Campaigns: What determines army ability to avoid battle?

[istari6]

First, let me compliment the team on a fantastic game! Been on a two month journey learning all about the period thanks to Pike & Shot.

I've now completed all the included scenarios, and have embarked on my first campaign: the English Civil War.

However, I'm becoming quite frustrated by the Royalists ability to continually slip away from my advances, while I seem to get pinned and brought to battle on unfavorable terms surprisingly often. The latest example (which prompts this post) is trying to finish off the main Royalist army in Northumbria.

The Royalists hold only Yorkshire, Northumbria and Cornwall in late 1645. I have one medium-sized army in Cumbria and my main force laying siege in Yorkshire. The last Royalist force is in Yorkshire. I attack north into Yorkshire, figuring the Royalists will either give battle or retreat into Scotland where they'll be pinned.

Instead, they slip by me and move into Yorkshire as I pass them into Northumbria. I have AP points remaining, so I attack southwards. Again they slip past me going north. We've ended up swapping positions twice. I'm out of AP (having combined forces), so I click End Turn. The Royalists immediately attack sideways into Cumbria where my lesser force is sitting. They have an advantage against this secondary army of something like 1300 to 600 points, yet bam, I'm forced to fight. This pattern of chasing a will-o-wisp on my turn, then immediately having my weaker force get attacked and forced to fight on the AI's turn is maddening.

Is there some logic to understand when the AI can refuse battle and when I cannot? Does it have to do with % of the force which is cavalry (assuming higher #s of cavalry would give both greater warning of oncoming enemy and serve as a covering force to allow the foot to slip away). Or some other dynamic?

I appreciate the campaign for the ability to raise one's own armies and the accumulating effect of victories and defeats on elan and experience. But the slipperiness of AI armies and my inability to benefit from the same has happened multiple times and is really souring the experience.

Any advice is welcomed!

Chris

[rbodleyscott]

The Royalists immediately attack sideways into Cumbria where my lesser force is sitting. They have an advantage against this secondary army of something like 1300 to 600 points, yet bam, I'm forced to fight.

These odds cannot be possibly be correct (unless you were the attacker). Player armies follow exactly the same rules and decision-making process as AI armies as to when they retreat, and are hence just as slippery as AI armies. If the odds were really 1300 to 600 points against you when the enemy invaded Cumbria, your army would certainly have retreated. I wonder if you did not take into account the up to 300 points of local units that may have been added to your army prior to the battle?

However, although both sides are equally capable of retreating when they deem the odds to be beyond their capability to overcome, those acceptable odds depend on the difficulty level set. (Player: AI. Odds include local units added to either army. Odds are in points, not men) :

o Captain: 100:105 (but your units get an additional quality tweak in their favour, and the enemy get a quality tweak against them)
o Colonel: 100:105
o Sergeant Major General: 100:118
o Lieutenant general 100:134
o Captain General: 100:134 (but the enemy units get an additional quality tweak in their favour, and your troops get a quality tweak against them)

These are the same odds as you will get on these difficulty levels in standalone skirmishes.

There is some leeway on these odds because either side may be working on inaccurate scouting estimates (which can be within 75% and 133% of the true figure), and because both AI and player armies will sometimes engage at slightly worse odds, or adopt a defensive position at even worse odds.

As you can see the AI usually seeks a moderate points advantage before it is willing to fight, on the assumption that the player is capable of winning at least 50% of the battles at those odds - because that is the difficulty level the player has set. This is because an AI can never be as good as an experienced human player, and this becomes more apparent the more experienced you get with the game. The AI therefore seeks greater odds to compensate for its disadvantage, and retreats if it cannot match them - as a human general would do if he knew that his opponent was superior.

Similarly, if you were playing on Sergeant Major General level, your armies would choose to retreat if the enemy have more than 18% advantage.

So if you set the difficulty level appropriate to your skill, the moderate odds in the AI's favour should only be enough to balance the game so that the battles remain challenging.

As a general rule, you are unlikely to win campaign battles on a difficulty higher than you can win skirmish battles, because the AIs preferred odds ratios are the same.

To summarise: Both player and AI armies are equally "slippery". The only difference is the odds at which they will decide to "slip" away - and those odds are under your control by setting the difficulty level.

The AI may seem more slippery if you constantly attack it with armies that it has no chance of beating. If you do this, it will always retreat. You have to give your opponent some hope of victory if you want them to be willing to accept battle. Because the AI tends to attack with an economy of force rather than overwhelming odds, your army is perhaps less likely to decide to retreat when the AI attacks.

Note also that neither AI nor player armies can retreat from a confrontation if they are attacking, unless the opponent adopts a defensive position that their general deems too strong to attack. So if you attack a province at unfavourable odds you are stuck with fighting the battle. The AI won't attack at unfavourable odds unless its scouting estimates are wrong.

[istari6]

Richard,

Thank you for your very detailed reply.

I went back and double-checked the map and you're absolutely correct. My Parliamentarians had 664 points in Cumbria and the Royalists had 1312 in Northumbria at the end of the turn. However, the Royalists were split in two armies. At the start of their turn, they attacked with the larger field army at 999 points, and once engaged, my army had grown to 813 points (due to local militias, etc.). I had been remembering the point differentials at the end of the turn, not accounting for the splitting nor the local reinforcements.

But the bigger point is that I was operating under a misunderstanding of strategic map gameplay. I was trying to outmaneuver the Royalists and concentrate strength against weakness. I didn't realize that the game is designed to avoid engagements until the odds are mildly in the AI's favor, no matter the terrain, mix of cavalry, experience of armies, etc. I'm not complaining - if the goal is setting up interesting tactical battles, these formulas make sense. I just didn't understand that the system was biased to generate engagements where I'm almost always mildly outnumbered. That was what was frustrating. I kept trying to setup Second Newbury and the King kept slipping away .

This is the most fun I've had in wargaming in years, thanks again for delivering such an excellent game.

[rbodleyscott]

But the bigger point is that I was operating under a misunderstanding of strategic map gameplay. I was trying to outmaneuver the Royalists and concentrate strength against weakness. I didn't realize that the game is designed to avoid engagements until the odds are mildly in the AI's favor, no matter the terrain, mix of cavalry, experience of armies, etc. I'm not complaining - if the goal is setting up interesting tactical battles, these formulas make sense.

Precisely so. This design decision was driven by annoyance at games where you are forced to fight out lots of battles that are foregone conclusions, but you daren't let the game auto-resolve them because the auto-resolve algorithm gives you much worse results than playing them out.

Of course it is also quite historical. It was rarely possible to force a general engagement against an unwilling foe. The English Civil War would have been a lot shorter if it was.

Although the King was caught at 2nd Newbury his forces were sufficient to hold the Parliamentarians to a standstill (and inflict higher losses on them than his own army suffered) and he managed to retreat handily enough the following night. This would suggest that the Royalist army was not that badly outmatched. This sequence of events could easily happen in the game, particularly if the King's scouting estimates underestimated the Parliamentarian presence. It is unlikely that the King would have fought the battle at all if the odds had been worse - he would have retreated, suffering some losses in the retreat. This is what the game does.

Many of the major Royalist defeats in the ECW (e.g. Marston Moor, Naseby) were due to attacking when they probably should not have done (at least from a tactical point of view, even if strategic considerations dictated otherwise). I would be hard-pushed to think of any (of significant size) that were due to being "caught" and forced to do battle at unsurvivable odds.

The strength against weakness approach works fine in the game for conquering territory, but not for forcing a field battle.

The Parliamentarian surrender at Lostwithiel is perhaps an exception, but the Parliamentarian force had painted themselves into a corner in the hope of being taken off by the fleet, which never arrived, and if they had shown a bit more initiative they could probably have broken out - as demonstrated by the Lostwithiel scenario in the game.

The strength against weakness approach works fine in the game for conquering territory, but not for forcing a field battle.

[istari6]

Oh, I agree on average this makes for a better game. Still, after the nail-biting prebuilt scenarios where I was almost always fighting outnumbered, there was a hunger to have at least one or two battles where I could just crush the enemy army thanks to superior maneuvering at the operational level. But now that I understand the game logic (and the rationale behind it), this is fine.

As for Second Newbury, it's true the King slipped away in the night, but my understanding is that there is considerable question about the Parliamentary leader's competence (or alternately, their motivations) in screwing up the "hammer and anvil" on the battlefield, and leaving the road entirely uncovered at night. I was hoping to fight Second Newbury as Cromwell had envisioned it . You do make a fair point that the P&S engine doesn't let the enemy escape scot-free, with a modest level of casualties/desertions inflicted during the chase.

I don't think I'd realized that the ability to refuse an engagement was a general feature of 16th & 17th century warfare. Interesting. So is it fair to say that the art of warfare at the operational level involved trying to keep the enemy retreating and gobbling up territory until reaching a point where either they HAD to defend a key point (hypothetically, London for the Parliamentary side), or they felt they could stand and win a fight (e.g. Naseby for the Royalists)?

Recognize that the P&S engine is tactical, and the campaign layer is mostly designed to help setup interesting tactical battles. However, would be interesting in a future version to have strongpoints or key prestige locations that had to be defended (Bristol, Hull, London) or suffer major loss of income or will to fight in the troops. Would create some anchors and add interesting decisions to the otherwise free-form swirling on the campaign map.

Cheers!

Chris

[istari6]

BTW - Pike & Shot fired my interest in the period, so I went out and bought the FOG Renaissance, Trade & Treachery and War of Religion books. For other folks on this forum, I'd particularly recommend the latter two for helping add a lot of historical information about these armies at a reasonable price.