Slitherine

MikeC_81 wrote: ↑Tue Jun 09, 2020 2:39 am

Cunningcairn wrote: ↑Tue Jun 09, 2020 12:48 am Mike all that has been said is obviously true with regards statistical phenomena. The phenomena that you experience would not be that noticeable if the probability of the events occurring was higher. The underlying problem remains that the probability attributed to certain events is too high. Play the mod as the clumping phenomena has not occurred but in saying that I don't believe enough games have been played to be certain.

Let's be more precise in our wording so we don't keep talking past each other, yes? Why is this an "underlying problem"? No one has yet managed to make the case that it is a "problem".

If it isn't a problem then you have a game mechanic preference issue which isn't so much a problem as you want the game to do something else due to taste. This isn't a concrete issue like a clear mathematical discrepancy in point costing or closing up ZoC loopholes in existing rules. This is a game design issue and if RBS is firm on this, you should respect that and stop making it out like it is anything other what it is. We all have things we don't like about the game. Schweet and Pete are making a cool mod and you seem to have played it quite a bit and enjoy it. That is cool too and when the Mod tournament starts, if everyone flocks to that instead of Vanilla then great. If not, then oh well.

MikeMarchant wrote: ↑Tue Jun 09, 2020 5:42 pm Winning a game when your opponent has outmanoeuvered you and has done evrerything necessary to deserve a win, is a problem.
Losing a game similarly, is a problem.
Coming away from the game have invested time and effort and not had an enjoyable experience is a problem.
Being so fed up with it that you want to never play the game again, is a problem.

None of those problems are explained away by discussions of statistics and probability. If a game fails to reward good play often enough, there is clearly a problem wth the game. Although we might disagree over 'how frequently' constitutes a problem, there is cleary a sufficently large number of people who feel disaffected in thsi way for it to be taken seriously.


Best Wishes

Mike

devoncop wrote: ↑Tue Jun 09, 2020 6:14 pm Not wanting to interrupt this interesting exchange of views I would point out to Mike (Marchant rather than Mike C ) that coming up to three years after release FoG2 has a "Very Positive" rating on Steam from 385 reviews and the same rating from 15 recent reviews.

I am not sure what numbers are involved with the "large number of people" who are allegedly dissatisfied but I doubt more than 10 people could be named so there is a certain amount of personal bias being transferred across in analysing player opinion I would suggest unless there has been an independent poll conducted of players I am unaware of.....

paulmcneil wrote: ↑Tue Jun 09, 2020 6:42 am I've found that paying much notice to the predictions of combat outcome is very similar to someone telling a six foot man that it's safe to wade across a river because it's only four feet deep on average. (Metric measurements are also available).

Cunningcairn wrote: ↑Tue Jun 09, 2020 6:56 am Snugglebunny, and this not an attack on him in any way, posted a few posts saying that he had just finished playing X number of games and that he had only had Y number of rallies. At the same time he posted this I was recording an average of 3 rallies per turn in the 4 games I was playing and logging. To be more accurate these rallies were logged in the end phase of battles where I had routed large number of troops. In one of these games my opponent had conceded yet 5 turns later won the game. There is a greater frustration out there than is realised because a lot of players cannot be bothered even visiting the forum let alone commenting. The mods do remove a lot of the perceived issues being discussed at the moment. I don't want to argue about this as there is no need. The proof is there for everyone to try. Finally I don't think anyone wants this to be a competition between the vanilla and the mod and if push came to shove I would support vanilla if I thought it was splitting the community.

Cunningcairn wrote: ↑Tue Jun 09, 2020 7:25 am Now to answer your question. The underlying problem is the problem that events that should be extremely rare are occurring very frequently. Making the case for stating a fact? What has been said is what has happened. The standard answer after the initial derogatory remarks implying the complainant is fabricating the story is that it is statistical clumping or some other statistical phenomena. That is extremely rude and uncalled for. There are many here that understand statistics, maths and many other things. Yes I fully understand that these events have not been experienced by everyone as yet but it is happening to many players. None of us know how many games are being played by players complaining or defending the issue so to argue about it is a statistical nonsense. What is real is a number of players who play a lot of games are experiencing these issues. If nothing is going to be done about that is fine until a new player raises the issue again. We then go through the same cycle again and again and again. I am now tired of it and Mike I really don't have anything against you and don't want to argue with you. You have done great things for the game and I have much respect for that. Join the league again and let's leave the conflict for the FOG battlefield.

stockwellpete wrote: ↑Tue Jun 09, 2020 9:19 am That's not going to happen. The mod is not being set up to challenge the vanilla game, it is just offering an alternative for those who are ready to try something a bit different every now and then. I will not be organising any separate tournaments with the mod and all my focus will be on running the vanilla FOG2DL and KO Tournaments. Hopefully, anyone who decides to set up a small tournament using the mod will time it so that it mostly takes place in the gaps between the FOG2DL seasons (i.e. in January, May and September).

MikeMarchant wrote: ↑Tue Jun 09, 2020 5:42 pm If a game fails to reward good play often enough, there is clearly a problem wth the game. Although we might disagree over 'how frequently' constitutes a problem, there is cleary a sufficently large number of people who feel disaffected in thsi way for it to be taken seriously.

MikeC_81 wrote: ↑Thu Jun 04, 2020 10:43 pm

Cunningcairn wrote: ↑Thu Jun 04, 2020 10:23 pm Mike I don't think anyone is suggesting that good players win most of the time against lesser opponents despite these rules. Read Stockwellpete's posts above. Every time this is brought up the argument against a change is that Pantherboy, Dkalenda and Nosy_Rat still win most games. It isn't relevant. What is wrong with making the probability of a double drop less than it is at the moment? What is wrong with dropping the probability of units rallying when not in command range of a general? Why should a rallied unit out of sight at the other end of the battlefield save a defeated army?

Nothing is wrong with any of those suggestions and I did read Pete's as well as everyone else's posts carefully. The reason why Pantherboy and friends always get brought up is because the topic invariably begins with some form of 'I lost x game because y unit double dropped!' with the usual unsaid implication being that they were somehow robbed and proceeds down the well-trodden path of 'there is too much RNG negating skill'. Paul straight up said this. Snuggles quoted it. The standard response at this point is that if RNG is negating skill, then the Pantherboy phenomenon needs to be explained.

There are 3 possible explanations.

1. RNG is not enough to override skill gaps as the primary determinant of games so the original assertion is false. I tend to fall into this camp.

2. RNG is enough to override skill gaps as the primary determinant of games and pure luck is responsible for this convergence of players repeatedly winning tournaments or placing very highly. I find this unlikely however I do agree that you cannot be unlucky and still win a tournament unless you are much better than your peers).

3. RNG is enough to override skill gaps as the primary determinant of games and these players are getting to the top consistently by save-scumming to mitigate bad RNG when it happens. I find this answer to be unlikely although imo, from what is known about the system, it is my belief you can save scum with enough regularity to affect the outcomes of games without being flagged.

You get to pick which option you want to believe in.

However, you can be firmly in camp 1 and still be against double drops. You simply might not like where the RNG is being served up and that is a game design issue and you can keep articulating to RBS if you want. Or you can do what Pete and Schweet is doing and make a mod and play that. Or you can find another game where the RNG is to your liking. All those options are available but if you are in camp 1 and believe skill matters, don't say things like this:

paulmcneil wrote: ↑Wed Jun 03, 2020 10:27 pm Combination of double drops and lopsided rallies from rout decides a huge number of games, de-skills and passes outcome to chance.

MikeC_81 wrote: ↑Tue Jun 09, 2020 8:30 pm

Cunningcairn wrote: ↑Tue Jun 09, 2020 6:56 am Snugglebunny, and this not an attack on him in any way, posted a few posts saying that he had just finished playing X number of games and that he had only had Y number of rallies. At the same time he posted this I was recording an average of 3 rallies per turn in the 4 games I was playing and logging. To be more accurate these rallies were logged in the end phase of battles where I had routed large number of troops. In one of these games my opponent had conceded yet 5 turns later won the game. There is a greater frustration out there than is realised because a lot of players cannot be bothered even visiting the forum let alone commenting. The mods do remove a lot of the perceived issues being discussed at the moment. I don't want to argue about this as there is no need. The proof is there for everyone to try. Finally I don't think anyone wants this to be a competition between the vanilla and the mod and if push came to shove I would support vanilla if I thought it was splitting the community.

I think Snuggles' experience in contrast to your own is precisely what I have been trying to point out. There will be people who experience far more than the normal distribution of a particular event and people who experience far less of the same event. For whatever reason, you seem to be severely unlucky. We commented on this the last time we played in the DL when my offensive spears were charging your Lancers in a most futile manner trying to escape the ZoC locks when you replied that most of your games somehow saw spearmen breaking your lancers despite charging into them and it was a refreshing change to see my spearmen do nothing.

Cunningcairn wrote: ↑Tue Jun 09, 2020 7:25 am Now to answer your question. The underlying problem is the problem that events that should be extremely rare are occurring very frequently. Making the case for stating a fact? What has been said is what has happened. The standard answer after the initial derogatory remarks implying the complainant is fabricating the story is that it is statistical clumping or some other statistical phenomena. That is extremely rude and uncalled for. There are many here that understand statistics, maths and many other things. Yes I fully understand that these events have not been experienced by everyone as yet but it is happening to many players. None of us know how many games are being played by players complaining or defending the issue so to argue about it is a statistical nonsense. What is real is a number of players who play a lot of games are experiencing these issues. If nothing is going to be done about that is fine until a new player raises the issue again. We then go through the same cycle again and again and again. I am now tired of it and Mike I really don't have anything against you and don't want to argue with you. You have done great things for the game and I have much respect for that. Join the league again and let's leave the conflict for the FOG battlefield.

This next part is very important to what I have been saying.

I never thought we were arguing. A difference of opinion or views isn't an argument. The only thing I will say is that we have gone over the possibility of the RNG generator being broken before with the Double Drop tests and despite a ton of tests run by Pete, we could not find the anomaly. I think is important to separate is the stance of these RNG events "removing skill from the game" (as in overriding skill as charged by many players like Mike M most recently) vs a stance that these events are annoying to deal with and unsatisfying gameplay-wise. If it is the latter, then I have zero problems with it. I only ever comment when, in the course of players trying to effect change in the game, they bleed their arguments to the former or outright state it as such (like Mike M and Paul M did within the past 5 pages) to try and strengthen their argument for a change in the rules. That is a far more serious charge to the integrity of the game and requires them to produce much more evidence and explain the continued and sustained success of certain players.

As for returning to the league, that will probably wait until RBS puts out a new DLC for us. I play random pickups for fun when I see interesting armies on the menu or do beta testing when an interesting change occurs. I am just tired of seeing "camping" in rough terrain or extreme fortress defensive play all over a 10 dollar voucher. Also a lot of times I may have only one or two armies I really want to play but the nature of army selection means I might not get it and now I am obligated to play 9 games with something I didn't really want to play, to begin with.

devoncop wrote: ↑Wed Jun 10, 2020 9:11 am So vocal minorities "won the vote for common man and women"....they "abolished slavery and established rights in many areas. You agree with these measures so are happy to give vocal minorities the credit for bringing them about.

Where vocal minorities "influenced directions we might not approve of" this is apparently "not the consequence of being a vocal minority"

Convenient

devoncop wrote: ↑Wed Jun 10, 2020 2:42 pm Fair enough.

So we are left with the position of the status quo without any evidence that beyond such a minority there is any demand to change that status quo.

After all, as someone who sets such store by "winning the vote for the common man and women" I am sure you would never want to ignore the majority opinion which according to the only gauge of opinion we have on the game (user reviews) remains very happy with the game mechanics as does that game creator......

Interesting stuff though.

MikeMarchant wrote: ↑Wed Jun 10, 2020 8:53 am Having thought on the issue of Poisson distribution, which I can't enageg with from a mathematical point of view but can from a philosiphical viewpoint, I ask the following: Statistical clumping is a phenomenon which can and does occur naturally. Statistical clumping is also a phenomenon which can and does occur as a result of a fault in the system responsible for generating the data. How, as an observor of the data, can one determine whether the statistical clumping is the result of a natural phenomenon or as a result of a fault in the system?

MikeMarchant wrote: ↑Wed Jun 10, 2020 8:53 am As Pete commented, by 'sufficiently large' I didn't mean to imply large, because I simply don't know how many people feel the same way I do. The reason that the game has such a large number of very positive reviews on Steam is because it's an excellent game, I am not arguing the contrary. I am simply arguing that it could, and should, be better. Having just spent a little time skimming through the reviews, I do notice that there are comments about the luck factor being too great, and comments about specific flaws in the game that I would agree with. Not all of the complaints about the luck factor are from negatvie reviews, some are also in the positive reviews. But, when all is said and done, we only know the views of those who speak out. We can neither of us claim that those who don't speak out are on our side.

Surely there's nothing intiniscally wrong with being a vocal minorty, is there? Vocal minorities won the vote for the common man and for women. Vocal minorities abolished slavery, established equal rights in many areas, won the right for same sex marriage, etc. Of course vocal minorities have also influenced directions we migh not approve of, but that is not the consequence of being a vocal minority.


Best Wishes

Mike

devoncop wrote: ↑Wed Jun 10, 2020 5:01 pm Supporting universal sufferage is certainly not unusual.

Supporting it and then wanting to subvert the result shouldn't be unusual either but your views on this matter are increasingly common too sadly.........

Still we shall agree to differ on this point and at least agree that FoG2 is a fine game (even though my record is on the poor side of mediocre....nothing to do with RNG by the way...just ineptitude)

All the best

Ian

MikeC_81 wrote: ↑Wed Jun 10, 2020 8:55 pm

MikeMarchant wrote: ↑Wed Jun 10, 2020 8:53 am Having thought on the issue of Poisson distribution, which I can't enageg with from a mathematical point of view but can from a philosiphical viewpoint, I ask the following: Statistical clumping is a phenomenon which can and does occur naturally. Statistical clumping is also a phenomenon which can and does occur as a result of a fault in the system responsible for generating the data. How, as an observor of the data, can one determine whether the statistical clumping is the result of a natural phenomenon or as a result of a fault in the system?

This is a complicated topic. It will never be possible to be sure whether any generator of random events or selection is truly random or not with 100% confidence. This is why you see polls conducted, even when meeting the strict scientific definition of a random selection of participants, always is qualified with the statement of 'accurate to within plus or minus x percentage points, y% of the time. Then we get into how most computer games generate random numbers. Most engines use some form of Psuedo Random Number Generator program or subroutine. It usually an arithmetic formula that spits out a number based on an initial seed parameter. It is called "psuedo" because the sequence of numbers that the formula spits out is entirely dependant on the initial seed value. Indeed you can reconstruct the entire sequence of numbers the formula spits out as long as you know the formula and have possession of the initial seed value. The seed value could be anything. For example, some PRNGs simply use the time and date when the request for a number or series of numbers is asked for as the initial seed value. I am unaware of how the PRNG works in FoG2 but it is most likely some tried and true formula.

It is important to note that they are not "truly" random since they are deterministic. You literally feed values into the generator and you know exactly what comes out as long as you know how it works. What is important to note though is that they are generally good enough with respect to what games are trying to accomplish. A lot of FoG2 is dice rolling with the exception of the rather opaque and complex combat calculator. But for the purposes of the discussion at hand, rallying troops is literally rolling two 2d6 and seeing if it meets or exceeds a required number based on various factors. The expected frequency of outcomes for 2d6 is child's play and everyone in this forum knows it (1/36 for 2 and 12, 2/36 for 3 and 11, etc etc). If you believe the PRNG is broken in this game, you can set up very simple tests for this, log results, and compare them against the distribution of outcomes. The Law of Large Numbers will kick in very quickly and expose any weakness in either the PRNG or the code using the PRNG (ex say RBS messed up and things are rallying on rolls of 3 rather than 4).

I think 12 months ago we had another RNG discussion where it was charged that Double Drops were occurring more often than the underlying formula should say it should be happening. I sat down with the rules and figured out what the percentages should be and Pete provided the raw data by smashing hundreds of Warbands against each other. The result was that there was no noticeable weakness in either the PRNG generator or the morale calculation code. You can find my conclusion here. viewtopic.php?p=775756#p775756

If there was a weakness in the PRNG, it should have shown up in these tests but it doesn't. We rapidly got results that converged to the best theoretical results we could come up with, given win/draw/loss percentages are not 100% known and only approximated through empirical means. I find it unlikely that FoG2 would randomly use two different PRNGs in the game so that half is safe. If rallies are happening too often beyond what the code says should happen then the error would be with the rally calculation. You can sit down and test this too. Ask Pete to whip up a scenario for you and simply grind the numbers and see if they fit.

PRNGs do have weaknesses in that they are not truly random and for people in the scientific fields with super-precise measurements and predictions, they have to rely on a very solid one, and their work needs to be rigorously tested by others to make sure PRNG weakness didn't taint results. But for the purposes of rolling two dice, inherent PRNG misbehaviour is a nonissue.

MikeMarchant wrote: ↑Wed Jun 10, 2020 8:53 am As Pete commented, by 'sufficiently large' I didn't mean to imply large, because I simply don't know how many people feel the same way I do. The reason that the game has such a large number of very positive reviews on Steam is because it's an excellent game, I am not arguing the contrary. I am simply arguing that it could, and should, be better. Having just spent a little time skimming through the reviews, I do notice that there are comments about the luck factor being too great, and comments about specific flaws in the game that I would agree with. Not all of the complaints about the luck factor are from negatvie reviews, some are also in the positive reviews. But, when all is said and done, we only know the views of those who speak out. We can neither of us claim that those who don't speak out are on our side.

Surely there's nothing intiniscally wrong with being a vocal minorty, is there? Vocal minorities won the vote for the common man and for women. Vocal minorities abolished slavery, established equal rights in many areas, won the right for same sex marriage, etc. Of course vocal minorities have also influenced directions we migh not approve of, but that is not the consequence of being a vocal minority.


Best Wishes

Mike

You are trying to equate vocal minorities with respect to moral judgements vs empirical truth? There is a vocal minority for the earth being flat too!

General Shapur wrote: ↑Thu Jun 11, 2020 4:17 am The double drops seem to plague pictish troops and those warbands - probably other things too. So much so I command them less aggressively than they really would have been, hanging the army back until the last half dozen turns if possible. What I'd like is for double drops to be disabled on the charge in non-raw troops - this would encourage more aggressive gameplay. Less of an issue in large games, but one double drop can ruin a medium or small game I think.

General Shapur wrote: ↑Thu Jun 11, 2020 4:17 am The double drops seem to plague pictish troops and those warbands - probably other things too. So much so I command them less aggressively than they really would have been, hanging the army back until the last half dozen turns if possible. What I'd like is for double drops to be disabled on the charge in non-raw troops - this would encourage more aggressive gameplay. Less of an issue in large games, but one double drop can ruin a medium or small game I think.

paulmcneil wrote: ↑Thu Jun 11, 2020 1:38 pm Anecdotally (so I guess that I'll be told that I'm mistaken or exaggerating) I see Warbands getting frequent double drops, more so than other troop types

General Shapur wrote: ↑Thu Jun 11, 2020 4:17 am The double drops seem to plague pictish troops and those warbands - probably other things too. So much so I command them less aggressively than they really would have been, hanging the army back until the last half dozen turns if possible. What I'd like is for double drops to be disabled on the charge in non-raw troops - this would encourage more aggressive gameplay. Less of an issue in large games, but one double drop can ruin a medium or small game I think.