Multiplayer Spire of Eternity feedback

[KarlD]

Are there any math/probability people here? I have a question for you. Below is a simulation of the negotiation process in the Spire. It assumes 8 items labeled A-H are offered. On the left are what items are available to be offered at the start of each offer round and the right shows the results of that offer round. After round 2 only Spirit 2 is unsatisfied. Out of the 4 possibilities (CDGH) for round 3 I chose C because in round 1 C was in the wrong position. I understand that in Round 2 Spirit 1 needed C and I could not conclude that Spirit 2 also needed C, but I have this feeling that there is a higher probability that Spirit 2 needed C compared to D, G, or H. Does anybody know if I am right in thinking that?

Offer Round	Items Available to Offer	Spirit 1	Spirit 2	Spirit 3	Spirit 4	Spirit 5
1	ABCDEFGH	A-NOT NEED	B-NOT NEED	C-WRONG	D-MATCH	E-NOT NEED
2	CD FGH	C-MATCH	F-NOT NEED	G-MATCH	D-MATCH	H-MATCH
3	CD GH	C-MATCH	C-MATCH	G-MATCH	D-MATCH	H-MATCH

 

[Pheryll]

Assuming all the spirits pick randomly with equal chance of each good, then all four goods that could be selected in your third round are still just as possible (25%) from the information you received from the previous two rounds. There is no reason to suspect a favoritism toward choice C.

 

[DeletedUser12171]

What you wrote is not entirely true with regards to the Spire because there isn't really a "full price" that you pay all the time, every time. The process of convincing the spirits could yield a very cheap negotiation if you guess right on the first try, or very expensive if you somehow managed to offer goods so that everything was in the wrong position or not needed and you had to keep offering 5 sets of items and ended up needing to spend 75 diamonds.

It doesn't detract from my point that fighting gives you a better chance of solving an encounter at a lower cost

I've fully completed the Spire every week since the second week after it was released and I'll tell you that getting a negotiation "right on the first try", while theoretically possible, in reality never happens, at least to me. Fighting is a much more reliable way of solving the encounter cheaply.

That being said, fighting in the Spire is indeed harder than Tourney or Scouted Provinces, because the enemy matchups are often stronger. Manual fighting has more impact here than anywhere else. However, autofight in Spire is still viable and yields decent results. The key is in choosing your fights. I negotiate in Spire only in these two scenarios:

1. I have a high chance of losing many troops or losing the battle entirely
2. I don't have enough squads to fill 5 slots

A 3rd (less important) scenario where I nego is when the goods needed are those I have plenty of, i.e. T1-3 goods. Orcs especially, I don't like to use for negotiation because they are so slowly produced. And Sentient goods - wait till you have to research and upgrade in chapter 15 and you'll not want to negotiate with those in the Spire

 

[KarlD]

Assuming all the spirits pick randomly with equal chance of each good, then all four goods that could be selected in your third round are still just as possible (25%) from the information you received from the previous two rounds. There is no reason to suspect a favoritism toward choice C.

I totally get what you are saying and I am not even disagreeing with you, but consider this. What if instead of offering C to Spirit 1 in Round 2 I offered it D and got the result "D-NOT NEED". Spirit 1 still needs C, but we just don't know it yet. We would now say that the probability that Spirit 2 needed C was 50% and it was 50% that Spirit 1 needed C. It just seems odd to me that because I know LESS about what spirit 1 needs after round 2 that I somehow know MORE about what Spirit 2 needs.

 

[Deleted User - 3932582]

It just seems odd to me that because I know LESS about what spirit 1 needs after round 2 that I somehow know MORE about what Spirit 2 needs.

What's so odd about that? This setup is also a reason why if you get MATCH-MATCH-MATCH-WRONG-NONEED in the first round, you don't want to offer correct good to the last ghost in the second round even though you 100% know what that is (assuming still multiple options for #4). Because doing so won't give you any new information, and you effectively only have 1 slot on round 2 that will provide new information, and that may not be enough. You want to offer two different goods in round 2 - none if which is a correct one and you know it. Doing so will narrow down your outcomes for round 3 much more effectively.

 

[KarlD]

It doesn't detract from my point that fighting gives you a better chance of solving an encounter at a lower cost

I really wasn't trying to detract from your point. I'm a fighter and I'm convinced that the Spire is 10 times easier for me to get through because I manual fight my way through and rarely negotiate or autofight. I was just making a relatively minor point that it's not quite as cut and dry in the Spire as the others because of the variability of the negotiating process. So if I were psychic and could guess what the spirits wanted on the first try I would cater a lot more because it would be faster and frequently less expensive than fighting. Of course, as you said, it rarely happens that I guess right on the first try. I did it once when there's only 3 items to choose from.

I actually am in Chapter 15 and I am loathe to negotiate with my T6 goods. My boosted turned out to be the one that like 90% of the world is also boosted in so trading for other T6 goods is often slow and expensive so I will try very hard to avoid negotiating with those. I produce other goods in sufficient quantities that I'm not especially concerned about them. Fortunately I see the light at the end of the tunnel for getting through Chapter 15 so will soon greatly reduce my T6 issues.

 

[KarlD]

What's so odd about that? This setup is also a reason why if you get MATCH-MATCH-MATCH-WRONG-NONEED in the first round, you don't want to offer correct good to the last ghost in the second round even though you 100% know what that is (assuming still multiple options for #4). Because doing so won't give you any new information, and you effectively only have 1 slot on round 2 that will provide new information, and that may not be enough. You want to offer two different goods in round 2 - none if which is a correct one and you know it. Doing so will narrow down your outcomes for round 3 much more effectively.

1) Your reasoning is correct. But, statistically speaking, it will cost you more goods while, statistically speaking, save you diamonds. One may or may not value the diamonds over the goods.

2) I understand what you're saying, but I don't feel it addresses my situation because in my example I seemingly know MORE about one spirit just by virtue of the fact that I know less about another spirit. When I didn't know what Spirit 1 needed there was a 50% chance that Spirit 2 needed C. But by gaining the information that Spirit 1 needed C, I seemingly now know less about Spirit 2 as there is now only a 25% chance that Spirit 2 needs C. I'm not talking about tactics for how to more effectively gain more information as your example gives. In my example I actually learned something about Spirit 1 and that very knowledge seemingly made me know less than I knew before about Spirit 2.

 

[Pheryll]

2) I understand what you're saying, but I don't feel it addresses my situation because in my example I seemingly know MORE about one spirit just by virtue of the fact that I know less about another spirit. When I didn't know what Spirit 1 needed there was a 50% chance that Spirit 2 needed C. But by gaining the information that Spirit 1 needed C, I seemingly now know less about Spirit 2 as there is now only a 25% chance that Spirit 2 needs C. I'm not talking about tactics for how to more effectively gain more information as your example gives. In my example I actually learned something about Spirit 1 and that very knowledge seemingly made me know less than I knew before about Spirit 2.

Odds can change with the introduction of new information. Suppose the situation of a good prize being behind one door and bad prizes for each of the other two doors. You select a door and another door with a bad prize is shown to you. You are given the option to change your guess or leave it on the current door. This is the classic Monty Hall setup. If you knew that the structure of the game was that a bad door that you did not choose would always be opened, then your odds are 2/3 to switch and 1/3 to stay the same to get the good prize. However, if you are open to any strategy being in play, allowing the bad door revealed to be a random opening or based upon some other strategy, then the odds are 50% to switch 50% to stay the same. Knowing the strategy of the game changes the probabilities of the two doors, not because information is lost, but rather because information is gained. But when information is gained, you see that one door goes down in probability while another goes up. So it is in no way given that adding information will increase probabilities.

 

[Pheryll]

When I didn't know what Spirit 1 needed there was a 50% chance that Spirit 2 needed C.

Technically this was 4/7 chance. That is after accommodating spirits 3, 4, and 5 in the second round but before accommodating spirit one you had 7 possibilities each with equal chance for the goods for spirit 1 and 2. These are:

C-C
C-D
C-G
C-H
D-C
G-C
H-C

By learning that spirit 1 took C you eliminated three of the possibilites leaving the remaining 4 each with equal chance.

 

[Mykan]

Are there any math/probability people here? I have a question for you. Below is a simulation of the negotiation process in the Spire. It assumes 8 items labeled A-H are offered. On the left are what items are available to be offered at the start of each offer round and the right shows the results of that offer round. After round 2 only Spirit 2 is unsatisfied. Out of the 4 possibilities (CDGH) for round 3 I chose C because in round 1 C was in the wrong position. I understand that in Round 2 Spirit 1 needed C and I could not conclude that Spirit 2 also needed C, but I have this feeling that there is a higher probability that Spirit 2 needed C compared to D, G, or H. Does anybody know if I am right in thinking that?

Offer Round	Items Available to Offer	Spirit 1	Spirit 2	Spirit 3	Spirit 4	Spirit 5
1	ABCDEFGH	A-NOT NEED	B-NOT NEED	C-WRONG	D-MATCH	E-NOT NEED
2	CD FGH	C-MATCH	F-NOT NEED	G-MATCH	D-MATCH	H-MATCH
3	CD GH	C-MATCH	C-MATCH	G-MATCH	D-MATCH	H-MATCH

Assuming all the spirits pick randomly with equal chance of each good, then all four goods that could be selected in your third round are still just as possible (25%) from the information you received from the previous two rounds. There is no reason to suspect a favoritism toward choice C.

While the probability is as stated above the observation of myself and other fellowship members is there is a favourtism towards the first available resource. The number of times I take a guess based on that and get it right is a lot higher than the actual odds of it being correct.

Someone else in our fellowship also notes with their observation that their is a favourtism towards goods that have had multiple ghosts accept them

I've fully completed the Spire every week since the second week after it was released and I'll tell you that getting a negotiation "right on the first try", while theoretically possible, in reality never happens, at least to me. Fighting is a much more reliable way of solving the encounter cheaply.

It does happen but it is very rare, I have had it once or twice and our arch got it once in the last week. Anyone negotiating though shouldn't be basing cost on getting it right first try but more aiming to get it right on the first 3 guesses. It gets very expensive when you have reset the game 5 times before getting it right (so likely 10-15 guesses, not counting diamond guesses)

I really wasn't trying to detract from your point. I'm a fighter and I'm convinced that the Spire is 10 times easier for me to get through because I manual fight my way through and rarely negotiate or autofight. I was just making a relatively minor point that it's not quite as cut and dry in the Spire as the others because of the variability of the negotiating process. So if I were psychic and could guess what the spirits wanted on the first try I would cater a lot more because it would be faster and frequently less expensive than fighting. Of course, as you said, it rarely happens that I guess right on the first try. I did it once when there's only 3 items to choose from.

catering is almost never less expensive then fighting because both cost supplies (and gold for goods) and the supply input and the quantity required is almost always more than the cost of troop production. Where diplomacy is a good option in spire is when you have a stash of those resources. For me as an example, I am end tech so goods, coins and sentient goods are very "cheap" in terms I have no real use for those. I almost have an out of control orc population so those are effectively a free guess for me. Seeds are the only resource I have to be more mindful of but again gaining seeds is free. So for an end tech person negotiating might be cheaper but only in terms of the uses of those resources compared to what they might do with the troops saved.

A person doesn't need to be end tech for that scenario to be true, I have seen similar on chapter 3 and chapter 5 towns however those towns stop at map 1 and map 2 respectively as that is their sustainable point. In those towns cases making the troops is much harder due to very short barracks queues and an inability to check the game every 1-2hrs all the time to keep barracks producing.

 

[Pheryll]

While the probability is as stated above the observation of myself and other fellowship members is there is a favourtism towards the first available resource. The number of times I take a guess based on that and get it right is a lot higher than the actual odds of it being correct.

You really would need to gather significant data to make that claim. For example, while this case with spirit 2 choosing C may seem to support this conclusion, one would also have to factor in all the spirits ignoring A and B. So many data sets that may seem to show favoritism in a certain direction may actually show the opposite.

 

[Deleted User - 3932582]

1) Your reasoning is correct. But, statistically speaking, it will cost you more goods while, statistically speaking, save you diamonds. One may or may not value the diamonds over the goods.

Only if you always close failed negotiations with diamonds. I don't (outside of a couple of late stage bosses), I just exit and restart negotiations. So this actually saves me goods. But indeed, this is because I value diamonds much higher than goods. With the amounts involved, I don't think it is particularly controversial (25+ diamonds vs a few thousands of goods? yes please!).

Pheryll already ninjaed me on the probabilities. Having said that, in your original case after round 2 you're left with

C-C
C-D
C-G
C-H

equally likely options, so 25% chance of guessing it right on your 3rd attempt. If you do D for ghost #1 on your second, then you're left with

C-C
C-G
C-H
G-C
H-C

so only 20% chance of guessing it right on your 3rd attempt. The fact that each of #1 and #2 have 60% chance of wanting C does not help you as C-C is just as likely as any other combo. If you're willing to go extra round (with extra information) then I am pretty sure that you can finish second variant in round 3 and 4, while it is certainly not guaranteed in the first case.

While the probability is as stated above the observation of myself and other fellowship members is there is a favourtism towards the first available resource. The number of times I take a guess based on that and get it right is a lot higher than the actual odds of it being correct.

Someone else in our fellowship also notes with their observation that their is a favourtism towards goods that have had multiple ghosts accept them

I think this is likely confirmation bias. I did run notes for quite a while, and noticed no such thing. Not saying it is not there, but my data certainly did not support that.

It does happen but it is very rare, I have had it once or twice and our arch got it once in the last week.

I have had it several times in between 2 cities. With 3 or 4 goods in rotation your chance if hitting it right in 1 go is 0.1-0.4% (1:243 and 1:1024).

 

[KarlD]

catering is almost never less expensive then fighting because both cost supplies (and gold for goods) and the supply input and the quantity required is almost always more than the cost of troop production. Where diplomacy is a good option in spire is when you have a stash of those resources. For me as an example, I am end tech so goods, coins and sentient goods are very "cheap" in terms I have no real use for those. I almost have an out of control orc population so those are effectively a free guess for me. Seeds are the only resource I have to be more mindful of but again gaining seeds is free. So for an end tech person negotiating might be cheaper but only in terms of the uses of those resources compared to what they might do with the troops saved.

I was saying I'd negotiate in the Spire if I could do it on the first guess with psychic powers because the quantities are actually pretty low if I can do it in one guess and there is significant saving in time compared to my manual fighting. Sure there will be times when I could fight and lose no troops, but we all make that trade of time vs cost when we choose to autofight vs manual fights when tourneying. In a recent tournament I was finding I could lose 1/2 a squadron manual fighting and 2 squadrons with autofight, yet I still did quite a bit of autofighting. It's not just about the direct cost. I haven't measured but I can easily imagine me spending 10 minutes on a multiwave fight. I'm a slow, but reasonably effective fighter. So the cost of paying one round of 5 goods for each Spire encounter seems very cheap to me. Also I am a near end tech person so most goods except T6 are pretty cheap to me.

 

[Ashrem]

getting a negotiation "right on the first try", while theoretically possible, in reality never happens, at least to me.

I imagine you're right.

With 1/243 chance of getting it right on the first try, and one spire per week, that means (on average) you should get it right about once every 4.8 years, and every other attempt during those 4.8 years will have at least one error.

 

[Deleted User - 3932582]

I imagine you're right.

With 1/243 chance of getting it right on the first try, and one spire per week, that means (on average) you should get it right about once every 4.8 years, and every other attempt during those 4.8 years will have at least one error.

Well, as a sample size of 1, I can tell you that between 2 cities I've got successful 1-round negotiations several times. At least 2 times, probably 3. Not all were in 3-goods rounds. So it is certainly possible and does happen (but I have also completed Spire several times using strictly negotiation).

 

[Ashrem]

Well, as a sample size of 1, I can tell you that between 2 cities I've got successful 1-round negotiations several times. At least 2 times, probably 3. Not all were in 3-goods rounds. So it is certainly possible and does happen (but I have also completed Spire several times using strictly negotiation).

Because I didn't read carefully enough, I restricted myself to the first encounter in the spire with its three options to offer to five spirits. As soon as you move past the first encounter, your odds for any given encounter will worsen, but your overall odds will be higher. Roughly doubled if you do two encounters every week, and by the time you hit the first door, it should probably be happening about once every year and a half.

 

[KarlD]

Thanks everybody for your responses to my probability question. I much appreciate it. They have indeed helped clear up some of my confusion.

 

[neeronie]

So true, and more then likely you are low on troops, too, if you are anything like me. My unit size is so big it takes a good while to rebuild. Even if you have enchanted the brown bear but don't have time to run them through with time boosts. So, you might win CC, to make pet food to fed the bear to make the troops to win some CC to make the pet food to fed the bear to make the troops, It's like a dance. In addition, it would seem that you only get the use of a % of the units you have. I was to lazy and/or busy to work that out but that what I have observed.