PBEM Offline (More) Options

[kemal81]

Here it is another example. The battleship is picked before the carrier. You can also see my opponent is offline.

Like transports carrying a full load, the game sees the aircraft carrier and the fighters as a single unit when determining the cheapest unit. Battleship is worth 24 and the combination of AC & fighters is 34. Thus the battleship is killed off first.

So you are saying algorithm itself is wrong. Cause in a defensive position loaded ac is completely different than loaded transport. With loaded ac the fighters can still fight the battle even if they belong to other power. And if the ac is sunk the fighters can fly to adjacent friendly territory or friendly ac if available. While with the transports the land units can't fight, can't take additional hits and if the transport is sunk then they will die for sure. I am sorry but the analogy here is plain wrong.

[RobEng]

We have debunked the hypothesis that the combined value of the carrier and the fighter devalued the battleship in either strength or cost. We still do not have an explanation to why it chose the fighters over the battleship as the carrier is both cheaper and weaker than the battleship. In our testing, in a similar attack the server took the casualties correctly with the carrier going down before the battleship.

Looking at your screenshot did give me an idea though to make PBE defender casualty orders better. I emphasize the word better and am not saying this is perfect. We are not going to have a customizable casualty order but I was thinking if there was a simplified single criteria to determine whether to take the cheapest or the weakest units? It seems to me whether you think you are going to survive would be the primary determination. I should only be thinking of the cost if it is likely I will survive the battle. If all my units are going to die, I just want to take as many enemy with me. With that in mind, I was thinking how about a casualty system where:

• If the defender is currently weaker, it should take weaker units
• If the defender is currently stronger, it should take cheaper units

I am posting this out here for folks to shoot holes in it or to build on the idea. It will not capture every factor for determining casualty order but it is seems to be an improvement over deciding cheapest or weakest what we currently have. Determining who is ahead in a battle might be a challenge. The criteria for determining which to go is best if it is a mathematical value.

What do you think? Too simple? I don't play too much Axis but I think I have something here.

[gingerbenjamin]

Rob, this is a nice idea, but I still think you should work on the option for people to logon to take casualties. The reason why is I think this would be easy to program (am I wrong?), there are so many combinations where of fleet battles where you need a specific outcome, but they are in fact rare in games so adding this very easy to understand feature will not delay most games.

As I said, the idea is nice, but covering 100% of the scenarios that people might want is hard. Often a tranny surviving is the most important but it is always cheapest and weakest......

[RobEng]

Rob, this is a nice idea, but I still think you should work on the option for people to logon to take casualties. The reason why is I think this would be easy to program (am I wrong?), there are so many combinations where of fleet battles where you need a specific outcome, but they are in fact rare in games so adding this very easy to understand feature will not delay most games.

As I said, the idea is nice, but covering 100% of the scenarios that people might want is hard. Often a tranny surviving is the most important but it is always cheapest and weakest......

This option would effectively force opponents to be logged in at the same time or breakdown the combat rounds to requiring multiple log ins. Were you envisioning being able to designate particular territories "as must get casualties from player" zones? I am thinking that might not be easy to do within a game because what you are suggesting is changing how the game server runs the actual game (forcing the defender to log on to designate the casualties). I may be wrong but I am guessing that to make this happen it would have to be set at the start of the game creation and effect all combats. Even if you just did it for naval battles, the primary source of this problem, I am guessing that this would lengthen games by 50% because unless we were on at the same time, you would take your turn, 1st round of combat you attack, then I need to log on to take my casualties, make my attacks, then you will have to log on to take your casualties. This adds 2 extra log on instances per round of combat.

In what situations other than amphibious landings would you want to keep a transport?

[gingerbenjamin]

Rob,

not suggesting you should attach this feature to a particular territory, but make it a general option for fleets - assuming that it only applies when the defender has a real choice. Hence it would not apply to most of the opening battles sz 12, 15, tranny in 59, japan sub etc.., though these are critical battles all the choice is with the attacker.

In general, there are (at most) 0-2 fleet battles a game where defender really has a stack and a choice. One in the atlantic, one in the pacific. After that, nothing usually happens (e.g., Germany rarely wants to wipe out UK fleet twice, given that objectice is moscow).

I could be wrong....looking for other players to chip in here.......

Ben.

[rem400]

It sounds as if the "play live" OOL option is out of the question because the cost of implementing such a system and the added time to play out these battles far outweighs the benefit in the 1 or 2 naval battles in an even smaller subset of games where this would apply. Although not perfect, the current option of cheaper/weaker is ok despite its limitations. We all know how those limitations impact these battles.

Rob, I'm thinking that, absent any practical experience with the change, it's hard to see how a change in the OOL would result in any different results from the current cheaper/weaker option for the defender. It does take some thought to remember to toggle the option. (The submerge sub option is a whole different can of worms.)

[kemal81]

It sounds as if the "play live" OOL option is out of the question

It is for me. I simply don't have blocks of hours to play this game. I also disagree with that it matters only for couple naval battles. I had lots of land battles that it did matter. When the bomber should be killed is always a problem. I usually want the bomber to be killed before the fighters and after the land units, sometimes even before tanks. I almost lost a game because i couldn't pick the order I wanted. Another example is multinational kills which I have talked about before. And its a problem for me in every game when I am Germany as I want dd and transport killed before the subs, and the surviving 1 or 2 submarines to submerge. Sometimes also i can't move my Japanese or American fleets to where I want because the options are very limited. So I wait a turn. I don't understand why can't we use the code system I proposed so we can have different killing order for different territories. If it can't be done I think Rob's idea would be better than the current. Also if we can have subs last and submerge instantly that would be really good, so that at least it will fix my (and probably others') Geman naval opening.

[bmaster]

In what situations other than amphibious landings would you want to keep a transport?

Transports are almost always more valualbe to keep alive than subs, IMO, except in the case of a very close naval battle involving lots of ships and few air. Not all players would agree with this, but transports are the real strategic piece and subs comparatively irrelevant. Transports project threats to land targets, its important that they be in position, and they can threaten air units as well as subs cannot.

The order of loss options now seem to do a bad job of allowing defenders to opt to defend transports over subs, especially vs air attacks.

[Zhukov4]

It may be that OOL problems only effect a small percentage of games, but it always has the potential to effect a game. So with imperfect OOL, some aren't going to be satisfied and won't want to use PBE for competitive game play.

I'd note that if you're concerned about an OOL issue ruining a game, you can always hold off and not finish your turn, and insist that your opponent come on and play the round live so you can select your casualties.

Since the computer chooses casualties in the PBE format, there's no easy solution. It might be worthwhile to create a pop-up OOL interface that you could turn on or off as you like. But it sounds like a tough challenge, because in some battles the optimal OOL selection changes from round to round, depending on dice outcomes.

One possible solution is create an option where the attacker gets to choose his opponent's casualties when the opponent is not online. That's how OOL works in other PBEM formats. In cases where the OOL isn't obvious, you just email your opponent and ask for their selection.

Gingerbenjamin's solution is also good. With Ginger's solution (as I perceive it), you'd have an option "choose own OOL"... With this option on, the attacker should choose to resolve the problematic battle(s) first. After the problem battle is resolved, then the defender turns the 'select own OOL' button off, and play resumes as normal.

[RobEng]

Yes, a customizable OOL is off the table at this point, particularly if you want to have set for different battles in the same turn. We could probably include a "protect transports" or "sacrifice bombers" options but it would apply to all of your battles. Would those be better than my attempt to make a series of mathematical or "if x then" formulas to come up with a better auto-OOL option than either kill cheapest units or kill weakest units?

[bmaster]

if "protect transports" meant something uncontraversial like always losing subs before transports, I'd consider implementing that as the new general rule without any buttons.

As a button, "protect transports" that simply gave them higher value than subs in keeping alive would be my preference. There are other circumstances where a button to keep transports alive more than any other boat would be useful, but the vs subs piece would be the biggest thing IMO.

b

[whooper]

I think an option to choose subs over transports would be nice. (I do think that it should be a checkbox option and not turned on by default.)

It's a rare case when I would want to choose anything other than a sub over a transport. (Yes, it does happen, just not that often.)

To an extent, you just have to accept the causualty selection in a PBEM game as just part of the game. It's not perfect, but those are the rules and you have to adjust your tactics accordingly. (I'm not saying don't try to improve it, but keep the expectations realistic.)

[kemal81]

This thread is starting to become productive. I may not get exactly what I wanted but there are really good ideas coming up. Keep it up guys.

[gingerbenjamin]

Just some more evidence to show OOL is broken at the moment. Inexplicably takes two hits on BS but does not kill destroyer. I don't think there are any logical explanations for this (unlike with the carrier example, where we though carrier & planes might be grouped in calc and assessed as being more valuable).

And after I said I hadn't seen these problems....now I see them often lol!!!

BS first.jpg (426.51 KiB) Viewed 3038 times

[DocX]

I attacked Japan's 2BB 2AC and 4fig in zone 13 with BB, 2 cruisers, 2 destroyer and 3 fig. Got lucky and got 8 hit on the first wave and cpu choose 4 fig and 4 hits on BB, leaving the 2 AC. Was playing PBEM and one of my opponent was online during the time, but he was playing Germany so he couldn't choose.