Strategic bombing problem

Fury Games has now signed with Matrix Games, and we are working together on the next Strategic Command. Will use the Slitherine PBEM++ server for asynchronous multi-player.

Moderators: MOD_Strategic_Command_3, Fury Software

User avatar
Jim D Burns
Posts: 3980
Joined: Mon Feb 25, 2002 6:00 pm
Location: Salida, CA.

Strategic bombing problem

Post by Jim D Burns »

I was watching the recent lets play videos linked here tm.asp?m=4158565, I noticed that the AI allies are conducting strategic warfare via bombings in 1940 while the full force of the German fighter arm is in the low countries. I also note it appears the AI does nothing else, no attacks at all vs. the German ground targets or airforce. Would it be possible to script the AI to not conduct strategic bombing at all until mid 43?

The cost in rebuilding their bombers and fighters far outweighs any advantage blasting a few German build points might bring. And the loss in new unit production due to the massive amount of points the AI is pouring into repairing its airforce will almost certainly lead to the fall of Suez.

By mid 43 the US should be in the war in strength and their added fighters would go a long way to making the strategic game start to cost the Germans more production overall instead of it costing the allies more. As it stands now the AI allies are losing the game for themselves due to the boneheaded strategic bombing in 1940. Unless the AI gets huge production cheats I see no way for them to overcome the losses they are taking from all the repairs they must be doing every other turn.

Jim
User avatar
Hubert Cater
Posts: 5862
Joined: Mon Jul 22, 2013 11:42 am
Contact:

RE: Strategic bombing problem

Post by Hubert Cater »

Hi Jim,

In early 1940 the British AF does not possess Tactical Bombers so the lack of strikes against ground targets or air units are to be expected. Fighters don't typically attack one another as direct targets, but rather only in the air as escorts and interceptors, which is what you are seeing when the Bombers make their strategic bombing runs.

As for the Bombers this is a bit of a tricky one as you are right that the potential losses are not an ideal return on their attacking investments... however it is a tough balance to get just right as the optimal play would indeed be for the AI to avoid these attacks until it was more prepared, but then we run the risk of players feeling the AI is not really doing much and that the game then becomes uninteresting. For now it takes losses but may at the same time act as a bit of a nuisance or slow down to the Axis AF if they too suffer losses and need to be repaired as they advance towards Paris.

Another example of this was how we originally coded the AI in France to play, where it originally had an optimal strategy of stubborn defense with minimal counter attacks. Idea being that if the AI positioned itself behind rivers and other key defensive positions and was able to entrench a bit, it was more work for the Axis to break through, potentially increased their losses, and prolonged the inevitable fall of France.

However during initial Beta testing, many felt that with the lack of counter attacks and some aggressiveness by the French AI, the battle for France was not as interesting as it could be.

As a result, and although we now run the risk of having the AI interpreted as slighter weaker than ideal, the French AI is now playing a bit more freely with counter attacks (often to its own detriment) and may even leave gaps in its lines. But the feedback is it feels more historical and allows for more of a Blitzkrieg on the German side which in the end is not a bad thing to have.

It's a tough balance to get just right as some players want the AI to take full advantage of its edge when it comes to understanding all the numbers in the game and play more efficiently than a human, while others hope for game play that mimics more of history.

Hubert
User avatar
Jim D Burns
Posts: 3980
Joined: Mon Feb 25, 2002 6:00 pm
Location: Salida, CA.

RE: Strategic bombing problem

Post by Jim D Burns »

Personally I say make the AI as good as possible, forget trying to make it feel busier. After all, your game AI will be judged on how hard it is to beat, not on whether it moves stuff around a lot and makes ill advised attacks.

AI's in PC games are bad enough already, intentionally making it worse just to make the game feel busy seems like poor game design to me. Perhaps have a separate 'optimal AI' scenario for players who want a challenge. Then warn players in the scenario text that the game will be less active as the AI will choose entrenchment over suicide attacks to preserve its fighting strength.

Jim
Biker1984
Posts: 27
Joined: Wed Oct 12, 2016 4:08 pm

RE: Strategic bombing problem

Post by Biker1984 »

Hi Hubert,

First of all, thanks for the detailed opinion and designthoughts you are addressing. I'm currently following the Beta videos of Paradogs Gamer and I'm happily surprised that the French AI is attacking and making it challenging for the German (player).
Also that the UK is redrawing the BEF back as the tide begins to change which is very positive.
I real think that I will add this title to my collection as these games are pretty rare.
What I noticed is during a playthrough of the Beta that the heavy panzers of the French in Rouen in May 1940 are not moving or attacking any weakpoints. Is there any indication that this will be addressed?
Also it appears that the French are staying a long time at the Magino line with there full force and planes/HQ are very near the front which makes exploitation (and direct attacking) of these units after the passing the first line pretty easy.
The naval warfare and the British AI look very good.

Best Regards and good luck with your release,

Raoul
User avatar
Christolos
Posts: 980
Joined: Wed Apr 23, 2014 10:45 pm
Location: Montreal, Canada

RE: Strategic bombing problem

Post by Christolos »

Hi,

I too was a little surprised (following the Beta videos of Paradogs Gamer) that the French armoured unit did move and/or attack.

I would have thought that the German corps at strength 5 North West of Paris would have been an easy target to eliminate and possibly opening up an attack on the nearby German artillery unit...


Image

Don't get me wrong though, I fully understand the issues of balancing aggressive play and all that and I am still eagerly awaiting release of this great game!

Another thing I noticed is that while it is possible to conquer Poland in as little as two turns (with 3 being more likely), it seems that an early victory against France (as per the historical surrender of June 22, 1940) is less achievable. I'm not sure though whether this could necessarily be attributed to a tough playing AI or the fact that the turns seem to advance rather quickly (1 week per each side). With this in mind, I am wondering how long the campaign could theoretically last. In other words, could the game be played beyond 1945, and if so, when would it end if no side was clearly victorious?

C
Attachments
Untitled.jpg
Untitled.jpg (188.22 KiB) Viewed 106 times
“Excellence is never an accident. It is always the result of high intention, sincere effort, and intelligent execution; it represents the wise choice of many alternatives - choice, not chance, determines your destiny.”

-Aristotle-
Biker1984
Posts: 27
Joined: Wed Oct 12, 2016 4:08 pm

RE: Strategic bombing problem

Post by Biker1984 »

ORIGINAL: CC1

Another thing I noticed is that while it is possible to conquer Poland in as little as two turns (with 3 being more likely), it seems that an early victory against France (as per the historical surrender of June 22, 1940) is less achievable. I'm not sure though whether this could necessarily be attributed to a tough playing AI or the fact that the turns seem to advance rather quickly (1 week per each side). With this in mind, I am wondering how long the campaign could theoretically last. In other words, could the game be played beyond 1945, and if so, when would it end if no side was clearly victorious?

C

Thanks for explaining my point with a picture, this is much appreciated.
I'm not completely convinced that the moves of the player we are discussing are the most optimal moves. For example I have not seen him once using "Forced March", which is results that ground units have double movement. This resulted in a significant delay of operation Fall Gelb.
I'm also watching Night Phoenix (he managed to capture Poland easily in 2 turns) and invests more in single units, is more aggressive in attacking the convoy lines (I'm still considering if that would be a good approach) and was completely prepared in March 23, 1940. In his last video it is currently April and the Netherlands and Belgium are both out of the war. This would be a more preferable approach IMO.
User avatar
Toby42
Posts: 1629
Joined: Sat Aug 09, 2003 11:34 pm
Location: Central Florida

RE: Strategic bombing problem

Post by Toby42 »

I'm little put off by the turns advancing so quickly? I'm not sure that I understand that?
Tony
User avatar
Hubert Cater
Posts: 5862
Joined: Mon Jul 22, 2013 11:42 am
Contact:

RE: Strategic bombing problem

Post by Hubert Cater »

ORIGINAL: Jim D Burns

Personally I say make the AI as good as possible, forget trying to make it feel busier. After all, your game AI will be judged on how hard it is to beat, not on whether it moves stuff around a lot and makes ill advised attacks.

AI's in PC games are bad enough already, intentionally making it worse just to make the game feel busy seems like poor game design to me. Perhaps have a separate 'optimal AI' scenario for players who want a challenge. Then warn players in the scenario text that the game will be less active as the AI will choose entrenchment over suicide attacks to preserve its fighting strength.

Jim

Hi Jim,

Thanks again for your feedback and in response I would say that we are not necessarily trying to intentionally make it worse, just more interesting while still maintaining as much quality as possible.

For example, and as mentioned above, our original French AI did a really good job of continually lining up behind rivers and forcing a steady grind for the Axis towards Paris. It was optimal and did take full advantage of the AI's number crunching ability (as it rarely if at all counter attacked due to the disadvantage of most attacks), but it was not exactly historical and the feedback was that it felt more like WWI than WWII.

For new players to the series during testing, while the AI played technically very well, this resulted in a bit of a turn off and probably rightly so.

It wasn't easy for me to let go of the optimal scripts I had crafted, believe me!, but the arguments against it I now firmly believe to be correct. Reason being is that my thinking on what made an exceptional AI shifted from one that played purely technically well, to one that not only played well but also met majority expectations. Essentially if the the AI feels more engaging and there is only a slight drop in overall effectiveness then this was probably a good thing.

Another example was how we originally scripted the USSR and it was much of the same, continual withdrawal using rivers and superior supply to its advantage (due to scorched earth) and that it would only counter attack once the optimal shift in overall strength balance occurred deep in the USSR.

Technically it did a great job, but it resulted in almost 18 months of a withdrawl/reinforcement of Soviet units with little to no counter attacks by the Soviet AI in that time. Now, and while still not perfect, there is a bit more ebb and flow and players will experience some counter attacks during the time of organized withdarawl/defense. Again, while this might be at times a bit detrimental in a purely technical sense, the game flow is improved and feels much more interesting/historical for the player.

One additional item of consideration is that these examples are very specific AI behaviors and not a reflection of the AI as a whole. Essentially we can craft the AI for different theaters, and for offensive and defensive situations as needed. The French AI will be noticeably different than the UK/US AI in North Africa, or the Soviet AI in 1943, or the German AI from 1939 to its high point in mid to late 1942.

If in the end if we can adjust the early UK and French AI to meet majority expectations at the cost of perhaps Paris falling a few turns earlier it is a compromise we are willing to make etc. But that being said there is always room for improvement and we are continually adjusting and tweaking as needed.

Hubert

User avatar
Hubert Cater
Posts: 5862
Joined: Mon Jul 22, 2013 11:42 am
Contact:

RE: Strategic bombing problem

Post by Hubert Cater »

ORIGINAL: Biker1984

Hi Hubert,

First of all, thanks for the detailed opinion and designthoughts you are addressing. I'm currently following the Beta videos of Paradogs Gamer and I'm happily surprised that the French AI is attacking and making it challenging for the German (player).
Also that the UK is redrawing the BEF back as the tide begins to change which is very positive.
I real think that I will add this title to my collection as these games are pretty rare.
What I noticed is during a playthrough of the Beta that the heavy panzers of the French in Rouen in May 1940 are not moving or attacking any weakpoints. Is there any indication that this will be addressed?
Also it appears that the French are staying a long time at the Magino line with there full force and planes/HQ are very near the front which makes exploitation (and direct attacking) of these units after the passing the first line pretty easy.
The naval warfare and the British AI look very good.

Best Regards and good luck with your release,

Raoul

Thanks Raoul and as a general rule the AI does not leave a city/resource unless enemy units are not in range to capture it, or only if another unit is available to replace it as a garrison to hold this city from enemy attack. We can override this behaviour in some instances but for the most part it works well throughout the map.

Further away from Paris the French AI is currently set to be more dynamic and you can expect more attacks and counter attacks, but keeping in mind the desired balance I described above to Jim, closer to Paris, French units are more resolved to hold fast in position and force the attackers to grind their way forward. Often this is advantageous to do so as it will be behind a river, entrenched, and with some of the new zone of control rules in place it makes forward progress that much more difficult for the Axis as they tend to have limited mobility. Rooting out a stubborn defender in an attempt to break through a defensive line can then be much more challenging for the Axis.

Additionally, the typical timing of the arrival of the French armor in this game usually happens once German units are in striking range of Paris, so more often than not the French armor will be holding a city/resource position, rather than behaving more dynamically as it did in the earlier part of the French campaign, and as seen in the screenshot above this area provides many opportunities to be behind a river for optimal defense. The AI is also scripted to attempt to swap out the weaker "at start" unit in Paris with a stronger unit and this is why often you'll see the armor unit make its way there or an Army and so on. This part of the AI mirrors what most human players will do at this point and that is to try and make the Axis work for every inch of French soil in the capture of Paris.

Also as mentioned above, this is not necessarily standard AI behavior to be expected for armored units throughout the game, but rather specific fine tuned behavior for this part of the war only.

For the Maginot units, this too is a careful balance as those positions are very good defensive positions and the goal here is to attempt to have the AI hold them as long as possible before withdrawl to a defensive line near Paris. Again, this is tricky since if they leave too soon that particular front collapses quite quickly making the drive to Paris easier, but if they attempt to leave too late they might be limited in their fighting withdrawl options as the rail lines may have been cut or their is insufficient funds to quickly operate them back to the newly desired defensive positions.

As a result, you'll notice some variety in the AI behavior here (from game to game) as it is mostly reactionary and depends more upon the effectiveness of the Axis assault and the reality of the battlefield than anything else. All this means is sometimes you'll see a very effective withdrawl from the Maginot, and at other times not if it is not possible for the AI to do so.


User avatar
Hubert Cater
Posts: 5862
Joined: Mon Jul 22, 2013 11:42 am
Contact:

RE: Strategic bombing problem

Post by Hubert Cater »

ORIGINAL: Biker1984
ORIGINAL: CC1

Another thing I noticed is that while it is possible to conquer Poland in as little as two turns (with 3 being more likely), it seems that an early victory against France (as per the historical surrender of June 22, 1940) is less achievable. I'm not sure though whether this could necessarily be attributed to a tough playing AI or the fact that the turns seem to advance rather quickly (1 week per each side). With this in mind, I am wondering how long the campaign could theoretically last. In other words, could the game be played beyond 1945, and if so, when would it end if no side was clearly victorious?

C

Thanks for explaining my point with a picture, this is much appreciated.
I'm not completely convinced that the moves of the player we are discussing are the most optimal moves. For example I have not seen him once using "Forced March", which is results that ground units have double movement. This resulted in a significant delay of operation Fall Gelb.
I'm also watching Night Phoenix (he managed to capture Poland easily in 2 turns) and invests more in single units, is more aggressive in attacking the convoy lines (I'm still considering if that would be a good approach) and was completely prepared in March 23, 1940. In his last video it is currently April and the Netherlands and Belgium are both out of the war. This would be a more preferable approach IMO.

From experience, and from what we've seen from testers Paris typically falls anywhere between July and September depending on tactics, experience level, and the number of times you've played the game as over time players generally improve quite a bit after they get used to the full SC system.

Essentially, on the first time through it may feel that the turns accelerate more quickly than expected, but once the system beds down for you, the turn lengths and the overall timing of the entire war will feel just about right, i.e. by the time you finish off France your preparations and repositioning for things like Sealion, Yugoslavia and/or the planned invasion of the USSR will be on or very near the historical timeline. This continues on throughout the war and on average it works out quite well and close to history as we've spent a lot of time fine tuning it.

One other consideration is that in this game the player is typically required to capture Paris in order for France to fully capitulate. Whereas if you compare this to the time you can likely reach the outskirts of Paris, and in many play throughs this will be June, then the argument can be made that the timeline of the game does indeed match the historical timeline here as well.

Of course we also tie in French surrender to their National Morale, which in our game gives the French player a bit more control over their fate as opposed to just replaying history, but it also means that they are only likely to surrender once you take Paris and this can add a few more turns. But again this is a part of the give and take we present in trying to balance out the historical imperatives versus optimal game play for both sides.

But that being said, we do have at least one Beta tester that is able to take Paris and force a French surrender in May, and no matter what AI strategy we use against his game play. His diverse use of tactics is, I would say, second to none, and often provides us with not only excellent feedback but some of our greatest design and balance challenges as he represents the other extreme of player type you'll find for this game.
User avatar
Hubert Cater
Posts: 5862
Joined: Mon Jul 22, 2013 11:42 am
Contact:

RE: Strategic bombing problem

Post by Hubert Cater »

Hi CC1, Treale,

Hopefully my other responses in this thread have answered most if not all of your questions/concerns here [:)]


Hubert
User avatar
Christolos
Posts: 980
Joined: Wed Apr 23, 2014 10:45 pm
Location: Montreal, Canada

RE: Strategic bombing problem

Post by Christolos »

ORIGINAL: Hubert Cater

Hi CC1, Treale,

Hopefully my other responses in this thread have answered most if not all of your questions/concerns here [:)]


Hubert

Many thanks Hubert for your very thoughtful and informative responses!

I am more excited than ever for the release date![:)]

C
“Excellence is never an accident. It is always the result of high intention, sincere effort, and intelligent execution; it represents the wise choice of many alternatives - choice, not chance, determines your destiny.”

-Aristotle-
User avatar
Hubert Cater
Posts: 5862
Joined: Mon Jul 22, 2013 11:42 am
Contact:

RE: Strategic bombing problem

Post by Hubert Cater »

ORIGINAL: Treale

I'm little put off by the turns advancing so quickly? I'm not sure that I understand that?

The game uses a system of longer turns in the Summer, medium length turns in the Spring and Fall, and shorter length in the Winter. The design idea here was that since most action, on average and over the course of the war, occurred in the Summer and less in the Winter, we wanted this to be well represented in game.

However, for a year like the Spring of 1940, where Germany conquered the Low Countries and France in 6 weeks, having a few less turns in the Spring can be a bit challenging, but later in the war, a shorter set of turns in the Spring is not as critical.

Again for us it is all about finding the right balance throughout the war on overall turn lengths and timelines with the universal turn length system we adopted for the game. For example, we not only need to keep in mind what needs to happen in the Spring of 1940, but also what happens in the Summer, Fall and Winter of 1941 and so on.

Hope this helps,
Hubert
Biker1984
Posts: 27
Joined: Wed Oct 12, 2016 4:08 pm

RE: Strategic bombing problem

Post by Biker1984 »

After looking at the video today from Paradogs Gamer I think the naval AI focus a lot of destroying single ships which is good to see and the most beneficial. After this I had 2 questions regarding the Mediterranean theater:

1) Will the Uk move ships from the Uk to the Mediterranean to strengthen them self if the fleet is losing the Mediterranean?
2) After the UK 7th armor was created it was presented with a lot of upgrades ( I believe 4), I also like how it was setup to the front. The only thing was that it would not engage the Italians which had no upgrades in 2 turns (meaning 4 strikes). This seems (like the armor in France) that it missed an opportunity here to bring heavy casualties to the Italians making them fall back like in the real war without the Africa corps. Any reason why the 7th armor did not attack here?

Thanks a lot for any feedback. I'm reallyu looking forward to the game.

Raoul
User avatar
Hubert Cater
Posts: 5862
Joined: Mon Jul 22, 2013 11:42 am
Contact:

RE: Strategic bombing problem

Post by Hubert Cater »

Hi Raoul,

1) It does send some initial ships to the Med at the beginning of the war, but currently it doesn't send any further down in case it begins to lose badly. Part of the reason is that there is always the risk of a Sealion, so I've opted to have the UK AI keep a minimum number of ships at home, and the Kriegsmarine is well represented in this game as well and the UK will need every ship it has to win the Battle of the Atlantic. Lastly, and if the thought was to send some down after the threat of Sealion has passed, usually if the UK navy loses in the Med, then Egypt shortly follows. At that point there is little reason for the UK navy to risk more action in the Med, and typically the remnants in the Med are sent back to the UK either way if the Suez is lost.

2) This was an interesting one that I had to recreate on my end to really understand what happened, and what seemed to have occured was that in order for the 7th Armor to reach that particular position at the front, i.e. from where it initially deployed in Egypt, it needed to force march due to the zone of control the Italian units were exerting over that position. As a result there was no attack as units cannot attack after a forced march and on the subsequent turn there was no attack once again as 7th armor lost too much morale from the forced march. On my end, by the time the turn returned back to the Allied side 7th armor was at 35% morale versus around 70% morale for the Italian armor. Checking the attack predictions gave me a 2:2 but the AI also checks to get a rough idea what the return damage will be from a counter attack on a subsequent turn to get a better idea if an attack on this turn is actually a good idea or not. On the Axis side the Italian armor has an advantage of 3:2 so all of this combined adds up to no attack for the 7th armor for at least 3 turns. Perhaps once the skies clear and the UK air hits a few frontline targets and weakens them this might change the likelihood of a 7th armor attack, but from the video there is currently a sandstorm so we'll have to see what happens once the skies clear.

As a side note I did look into the situation of the French armor in Rouen and that particular spot, as well as Reims and Chalons, are currently set to have the AI simply hold those positions and not counter attack at all. Again part of the reason here is that counter attacks from those positions, especially Rouen as it is surrounded by a river, typically make things easier for the Axis, but I've decided to lift this restriction on these locations in case the French armor does position itself in one of these spots. So now if there is potential combat odds in its favor, and it is in one of these locations, it may now counter attack.
User avatar
Hubert Cater
Posts: 5862
Joined: Mon Jul 22, 2013 11:42 am
Contact:

RE: Strategic bombing problem

Post by Hubert Cater »

Forgot to mention one extra tidbit... the UK also only has 1 strike for its Tanks, i.e. the same as the Italians. This will only change once they achieve the Armored Warfare research level and then they'll match the Germans at 2 strikes.
Biker1984
Posts: 27
Joined: Wed Oct 12, 2016 4:08 pm

RE: Strategic bombing problem

Post by Biker1984 »

Thanks alot for this fantasic and detailed reply. This explains it well.
apec
Posts: 84
Joined: Fri May 01, 2015 10:55 am

RE: Strategic bombing problem

Post by apec »

ORIGINAL: Hubert Cater

Forgot to mention one extra tidbit... the UK also only has 1 strike for its Tanks, i.e. the same as the Italians. This will only change once they achieve the Armored Warfare research level and then they'll match the Germans at 2 strikes.

Hubert, is the Armored Warfare tech a general prerequisite for Tanks having 2 strikes or it is just for UK? In addition, what about the mechanized corps, is Armoured Warfare needed for them as well?

Thank you in advance for any info you can share on this,

regards,


AlbertN
Posts: 4201
Joined: Tue Oct 05, 2010 3:44 pm
Location: Italy

RE: Strategic bombing problem

Post by AlbertN »

I am quite worried of the "quick turns" too. Especially if the Fall of France is so behind on regular schedule, probably it also means Barbarossa will always be in '42.

And obviously the fact in 1940 there is strategic bombing which already "works". Before 1942 the strategic bombing effect of the RAF was quite negligible - and certainly their planes did not had escort range to reach Germany.
Biker1984
Posts: 27
Joined: Wed Oct 12, 2016 4:08 pm

RE: Strategic bombing problem

Post by Biker1984 »

ORIGINAL: Cohen

I am quite worried of the "quick turns" too. Especially if the Fall of France is so behind on regular schedule, probably it also means Barbarossa will always be in '42.

I really like the concept of having 2 weeks during summer times; this is also the challenge to have your troops prepared and not being to able attack every were. I believe the concept is similar to other grand strategy WWII turnbased games from Matrix/Slitherine.
I also like the current strong UK fleet in the Atlantic and that the UK is using a lot of aircraft which gives historical feeling making it a Sealion or early Barbarossa choice rather then having a high chance of doing them both.
I would rather have the strategic railroad movement cost reduced, so you are able to strategically redeploy you forces more. In the video of Night Phoenix you can see that Paris could be taken before the original date. So I will follow him to see how fast he can get his forces up to full strength for Barbarrosa June 1941.
Post Reply

Return to “Strategic Command WWII War in Europe”