Combatintman

I ended the last post talking about compromises, you will see that I will bring up compromises and workarounds in most of my posts rather than devote a whole section or post to them. The ability to adjust when making missions is a key skill in mission design and I hope that you will see things throughout this tutorial that will assist you in the process.
 
This post will mostly deal with the geographical area in which the mission is set and the ‘frame’ for the mission.
 
Obviously then it would be helpful if we identified the terrain over which the mission is to be fought.  This means going back to some research and cracking open your mapping application(s) of choice. Fortunately again, Mr Bin Laden’s notoriety means that some helpful person has actually plotted the compound in Google Maps here …
 
https://www.google.com.au/maps/place/Osama+bin+Laden's+compound+in+Abbottabad,+Abbottabad,+Pakistan/@34.1688404,73.2420684,273m/data=!3m1!1e3!4m5!3m4!1s0x38de31bac2c3add9:0x8243435e0624afa9!8m2!3d34.1693431!4d73.2424638
 
This is extremely useful but I prefer to use Google Earth as my programme of choice so I need to do a bit of imagery analysis to find the location
 

 
Having found it I have place marked it so I can find it later as shown in the screenie above.
 
Also note that I have used the historical imagery tool to show the place as it looked in 2010 because if you look at current imagery the place has changed dramatically ….
 

 
So time for more map study.  The initial thought is that this could be a bit of a tricky one to pull off as a workable mission mainly because of the proximity of civilian buildings which could complicate Blue player options or make the narrative unbelievable.  But rather than give up yet I need to zoom out to see whether I can fit something workable into the CMSF playing area (max 4km x 4km) and I am also looking for the Pakistani military facilities which I know from my research to be in the area.
 
So here is what I found …
 

 
This is a bit rough and ready but if you zoom in on Google Earth you can see that it is a huge military facility.  For the purposes of the mission, because we’re not going to storm the facility it is unimportant, what I am looking for is something to support a narrative of the Pakistani Army reacting to the US raid which was the contingency planning insisted on by President Obama.  So I’m just confirming its existence and how far away from the compound it is.
 
I have also highlighted two areas as ‘Possible Infil/Exfil Locations’.  Here I’m thinking about a concept of the SEALs, having executed the mission being required to withdraw to an LZ for pickup.  The two highlighted areas are fields surrounded by trees which tactically would support helos landing in the middle and provide a defensible perimeter.  I may not employ this in the final mission but I want to examine the possibility.  It also helps frame the size of the map I will need to create in the editor.  The deduction is that I can probably work with something in the 1km x 1km range which means that this is eminently achievable mapping-wise without compromising too much on cropping distances or replicating the real ground.  This is a key point for me when designing ‘historical’ or ‘semi-historical’ missions (this one will be what I call ‘semi-historical’).
 
Having ticked the ‘doable’ box for ground, let’s return to wider feasibility issues …
 
Force ratios are about 1:1 which works.
 

 
As you can see from the above graphic of the raid, the Blue force insert is a problem because I have no helicopters to play with so I will have to just put Blue in suitable starting locations and configurations to simulate exiting the helicopters on the LZ.  I may have to compromise on historical accuracy and have the Blue Force setup as if it had been inserted by helo a few hundred metres away (this is where my ‘Possible Infil/Exfil Locations’ come in handy).
 
My nearest match for SEALs is going to be USMC and I have to accept that I cannot model the actual weapon systems and the dog is going to be a non-starter.  Neither are deal breakers.  I also need to include breaching charges in the Blue Force composition.
 
The PAKMIL facility is about 2km from the compound by road which works (when reaction times are factored in) in terms of a ‘fight their way out’ scenario.
 
The proximity of civilian houses presents challenges in terms of collateral damage for the Blue Player and also provide potential escape routes for Osama Bin Laden should I choose to have ‘escape’ as one of the Red AI Plans.
 
Both ‘Possible Infil/Exfil Locations’ are approximately 600m from the compound making both infil and exfil aspects of the mission workable.
 
There are no Pakistani forces in CMSF but I can use Syrians to simulate them.
 
Overall the ground offers sufficient cover and concealment and routes to give the Blue Player flexibility to adopt multiple COAs (Courses of Action).
 
The 30 minute flight time from the remote LZ for the Blue QRF is a workable factor if I want to include this element in the mission.
 
The compound is an irregular shape, made up of many structures and measures roughly 90 x 50m.  I might have to play around with the dimensions a bit but the takeaway is that it will provide assaulting challenges to the Blue Player.
 
So the sum of the parts is that potentially this is doable but I need to work out how the infil and assault are going to play out.  I would like to replicate a direct insert onto the target but need to know if it will work.  This means I will have to test this early on in the piece and as a result I can do this by building the compound and immediate surrounds. If the test works I can continue building the rest of the map.
 
The next post will show some of those map creation steps.

Ok a couple of people (yes you @Borg ) have been screaming out for a ‘Special Ops Mission’.  So I thought – hell why not?  But there is a twist to this … it is sort of a scenario design tutorial to encourage new mission design within CMSF.  As I hope to show, this isn’t difficult ... although this will be a long journey for sure.
 
First up it is worth quoting the great JonS in his scenario design manual …
 
‘Scenario design is an intensely personal and creative activity, and these posts are not in any way intended to be prescriptive, or a set of rules that must be adhered to. Instead it's an outline of how I go about this, and some guidelines that you might choose to follow. To misquote German doctrine; scenario design is an art, a free and creative activity, and each designer needs to find their own way of skinning these cats’.
 
So first up is the scenario concept or the narrative.  You’ve already got the concept … I thought ‘hell why not?’ to doing a Special Ops Mission.  Well if you’re going to do one, you might as well do a famous one so we’re going to do the kill/capture operation against Osama Bin Laden by US Navy SEALs on 01-02 May 2011.  The advantage here is that unlike most SOF operations, this one is pretty well documented which makes the research easier.
 
Whatever the concept or narrative it needs to be a fairly strong one but open to adjusting as you go through because as we will no doubt see, there is only so much you can do with the mission editor so compromises ultimately have to be made.
 
Basically then my narrative is that I want to create an approximate representation of the events of that night and throw in some plausible challenges along the way.
 
At this point it is probably worth stating roughly what happened and this is a Segway into the next step which is research (although there is an argument that the research quite often drives the original scenario concept – neither is right or wrong and the process isn’t entirely linear). Here is the summary view from Wikipedia
 
‘Osama bin Laden, the founder and head of the Islamist group Al-Qaeda, was killed in Pakistan on May 2, 2011, shortly after 1:00 am PKT (20:00 UTC, May 1) by United States Navy SEALs of the U.S. Naval Special Warfare Development Group (also known as DEVGRU or SEAL Team Six). The operation, code-named Operation Neptune Spear, was carried out in a Central Intelligence Agency-led operation. In addition to DEVGRU, participating units included the United States Army Special Operations Command's 160th Special Operations Aviation Regiment (Airborne) and CIA operatives. The raid on bin Laden's compound in Abbottabad, Pakistan, was launched from Afghanistan. U.S. military officials said that after the raid, U.S. forces took bin Laden's body to Afghanistan for identification, then buried him at sea within 24 hours of his death in accordance with Islamic tradition. According to a Pakistani official, the United States had direct evidence that Inter-Services Intelligence (ISI) chief, Lt. Gen. Ahmad Shuja Pasha, knew of bin Laden's presence in Abbottabad’.
 
So from the above I can draw out a couple of pointers which will become useful when I’m constructing the scenario.
 
Time 0100
Date 02 May 11
Location Abbottabad
Blue Force SEAL Team 6
Rather than quote the whole Wikipedia article (you can read it here … https://en.wikipedia.org/wiki/Death_of_Osama_bin_Laden)
 
I’ll just extract some other snippets that may be of use.
 
Overall Force Strength was 79 operators and CIA personnel 6 helicopters and a dog. The raid was carried out by two dozen operators with another two dozen operators waiting at a deserted LZ to act as a reaction force if the Pakistani Army intervened.
 
Inside the compound were 22 residents.
 
Further quote from Wikipedia …
 
‘When the National Security Council (NSC) met again on April 19, Obama gave provisional approval for the helicopter raid. But as he worried that the plan for dealing with the Pakistanis was too uncertain, Obama asked Admiral McRaven to equip the team to fight its way out if necessary’
 
The raid launched from Jalalabad in Afghanistan with a flight time of 90 minutes to the target and a QRF was positioned 2/3rds of the distance away (ie 30 minutes flight time away).
 
And more from Wikipedia about the execution of the mission
 
‘As they hovered above the target, however, the first helicopter experienced a hazardous airflow condition known as a vortex ring state. This was aggravated by higher than expected air temperature ("a so-called 'hot and high' environment’) and the high compound walls, which stopped the rotor downwash from diffusing. The helicopter's tail grazed one of the compound's walls, damaging its tail rotor, and the helicopter rolled onto its side. The pilot quickly buried the aircraft's nose to keep it from tipping over. None of the SEALs, crew and pilots on the helicopter were seriously injured in the soft crash landing, which ended with it pitched at a 45-degree angle resting against the wall. The other helicopter landed outside the compound and the SEALs scaled the walls to get inside. The SEALs advanced into the house, breaching walls and doors with explosives.’
 
Other useful bits and pieces from the article …
 

 
A rather fancy schematic of the compound
 
So this is all great stuff and I could take the research to deeper levels but before I do it is worth taking a moment to see if this passes the feasibility test as a mission.  Knowledge of the editor helps at this point so if you are unfamiliar you just have to kick things around and setup little tests.  It might be that those tests rule out the mission altogether because it is not achievable in the editor or that the mission is unlikely to be a fun experience for the player.  JonS frequently says that many historical missions do not make fun CM missions.
 
It is also worth thinking about what compromises or adjustments may be required to make your concept a workable mission and I'll discuss these as I go. Next time around I will cover the ground and some other factors in a bit more detail.
 
 

Or here ...
 

Only in the sense that any casualties incurred on the player's force as a whole have an effect on the victory conditions of that individual campaign scenario (e.g., a friendly force condition/casualty parameter or if those units have been set as unit objectives) which leads to being dumped out of the campaign or a campaign branch as dictated by the campaign script.  In the broad and generic sense of @Ithikial_AU's statement - it is correct and consistent with the manual which states that non-core units are not tracked in a campaign.

Or here ...
 

A quick check (CMFB) suggests the Mk-IV does:

Or put another way - like 90% of my AI plans - I love not urinating around with the exit before/exit after timings if I can avoid it.

No dramas - although it did remind me of an old Army adage "don't volunteer" - it turned out to be quite an epic undertaking but it also gave me an appreciation of the level of detail and research expected by Battlefront, which I guess is why serious wargamers love CM titles.

@Sgt.Squarehead - as I've said before I think, stop noodling around with your stuff.  There comes a point where chasing the 1% is counter-productive.  Tidy them all up, write the briefings and square the maps away and kick them out of the door.  One released and 'good enough for release' product is better than 200 sat on your hard drive.
Taking Ap Bac or more correctly, 'A Miserable Damn Performance' as an example - it's done ... is it absolutely perfect - probably not, I don't know yet because it will bundle with the next H&E build so people haven't had a chance to play it and give their feedback.  However, despite the fact I felt that it needed a couple more play throughs and there were a couple of things that I felt could be changed, those changes and tests would only be marginal improvements.
The important thing is that the scenario fits the vision (which is why I banged on about vision in my Neptune Spear scenario tutorial) that I had in terms of its look, feel and outcomes.  Whether AI Group 1 plan could be tinkered with so that it moves between order 3 and 4 at 4:30 instead of 4:15 because it would be slightly more realistic or not is something I'm over.  The player generally won't notice the nuance if the overall experience is good.
It was gruelling getting to that point - as I recall, you said something like - 'flip - how did you knock the map out so quickly?'  The answer was simple - real world map, focus on getting it done and again I wasn't going to worry too much about whether a bump in the ground here or a tree there looked better or would create some sort of challenge or not.  The joy of real maps is that you don't need to think about the ground - you just trace the map.  
The graft was testing the thing because each test was about a week's worth of effort.  I reached out to our mutual friend @MOS:96B2P only when I had a pretty good idea that the thing worked and had a checklist of things that I wanted an opinion on.  He graciously spent about a week going through it and gave me some helpful suggestions.  From there, I incorporated most of those suggestions and tested again.  Then I had to deal with a couple of changes in the mod build and tested again.  By then, although not totally happy, I figured that it was good enough and I am confident that it is. 
I say the above, which is why I reached out to MOS, was because my vision was to replicate the frustrations and frictions of that day and I wanted to confirm that I'd got that right.  I'm pretty sure that some people will hate the scenario for this reason but I know that I've delivered the vision and I explain this in the designer notes.  I am happy enough that, despite some imperfections, I can explain to anyone who plays it - that:
I put a bit of thought and effort into it I know how the thing works I know how it is meant to work I know why it is meant to work that way. I think you just need the same confidence to accept that 'good enough is good enough'.
 

No dramas - although it did remind me of an old Army adage "don't volunteer" - it turned out to be quite an epic undertaking but it also gave me an appreciation of the level of detail and research expected by Battlefront, which I guess is why serious wargamers love CM titles.

I get the impression you come back from the shops like this ...

I know warrant officers who would have had you on repeat restriction of privileges for your love of scruffy clothing and otherwise ... 😉
Just to add - professional soldiers look after their kit and the raggety-taggety look can be overdone and perhaps is not so common in certain environments as many may think ...

Same uniform same tour ...

I get the impression you come back from the shops like this ...

I know warrant officers who would have had you on repeat restriction of privileges for your love of scruffy clothing and otherwise ... 😉
Just to add - professional soldiers look after their kit and the raggety-taggety look can be overdone and perhaps is not so common in certain environments as many may think ...

Same uniform same tour ...

Or put another way - like 90% of my AI plans - I love not urinating around with the exit before/exit after timings if I can avoid it.

7.62mm NATO when fired from a GPMG will go through breeze blocks and bricks in heartbeat - so some small-arms fire would be expected to drop people in buildings constructed of those materials.

I did the TO&Es for the British units which were derived from the official publications and manuals for the 2008 timeframe.  The implementation of the TO&E to fit CMSF structures inevitably means that there are compromises with groupings here and there but, apart from the NBC element which Steve did not want included because it was too specialist, the recce regiment is as accurate a representation of this unit.
Ground surveillance radars, UAVs and EW teams were not widely deployed in the British Army at the time and, possibly with the exception of Ground Surveillance Radars (I'm working from memory here), not part of the establishment tables for this formation.  The teams you see in the surveillance squadron were primarily trained and equipped to conduct dismounted reconnaissance, be that close target recce or sitting in OPs mostly employing the mark 1 eyeball, binos and a radio.

I get the impression you come back from the shops like this ...

I know warrant officers who would have had you on repeat restriction of privileges for your love of scruffy clothing and otherwise ... 😉
Just to add - professional soldiers look after their kit and the raggety-taggety look can be overdone and perhaps is not so common in certain environments as many may think ...

Same uniform same tour ...

Obviously not for me then 😏

I get the impression you come back from the shops like this ...

I know warrant officers who would have had you on repeat restriction of privileges for your love of scruffy clothing and otherwise ... 😉
Just to add - professional soldiers look after their kit and the raggety-taggety look can be overdone and perhaps is not so common in certain environments as many may think ...

Same uniform same tour ...

DW makes good stuff

For those not familiar with the 'yellow handbag' cultural reference ...
https://collection.nam.ac.uk/detail.php?acc=2018-04-3-1
 
Made up that this is from the National Army Museum, although in my day 4 Armoured Division was Herford-based.

Although this doesn't focus on the battle for Berlin - possibly of interest to eastern front fans:
https://www.bbc.co.uk/programmes/m000p9tc
 

There are a number of aspects to this which I probably won't cover off entirely but here goes ...
In the early CMx2 days, a lot of people complained vociferously and frequently usually in threads called 'Extra Time for Scenarios' about early AI surrenders along similar lines to your original post i.e., 'I'd tee'd up the perfect plan to nail the enemy only to have the scenario end' - or as you put it ...
"Well I was all lined up for the final strike on the German Tigers. Been getting everyone in position for over 10 minutes."
It took people a while to work out that if you added a few extra units that the player never saw, you could create a scenario that would allow the player to execute their 'perfect' plan.  That is one of the reasons that the trick is widely used by scenario designers.
From a design point of view, I admit it probably isn't perfect and maybe the game should take into account other factors such as objectives held or the morale state of the opposition, although of course with the latter - that only becomes apparent at end game and still leads to the frustration at not being able to execute the 'perfect' plan.  Having a morale dimension would also completely kill certain scenarios at birth - my Ap Bac Scenario for the Heaven and Earth module absolutely would not work if morale played strongly into the equation as about 70% of the total ARVN force have shocking motivational and leadership factors.
Conversely to early surrenders, players don't like schlepping across the map for half an hour having knocked over the enemy and secured all of the objectives at the end of a mission to root out a single unit in a tactically insignificant location, and slot it, in order to achieve a victory.
FWIW, my scenario design philosophy has always been about the narrative and I am mindful of the fact that anybody who elects to download and play a scenario is making a choice.  This means you have to accept that it is transactional between player and designer and ultimately you have to try as best as you can to satisfy player needs.
That means your narrative and design concept has to either hit, or balance some quite often divergent factors and of course no scenario is ever going to tick everyone's boxes, but the closer you get to achieving that aim the better.  As a simple example of this - if I call a scenario 'the Battle of Waterloo' then it ought to include things like La Haye Sainte, an allied defending force, a French attacking force and some Germans rolling up late on in the game because that is what the player is expecting.  I would also do my best within the editor to make sure that the map and forces resemble the real thing as well as ensuring that the moving parts in the AI plan bear some resemblance to real events.  If I fail to do this adequately, the transaction between designer and player fails because the player is expecting to refight Waterloo.  Similarly, if I call a mission (or have as its premise) 'Platoon attack' I would design that mission so that the player gets to execute a platoon attack and can actually culminate that attack.  This means that the enemy doesn't surrender when it loses a handful of soldiers.
Since I learned the 'reinforcements that never arrive' trick I have found that it gives me more control over achieving my scenario narratives.  It certainly isn't the only trick by the way but it is one that I consistently use with what I would like to think are fairly successful results.
The transactional piece for me looks vaguely like this:
Players want to win. Players want to feel challenged. Players want to feel that they are a commander in a realistic environment. CM players expect scenarios modelled on real engagements to bear some resemblance to them. To achieve that, you give them the tools in terms of time, forces available to them (e.g., sappers if there are minefields), clear mission command-style orders, a strong narrative, a realistic setting and you set the objectives/VPs accordingly.
It is usually the latter that is the most nuanced mainly because objectives and VPs are how victory is judged and what the victory-focused player will therefore judge the outcome.  It is also nuanced in terms of surrender points - I'm happy that the AI will trigger a surrender at the point where the player has closed in on the last objective but will make absolutely sure that it doesn't happen before that because the player rightfully feels cheated.  The same is true of unit objectives of any description - you can't give a player the goal of 'destroying all the Tigers' and then have the AI surrender before the player has even seen them.
Your comment about reaching the point where you'd expect them to surrender is absolutely valid but is subjective - some will call a scenario/surrender unrealistic if it doesn't happen at the point at which the force suffers 30% casualties - CM has it modelled at around 60% - yet this didn't seem right to you based on the fact that some gucci armour was on the map in the engagement you were fighting.  There is a lot of truth in what @sburkesaid - you only know you were 'cheated' after the scenario ended and to a large extent are judging based on that godlike information given to you at ceasefire.  I recall testing one of @George MC's scenarios recently and had got to a point in the scenario where I thought - "I'm not getting anywhere here, I'll hit surrender and see if I can edge a win," - an excellent example of how friction and the essence of manouevre warfare works - shattering the enemy's will (or in this instance mine) had come into play.  As it happens, I thought - "I'm testing this scenario so I should play it through in order to give the best feedback" and about three turns later I got the sense that the enemy had culminated and that I could actually continue with a reasonable chance of achieving the set objectives.  About five turns later, the enemy surrendered.  In short - a brilliant example of scenario design which ticked most of the transactional boxes for me.
To conclude - balancing these factors and explaining it in the narrative to manage the player's expectations and then testing the scenario to make sure that the thing works is key to the whole scenario design piece.  Adding a few extra dudes that the player doesn't see to stave off the early surrender is a proven and effective tool that people use to best deliver a 'really good scenario which I recommend' comment versus 'the AI surrendered early' comment.  Not perfect of course and it would be better if the designer could control the ratios and maybe use a combination of tools.

@waffelmann - no dramas here, I've always been happy with the thing coming out when it comes out approach.

Sorry mate - forgot to add, I like what you're doing with this and there is real creativity and attention to detail.

There are a number of aspects to this which I probably won't cover off entirely but here goes ...
In the early CMx2 days, a lot of people complained vociferously and frequently usually in threads called 'Extra Time for Scenarios' about early AI surrenders along similar lines to your original post i.e., 'I'd tee'd up the perfect plan to nail the enemy only to have the scenario end' - or as you put it ...
"Well I was all lined up for the final strike on the German Tigers. Been getting everyone in position for over 10 minutes."
It took people a while to work out that if you added a few extra units that the player never saw, you could create a scenario that would allow the player to execute their 'perfect' plan.  That is one of the reasons that the trick is widely used by scenario designers.
From a design point of view, I admit it probably isn't perfect and maybe the game should take into account other factors such as objectives held or the morale state of the opposition, although of course with the latter - that only becomes apparent at end game and still leads to the frustration at not being able to execute the 'perfect' plan.  Having a morale dimension would also completely kill certain scenarios at birth - my Ap Bac Scenario for the Heaven and Earth module absolutely would not work if morale played strongly into the equation as about 70% of the total ARVN force have shocking motivational and leadership factors.
Conversely to early surrenders, players don't like schlepping across the map for half an hour having knocked over the enemy and secured all of the objectives at the end of a mission to root out a single unit in a tactically insignificant location, and slot it, in order to achieve a victory.
FWIW, my scenario design philosophy has always been about the narrative and I am mindful of the fact that anybody who elects to download and play a scenario is making a choice.  This means you have to accept that it is transactional between player and designer and ultimately you have to try as best as you can to satisfy player needs.
That means your narrative and design concept has to either hit, or balance some quite often divergent factors and of course no scenario is ever going to tick everyone's boxes, but the closer you get to achieving that aim the better.  As a simple example of this - if I call a scenario 'the Battle of Waterloo' then it ought to include things like La Haye Sainte, an allied defending force, a French attacking force and some Germans rolling up late on in the game because that is what the player is expecting.  I would also do my best within the editor to make sure that the map and forces resemble the real thing as well as ensuring that the moving parts in the AI plan bear some resemblance to real events.  If I fail to do this adequately, the transaction between designer and player fails because the player is expecting to refight Waterloo.  Similarly, if I call a mission (or have as its premise) 'Platoon attack' I would design that mission so that the player gets to execute a platoon attack and can actually culminate that attack.  This means that the enemy doesn't surrender when it loses a handful of soldiers.
Since I learned the 'reinforcements that never arrive' trick I have found that it gives me more control over achieving my scenario narratives.  It certainly isn't the only trick by the way but it is one that I consistently use with what I would like to think are fairly successful results.
The transactional piece for me looks vaguely like this:
Players want to win. Players want to feel challenged. Players want to feel that they are a commander in a realistic environment. CM players expect scenarios modelled on real engagements to bear some resemblance to them. To achieve that, you give them the tools in terms of time, forces available to them (e.g., sappers if there are minefields), clear mission command-style orders, a strong narrative, a realistic setting and you set the objectives/VPs accordingly.
It is usually the latter that is the most nuanced mainly because objectives and VPs are how victory is judged and what the victory-focused player will therefore judge the outcome.  It is also nuanced in terms of surrender points - I'm happy that the AI will trigger a surrender at the point where the player has closed in on the last objective but will make absolutely sure that it doesn't happen before that because the player rightfully feels cheated.  The same is true of unit objectives of any description - you can't give a player the goal of 'destroying all the Tigers' and then have the AI surrender before the player has even seen them.
Your comment about reaching the point where you'd expect them to surrender is absolutely valid but is subjective - some will call a scenario/surrender unrealistic if it doesn't happen at the point at which the force suffers 30% casualties - CM has it modelled at around 60% - yet this didn't seem right to you based on the fact that some gucci armour was on the map in the engagement you were fighting.  There is a lot of truth in what @sburkesaid - you only know you were 'cheated' after the scenario ended and to a large extent are judging based on that godlike information given to you at ceasefire.  I recall testing one of @George MC's scenarios recently and had got to a point in the scenario where I thought - "I'm not getting anywhere here, I'll hit surrender and see if I can edge a win," - an excellent example of how friction and the essence of manouevre warfare works - shattering the enemy's will (or in this instance mine) had come into play.  As it happens, I thought - "I'm testing this scenario so I should play it through in order to give the best feedback" and about three turns later I got the sense that the enemy had culminated and that I could actually continue with a reasonable chance of achieving the set objectives.  About five turns later, the enemy surrendered.  In short - a brilliant example of scenario design which ticked most of the transactional boxes for me.
To conclude - balancing these factors and explaining it in the narrative to manage the player's expectations and then testing the scenario to make sure that the thing works is key to the whole scenario design piece.  Adding a few extra dudes that the player doesn't see to stave off the early surrender is a proven and effective tool that people use to best deliver a 'really good scenario which I recommend' comment versus 'the AI surrendered early' comment.  Not perfect of course and it would be better if the designer could control the ratios and maybe use a combination of tools.