". I've decided to use the small 'Fields' map as a test bed because the size is more managable for testing with small numbers, and the possibilities of action are reduced as well due to same. This means I can test the systems in scale.

When switching this kind of system (dynamic mission response) to a larger map, the work load doesn't increase on a 1:1 basis due to the exponential increase in possibilities. If two airfields per side becomes four, now instead of four possible bombing runs (A to C, B to C, A to D, B to D) you have a possible sixteen routes (A to E, F, G, and H / B to E, F, G, and H / C to E, F, G, and H / D to E, F, G, and H). Multiplied by the number of different assets beyond airfields (supply points, armour concentrations, town garrisons, etc.) and what seems like a relatively small map increase becomes a mountain of work.

So by testing with the tiny 'Fields' map, I'll be able to make significant revisions faster and iterate more quickly Iteration is key in developing a complex system. Even if something works okay, there's nothing to say that a different way couldn't work better. Constantly making quick changes in a 'what if' sort of way allows you hone in on the best design.

The placement of objects, and creating the templates to be able to load them quickly and efficiently, is an interesting design stage. My recent experience with the very well done Storm of War campaign run by 5./JG27 and ACG has helped greatly with the problems in such work.

You have to make sure a target is destroyable, but not TOO destroyable. How destroyable is fun? Is it more fun to have to make multiple raids on a target to destroy it? Is it better to have it set up so that an expert strike will destroy it, but novice runs will require more attempts? How much AA fire is too much AA fire? Lots of interesting puzzles.

One thing I'm working on is how to reveal that a target has been destroyed. Traditionally, the players receive a message that informs them that the target has been destroyed relatively soon after the bombs go off (if not instantaneously). This has several effects on gameplay:

First, the players immediately know that no more action is required. Defenders stop defending and attackers stop attacking. The target and the target area becomes a non-event point and the map (and the possibilities) is essentially shrunk in size.

Second, an instant message encourages 'suicide' runs. When there's a concrete indicator of whether or not you get the target or not, it seems to be that players are far more likely to 'go for broke' on targets with very little concern for if they survive or not. This comes out of a phenomenon that hard wires us to look for validation or reward. The 'DESTROYED!' message is pleasing to us and we want it.

More thoughts to come later in part two... (I've got to run for now!)
...
And now to continue!

By removing the instant message from the target, we can create a different sort of battle zone.

Recon flights become incredibly important not simply as a way of locating a target, but of also identifying if the target has been significantly damaged. Poor recon can mean throwing bombers at a target that has already suffered significant damage, or NOT throwing bombers at a target that has received only light damage.

Even just locating targets becomes different. In a spread out armoured advance, if 6 of 10 tanks must be destroyed to knock out the advance as soon as the sixth tank is destroyed, the rest of the armoured advance is ignored if instant messages are on. With them turned off, perhaps the 7th tank is spotted while the 6th tank is being destroyed. Now it's critical that the attacking aircraft identify just how big the attack is before deciding if they've done enough damage to it.

We also touch on the idea of partial results. Instead of a polar opposite result (alive or dead) we can open the infinite results of an analog scale. Partly damaged? Heavily damaged? Routed? Utterly destroyed? By leaving out any pre-keyed destruction quota, results from ground attacks are free to be more organic. This works in a situation in which someone can observe and report on the targets (the role of the warlord) and make judgements based on the attack. Perhaps the attacking aircraft sent a wave that destroyed four tanks in seconds before being engaged by fighter cover. Perhaps the speed and rapidity of that attack causes moral to fail and the attack stalls. Unencumbered by the destruction trigger, we can make a more compelling story from the game.

We've been trained to want iron-clad results. Did I destroy the target or didn't I? 1 or 0. I do not find anything poetic or interesting in that view. Many of us play the game as a way of trying to experience a small sliver of the conflict that embroiled Europe for so many years in the late '30s and 40s. There was very little about the war that was so cut and dry as destroyed / not destroyed.

In building the map, I have placed static cameras over all fixed targets (airfields, support facilities, etc.) in order that the status of these objectives can be observed when required to determine their current state. I'm intentionally leaving the condition for destroying targets vague and uncodified.

And now we get to one of the key parts of Project "Rhubarb" and something that will, I'm hoping, help the community develop something which sometimes seems to be very much lacking: trust.

We don't, seemingly, trust each other. Your aircraft are better. No, yours are. You're making missions to suit your side intentionally. You're not interested in seriously considering our recommendations. On and on and on... .

With Project "Rhubarb" TRUST is so important. Maybe not as much if run as a co-op campaign, but in the eventual move to VS. campaigns, trusting the Warlord's decisions will be of the utmost importance.

The ultimate goal of the Warlord is not to 'win the war'. Certainly he is pursuing that end as cleverly as he can with the assets he has, but the Warlord's true goal is to give the players a compelling narrative. It's easy to say that the players need to trust that the warlord can count properly and will get the destruction values right. That's not what I'm looking at. I'm interested in making sure that the players trust the warlord to make the right calls to provide a fun, interesting, and challenging game experience that is rewarding to players as lovers of video gaming AND history.

When I have to make the call of whether or not the players destroyed a town's garrison to allow them to take the town, but they only hit roughly half of the enemy's units, I have to decide whether or not allowing them to take the town in that turn is more or less fun, interesting, and challenging to the players.

The players need to trust that if I decide no, the town was held and their attack fails that I am making that decision in their interest, not mine. There is no room to debate. "No, we got that target! It's not fair. You're not being fair." That sort of response is exactly the kind of mistrust we have of each other in the community that prevents us from doing the more interesting types of games.

My hope is that in the above situation, instead of "It's not fair" I hear this instead: "Cool, this is becoming a huge battle. The Battle for St. Omer!" Because, if the situation above happens, that's probably exactly what I'm trying to do; tell the story of the Battle for St. Omer.

And let me tell you, that Battle for St. Omer is going to be EPIC.

Hi Wolv,

I expected some info about Sunday's testing but have got nothing. Pitty I would haven taken part as announced. Later at night I spotted some of you on TS (but it was too late). How was it going ? Any news?

No.310_cibule wrote:Hi Wolv,

I expected some info about Sunday's testing but have got nothing. Pitty I would haven taken part as announced. Later at night I spotted some of you on TS (but it was too late). How was it going ? Any news?

Hi Cibule,

Sorry about that. We had 3 pilots to test the overall air combat aspect of the system and it went well in general. A few things to sort out (numbers of aircraft, how the AA is set up at airfields and other strategic positions, etc.) but the general concept is still solid.

RED received moderate damage to its bomber airfield, BLUE took large losses in aircraft. RED attempted to bomb a BLUE airfield, but the raid was dispersed with one Wellington being lost and the other two returning to home damaged.

RED managed to strafe an airfield and take out a few more aircraft that way.

Next week I hope to introduce a little bit of the ground war as well.

One thing we came away with from today is that the number of players is important. With 3 players, the battlefield wasn't dynamic enough. Squadrons need to work like squadrons in that there needs to be some people at the 'readiness' status to be able to take on new raids. It was the first time I ever got the feeling (in terms of campaigns that we've done with Cliffs of Dover) that the squadron 'scramble' would have actually been very important rather than just having everyone take off at the same time when the campaign started.

We'll be testing again same time next week (2:00 EST) so if you can join us, we'll have you.

Sounds very impresive. Thx for the info. I'll try to be part of the testing next Sunday.

Project testing to begin shortly (Sunday 12th of May). We will be using the ACG teamspeak server for communication. Find us there!

May I add this sounds amazing to the others who have.

Hi Wolv,

It really sounds fantastic.

As we have just finished SoW campaign (flown on Sundays) I wonder if there is a chance to take part in your great project. Any idea at what day time is it going to be flown?

Unfortunately, I got very frustrated in testing the AI ground and air units and suspended work on Project "Rhubarb" for quite a while. It's shocking to me now, but these test updates are now exactly a year old!

I'd still like to do something similar to this idea. At the moment, I'm focused on tweaking and improving the current mission on our server as it seems to be gaining some popularity with some members of the community.

If I can get some more time to do development, I'll look into this system again and try to work something out that would be playable.

I have experience in being a part of running virtual wars - ran one for 10+ years with AKA_Scorp - we started out manually building them, then shifted to use IL2 DCG - then eventually spun that off into our own creation.

In your last diary I see the frustration of the ground war. My take on this is to be creative and not have the traditional 'tanks roll into the cities, battle it out and take the city'. BoB offers some unique opportunities.

I can see this in some of the ATAG missions - where we take out airfields, strikes against radar towers, shipping raids, etc...

In battles on French soil, I view it more of the same way I would of a 'risk' game. Adjacent 'towns' have tanks, they don't have to roll down the roads - they can be defensive posts that if human pilots can bomb and take them out, and knock that defense out, then the tanks in the adjacent towns move forward.

What I like what you have here is the ability to have AI play a smart roll - where pilots can call them in to assist, you can have a lower amount of humans and yet still have commanders calling in AI fighters/strikes forces, etc...

Basically I'd minimize the need to have as much of a moving ground war - 'simulate' that, put the focus more on controlling the airforces, but have rules around taking territory....

Just some initial thoughts - as I'm very interested based on the past to see a system form, and I think your the right person to do it

in the late phases of the war, I started to create a 'supply' capability - whereas it gave bombers targets outside of needle in the haystack, on the deck attacks.

Basically 'factories' and other targets represented overall 'supply capability' - by knocking those out, it would have slow down the advancing forces.

These targets were especially designed to be level bombed (by tons of lower altitude flack, making it nearly suicidal to come on the deck)

Also, the ability to knockout forward airfields by bombing them.

This promoted bombing to a 1st class citizen, as typically all the fighter jocks just want to circle fight the whole time. Now they need to defend, help escort, etc.. if they want their side to 'win'.

To me, even an airkill was nothing more than a 'defense' - so I wanted more and more to downgrade that so we don't have 'aces' game

We ran persona's and 'DiD' concepts during this time as well - to promote people wanting to stay alive, but being passive won't win wars, so it creates a double edged sword