The RFB Development Discussion Thread

[Beery]

Here's the place to discuss issues relating to the development of RFB.

I'll start the ball rolling.

Issue 1: I'm considering crew ratings for the Japanese escorts. The problem is that the game is overstating the chances of getting killed by escorts by a very wide margin - currently the game's survival rate is worse than it was for German U-boats in the Atlantic and that's way way waaay off. A whole lot of SH4's destroyer crew ratings are set to 'elite' and the abilities for the escorts seem virtually unchanged from SH3 - i.e. if you're faced by more than one elite it means certain death no matter how deep you dive or how good your escape tactics are. It appears that elites cannot lose you once they have your signal locked. I'm considering doing what we did in RUb and simply replacing all level 4 (elites) with level 3 abilities. Does anyone who uses RFB have anything to say on this? Would it be better to reduce ratings across the board (i.e. 1 to 0, 2 to 1, 3 to 2 and 4 to 3)? Also, how were Japanese crews in real life? In RUb we left in a few level 4s because the Allies did indeed have seriously dangerous hunter-killers who were excellent at their job, but I'm not sure if the Japanese had such units. The high survival rate of real US subs argues against ANY level 4 escorts, as the game's level 3 AI will occasionally kill us.

Issue 2: regarding contact reports, I'm thinking of instituting an 'across the board' adjustment so that contacts are reported only 20% of the time (just as they were in RUb). The problem is that at the moment there's no use in stopping every now and then to check for sound contacts because the game is set to report contacts pretty much 100% of the time. With 20% reporting we'd have more incentive to play the game while transiting to our destination. Right now it's just a matter of taking the game to high TC, toggling between surface at night and periscope depth during the day (to avoid air attack) and waiting for the reports to come in. This results in boring play and high tonnage gained for no real effort.

[tater]

1. I haven't really seen this effect, I very rarely get DCed at all, maybe it's my sneaky nature. It's probably not a bad idea though, to consider AI 3 as "normal" for Fleet DDs. A few 4s here and there. I'd not use less than 2 for any warship normally, at least not one at sea... training in the Inland Sea or Sea of Japan, maybe .

2. By this you mean the contact reports that are part of the "Group Properties," correct? I agree they need a massive lowering, I have already done so in my experiments. 20% might be too high if you change nothing else about the layers since the chance of a mission spawning for many of those is 70% every 140 hours. I dropped the contact report %, the % for many to spawn in the fisrt place, and I increased the time between attempts to respawn the convoy/TF/etc. In the case of TFs, I looked at the RT time to the point the mission starts, added a week or two, and used that.

[Beery]

Originally Posted by tater

1. I haven't really seen this effect, I very rarely get DCed at all.

Attacking task forces using the unmodded game I found that almost all of them are set up with 15 escorts - all level 4 - with a 30% spawn rate. This means that each task force has about 5 elite escorts. I've never been able to escape when two or more of these guys have latched onto me. On the other hand I've yet to test going under a layer - that might work. If it doesn't though, I'm probably going to reduce them to CrewRating=3 because at 4 these guys are virtually certain death, and I'm of the opinion that if a sub got deep it should be virtually undetectable, even by elite crews.

2. By this you mean the contact reports that are part of the "Group Properties," correct? I agree they need a massive lowering, I have already done so in my experiments. 20% might be too high if you change nothing else about the layers since the chance of a mission spawning for many of those is 70% every 140 hours. I dropped the contact report %, the % for many to spawn in the fisrt place, and I increased the time between attempts to respawn the convoy/TF/etc. In the case of TFs, I looked at the RT time to the point the mission starts, added a week or two, and used that.

I tried reducing the spawn rate but I wasn't all that happy with the result. If I can reduce the chances of finding ships I'm happy to leave the populations high. What I'm really trying to do here is make the player work a bit more for contacts - I usually find that when everything is handed to the player on a platter it reduces the fun factor, and unfortunately in the unmodded game contacts are definitely handed to the player. Adjusting the rate of respawn is an interesting idea that I hadn't considered.

[Redwine]

Originally Posted by Beery

Here's the place to discuss issues relating to the development of RFB.

I'll start the ball rolling.

Issue 1: I'm considering crew ratings for the Japanese escorts. The problem is that the game is overstating the chances of getting killed by escorts by a very wide margin - currently the game's survival rate is worse than it was for German U-boats in the Atlantic and that's way way waaay off. A whole lot of SH4's destroyer crew ratings are set to 'elite' and the abilities for the escorts seem virtually unchanged from SH3 - i.e. if you're faced by more than one elite it means certain death no matter how deep you dive or how good your escape tactics are.

The survival probability in this game must be about 83%, the life was much more easy for USA crews than for German crews.

I note some DDs are too weak, and anothers like Akizuki are so dangerous.

Did you tryied the setting into my DDs Detection Mod for SH III ?

May be it can help....

[tedhealy]

Originally Posted by Beery

Here's the place to discuss issues relating to the development of RFB.

I'll start the ball rolling.

Issue 1: I'm considering crew ratings for the Japanese escorts. The problem is that the game is overstating the chances of getting killed by escorts by a very wide margin - currently the game's survival rate is worse than it was for German U-boats in the Atlantic and that's way way waaay off. A whole lot of SH4's destroyer crew ratings are set to 'elite' and the abilities for the escorts seem virtually unchanged from SH3 - i.e. if you're faced by more than one elite it means certain death no matter how deep you dive or how good your escape tactics are. It appears that elites cannot lose you once they have your signal locked. I'm considering doing what we did in RUb and simply replacing all level 4 (elites) with level 3 abilities. Does anyone who uses RFB have anything to say on this? Would it be better to reduce ratings across the board (i.e. 1 to 0, 2 to 1, 3 to 2 and 4 to 3)? Also, how were Japanese crews in real life? In RUb we left in a few level 4s because the Allies did indeed have seriously dangerous hunter-killers who were excellent at their job, but I'm not sure if the Japanese had such units. The high survival rate of real US subs argues against ANY level 4 escorts, as the game's level 3 AI will occasionally kill us.

Issue 2: regarding contact reports, I'm thinking of instituting an 'across the board' adjustment so that contacts are reported only 20% of the time (just as they were in RUb). The problem is that at the moment there's no use in stopping every now and then to check for sound contacts because the game is set to report contacts pretty much 100% of the time. With 20% reporting we'd have more incentive to play the game while transiting to our destination. Right now it's just a matter of taking the game to high TC, toggling between surface at night and periscope depth during the day (to avoid air attack) and waiting for the reports to come in. This results in boring play and high tonnage gained for no real effort.

I agree with number 2. Cut the contact reports. I haven't really played enough to make informed comments on number 1, but I will anyways Anytime I've had the depth, I've always been able to get under a thermal layer and get away. The only times I've been killed lately is in shallows or pushing my luck at periscope depth trying to get that last salvo of torpedoes off before that DD 350 yards away runs me over. But like I said, I haven't played enough to consider my recent experiences the norm for me yet (career on very hard if that makes any difference).

[Beery]

Originally Posted by tedhealy

...career on very hard if that makes any difference.

By the way what does the career difficulty actually do? I've never been able to figure this out and it's not mentioned in the manual at all. Does it increase AI effectiveness at all? Does increasing it increase the chances of retiring earlier? Does anyone know for sure what it does? I always set it to 'Normal' because I haven't got a clue what it's for.

[akdavis]

I'm pretty sure it just sets the difficulty options.

I disagree on the "survival rate" issue. By the simple nature of it being a game, players are going to have many more opportunities to take risks and are going to take those risks out of all proportion to reality. There are also so many other aspects of the AI that are deficient and can be exploited that some redress must be made in other areas. Also, I'm not of the opinion that you should always be able to escape once detected. There were US subs in the war that weren't able to escape after being detected. We just don't get to read their stories.

I think that using historical survival rates as the groundwork for realism is going to prove very problematic. There were so many other factors that affected historical survival rates that is not going to be a good yardstick to measure realism against.

[Beery]

I'm not saying that players should always be able to escape. I'm just saying that it shouldn't always be impossible to escape from certain AI levels - and that seems to be the case right now when it comes to level 4 AI. No real Japanese escort was equipped with super sonar that could pinpoint a sub 400ft beneath the surface, and the game's AI also should not have such an unrealistic advantage. Players will get killed often enough from the threats that exist in the game without having a guaranteed killer AI running around.

As for it not being a good idea to simulate actual survivability, I disagree. I did it with RUb and RUb was the most popular mod for SH3 among players who strove to play historically. I see no point in making a mod that factors in a player's tendency to play carelessly - my mods aren't intended for such players. Players of RUb and RFB should always play as if their lives depend on the outcome. That's the only way these mods can be played properly. Anyway, real skippers sometimes took suicidal risks, so I think the notion that we all play too carelessly to be able to simulate reality is a false one. Real skippers were brave and went into harm's way at probably the same rate a careful player of SH4 does.

[Coolhand01]

Berry,

If you could when looking at the AI levels also check the merchants. With the present rate of fire for the deck gun I find myself being out-gunned by about 4 to 1 against an armed merchant. SOB's put a nice hole through my pressure hull last night and took out all 4 diesels with one shot. THX......CH

[akdavis]

Originally Posted by Beery

As for it not being a good idea to simulate actual survivability, I disagree. I did it with RUb and RUb was the most popular mod for SH3 among players who strove to play historically. I see no point in making a mod that factors in a player's tendency to play carelessly - my mods aren't intended for such players. Players of RUb and RFB should always play as if their lives depend on the outcome. That's the only way these mods can be played properly. Anyway, real skippers sometimes took suicidal risks, so I think the notion that we all play too carelessly to be able to simulate reality is a false one. Real skippers were brave and went into harm's way at probably the same rate a careful player of SH4 does.

I'm not saying it shouldn't be a factor, but I would not use it as the baseline for realism decisions (i.e. will this or will this not cause an overall improvement in survival rates). I'm also not saying artificial difficulty should be inserted wherever possible, but some heightened difficulty in specific areas compensates for much greater ease than real life in other areas. Take for example your average Japanese task force. A powerful, well-escorted task force that has had its AI tuned in such a way to bring about a global reduction in loss rate will present an incredibly easy target for a player, as it will not be zig-zagging and will be behaving in other ways divergent from reality.

Furthermore, overall historical survival rate is based on a global tally of all subs and all missions, a reality not likely to be reflected in the game. While the survival rate of patrols to some areas was 100%, the survival rate for patrols to other areas was disastrously low. Take, for example, the suspension of patrols to the Sea of Japan in 1943. Overall global survival rate just doesn't tell the whole story. The U-boat campaign was a completely different equation with a whole different set of factors affecting survival rate. I don't think the two are directly comparable in terms of realism within the game.