quote:


And A-bombs are a huge liability for Allied victory if more than two are used. It's easy for the Allies to move from a win to a draw, or a draw to a loss, by using more than two. See Victory Conditions section of the manual.

quote:

ORIGINAL: erstad

Our favorite member of the species alces twincitiesus and I had some brief comments on victory conditions and a-bombs in another thread, but I wanted to pen a small novella without hijacking the other thread (any further [8|]) so I'm continuing the discussion here:

quote:


And A-bombs are a huge liability for Allied victory if more than two are used. It's easy for the Allies to move from a win to a draw, or a draw to a loss, by using more than two. See Victory Conditions section of the manual.

True, but they are a huge VP generator for the allies.

I've only dropped them in AI GCs. In three GCs I dropped at minimum two, and at max, with a GC I allowed to go to engine shut-down in 1946 even though the Allies had AV earlier, I dropped six. I've never seen them be huge VP generators. The engine doesn't have code to treat them as A-bombs, with radiation and flash. It treats them as simply huge chemical bombs with immense penetration (30,000 I think from memory.) They take out large amounts of random industry, similar to a very large Manpower attack, but they don't take out LCUs or planes on the ground.

I don't have the old combat reports any more, but I've never gotten anything remotely like the VP levels you report. Couple of thousand each, maybe. Not 65k in total, even with six. If someone has live combat reports I'd like to see them. Maybe I picked the wrong targets. Although Tokyo was hit in each game at least once.

For example, in my game with aztez (literally within days of reaching the end), I have ~97K points and he has 170K points, we've been oscillating around the 1.75 threshold for a couple week. (although I'm not sure aztez cares much about the points). Of that 170K, approximately 65K are a-bomb (roughly added from tracker on the days when things went really south).

If this is true then this acts as a driver for what you say below. I don't disagree with what you say below IF you get these kinds of VP results from A-bombs. I just never have.

So with the a-bombs, we will either end at Jap marginal (if the ratio goes back above 1.75, so allied decisive + 2 shifts + 1 shift) or Jap decisive (if the ratio stays below 1.75).

without the abombs, we would easily be inside the 1.24 range for a draw, and it would be for sure a Jap decisive (draw + 2 shifts).

This also may be true. I can't say without knowing what he did with conventional start bombing in the HI.

So the abombs are giving a fighting chance for a Jap marginal. (And if my opponent really cared about points instead of scrapping to the end there's a couple of things that could have been avoided that would have pretty much guarantee only a jap marginal victory.)

This is also hard to parse without more. I play for VPs. I understand that others don't, although I find it hard to understand how there's a middle ground.

If he had "merely" another 25K points, which seems in the realm of possibility for someone in fighter range of the home islands, then with abombs the game would end early with allied autovictory, adjusted down one to allied marginal; whereas without abombs the clock would run out and it would be a draw at best, and more likely a japanese marginal.

OK, I get that. But if an Allied player doesn't drive for AV, but instead backs off to let things ride without winning "just to see", that's not really an A-bomb or a victory conditions issue. The victory conditions seem to assume--correctly I think--that a player will seek to win as soon as it's feasible. I can see a player, in this case the Allies, waiting to force AV until the correlation of forces allow him to drive for a decisive rather than a marginal, but the conditions don't assume the Allies will wait forever just to see 1946. I don't think any set of V. conditions could handle both of those cases well.

So there are definitely cases where the abombs actually tip the scale a level, or even two, in the allied favor.

If, a big if in my experience, the A-bomb allows the Allied VP total to leapfrog then yes, it could happen that way. As I said, I've never seen these sorts of VP numbers accrue from their use.

Of course, if the allies are on track to autovictory without the third abomb, then they should refrain, because then it hurts. I won an allied autovictory in AE before abombs became available, so it can happen.

So my thoughts are that the third (and following) bombs are dropped if
- They are expected to push the allies over autovictory, and autovictory won't be achieved without them, OR
- Things are headed to a draw (pre-shift), and the a-bombs can push up to an allied decisive (pre-shift)

They would not be dropped if
- Allied autovictory is in the cards without them

It theoretically wouldn't matter much if
- Things are headed for an allied marginal victory (pre-shift). the abombs would presumably shift to an allied decisive which would then shift back.

Or, if my experience is more normal, they could fail to jump to decisive but still impose the penalty, in which case the Allied player just stepped in a big pile of "uh oh."

The good news is you have several years before being faced with that decision vs. 1EyedJacks!

I fully expect the Japanese to run away like little girls long before Oppenheimer gets to recite the famous phrase. [8D]

And of course, the overarching question is whether the victory conditions are properly balanced. They aren't, of course, it's only a question of degree (and direction). A game of this scope and breadth could never be playtested adequately to ensure the victory conditions are "fair." (not to mention constant changes since release, think about all the unit data and gameplay changes since the first release)

Here I disagree too. I don't think anything in the patches or OOB changes has fundamentally altered the game balance. I think the Victory Conditions are extremely fair IF, a big if, both players observe their assumptions. For the most part that means playing the game as a game and not starting o9ut saying things like "I play for the experience, not VPs." If you want to play for the experience fine, but don't blame the victory conditions if you never get to the destination. Not saying you personally, but in general.

quote:

And of course, the overarching question is whether the victory conditions are properly balanced. They aren't, of course, it's only a question of degree (and direction). A game of this scope and breadth could never be playtested adequately to ensure the victory conditions are "fair." (not to mention constant changes since release, think about all the unit data and gameplay changes since the first release)

Here I disagree too. I don't think anything in the patches or OOB changes has fundamentally altered the game balance. I think the Victory Conditions are extremely fair IF, a big if, both players observe their assumptions.

quote:

I don't even bother to pay attention to the game's victory point tallies, if you know what you're doing, you know whether you're winning or not. The game vps are worse than irrelevant. Ignore.

quote:

ORIGINAL: erstad

quote:

And of course, the overarching question is whether the victory conditions are properly balanced. They aren't, of course, it's only a question of degree (and direction). A game of this scope and breadth could never be playtested adequately to ensure the victory conditions are "fair." (not to mention constant changes since release, think about all the unit data and gameplay changes since the first release)

Here I disagree too. I don't think anything in the patches or OOB changes has fundamentally altered the game balance. I think the Victory Conditions are extremely fair IF, a big if, both players observe their assumptions.

What I mean by this is that in a game the size of Witp, there can't possibly have been sufficient PBEM playtesting to ensure that the victory point ratio thresholds are fair. (By fair, I mean that two competent opponents of equal skill would over an infinite number of games get a balanced set of results, e.g., all draws, or an equal number of allied/japanese victories of the same magnitudes). Is 1.75x the right threshhold? Does the autovictory rule of 4x, 3x, 2x by year plus the "run-out" rules give the allies and japanese an equal chance of achieving a decisive victory? etc.

I'm not arguing that the bias is necessarily to Japan or to the Allies (I have my suspicion, but no real data), just that it is incredibly unlikely that it's spot on given that it's impossible to PBEM playtest. (And, since my experience is that VPs are dominated by losses, it doesn't seem like AI to AI autoplays are going to mimic human play well).

You might be right that the subsequent updates haven't changed the victory levels much, although a few things that can be pretty significant (versus the initial release) are the toned down artillery and increased IJA China garrisons, both of which make it (appropriately) harder for Japan to steamroll China; and the loss of drop tanks on Judys (which makes a huge difference when IJN is lucky enough to get a 8 hex air exchange). But in the grand scheme of things maybe these are small potatoes.

I'll dig up some representative a-bomb results for comparison. Have any of yours still? given the dramatic differences we see, and given we are both competent observers, I almost wonder if the AI difficulty settings modulate a-bomb results (or perhaps strat bombing more generally and it just carries over to the abombs)

By the way, i do agree the military effects are not that severe.

quote:

ORIGINAL: erstad

quote:

And of course, the overarching question is whether the victory conditions are properly balanced. They aren't, of course, it's only a question of degree (and direction). A game of this scope and breadth could never be playtested adequately to ensure the victory conditions are "fair." (not to mention constant changes since release, think about all the unit data and gameplay changes since the first release)

Here I disagree too. I don't think anything in the patches or OOB changes has fundamentally altered the game balance. I think the Victory Conditions are extremely fair IF, a big if, both players observe their assumptions.

What I mean by this is that in a game the size of Witp, there can't possibly have been sufficient PBEM playtesting to ensure that the victory point ratio thresholds are fair.

You may be right, but in an asymmetric game like this I'm always suspicious of words like "fair." A subjective word. What is fair is that in every game the OOBs and the map don't change in the same scenario. The victory conditions don't change. The time limits, same. The phasing, same. Both sides know these things up front, or can easily look them up. Should a Japanese fighter be worth 1.276x as many VPs as a destroyed Allied? I think that's what you're asking. And I don't know. I also don't think many PBEM trials will tell you. There are no equally matched PBEM players, or any way to really define what that means. They're humans, not widgets. You could accumulate a large statistical DB of PBEM games, but that doesn't mean two newbies fit. Or that two vets, getting advantageous rolls in the areas of the game they're better at one time, and not in the next game, have the same outcomes.

(By fair, I mean that two competent opponents of equal skill would over an infinite number of games get a balanced set of results, e.g., all draws, or an equal number of allied/japanese victories of the same magnitudes).

Yeah, but you can never achieve this test in RL, even in infinite trials. "Competent opponents" and "equal skill" are fairly meaningless phrases in this context. Is Nemo more competent than jrcar? Is CR more than Greyjoy? What are the criteria?

Is 1.75x the right threshhold? Does the autovictory rule of 4x, 3x, 2x by year plus the "run-out" rules give the allies and japanese an equal chance of achieving a decisive victory? etc.

OK, those are very fair questions. When setting t hem I assume GG and Co. had to somehow weigh the advantages Japanese production control gave to that side, as well as a host of other things. It's long been commented on that planes are way, way too valuable versus ships if the manhours and resources invested in production are the measures. A pretty large merchant ship versus a fighter? Look at the VPs. Nuts. But it's equal between the sides, so the absolutes wash out. It's the relative that matters.

I'm not arguing that the bias is necessarily to Japan or to the Allies (I have my suspicion, but no real data), just that it is incredibly unlikely that it's spot on given that it's impossible to PBEM playtest. (And, since my experience is that VPs are dominated by losses, it doesn't seem like AI to AI autoplays are going to mimic human play well).

I think a lot of it was simply GG's decades of experience designing.

I'd disagree a bit on VPs and splits. In many games you're right, but in the design it doesn't have to be the case that losses dominate VPs. We see a lot of bash-a-thon games with little maneuver or subtlety. A lot of players don't play for territory and PTO WWII was perhaps the greatest territory-centric modern war ever. There are key bases that, when combined with the construction multipliers, are vast storehouses of VPs. But a lot of players don't take and build them.

quote:

ORIGINAL: erstad

Here's a before/after industry screen from the first abomb attack in a PBEM.

Manpower - 4 damaged, 9 destroyed: 188 VP
HI - 39 damaged, 61 destroyed: 1298 VP
Naval shipyard - 45 damaged, 55 destroyed: 1190 VP
Vehicle - 28 damaged, 14 destroyed: 336 VP
Kawasaki Ha-60 - 47 damaged, 53 destroyed: 1154 VP

total of about 4100 VP from one Abomb. Did all the math in my head so might not be 100% accurate but should be close.

Been hit by 12 abombs total. For two of them, tracker showed over 11K loss for the day almost all of which has to be abomb related, more typically it's in the 5-7K range (which would largely, but not necessarily all, be abomb)





[image]local://upfiles/13407/7F10853F4029432CAB97BD5B11CC0AB7.jpg[/image]

quote:

ORIGINAL: geofflambert

I don't even bother to pay attention to the game's victory point tallies, if you know what you're doing, you know whether you're winning or not. The game vps are worse than irrelevant. Ignore.

quote:

ORIGINAL: Bullwinkle58
What do you call the manager of the St. L. Cards when he tells the media "We stole four bases tonight, got some good hits, tried out our new bonus baby on the mound, and nobody got hurt. So we lost 12-3. We had fun. I never pay attention to the score anyway." What do you call that guy? I'll tell you--unemployed.

quote:

ORIGINAL: Bullwinkle58

OK, after more searching than I wanted to do, I found one A-bomb mission.

Some Manpower hits, some HI, some LI, some Resources, Oscar plant, Helen plant. I may have bombed Tokyo before this; I had been bombing the HI in increasing volume since autumn 1944. The AI might have repaired too. I don't know. But this is roughly a typical A-bomb mission I've seen playing the AI. I doubt the code is any different for PBEM, but I can't say for sure.

quote:

ORIGINAL: Bullwinkle58


Long ago Michael outlined how the code works to provide A-bombs. I don't remember the details, but I think there's random chances every month after 8/45 of getting one. To get 12 you have to be very deep into 1946 as well as have the Allied player get lucky.

quote:

I vaguely recall this. If memory suits, the display (image) doesn't show the destroyed units, just the total remaining and the number that are damaged. The destroyed units are 'where it's at' for VP generation of an atomic bomb hit.

quote:

ORIGINAL: Bullwinkle58

quote:

ORIGINAL: Bullwinkle58


Long ago Michael outlined how the code works to provide A-bombs. I don't remember the details, but I think there's random chances every month after 8/45 of getting one. To get 12 you have to be very deep into 1946 as well as have the Allied player get lucky.

I was curious about this and it wasn't hard to find.

"There is usually a build rate of 1 per 30 days - that is, a chance of getting a bomb of 1 in 30.
A roll of bad dice could mean none in a month.

I have a save game dated Jul 27,1945 and it has 2 atomic bombs in the pool.

_____________________________

Michael "

http://www.matrixgames.com/forums/fb.asp?m=3255077

quote:

ORIGINAL: erstad

Yup, that would suggest the 12 I (and the other poster) saw might be a little high, but with 1 every 30 days on average, 9-10 might be common. That wouldn't qualitatively change my comments on when Allies should/should not use them, but does somewhat narrow the VP ranges where they are advantageous to use (beyond the first two free ones, there's never a gameplay reason not to drop those).
I assume Michaelm means a 1/30 chance per day.

quote:

ORIGINAL: Bullwinkle58

quote:

ORIGINAL: erstad

Yup, that would suggest the 12 I (and the other poster) saw might be a little high, but with 1 every 30 days on average, 9-10 might be common. That wouldn't qualitatively change my comments on when Allies should/should not use them, but does somewhat narrow the VP ranges where they are advantageous to use (beyond the first two free ones, there's never a gameplay reason not to drop those).
I assume Michaelm means a 1/30 chance per day.

Long time since I took Probability, but a 1/30 chance per day, if the trials are independent, should almost never give a player 10 bombs. I suspect the trials aren't independent though, with some function built in to increase odds of generation in the late month, or maybe give the following month a boost if the previous didn't generate.

In the bold part, I am perhaps missing your meaning, but dropping 3 or more is a terrible decision if the Allied player is about to achieve auto-victory. It takes the win from Decisive in all cases, to Marginal.

Looking at your just-concluded game your opponent was so far behind the "normal" curve in VPs that dropping 12 didn't hurt him. Didn't help him either as he didn't get auto-vic. What was the final Japanese win given the penalty and VP totals? He didn't post that screen.

quote:

ORIGINAL: erstad

quote:

ORIGINAL: Bullwinkle58

quote:

ORIGINAL: erstad

Yup, that would suggest the 12 I (and the other poster) saw might be a little high, but with 1 every 30 days on average, 9-10 might be common. That wouldn't qualitatively change my comments on when Allies should/should not use them, but does somewhat narrow the VP ranges where they are advantageous to use (beyond the first two free ones, there's never a gameplay reason not to drop those).
I assume Michaelm means a 1/30 chance per day.

Long time since I took Probability, but a 1/30 chance per day, if the trials are independent, should almost never give a player 10 bombs. I suspect the trials aren't independent though, with some function built in to increase odds of generation in the late month, or maybe give the following month a boost if the previous didn't generate.

In the bold part, I am perhaps missing your meaning, but dropping 3 or more is a terrible decision if the Allied player is about to achieve auto-victory. It takes the win from Decisive in all cases, to Marginal.

Looking at your just-concluded game your opponent was so far behind the "normal" curve in VPs that dropping 12 didn't hurt him. Didn't help him either as he didn't get auto-vic. What was the final Japanese win given the penalty and VP totals? He didn't post that screen.

My wording you bolded might have been obtuse. What I was trying to say is "from a gameplay standpoint, one would always drop the first two a-bombs because they are 'free'. The decision comes when one drops (or not) the third"

With about 300 days, a 1/30 probability would give an expected value of 10. In my copious free-time I'll build/post a probability table for different numbers of a-bombs. Math is fun!

I'll post some stats later, probably not till the weekend.