CMBN games seem to be more imbalanced...

[GreenAsJade]

My sense, so far, has been that there are fewer closely fought battles in CMBN. You either win decisively or loose disasterously (generalisation) has been my impression.

I thought y'all might be interested in some stats from We Band Of Brothers game database:

http://gregories.net/bobster/bobster.cgi?function=scenario-spread-histogram

GaJ

[ChrisND]

CMAK has an exponentially larger sample size than CMBN.

[Wiggum]

I had some really close fights with last minute wins.

[Ranger33]

All of my MP matches have resulted in both sides knocking the crap out of each other in the first thirty minutes or so, and then whoever maneuvers their remaining forces best usually wins.

[George MC]

Another elements to take into consideration is PBEMs in CMX1 tended to be more binary i.e. catch the flags. Win conditions in CMBN can be/are more subtle.

Another factor is how many of these players are used to playing CMBN? I suspect if you are not used to the game engine you can very quickly make mistakes that can cost the player the game big style.

But as ND said they are very differant sample sizes. Would be worth comparing say in a year when I would expect there to be significantly more PBEM games to have been completed.

Cheers fur noo

George

[GreenAsJade]

CMAK has an exponentially larger sample size than CMBN.

Thanks for pointing out the bleeding obvious.

What is your point?

Do you think that despite annecdotal evidence in this direction, tabling some actual data, scanty as it is, is a bad move?

Or maybe you'd like an actual conversation here in the forum where we converse about CMBN and our experience with it??

I happen to think it's interesting.

Are you willing to hazard an opinion about what the trend will be as more data points come?

Do you think that scenario designers will get better, and that this will deliver a better percentage of close-fought matches?

Do you think that scenario designers are fine, and we're just looking at a random small sample?

GaJ

[GreenAsJade]

Another elements to take into consideration is PBEMs in CMX1 tended to be more binary i.e. catch the flags. Win conditions in CMBN can be/are more subtle.

This is what I thought, too. So I was expecting CMBN to be _more closely fought_, rather than having apparently more binary outcomes.

Another factor is how many of these players are used to playing CMBN? I suspect if you are not used to the game engine you can very quickly make mistakes that can cost the player the game big style.

I reckon this, combined with scenario design "inexperience" is surely a contributing factor.

But I also have a suspicion that the current mechanics of CMBN are contributing. In my limited experience, the CMBN battlefied seems way more harsh. There is just hardly any cover anywhere. This appears to mean that the guy who spots the other guy's bulk of force first and manages to shoot at them gets rolling advantage. Once you're on the back foot, there's no-where to hide...

GaJ

[TheVulture]

Maybe a contributing factor is scenario length. The vast majority of CMx1 scenarios in my experience were decided by tallying up the points once time ran out - pretty much every scenario I can remember had variable time and I used up the full amount of variable time without one side surrendering.

The majorty of CMx2 scenarios on the other hand are much longer (30 minutes would be a long fight in CMx1 and an alarmingly short one in CMx2). And most scenarios don't get close to running out of time; one side or the other has a forced surrender at some point, at which point the winner gets to own all objectives.

So in CMx1 battles are stopped before one side gets crushed, casualties haven't yet snowballed, and terrain objectives (flags) might well be contested. CMx2 battles are fought to decision where the winning side gets to mop up the remaining enemy (causing very lop-sided casualty figures in the mopping up phase) and taking all terrain objectives when the enemy surrenders. That last phase of the battle (going from superiority through to surrender) is missing from CMx1 on the whole and is the part that racks up the score in fabour of the winner.

Or such is my guess.

[GreenAsJade]

That's a good point.

Suppose this was the "only" factor. What should we do about it?

I'm coming from the point of view that I don't really enjoy decisive battles. I want a close-fought battle. I guess others are like this too?

From this vantage point, does it mean that in fact CMBN games should have their length shortened, so that it's harder to get the objective and have time left over to wipe out the other guy as well? It might be as simple as that!

GaJ

[ChrisND]

Thanks for pointing out the bleeding obvious.

What is your point?

There is no need to get testy. My point is the "bleeding obvious", that the sample sizes are so disparate that the data isn't useful yet. I don't even know the nature of the scenarios used for that data.

It is quite possible that the nature of the game and the way scenario designers are setting points is making results tend to be extreme. But I wouldn't infer that from the histograms presented. I wouldn't infer anything at all from it.

[ChrisND]

From this vantage point, does it mean that in fact CMBN games should have their length shortened, so that it's harder to get the objective and have time left over to wipe out the other guy as well? It might be as simple as that!

That would be one way to do it. Another would be to have higher VP emphasis on inflicting casualties as opposed to controlling ground. This way a defender, for example, who loses his objectives but makes the attacker pay dearly for them could still get a lot of VP.

[GreenAsJade]

This latter idea doesn't ring so true to me. It seems to me that the decisive victories are happening where the attacker can gain the VP _and_ go on to wipe out the defender. So something needs to offset this "opportunity" to "get all the points". I could be wrong, but one feeling I have is that once you've lost some ground and your opponent knows where you are, it's hard to survive. There's so little real cover. So if you're on the back foot, you're on the road to total wipeout... it seems hard to "be fighting a brave hard-fought back foot defence".

Just my experience so far... it does seem like a shorter game length would go a ways to redressing...

GaJ

[ChrisND]

This latter idea doesn't ring so true to me. It seems to me that the decisive victories are happening where the attacker can gain the VP _and_ go on to wipe out the defender. So something needs to offset this "opportunity" to "get all the points".

I'd also add that the defender should in many cases be able to withdraw off the map. Combine this with a higher VP ratio on casualties inflicted for the defender, and I think that's one way it could work. It might not stop the attacker from winning, but it could blunt the victory.

Right now defenses are usually an all-or-nothing affair that are pretty extreme circumstances. Giving the defender the ability to quit the field would alleviate that.

[GreenAsJade]

Yeah, that's for sure!

that the sample sizes are so disparate that the data isn't useful yet.

I don't think disparate sample sizes is itself a problem. If there were 1000 CMBN games and 1,000,000 CMAK games, the comaprison would still be useful. The question is whether the "small size" in absolute terms of the CMBN data is "too small". A small size, in absolute terms, of the CMBN data set would certainly limits its usefulness. But there are 80 games in there: that's not ridiculously small. After 80 games, we can see that the "apparent trend" is for at best a flat distribution, maybe even favouring extreme results over even ones.

How many games would you want to see in the data-set before you started to thing that this result is statistically significant? I'm no mathemetician... but it seems to me that after 80 games, having 30+ of them with an 80% plus winning margin seems like a lot.

I don't even know the nature of the scenarios used for that data.

All the information is available at WeBoB Statistics Database. You need an account, which comes with WeBoB membership.

In very broad terms, these 80 games consist of the scenarios chosen for the H2H games that the WeBoB members have played since CMBN came out. They are a mixture of scenarios and QBs. Is there any other relevant "characterisation" or statistic that would add to your understanding?

GaJ

[LJFHutch]

I haven't played a lot of multiplayer but in my experience Combat Mission certainly does allow for extreme results, which is a good thing.

If you play really poorly you won't just lose a little bit like you might do in Total War or others, you'll get absolutely destroyed. I played a game the other day in hotseat and out of 160 soldiers and 2 tanks I had 1 dead and 1 wounded, with some very minor damage to one vehicle (tight bocage map), my opponent lost the majority of his forces and was pushed right back from the objectives after a few mistakes (poor tank positioning and a bodged flank assault, I also screened his advance with some mortars so I could get myself into a defensive position early on).

I would say the extreme results show just how deep combat mission is so I don't really consider it a problem.

[GreenAsJade]

If we take an observation, like "CMBN results in more unbalanced results", then one thing is clear: some people will think it's a problem and some will think it is not.

This is simply a matter of opinion. Especially in this particular case, I find if of little interest. I know that some people will like it when every game is a "grind the opponent into the dust". I know that I prefer tightly contested matches.

So for me, an interesting discussion is whether the observation itself is true or not.

Is it the case that CMBN battles really are by their nature going to be more imbalanced?

And also - suppose they are, and you're in the group that see it as a problem, what could we do about it?

If you agree that they're more imbalanced by the nature of the game, that's interesting.

Beyong that, if you don't think it's a problem, then ... that's your opinion, you won't be interested in the discussion about what to do about it

GaJ

[simmox]

the score in CMBN as compared to CM1,in giving a result,isnt a fair representation of the battle facts of a particular battle imo

comparing a total victory between the two games,say

ive had plenty of battles were ive actually taken more losses than my opponent and won a total

my latest game has been a real hard even slog,whomever has the strength left to push each other out of the OB will probably achieve a total victory,even though the battle was far more even.

in CM1 you can count the scoring as you go,you know you get 100 points for armor kills etc,you can play it like poker somewhat.

its already been said a few times,if your comparing it to CM1,your probably doing yourself a diservice,i speak with some experience here:)

Salute

[GreenAsJade]

The strange thing is that I was expecting the victory system in CMBN to deliver _more_ balanced results, assuming that scenario designers use it in a way that you can score points even if you loose men.

It doesn't seem to work that way as much as I thought!

GaJ

[LJFHutch]

Well I think there is a difference between "CMBN games are unbalanced" and "CMBN games can be unbalanced". Two players of equal skill should probably end up with a very even battle, I suppose I got mixed up a bit there thinking about the possibility of very wide gaps in end results.

I would also argue that the results which have one side utterly crushed are likely to be in tiny/small mixed battles, which even in bocage can depend quite heavily on armour (of which there is a very small amount of). In those cases the battle often comes down to who destroy's the other tank/s first, once you have a mixed force fighting an infantry force it's very uneven, throw in the fact that the best player is likely to end up with the surviving armour and I think that would more than explain those results.

So I would say the solution is most likely try to match player skill levels and play much bigger games where losing a few tanks isn't such a big deal.

(This is mostly coming from a QB perspective of course, though I suspect it would hold up to scenarios as well to some extent).

[GreenAsJade]

Well I think there is a difference between "CMBN games are unbalanced" and "CMBN games can be unbalanced".

I wasn't saying "CMBN games are unbalanced". No-one could support a generalisation like that. Neither was I saying CMBN games can be unbalanced. Obviously they can.

What I _was_ saying, and backing up with data, is that it appears that CMBN games TEND TO BE MUCH MORE unbalanced than CMAK. All other things being equal, with CMBN you get an extreme result way more often.

IE given a set of 80 random pairings of players, which presumably include some mismatches and some close matches, the outcome tends to be more towards the extremes than the middle.

This is the opposite of what appears to be the case with CMAK: even with a random spread of players, a large pool of CMAK games tend to deliver closely fought results, with decisive outcomes being far more rare.

Therefore, I don't think this is about "trying to match player skill levels". I think its about "overall", with a random mix of skill levels.

What we can see is that _despite_ a random mix of skill levels, CMAK delivers closely fought matches. That for me is a "good experience" - which makes me keenly interested to hear opinions why CMBN is not like this and what can be done about it.

GaJ

[womble]

I've only fought the AI, but I can see that several factors that have already been mentioned could easily produce the 'catastrophe curve' results being reported.

• Completeness of the battle produces complete victory. In CMx1, the battles seem to have been of a length such that in the face of even barely competent opposition, there wouldn't be time to take all the VLs.
  
• Having only one VL produces complete victory. Small maps which seem to be popular in BN have fewer VLs, and so holding those will increase the marging of victory.
  
• Large VLs produce total victory. Since you have to clear a VL to get its points, you have to kill or break a lot of troops in a large VL, increasing the VP 'take' for each VL. Some of the QB defense maps have a setup location that's basically the VL, so it becomes necessary to inflict massive losses to even get the VL into 'contested'.
  
• Small forces produce total victory. If one side kills the other's mobile armoured machine gun nests while preserving theirs, a win becomes faster, and the other factors requiring massive MDK become easier to achieve, and in small battles, even closely-matched players can lose their single tank/assault gun to the fortunes of combat, producing a lop-sided result.
  
• Scenarios produce total victory. Some scenarios that don't have the 'HvAI only' tag on them probably should have. Some people don't notice/care about that tag, if it is there.
  

[BletchleyGeek]

Is it the case that CMBN battles really are by their nature going to be more imbalanced?

And also - suppose they are, and you're in the group that see it as a problem, what could we do about it?

If you agree that they're more imbalanced by the nature of the game, that's interesting.

Beyong that, if you don't think it's a problem, then ... that's your opinion, you won't be interested in the discussion about what to do about it

Tricky question. I refrain from drawing conclusions like "the game engine is imbalanced" with the relatively small set of scenarios we have available. Which is a natural thing, it's been 3 months and a half since release, and we're all still struggling to master both the game and the scenario editor.

"Balance" can mean many things, not all of them equally enjoyable. I'd bet good money that on most of the past and existing CMBB ladders, the most often fought QB was a "Meeting Engagement" set in late 1944. That was "balanced" in the sense that both sides had access to equipment which was closely matched, had to maneuver for position and could shoot at each other from 500m and have a good chance to score a kill. The thing that really only mattered was the players' skill to take advantage of terrain and get into a dominating position and knowledge of available TOEs to optimize their force. I, personally, find that to be extremely boring.

Now there's the "interesting" kind of balanced scenarios, which might or not, include balanced forces, terrain and objectives. These are really hard to get well done. One of the hardest things, in my opinion, is to come up with forces that allow - not guarantee - to meet the "combat mission" objectives, or the other way around, come up with reasonable objectives for a given force. This is all about playing with the VP allocation.

Let me put an example of an scenario I am working on right now, so we have something concrete to discuss on.

The scenario concept is as follows. The mission for the US side is to move across a road to get to the form-up position for an attack. Time is limited: they have to make it on time, otherwise the mission - join up with other forces - will be an utter failure. Germans' mission is to interdict this march, causing as many casualties as possible and preventing the US forces to freely march.

For this concept, what kind of terrain makes sense? Obviously, it needs to have a road connecting US deployment zone with an appropiately set Exit Zone. I would like to encourage US players to come up with proper flank security, so I need to put terrain that might support a credible threat (say an elevation difference, foliage preventing LOS from the road but not the other way around, etc.).

What kind of forces make sense? I want this scenario to depict what I understand as a problem of maneuvering within the reach of the enemy, which is usually modeled by operational wargames with ZOC concept.

The German force should be a combat patrol force, light and not very big. Let's say, one German Fusilier squad, a FO, a Sniper and a couple or three Panzershcrek/Tank Hunter teams. It has off-map support from Company mortars and possibly Battalion Heavy Mortars.

The US force could be pretty much anything, let's say a full infantry Company mounted on trucks, with a small cavalry force as an advanced guard (one M8, one M20 and a jeep).

Now, what should be the time limit for the scenario? A good starting point could be the time a US truck needs to cross the map, on road, move order adding an slack proportional to the number of vehicles in the column.

What about the victory conditions? The German side is not interested - and shouldn't - in holding any ground at all. The point is to make losses and hinder column movement. Therefore, German VP's should be awarded by destroying US units and preserving its own force. US side Victory conditions would be defined in terms of how many units make it through within the time allotted.

Would this be balanced? Certainly not, US troops should come into the battlefield in a staggered manner, it's a column, and this would prevent to some extent pre-plotted barrages. Delaying the availability of artillery would also help. If TRP's were to be included, they would placed well away from US deployment zone (pre-registered targets are that, pre-registered).

Would I get VP's, timings and all that right, after some hotseat playtesting. Certainly not. I'd need feedback from people trying the scenario H2H. Changes in one aspect of the scenario would probably have an influence in other aspects, interlocked in a feedback loop. I'd need to iterate several times, and that would probably take months.

Welcome to the world of wargame design: it's hard and takes time to get it right More so when the designers are hobbyists working on this on their spare time.

Bottom Line

I just think there hasn't been enough time for the CM:BN scenario community to jointly - by designing and adjusting and by playing - develop scenarios with a quality near the best developed for CMx1.

[stoex]

GaJ,

Besides the points made by other people in this thread, I have a few of my own thoughts to add as to what we see in the histograms you originally linked to and why:

First off, you make the claim that both sets of data come from games played by people with varying skill levels. While that is certainly true of the CMBN games, it is probably less true of the CMAK games. After all, people playing CMAK on WeBoB are very likely, IMO, to be at least somewhat experienced in CMAK, if not very much so. After all, if they weren't, why would they still be playing CMAK, and why would they be WeBoB members in the first place? There is also no indication of the skill levels involved in the games recorded in the histogram, so your claim is at least unfounded. I believe it is realistic to attribute a fair level of skill to CMAK players on WeBoB simply on the basis of the fact that CMAK has been around for many years, and people who play H2H regularly in an organized fashion (like on WeBoB) are likely to be more skilled than people who don't. As for CMBN, I would hazard the opinion that no-one could yet be considered highly experienced in CMBN, as the game hasn't even been out for half a year yet, and it is still evolving in terms of patches and scenario design. This is witnessed by thread upon thread of people (even ones very versed in CMx1) asking all kinds of questions about how this feature or another works in CMBN. We all ain't seen the half of CMBN yet, to be frank. CMAK? No surprises anymore there, people who are still playing it now know exactly how it works. Compare this to sports of any kind: at the world championships, the best teams rarely beat each other with high margins, the games instead tend to be relatively close-fought.

This of course tends to skewer results for several reasons: One, people on a similar high skill level are more likely to produce closer results, obviously. People with very disparate skill levels (one thing we see more in CMBN than in CMAK) will obviously produce more extreme results. Now comes the big one: IMO people on similar low skill levels (which I believe we see a LOT of in CMBN currently) are not only likely to produce difficult-to-predict results, but also more extreme results on average, simply because it is much easier to exploit a really bad tactical mistake by your opponent than it is to avoid making one yourself, if you are not that good. And as opposed to most sports, combat losses tend to snowball, so often it will come down to who makes the first big mistake.

Another factor which may contribute to the results in the CMBN histogram (if they are actually statistically useful, which is debatable due to sample size) is that the higher fidelity of many aspects of modeling in CMBN compared to CMAK makes for a lot more possible outcomes of a situation and much less predictability. Plus, as I said, we are all still learning, which means we don't yet have a really good grasp for which situations are more or less predictable and why, at least in many cases. This results in the whole act of playing being more of a "gamble" at this point (than in CMAK), and we all know that gambling often produces extreme results...

What I mean is, there are far more possible game-changing moments in CMBN than in CMAK, and most players are not experienced enough yet in CMBN to correctly determine how to make those game-changing moments work in their favor. Thus, results will vary widely when compared to the orignal "plan" and unit-make-up.

Third, and this has been touched upon already by other posters but I think it can stand to be underlined once more: I am pretty sure (though I have no statistical evidence) that the average CMAK battle tends to be larger (number of units at least) than the average CMBN battle. And as others have said, smaller forces lead to more extreme results on average.

Wall of text end.

[womble]

Tricky question. I refrain from drawing conclusions like "the game engine is imbalanced" with the relatively small set of scenarios we have available. Which is a natural thing, it's been 3 months and a half since release, and we're all still struggling to master both the game and the scenario editor.

"Balance" can mean many things, not all of them equally enjoyable.

And in the context of this thread, it 'imbalanced' doesn't have any connotations of bias one way or the other, it's talking about the percieved extreme nature of many victories, so 'unbalanced' really means 'balanced on a knife edge'. The slightest nudge will make the result fall drastically one way or the other.

Not that anything you've said about the nature of "game balance" is untrue, it's just not quite what GaJ was concerned about at the top of the thread.

I would however disagree that we'll be able to "...develop scenarios with a quality near the best developed for CMx1." I'd say that it's possible with the broader set of tools available to develop scenarios even better than the best in x1

[GreenAsJade]

I really goofed using the word "imbalanced" actually. I'm glad womble could see what I mean: he expressed it well.

I meant that CMBN games are likely to deliver extreme, rather than closely fought results. This is not what balanced means at all, mea culpa. A scenario that delivers extreme results can be balanced! IE it's balanced because each side has an equal chance of winning.... it's just that when one side wins, they win with much more of a decisive victory than with CMAK.

Howver stoex observed something really relevant:

First off, you make the claim that both sets of data come from games played by people with varying skill levels. While that is certainly true of the CMBN games, it is probably less true of the CMAK games. After all, people playing CMAK on WeBoB are very likely, IMO, to be at least somewhat experienced in CMAK, if not very much so

I think that this is a very very likely explanation for a large part of what we are seeing.

So we can say "yep, you guys are experiencing much more extreme results right now: guess what, this is what happens when less experienced people play. So hang in there, learn to play the game, and play it with others who have learned, and the battles will become closer".

That's probably a lot true.

I can't help thinking that CMBN has factors that make things more extreme - the inability to escape being one that's been observed, and the lack of cover prevalent in many maps being another. And longer games being another.

Well, it sounds like many of these things will improve with time, it will be interesting to watch the stats

GaJ