Spearman vs. Tank Probability

[joncha]

I voted never, but I think it should be statistically possible.

That is, about as possible as as winning the lottery (the 1 in 20,963,833 odds for Canada's Lotto Super 7 sound about right).

jon.

[notyoueither]

Considering it is 16 to 2, or 8 to 1, old fashioned board game odds charts would have tanks stopped (not killed) by Spear happening frequently. Add in defence bonuses and you could get into odds where an exchange would not be unlikely.

I think the problem is the image, myself. It is too bad that the worker graphics upgrades with the ages was not extended to obsolete units. Call them militia if you want.

[Rasputin]

i guess if people fed u[p with this maybe the should [play Multipalyer more or try harder levels ....

[Mr. President]

Originally posted by notyoueither


I think the problem is the image, myself. It is too bad that the worker graphics upgrades with the ages was not extended to obsolete units. Call them militia if you want.

They can't. They have no imagination.

[joncha]

This "use your imagination" or "pretend they're militia" stuff has got to be the weakest argument in favour of keeping the current system.

There is a militia system. It's called conscript.

There are irregular soldiers. They're called guerillas.

If you like the current odds for advanced vs. obsolete units, by all means defend it. If you can rationalize away the problem, good on you.

Me, I happen to like the GPT bug in the unpatched C3C. I tell myself its the reverse side of corruption. Bribes for the emperor and the like. But you don't see me going around telling people who don't like it that they just need to use there imagination more.

jon.

[Zero]

id be suprised if elite fortified spearmen on mountainous fortress surrounded by rivers and radar tower lost to a mere tank.

[notyoueither]

Originally posted by joncha
This "use your imagination" or "pretend they're militia" stuff has got to be the weakest argument in favour of keeping the current system.

There is a militia system. It's called conscript.

There are irregular soldiers. They're called guerillas.

If you like the current odds for advanced vs. obsolete units, by all means defend it. If you can rationalize away the problem, good on you.

I didn't say the player should imagine it. I said that it is too bad that automatic upgrades when a unit is too old are not part of the design of the game. They should be, and then there would no longer be these squabbles over spearmen.

No matter how you stack the numbers, and they are only 8 to 1 at the moment, there is going to be a possibility of the lower value winning a combat. No matter how small you make that possability, within reason, it will still happen sometime to someone. The solution is then to change the design so that when that event does happen it does not offend the sensibilities of the player(s).

[Monsto]

what i think is worse than the spearman/tank scenario, is the apparent lack of randomness in combat.

Last night: i plopped a toe-hold city onto the babylonians continent, few turns later i had 4 vet spearmen +1 elite, vet bowman, city walls, barracks, temple the whole nine yards. here come babylonian swordsmen...

elite spearman lost to regular swordsman. Next spearman lost 3 and was promoted. the other spearmen all lost. . . to regular swordsmen...

so i was pissed. so i reload the game from right before that conflict, and everything happened the exact same way. same patter of health loss, same everything.

why have a random generation system for anything if it's not going to be used at the core of the game? and if the answer is "to keep ppl from cheating in single player" why not let ppl do what is fun to them?

some ppl might accept it as part of the game when 6 health deals out 21 highly defended health, but there are also people that, like me, don't find it fun when the statistically improbable happens regularly.

like the unfortified warrior on grassland getting ambushed by knights. the warrior should lose almost immediately. Wasn't that the whole point of putting in the hp system in civ 2? to minimize the effect of "lucky shots?" sure a 3 pt standing warrier might crack off a lucky defense against the knight, but statistically is he going to do it 5 times? in a row? wtF? and then couple that with the fact that if you reloaded to right before that fight, it would have happened the same way because combat isn't randomized... either that or my version is screwery.

[Rasputin]

not very random no ,

but remeber this, if you toss a coin it can come down heads or tails,,, 50 :50

if you toss it again it still 50:50

so no matter how many times you toss the coin, even if all previous tosses were heads, it still 50:50 .. so another heaqd could quite possible occur. that is the strang ething with randomness, it is often not as radndom as you would expect it to be.

[notyoueither]

Monsto, do you have Preserve Random Seed turned on? It is an option when you start a new game. If you do and you reload 1000 times from just before the combat, you will get the exact same results 1000 times.

[homan1983]

Well I think it was fairly good the way it was before, a simple x/y chance. Just because a swordsman has 3 attack, it doesnt mean it should always beat a warrior. it should win 3/4 and warrior win 1/4. And was 1/4 a warrior wins only takes 1 life --> this means it would need to win around 3 sets without dying:

1/4*1/4*1/4 (although chance is higher, this is just to illustrate) this is ~1/64 chance, whilst its rare, it does give a warrior a chance.

Look back at our history, when great battles were fought and the outcome was sometimes surprising, it is ridiculous to want the higher number to win all the time, this is a strategy game and its not designed like Age of empire games etc.

They have already made the game pretty bad by "calculating the outcome 4 times and looking at the most common result". This has greatly further reduced the change of a warrior vs swordsman(though it still happens) and such things.

The problem is we cna't use multiple units to attack at once, this means that they cant make a battleship for example to always and easily beat a cruiser (now with the 4x calculation the whole game has changed, more expensive units are much more valuable than equivalent shield units of other type.. MUCH MORE...)

[bvoncranium]

Zero asserted:
id be suprised if elite fortified spearmen on mountainous fortress surrounded by rivers and radar tower lost to a mere tank.

Actually, you should be surprised, as the odds are that the veteran tank will win 79.8% of the time (4 hp 16-strength attacker vs 5 hp defender with strength of 7).
Mathematically yours, bvc

[bvoncranium]

notyoueither stated twice:
No matter how you stack the numbers, and they are only 8 to 1 at the moment, there is going to be a possibility of the lower value winning a combat. No matter how small you make that possability, within reason, it will still happen sometime to someone. The solution is then to change the design so that when that event does happen it does not offend the sensibilities of the player(s).

I think that your conclusion is reasonable.
The numbers deserve correction though:
The base odds for each ROUND of the fight are determined using the ratio of 16 vs 2.2 (extra .2 comes from minimum 10% defense bonus (for flat terrain)).
So in the first round the Spear has a 2.2/18.2 = 12.09% chance of hitting the tank.
The odds are the same each round, until one of the two units fighting has lost all of its hp (life).
From this math the odds become very long with each successive round. For example, the odds of a Spear doing 4 straight hits on an attacking tank will be (0.12089)^4 = 0.00021 (.021%).
When all of the combinations and permutations of hits done to each unit are considered, the probability of a veteran Spear with no defensive bonuses defeating an attacking veteran tank are only 0.6% - much less than 1/(1+8).

Obviously the odds get a little closer as the defender adds bonuses (fortified, difficult terrain, city/fortress, river crossing, radar tower, etc.).

I expect that these odds are still long enough so as not to offend the sensibilities of most players. The probability is non-zero though; some people will always find that unacceptable.

- bvc

Postscript: Firaxis are apparently investigating problems with streaks of abnormally high or low numbers in the RNG. If true, this could cause worse results than the math would predict.

[Tall Stranger]

Originally posted by homan1983
They have already made the game pretty bad by "calculating the outcome 4 times and looking at the most common result". This has greatly further reduced the change of a warrior vs swordsman(though it still happens) and such things.

The problem is we cna't use multiple units to attack at once, this means that they cant make a battleship for example to always and easily beat a cruiser (now with the 4x calculation the whole game has changed, more expensive units are much more valuable than equivalent shield units of other type.. MUCH MORE...)

Actually, this 4 roll combat change was never added to the C3C patch. It WAS proposed, but was widely denounced by lots of people, both here and at CFC, for exactly the reasons you describe. A 4-roll system would greatly unbalance the combat system. For this reason, Firaxis decided NOT to implement it.

[bvoncranium]

skywalker stated:
I voted "always"

While it is understandable that skywalker would vote 'always', it is funny to see that at this point more people have voted that a vet Spear should defeat a vet Tank 'always' (9) than the number who have voted for 'never' (8).
Do we conclude:
a) people just don't grok the combat system
b) Apolyton posters are anti-establishment deviants
c) Apolyton posters have a sense of humor ; or
d) All of the above

I -bvc

Postscript: I see now that a couple more 'Nevers' have rolled in. Personally, I think neither answer is right... but who am I to question the democratic process as practised by Civ players?