Week-End with the scenario day - 16/05-17/05

new doodle

http://doodle.com/ayswgirn3tf7wuee

Have put an ACW scenario together for one of the Pipe Creek maps, which we can play if no-one wants to do anything else. It's suitable for a smallish number of players.

Martin (J)

Yesterday battle video

The Union objective was to take the town of Manchester and defeat any Reb force they encountered.  The Confederates had two strong divisions, and were to protect the town or, failing that, to withdraw to the east in good order.  One problem with this, as both teams quickly realised, was that Manchester itself was quite indefensible, being overlooked by high ground.

The Union had about 4,000 more men, although only one of their two corps was under any form of player control.  I was interested to see how the AI-controlled 2nd Corps would behave.  As it turned-out, they moved directly on Manchester and fought hard to secure it.  It might have been a fairer fight if I had tweaked the scenario in some way.  Would I have achieved the right effect by putting the AI corps under a less competent general?  Not sure.

I think both teams can take some credit.  The Union inflicted more casualties and took most of the town.  The Reb force was battered but still more or less intact, and kept control of the high ground overlooking it.

Overall a score-draw.

Martin (J)

This looks like an interesting scenario. How many players have you been?
This layout reminds me of a scenario setup I'm thinking about currently, where the forces start more spread out instead of all forces from one side at one spot.
@Martin: Tweaking a MP scenario is very limited. What you could do with AI forces is to place the AI army or corps commander far away from the AI division(s). This way the courier will cause a delay in activation of the AI forces.

I didnt have time to test this yet, but it would be interesting to see if AI forces stay where they start if a scenario has no objectives and the type is "defend". Kevin maybe?

"I think both teams can take some credit. The Union inflicted more casualties and took most of the town. The Reb force was battered but still more or less intact, and kept control of the high ground overlooking it.

Overall a score-draw."

You'd make a good politician Martin, that's a very rosy description of what was an almost crushing defeat. I'm pretty sure the Union held all the town and we certainly had the objective secure. Another ten minutes of play would have seen Berto finally pushed off that high ground and the SW side of town secure. I had a completely fresh brigade in my division, hardly fired a shot and the other two were in good order, no units lost.

I agree that the arrival of the "Shinkansen" AI corps did unbalance the game somewhat, but Mitra's intercepted note of "Berto - retreat! This is a massacre!" I think is evidence enough of how one-sided the fighting in town was.

In any case it was a great game, quite unusual in 'shape' and it was refreshing to see a corner of the map we don't usually fight over. We should encourage more scenarios that provoke actions in the lesser used areas of our 5 mile maps. The extreme north of the Fox's Gap map and the extreme NW of Crampton's Gap are two areas very underused. Perhaps even a 'trapped against the river' scenario on the west side of Antietam.

Berto - retreat! This is a massacre!"

Almost correct Digby my compliment massacro is similar but ritirati is not immediate to identify , in reality I told, seeing his losses, "Berto ritirati, vedo che ti massacrano" -> "Berto retreat, I see that they massacre you" (I know in english it will be correct "they are massacring you" but in italian daily speaking is usual to use present simple at place of present continuous, for a italian to ask "what do you do" and "what are you doing" is the same question if nothing is specified, also if grammatically is not correct (the correct is vedo che ti stanno massacrando (or stanno per massacrare), the present continous is italian is given with the verb stare (to stay) + gerundio (-ndo) (almost equivalent to -ing) mode or stare + per (for) + infinitive

Leffe7 wrote:

This looks like an interesting scenario. How many players have you been?
This layout reminds me of a scenario setup I'm thinking about currently, where the forces start more spread out instead of all forces from one side at one spot.
@Martin: Tweaking a MP scenario is very limited. What you could do with AI forces is to place the AI army or corps commander far away from the AI division(s). This way the courier will cause a delay in activation of the AI forces.

I didnt have time to test this yet, but it would be interesting to see if AI forces stay where they start if a scenario has no objectives and the type is "defend". Kevin maybe?

We had 5 players, as follows:

Union:
1st Corps of 2 medium divisions (human)
2nd Corps of 2 small divisions (AI)

Confederate:
Hindman’s corps of 2 large divisions (human)

As scenario designer I took a small brigade in one of the Confederate divisions, and was not involved in the Confederate planning.

Good idea re placing the AI army or corps commander far away from the AI division(s).  I'll definitely try that.

Martin (J)

Diggers wrote:

"I think both teams can take some credit. The Union inflicted more casualties and took most of the town. The Reb force was battered but still more or less intact, and kept control of the high ground overlooking it.

Overall a score-draw."

You'd make a good politician Martin, that's a very rosy description of what was an almost crushing defeat. I'm pretty sure the Union held all the town and we certainly had the objective secure. Another ten minutes of play would have seen Berto finally pushed off that high ground and the SW side of town secure. I had a completely fresh brigade in my division, hardly fired a shot and the other two were in good order, no units lost.

I agree that the arrival of the "Shinkansen" AI corps did unbalance the game somewhat, but Mitra's intercepted note of "Berto - retreat! This is a massacre!" I think is evidence enough of how one-sided the fighting in town was.

In any case it was a great game, quite unusual in 'shape' and it was refreshing to see a corner of the map we don't usually fight over. We should encourage more scenarios that provoke actions in the lesser used areas of our 5 mile maps. The extreme north of the Fox's Gap map and the extreme NW of Crampton's Gap are two areas very underused. Perhaps even a 'trapped against the river' scenario on the west side of Antietam.

I suspect we’re looking at this from different angles, Diggers.  I was intending my comments to relate to team performance.  Yours – I think – relate to the military result.  That was a clear success for the Union, who inflicted considerable damage on the Rebs and occupied the objective.  But it was only to be expected, as they had over 35% more men, and over 60% more artillery.  

So in making my evaluation I felt obliged to look at the wider situation, and take that inbalance into account.  The Rebs were battered but still in the field.  They still held the high ground overlooking the objective from two sides.  The heights to the E of town were not seriously threatened.  The Union did not hold all the town.  In fact towards the end of the game, elements of McClaws’ Division drove part of the Union of 2nd Corps right through the town from E to W, and caused some Union regiments to retreat in disorder - I know this because my brigade was there.  Now it may be that Johnson’s Division could not have held the heights S of the town for much longer, but then it didn’t need to, as night had fallen, precluding further Union attacks.  In reality, the Confederates would now have had the option to reorganise and recover stragglers, and if necessary retreat in good order.  I still feel this was a creditable performance from the outnumbered Rebs.

I think that also applies to the Union, who avoided falling into the initial Confederate trap, inflicted more losses, and took the objective.  What they did not do was drive the Rebs from the field.  But that would have involved attacking the heights E of town.  I can understand their reluctance to do so, as they would have probably lost a lot of men.  And the clock was always working against them, and the initial Reb attacks used up more time.  So I stand by my view.  Perhaps I should have phrased it differently and said honours were even?

This scenario was unbalanced.  I think many battles were, and don’t see that as a problem.  I’m glad you enjoyed the unusual ‘shape’.  I hope the Reb commanders did!  Like you, I would like to see more variety in our scenarios.

Incidentally, I did not know what the AI corps commander would do.  Another possibility was that he would only move sluggishly and with much prompting from his fellow human corps commander.  At the other extreme, I thought it was possible that he might not move at all.  Had this happened, the advancing Union would have had the odds heavily stacked against them instead.  They would have been outnumbered by 30%, and facing somewhat better quality troops – which was one of the advantages the Rebs had.  In which case I would have been deploying my political language in favour of the Union team   .

I would have been happy with any of these outcomes.  But if we do play unbalanced scenarios, we need to take account of that in evaluating performance, and look beyond losses and points.  I think we owe that to players on the weaker team who have struggled against the odds – especially if we want them to keep playing unbalanced games, which are perhaps uncommon in wargaming.

I’m less happy that the AI at corps level seems unlikely to behave any differently on another occasion, according to Stefan’s last post.  So my expected range of possible outcomes in reality did not exist.  Hopefully we’ll see more attention given to the corps and army commander AI in the new Waterloo engine.

Martin (J)

Bah. I have no interest in this palaver. We captured the town long enough to make off with the table service. The rest, mere details.
M.T. Georgia, Commanding

That's a challenge.  How can I possible put a positive political spin on such reprehensible behaviour?  Let me see...........

Notwithstanding his command duties, and at some personal risk, General Granger took care to secure plates, bowls, cups & saucers, and remove them to a place of safety.

That will be 50 dollars.  And don't tell me you can't afford it!

Martin (J)

Martin: I’m less happy that the AI at corps level seems unlikely to behave any differently on another occasion, according to Stefan’s last post. So my expected range of possible outcomes in reality did not exist. Hopefully we’ll see more attention given to the corps and army commander AI in the new Waterloo engine.

From my observations if the AI is told to ATTACK then they usually move all their forces to the (first) objective. However, they don't always engage all their forces once contact has been made (like in the scenario "An Unexpected Ally"). If you set the AI to HUNT em down, then you can never know what happens, as the AI usually only moves a part to the objective and other forces move somewhere else or stay in reserve.

So there is already quite a range of possible AI actions   

Yes 'Hunt Them Down' might be better.  But are you sure the AI takes account of the objective in this setting?  Or does it just go out and look for the enemy?  I can't remember.

What about setting the AI to 'Defend'.  Would we see a range of responses there, or will it still move its entire force, as in 'Attack', and only behave differently once it has got there?

Martin (J)

In a "Defend" game, the AI will keep it's units motionless. So if some units are well away from the objective, they will stay inactive until they are called up as reinforcements due to other defending forces being pummeled. But if there are two independent corps, as we had in the last battle, then the corps that is out of the fight will never move unless directly attacked.

In a "Hunt Then Down" game. the AI will usually send a unit to a point near the objective. The others will be sent to other supporting locations. Depending on the quality of the overall commander and a random factor, the distance between supporting units can be quite large. In the case of the last battle, the AI corps commander would send his divisions to random sectors on the map. If no enemy was found, the division commanders would alert the corps AI of this fact and the AI would likely send them to new destinations.

Uncle Billy wrote:

In a "Defend" game, the AI will keep it's units motionless.

Is that the case even if none of the units are near the objective?  In that case, there would be no-one to call for assistance.  Or will that never be the case due to the game deployment logic?

Thanks for comments.  Am taking notes so I don't have to ask all this again!

Martin (J)

These settings were developed with SP sandbox use in mind, and not Stefan's SG or MP play. So in the sandbox, the defending troops are always placed close to the objective and stay in place. If they are not nearby, then the AI troops will still stay in place but the objective will be unoccupied.

As an example, Jack and I usually play corps size co-op against the AI, with us attacking. We each have our own division and the AI takes a third. The AI side is always deployed near the objective and waits for us to show up. Once it spots us, it will move some of the troops and re-position itself for our attack. But most will not move. The friendly AI always orders all three divisions to the objective and to attack.

Occasionally we play the side that defends. Again our corps is placed close to the objective. The enemy AI will usually send its entire force against us. There are a couple of exceptions to this behavior. If the enemy AI has several small corps, it may not send all of them to attack, but hold one in reserve. If the objective is near the edge of the map and the enemy AI is initially positioned near the opposite edge, it may not move at all. It thinks it is already close by. It is a math bug in the game. The game bends the map into a cylinder sometimes so it thinks these opposite edges touch one another. This behavior can also be seen when the two armies are partially deployed on top of each other. It only happens when the objective is near an edge.

Thanks Uncle Billy. That's useful.

At some point, perhaps we should create a sticky using this sort of stuff, so future scenario designers don't have to keep asking these questions.

Martin (J)