Malsqueek

Join Date 09/2009

+38

[0.915] Argument for Scaling Experience by Trained Unit Size

or, "Why Bigger is always Better, and why it shouldn't be"

Inspired by discussion in Gfirefly's thread: https://forums.elementalgame.com/426436

During my recent game I started noticing that unit size seems to matter a whole hell of a lot more than unit experience. The discussion above made me curious enough to see exactly how the relationship works out between them all.

Methodology:

All units have 4 Hit Points per model and increase at a rate of 4 hps per model per level.

Accuracy starts at 60 and increases at a rate of 2 per level

Attack is equal to wielded weapon's attack times number of men in the unit

Probable Damage is a contrived number assuming perfect probability of hits occurring on units with no armor.

Trained Unit Levelling Diagrams

3 person unit up to level 20

	Health	Attack	Accuracy	Prob. Dmg.
1	12	24	0.6	14.4
2	24	24	0.62	14.88
3	36	24	0.64	15.36
4	48	24	0.66	15.84
5	60	24	0.68	16.32
6	72	24	0.7	16.8
7	84	24	0.72	17.28
8	96	24	0.74	17.76
9	108	24	0.76	18.24
10	120	24	0.78	18.72
11	132	24	0.8	19.2
12	144	24	0.82	19.68
13	156	24	0.84	20.16
14	168	24	0.86	20.64
15	180	24	0.88	21.12
16	192	24	0.9	21.6
17	204	24	0.92	22.08
18	216	24	0.94	22.56
19	228	24	0.96	23.04
20	240	24	0.98	23.52

5 person unit up to level 20

	Group (5 Man)
	Health	Attack	Accuracy	Prob. Dmg.
1	20	40	0.6	24
2	40	40	0.62	24.8
3	60	40	0.64	25.6
4	80	40	0.66	26.4
5	100	40	0.68	27.2
6	120	40	0.7	28
7	140	40	0.72	28.8
8	160	40	0.74	29.6
9	180	40	0.76	30.4
10	200	40	0.78	31.2
11	220	40	0.8	32
12	240	40	0.82	32.8
13	260	40	0.84	33.6
14	280	40	0.86	34.4
15	300	40	0.88	35.2
16	320	40	0.9	36
17	340	40	0.92	36.8
18	360	40	0.94	37.6
19	380	40	0.96	38.4
20	400	40	0.98	39.2

7 person unit up to level 20.

	Health	Attack	Accuracy	Prob. Dmg.
1	28	56	0.6	33.6
2	56	56	0.62	34.72
3	84	56	0.64	35.84
4	112	56	0.66	36.96
5	140	56	0.68	38.08
6	168	56	0.7	39.2
7	196	56	0.72	40.32
8	224	56	0.74	41.44
9	252	56	0.76	42.56
10	280	56	0.78	43.68
11	308	56	0.8	44.8
12	336	56	0.82	45.92
13	364	56	0.84	47.04
14	392	56	0.86	48.16
15	420	56	0.88	49.28
16	448	56	0.9	50.4
17	476	56	0.92	51.52
18	504	56	0.94	52.64
19	532	56	0.96	53.76
20	560	56	0.98	54.88

9 person unit up to level 20

	Health	Attack	Accuracy	Prob. Dmg.
1	36	72	0.6	43.2
2	72	72	0.62	44.64
3	108	72	0.64	46.08
4	144	72	0.66	47.52
5	180	72	0.68	48.96
6	216	72	0.7	50.4
7	252	72	0.72	51.84
8	288	72	0.74	53.28
9	324	72	0.76	54.72
10	360	72	0.78	56.16
11	396	72	0.8	57.6
12	432	72	0.82	59.04
13	468	72	0.84	60.48
14	504	72	0.86	61.92
15	540	72	0.88	63.36
16	576	72	0.9	64.8
17	612	72	0.92	66.24
18	648	72	0.94	67.68
19	684	72	0.96	69.12
20	720	72	0.98	70.56

Results:

As you can see, a third level unit of 3 is exactly the same as a 1st level unit of 9, except the unit of 9 has three times the damage output potential. In many ways, this is both sensible and fine. However, when you look at this from an opportunity standpoint. Getting into combat with smaller units EVER is a bad choice, because it takes the same amount of experience to level a unit of 3 as it does a unit of 9, with the added costs of losing units easier and more protracted combats due to significantly lessened damage outputs.

Even groups of 5 quickly outclass their one step lower counterparts, and given how high up the research tree the other two techs are, that sort of makes the 5 man groupings the optimal size of build. 3 man units are swiftly outclassed by even 5 man units (assuming a relatively average damage output, it takes a 3 man unit identically equipped until level 4 before they can reliably take on a unit just two larger because of return damage.

Because of this, even in the face of increasing upkeep, because of constantly topped out growth in cities, and especially because of only marginally larger build times, it is always always the best policy to create the largest group you can possibly field. The new size techs are not unlocking new "options" they are unlocking the new only choice.

Assessment:

It is my opinion, that because of the dramatic rate at which smaller units, identically equipped, get outclassed by the next larger size of unit, that there needs to be some benefit at all of creating smaller units later in the game, or we may as well not have them at all from the get go.

First off, smaller teams will be able to learn from their experiences more quickly, and provide "real-world" combat training to each and every member faster than larger units. Because of this, larger units should gain experience at a marginally slower rate than smaller units.
One of the key concepts in the original vision of this game was that it would be a viable strategy to field "small but elite forces as well as large armies of trash". In many ways, that is not currently represented by the gameplay, and having larger units level slower than smaller ones would encourage a wider variety of unit builds as tactically viable.
Because there are a couple ways to improve the rate at which your armies level, having larger, more valuable units levelling slower on their own values up not only all the abilities which give 10-15-25% exp bonus to stacked units, but dramatically improves the appearance of any and all buildings which improve a unit's level.

Recommendation:

I would like to see Units of 3 levelling at the rate they currently do, with units of 5 levelling at 85%, units of 7 at 70%, and units of 9 at 55% to keep the relative hit points by experience somewhat in line with one another while still allowing the player to benefit from the significant increases in damage output afforded by the larger units. Between that and the smaller units having more "Actions" as a collective, I think would wind up pretty well balanced over the long term.

In many ways I think there is an argument for capping units at level 10 to allow for champions to seem more important, but that is a totally different argument.

Locked Post +1 Karma

Advanced Search

Watch this post

Do not email me updates for this post Email me updates for this post

Successfully updated karma reason!

discodog

joasoze

Join Date 02/2006

+39

Good work there.

The easy solution would be allowing us to upgrade unit sizes. The groups with 3 members that I use early game, I get quite attached to and want to keep. I think we should (for a price) be allowed to upgrade, but that the group should go down two levels (or something like that) to indicate that you have added rookies to your experienced group.

Easy and elegant methinks

Successfully updated karma reason!

Heavenfall

Join Date 07/2008

+441

I second the need for experience to go way down if a unit is upgraded in size. If that was in, it would be neat to upgrade units in size. Without it, it is simply overpowered as hell.

Successfully updated karma reason!

i think this is a really important point and would help to balance the game. It would also be nice to have some traits for small units that would be used later in the game.

Successfully updated karma reason!

I second that suggestion:

Loose 2 levels = allows that unit to be upgraded into the next group size. (if researched)

Successfully updated karma reason!

I'm not outright opposed the ability to upgrade units, but I have two concerns about it.

First, if you can build a faster and easier unit, then upgrade it, what does that do to the production cost of just building the large unit?

Second, if you can build a smaller unit and just upgrade it, doesn't that somewhat defeat the purpose of installing a system to help value-up smaller units?

I'm not exactly sure how to reconcile the first one without putting them back in a training queue, but I'm pretty sure that the second one can't really be reconciled.

Successfully updated karma reason!

Agreed. Really like the thought process you put together here.

Following your presentation of Health/Attack/Dmg/Prob, the straight forward solution is to provide experiance on a 'per man' rate instead of on the 'per unit' rate.

Hence, a party of 3 would gain experaince faster than a group of 5 because fewer troops have to do more work. The flip side, of course, is that if you are able to put in the time into the larger groups, their combat output can become significantly better.

Ultimately it comes to understanding the tempo of time. Do you need troops now or later? Odds are you'll need both! But how much of each, and when?

Going this route really helps the smaller groups and parties....the troops don't become obsolete just because someone found a better tech.

I'm not for the 'upgrading' idea. Rather, I'd sooner have a 'merging' idea. Take 2 parties of 3 and merge them into a party of 5. You take a 1/6 penalty, but consolodate your troops into a larger stack. Something similar could be done for the other levels...

Successfully updated karma reason!

Quoting joasoze,

reply 1

Good work there.

The easy solution would be allowing us to upgrade unit sizes. The groups with 3 members that I use early game, I get quite attached to and want to keep. I think we should (for a price) be allowed to upgrade, but that the group should go down two levels (or something like that) to indicate that you have added rookies to your experienced group.

Easy and elegant methinks

Bingo!!!! This is all they need to do. I think the level penalty should be based on how much your upgrading. If you go from a 3 man to 5 man then you should loose 1 level. If you go from 3 to 9 then you loose 3 levels

Or even better you loose a percentage of EXP depending on the difference between your starting unit and the new unit, 3 man unit to a 5 man loose 10% exp while going from a 3 man to 9man would be 30% xp.

Oh and the upgrade time would be what ever it would have been upgrading the bigger unit minus the build time of the smaller unit. For example if it takes 6 turnes to build a 9 man units but only 1 turn to build 3 man unit then to upgrade it would take 5 turns and you could do the upgrade in two ways either go back to the nearest city and get in the que or just be in frendly territory and be stuck to the place on the map for the duration (but you can be attacked)

Something similar to GalCiv2 upgrade would be fine

Successfully updated karma reason!

Quoting GFireflyE,

reply 6

Agreed. Really like the thought process you put together here.
Hence, a party of 3 would gain experaince faster than a group of 5 because fewer troops have to do more work. The flip side, of course, is that if you are able to put in the time into the larger groups, their combat output can become significantly better.

Exactly the rationale behind the suggestion.

Quoting GFireflyE,

reply 6

This would be the equitable way of addressing the issue, by mooshing identical created units together and averaging the exp between them. Addresses XP, production costs, materials costs, and includes a loss mechanic in the form of "losing" a man in the process. Not my preferred outcome, but a totally workable one.

Bellack, your math for EXP loss is backwards. Going from 3 to 9 man units represents a 300% growth, so you would need to see a 66.6% loss of total exp for that method to not be broken, AND it does not account for the difference in production and resources form a 3 man to a 9 man. You are getting something for nothing.

Successfully updated karma reason!

Quoting Malsqueek,

reply 8

Quoting GFireflyE,
reply 6
Agreed. Really like the thought process you put together here.
Hence, a party of 3 would gain experaince faster than a group of 5 because fewer troops have to do more work. The flip side, of course, is that if you are able to put in the time into the larger groups, their combat output can become significantly better.

Exactly the rationale behind the suggestion.

Quoting GFireflyE,
reply 6

I'm not for the 'upgrading' idea. Rather, I'd sooner have a 'merging' idea. Take 2 parties of 3 and merge them into a party of 5. You take a 1/6 penalty, but consolodate your troops into a larger stack. Something similar could be done for the other levels...

This would be the equitable way of addressing the issue, by mooshing identical created units together and averaging the exp between them. Addresses XP, production costs, materials costs, and includes a loss mechanic in the form of "losing" a man in the process. Not my preferred outcome, but a totally workable one.

Bellack, your math for EXP loss is backwards. Going from 3 to 9 man units represents a 300% growth, so you would need to see a 66.6% loss of total exp for that method to not be broken, AND it does not account for the difference in production and resources form a 3 man to a 9 man. You are getting something for nothing.

I was just throwing number in there to illistrate the point. But of coarse input what ever numbers that would balance things out to avoid the something for nothing problem.

Successfully updated karma reason!

I think you need to add some attack and defense bonuses, even if its +1 / 3 levels or something like that.

Successfully updated karma reason!

Quoting UmbralAngel,

reply 10

I didn't know that accuracy went up with levels too but in general I don't think just HP and Accuracy is enough to justify having a level 20 unit with 3 figures vs a level 5 unit with 9 figures. I am pretty sure level 5 would win.

I think you need to add some attack and defense bonuses, even if its +1 / 3 levels or something like that.

It would need to be something like a 0.05 multiplier on Attack per level (essentially doubling the damage output at level 20). Putting a bonus on defense would be ludicrous though. Armors above leather are already essentially a guarantee of victory and they don't need to be improved in any way. In addition, making units with higher hit points and higher damage output ALSO harder to wound against an equivalently equipped but lower level unit would just add to the already tenuous balance in our unit construction.

Successfully updated karma reason!

I like all the ideas here.

Larger units should gain experience slower.

A Unit should be able to upgrade its size with the loss of some experience and levels and should require a minimal level such as 3 to 5. Upgrading should cost X amount of materials and the "potential" hit points should go up immediately, the hit points should then "heal" to their new high.

Identical units should be allow to merge into larger units. Without cost (other than the loss of one man) with experience of the new group being the average of the combining groups. The hit points should be the total of the two groups combining or the maximum for the new group.

Successfully updated karma reason!

Am I the only one who thinks that it's not a good idea to make all the group sizes have similar effects? That would only make the other choice (smaller units) the obvious choice instead. I mean, why would you ever bother researching and spending extra build time/upkeep, if you can just use smaller units to the same effect (the leveling will happen naturally if you use them)? The only reason to ever use bigger units is to sit them in cities as guards (because they probably won't gain much exp).

Even the argument that larger units would have higher attack isn't completely true due to weapons and traits that increases in effect "per level". For example, one of the best, and most often used weapons in the game is the lightning pike... with that, a group of 3, at lv 20 gets 60 lightning damage, while a group of 9, with similar hp is only level 7, has 63 lightning damage. There's a bit more difference when it comes to the piercing damage, but the difference in damage isn't nearly enough to make you ever want to bother with large groups. This is also one of the biggest problems with upgrading group size (you essentially spit out high level units that can take advantage of really powerful traits/abilities).

I'm not completely against the idea of making smaller units more valuable, but it needs to be something that doesn't try to mimic large units. For example, add a couple of powerful traits that reduces in power when more guys are in a unit, say... a "lone wolf" trait that boosts initiative and attack by 10-unit size. In order to avoid these traits from breaking early game, make them available when you research group size research, with the strongest ones (like lone wolf) only available when you research company. The goal here is to be able to design smaller units that are still viable when bigger units become available, not to make bigger units seem like a gigantic waste of time/resources.

Successfully updated karma reason!

By all means i would like to see some traits only available to small units sizes, and some traits available only to large unit sizes, as well as faction only traits. In addition to this, it would be neat if smaller unit sizes got xp faster than large unit sizes. I think with some rules like this you could make some really unique things in this game happen.

Successfully updated karma reason!

Quoting Kalin,

reply 13

That's a great idea, and really bears some investigation. Perhaps each unit size tech also unlocks a trait for each unit size below, where 3 man units eventually wind up with 3 traits to choose from highlighting various benefits of being a small, nimble, specialist unit, and so on, up to 9 person units whose benefit is that they are large.

I would still like to see larger units level slower, but with significant units size traits, that could easily go from a 50% penalty to a 25% penalty and still be pretty well balanced.

Successfully updated karma reason!

Playing devil's advocate here for the sake of thoroughness (and to ease my confusion), why would you want to buff small groups more when you have spent the time researching and building the larger groups? No big army of similarly equipped/trained units should lose to a smaller army (or single hero), though some element of that is important for the fun/fantasy feel that this game is chasing so diligently. The natural progression should be all about researching how training and command structures can lead to more efficient armies (ie, bigger ones that work better together in combat)--so where I am confused is why one would want to cling to the smaller units, except out of some wild fit of roleplaying. In my mind, smaller groups have outlived their frontline usefulness once the larger units take over (and there is already extra costs from having to train up those new groups--and upgrading army sizes like you do weapons is a bad idea, IMO, because as someone pointed out, it gives you way too much advantage for free); what I do is simply relegate those elite smaller units back to protecting my frontier cities (where they can still be useful against invaders in conjunction with the disposable city defenders).

In theory, I think if you ever stop researching and think, "I don't need to research this further because I get no added benefit," then the game progression is broken at that spot. Because progress of a civilization is always about finding ways to build bigger and better. Am I missing some important reason why bigger groups should be made less good/important by small groups?

Successfully updated karma reason!

Quoting Gorde,

reply 16

It is more that each step of size up wildly outclasses the previous.

It generally takes 2-3 extra levels to make them competitive at all, and a freshly built 9 man unit at level 1 flatly outclasses a level 9 3 man unit because of damage output alone.

Large units should absolutely have the potential to be better than smaller units of the same level and gear. However, there are time, money, and material constraints that can prevent you from making a 5 man unit, and if you cannot make the largest size, it is at this point better to not build anything at all. The people in this thread are either asking that you can moosh smaller units together so they are not useless, or that a mechanism be put in place to in any way help retain the mid and late game usefulness of your early game units. Not to make them EVEN, make them USEFUL.

The basic argument is: If every single step up the unit size chart flatly outclasses the previous step(s) by at least half with absolutely no down-side, what you have is not a choice. What you have is a new standard. Consider the impact of that across factions. One of the most important things you can do is learn to build 5 person units. The instant you do that, you outclass everyone that can't on a unit-to unit basis. Again, if it is not a CHOICE, then why is it made to look like one? Start everyone off with 9 man units and be done with it.

Successfully updated karma reason!

Quoting Malsqueek,

reply 17

I understand that people want an army to be useful throughout the entire game (which is the dominion of champions, really), but perhaps that isn't viable over other choices--and perhaps the strategic choice (of larger armies rather than better equipped/smaller armies) needs to be more of a commitment. The area that seems unpolished is that you can get army size upgrades simply during the course of researching improvements to your overall civilization. I really like that army size is separated from weapons/armor research, but perhaps it's not separated enough from other things that have benefit on their own (which would make you research the tech anyway).

Could we look at it from this angle: That there are not enough research branches--they pretty much all tie together within the 3 main categories, with the exception of the heroes/quests branch, which is a good example of how other areas of the game research could be handled. Personally, I think it's more satisfying to have many more research options, with each research option taking less time to learn (so there isn't this feel of must-have tech because the leap forward isn't so great). This lets you chase various branches of research and truly customize your kingdom's style through your choices of research (which should not be bundled together with other unrelated things; at least not as a rule). In our particular example, running up the army size branch of research would take time away from the other areas, like city production/food, weapons and armor research, and perhaps even the army experience branch. Then, if balanced well, someone who made the choice to focus on weapons and experience could have a 3-unit group that would actually outclass a 5-unit group, because of better weapons and higher levels (or they could turn them out faster to balance out their losses because their city research has been buffed instead).

If that's asking too much of the devs, perhaps a very conservative balance could be struck to make the gaps less obvious. However, if you tip the balance the other way, the gameplay will suffer more for it than it is now (churning out more small units because they are cheaper, faster, and almost as good). And then you might have the pioneer spam problem with military units. One thing we all can agree on is less options = less replay value for the game.

Successfully updated karma reason!

Quoting Gorde,

reply 18

That is also a pretty compelling concept, and one that would bear looking into. Essentially having two or three relatively distinct "routes" up a given tech tree, with some side routes that have prereqs. For instance, The Civ tech tree is ultimately about learning Production, Agriculture, and Glory (prestige, growth, etc). Perhaps those three trees would be better developed in a somewhat linear fashion, ultimately being researched with little required "cross talk" and then have sub-trees that have certain key points along them as prereqs. For instance, you need Agriculture (supply lines and feeding people) to research the 5-man tech, with that tech, and another up the food production line as prereqs for the 7 man tech.

It would also be worth running some numbers about which weapon and armor techs allow smaller units to compete with lower tech larger units (I may get to that later this week to see where that lies). I suppose if the choices were balanced more at the tech level, they would (largely) work out in the gameplay, returning ultimate benefit to the faction with the best resources and planning.

Successfully updated karma reason!

Really like the banter in comments #16 to #18. Really good points made on both sides. I few things to note in point form:

1) First, to know my position, I'm an advocate for not having the parties become obsolete as soon as a nation achieves groups.

2) I really like the idea of having parties be able to compete in some way.

3) I really like the idea of being able to focus on upgrades over logistics as a mechanism to balance...this JUST may be the correct answer.

4) I don't really like expanding the tech trees too much...that seems backpeddling back to WOM.

5) I really like making choices in the 3 different techs trees. Choosing between a strong deadend tech or a weak progression tech is what makes RPG!

Look forward to reading more on this topic!

Successfully updated karma reason!

The best way is to have traits that are only available under a certain size. For example:

Camouflage - only available for size <= 3, invisible to everyone who does not have an adjacent unit

Skirmisher - only available for size <= 5, immune to overwhelm

Slayer - only available for size <= 3, critical damage +100%, critical chance +50%

Dragon/Giant/Elemental slayer - only available for size <5, damage/critical bonus against monster type

Infiltrator - gets bonuses from hostile dominion

Etc, etc, etc...

And of course, all of these should need research, as not to have them available too early. Also, one could have extra trait slots for small units, also with research.

Successfully updated karma reason!

Yeah, Firefly, no one wants a research system that's impossibly long and complicated (at least not until the 2nd or 3rd installment of the game when it's all been refined and play-tested for 3+ years, heh). Let me illustrate a possible solution on this one issue: Let's say Cooperation has a dead-end branch that leads to Battlefield Tactics (5-unit), then Troop Coordination (with perhaps a prerequ of Supply Lines for 7-unit), ending with Large-scale Logistics (9-unit). These techs would cost exponentially greater amounts (in line with the big weapons upgrades, for example), but if players wanted to focus on army size at the cost of progress elsewhere, they could have their bigger groups sooner (without getting any other civic or weapon benefits from that research). That would be a choice, because other players would be getting magical equipment or more powerful cities during that same time (while keeping/making their elite 3-unit armies).

I know there are always negative ways to look at things, yet hopefully we can try to be constructive in offering potential solutions. I think the ideal way to handle research is to have meaningful interacting techs for every single area of gameplay, and research balance can go a long way to balancing the gameplay mechanics (through order sequence and cost); making your existing armies bigger (upgrading their size) feels wrong to me, like it removes all your costs of building them early (same as if you were able to add magical rings and other boosts to units who began fighting with sticks of wood).

Successfully updated karma reason!

Wouldn't a better solution be to just allow units to upgrade to larger groups as the technology becomes available? I can't see a simpler solution than that.

Successfully updated karma reason!

Quoting seanw3,

reply 23

Wouldn't a better solution be to just allow units to upgrade to larger groups as the technology becomes available? I can't see a simpler solution than that.

Stop with the easy solutions allready.

Sincerely
~ Kongdej

Successfully updated karma reason!

I don't think that's a "solution" at all. Neither the tech branching and the group upgrade suggestions do anything for smaller units, all it really does is reinforces that large units are the best, and once you research it, you should never even consider smaller units again, which IMO, is a waste of a concept. Only the traits suggestions or the original idea (reduced EXP gain on larger units) tries to address this. The idea is to make group size more of choice instead of a no-brainer (large units should be better, but there should be a role for smaller ones as well). Although, at least with the tech branching suggestion, you get some semblance of balance (assuming you CAN balance it, which is, admittedly, a pretty huge assumption).

Group upgrade, on the other hand, just favors large empires steam rolling small ones (because they will be the ones that can afford the upgrade costs), or favors players over AI (because the AI doesn't prioritize its best units very well). There's no reason to give either of those sides even more bonuses. Besides, it also has the problem of spitting out really high level guys, which there's no real game precedence for. For example, how do you calculate the worth of a level 15 guy? If you have traits, weapons and equipments that is increasingly effective per level, that lv15 guy you just spit out is incredibly valuable (and lets face it, you'd only upgrade if they were high level in the first place). This isn't to mention the already existing problems with the ability to bypass resource requirement via the upgrade system (IE: making cheap units and upgrade to reduce crystal costs). On top of it all, if you make the whole thing cost too much, no one would use it, so it just becomes wasted development time. The only argument FOR group upgrades is because players get attached to their units and want something to do with it instead of disbanding it or turn them into city guards. If that was the reason, then it would be better to do something along the line of a "retire" option instead of "disband" when in a city, that would turn them into troop trainers and give new units built from that city some exp bonus per level of the disbanded unit above a certain threshold (say 2 exp for every level above 5, to prevent cheesing the system).

Oh, and I think the + trait slot for smaller units idea would be pretty cool. Maybe make something that would let you set a default group size (you can't make bigger but you can make smaller) and then add +1 trait slot for the every size bellow 9. So group of 7 would have 4 trait slots, group of 5 would have 5 trait slots, etc... as long as you unlock the trait slots slowly via group size research (so that you can't design 6 trait 3 size units at the start of the game), it could be pretty interesting (assuming there are interesting traits to use in those slots too, of course).

Successfully updated karma reason!