I don't if I'm just being slow, but I've only very recently realised that there is a fairly serious bias towards both Combat skills and Melee trait dice in the published heroes. I had been musing over the fact that we generally seemed to get more than 4 Combat skills total when drawing random parties of heroes, and I finally decided to check. Numbers below. I haven't included the promo heroes, as I don't own any of them and can't be bothered trawling through the BGG image galleries for the stats:

Skills (total skills across all heroes):

JitD: 22 Combat 19 Subterfuge 19 Wizardry

WoD: 8 Combat 7 Subterfuge 3 Wizardry

AoD: 7 Combat 5 Subterfuge 6 Wizardry

ToI: 10 Combat 4 Subterfuge 4 Wizardry

Total: 47 Combat 35 Subterfuge 32 Wizardry

Trait Dice (total across all heroes):

JitD: 20 Melee 18 Ranged 22 Magic

WoD: 8 Melee 5 Ranged 5 Magic

AoD: 7 Melee 6 Ranged 5 Magic

ToI: 12 Melee 3 Ranged 4 Magic

Total: 47 Melee 32 Ranged 36 Magic

Is this news to anyone else? Tomb of Ice is a particularly egregious offender with an enormous melee bias: there are more combat skills among its heroes than Subterfuge and Wizardry combined, and the same is true of trait dice. The original game wasn't particularly biased, and Altar of Despair is also quite close, but even here there is a certain melee-favour.

I find this somewhat irritating- it means you'll see each Combat skill about 50% more often than a Wizardry skill.

Interestingly, in JitD the trait dice bias was actually most favourable to Magic attacks, which mirrored (though not on the same scale) the enormous bias towards magic weapons in the treasure decks.

I wonder if it's intentional- I suspect not.

This is news to me. I agree that it's probably not intentional, but I think it IS probably for the best. Ranged and Magic heroes are already proven room-clearers (particularly Magic heroes with their runes. ) Melee heroes tend to hit hard, but their lack of range and often low movement limits their potential.

I guess my point is, they don't really need more Ranged or Magic heroes. Having a larger number of Melee heroes ensures at least one player will have a tank option in his draws, and two or three decent non-Melee is enough to get by in a party.

Interesting observation. Could be partly that melee heroes seem to be more popular (I've seen several threads suggesting 2 melee heroes for RtL, and one poster advocates parties with 3-4 melee heroes in vanilla). As an alternative way of looking at it, here's my count of heroes categorized by their distribution of trait dice:

Pure Melee: 6 (8 splits where melee is highest)
Pure Ranged: 3 (6 splits where ranged is highest; note: Ronan can still roll 3 ranged dice with Pico)
Pure Magic: 6 (4 splits where magic is highest)

Melee/Ranged: 9
Melee/Magic: 6
Ranged/Magic: 4

Even Split: 4

Additionally, given the strong correlation between traits and conquest value, it may be interesting to note that there are:

12 heroes worth 4 conquest
16 heroes worth 3 conquest
9 heroes with 2 conquest
1 hero worth 1 conquest (Zyla)

As I've discussed before, most melee-primary heroes are worth 4 conquest, most ranged-primary heroes are worth 3, and most magic-primary heroes are worth 2.

I notice several things. First of all, pure ranged heroes are only half as common as pure melee or magic heroes; this may be partly due to the glut of 3-conquest heroes and FFG's unwillingness to make ranged heroes worth 2 or 4 conquest (they're much more willing to make melee or magic heroes worth 3), though it could also be that the designers just don't like pure ranged heroes for some reason. Regardless, this may contribute to the impression that ranged attacks don't deal enough damage, since heroes using a ranged weapon are much less likely to have 3 trait dice than heroes using a melee or magic weapon.

Next, I notice that melee traits are over-represented in two-trait splits by a significant margin (they should show up in 2/3 with a random distribution, but actually show up in over 4/5). This could be because ranged and magic attacks are more similar, making a ranged/magic split seem redundant.

Magic is under-represented in two-trait splits; I'm tempted to think this could be related to the fact that magic weapons usually require giving up heavy armor, making it hard to "dabble" in magic, yet magic is the only trait that shows up with only one die more often than it shows up with 2, suggesting that the designers actually like to use it as a hero's second trait, but don't like giving split traits to magic-primary heroes. Puzzling.

Does anyone reading this actually routinely play with completely random hero selection, rather than draw-3-pick-1 or something similar? If so, any observations about typical party composition?

Steve-O, your reasoning seems a trifle inconsistent. You argue that it's good that there are more melee heroes because ranged and magic heroes are better, but then you suggest that the players would deliberately choose melee heroes over ranged/magic, given the option.

My impression is that to begin with in JitD everything was fairly well evenly distributed. There were 6 melee orientated heroes, 6 ranged orientated, 6 magic orientated and two jack-of-all-trades. This fitted in with the general design of the whole game in the same way as treasures and monster dice shared similar distribution between tiers/levels/types.

As the game saw more play in practice (and I believe it became common for a party to have two melee heroes) there started to be more of a melee hero bias creeping in with the expansions even though th WoD treasures finished off the even distribution of shop items and treasures. There was a sense in which AoD still tried to follow these rules in the treasures a new treasure 'type' (knockback, daze and web) for each attack, but a gold treasure only for the melee type, which I suspect was down to having used up their card count for the expansion, just as the Lone Wendigo card had to be removed from ToI. Eventually with ToI, the design team, (which I seem to recall had somebody other than KW heavily involved) seemed to completely abandon the idea of any such order and just through in stuff they thought was cool.

One result of all this is that while a larger frequency of melee heroes means more two-melee parties, (although I'm sure somebody will disagree with that comment), there is a resulting infrequency of melee treasures. In our recent RtL, I found targeting the melee weapons with Crushing Blow effective because they were harder to replace for two heroes.

Antistone said:

Actually, I was suggesting that they would want to pick at least one melee hero, because a party without a tank is going to get hurt. I apologize if that wasn't clear. Having more melee heroes in the "deck" helps to ensure at least one will come up. Conversely, taking too many melee heroes seems like a bad idea to me, as it would limit damage potential. I say this without actually having tried a party with 3 or 4 melee heroes in it, mind you. Although, assuming you have 4 players using the draw 3 pick 1 method (which we definitely prefer) then I suppose the odds would be pretty good anyway, even if all three attack types were evenly represented.

As I said before, I agree with the OP that it probably wasn't a deliberate choice. At least for the base game, they were working with heroes already designed in Runebound and although not all of them are exact translations between the two systems, there is a pretty obvious effort to keep them thematically linked to their counterparts, I think.

There is no game-mechanical reason why "tough hero" should necessarily equate with "melee hero". Requiring a "tank", therefore, in no way justifies a melee bias in attack types. In any case, the concept of a "tank" doesn't really apply all that well in Descent in the first place. There are two skills, Taunt and Bear Tattoo (both Combat skills, granted), that actually give a hero any sort of real tanking ability- beyond that you can only rely on trying to keep the tougher heroes on the outside of the group so that spawns tend to have to go for them.

I'm really not sure what you mean about "limiting damage potential". Melee attacks pretty much universally deal the most damage at all levels. They're easily the best type of weapon among the shop and copper item decks. Granted, at Gold level, Magic weapons are more desirable because they have greater multi-target potential, but then at Gold level (at least in vanilla), your trait dice matter much less anyway, as they represent a much smaller proportion of total damage and you can probably afford a few Power potions.

In addition, there are very few Combat skills that actually work only with Melee attacks (about four, from memory), so the correlation between Melee attacks and Combat skills is slight.

Oh, and for the record, we do play "one-draw, no returns" for hero-selection.

YellowPebble said:

Well, based on the heroes FFG gives us, someone over there seems to think there's a reason (though maybe not a game-mechanical one). There are no 4-conquest heroes printed to date with zero melee dice, and all but one have at least two.

There is also the heavy armor/runes issue that makes magic heroes a bit dubious as "tough heroes".

Antistone said:

I know, but since I'm complaining about the design of the heroes anyway, I may as well complain about the melee=tough policy as well. :-)

Heavy armour/runes is a point well made, but firstly, that does only apply to Magic heroes, and if Ranged heroes were actually as tough on average as melee ones, I think most people would agree that 2/3 of heroes being "tanks" was enough.

Secondly, it doesn't make *that* much difference anyway. Being unable to wear heavy armour is, barring the occasional treasure draw, only a -1 armour penalty. In fact, it's not even that bad, since you might want to consider not wearing chainmail for the extra speed, or alternatively play with a Mage Staff (they are admittedly not very good under most circumstances).

In case you were still wondering, I've added in the skills and trait dice from the four promo heroes.

Skills (total skills across all heroes):

JitD: 22 Combat 19 Subterfuge 19 Wizardry

WoD: 8 Combat 7 Subterfuge 3 Wizardry

AoD: 7 Combat 5 Subterfuge 6 Wizardry

ToI: 10 Combat 4 Subterfuge 4 Wizardry

Promo: 4 Combat 4 Subterfuge 4 Wizardry

Total: 51 Combat 39 Subterfuge 36 Wizardry

Trait Dice (total across all heroes):

JitD: 20 Melee 18 Ranged 22 Magic

WoD: 8 Melee 5 Ranged 5 Magic

AoD: 7 Melee 6 Ranged 5 Magic

ToI: 12 Melee 3 Ranged 4 Magic

Promo: 5 Melee 3 Ranged 4 Magic

Total: 52 Melee 35 Ranged 40 Magic

Seems that afterward Ranged heroes get left a little further in the dust. It is of note though that one of the four is a pure ranged hero (worth four conquest), so that's something I guess.

Not sure if it changes anything, but if instead of looking at totals you look at distributions you get the following:

Melee:
- 7 heroes with 3+ dice
- 9 heroes with 2 dice
- 11 heroes with 1 die
- 15 heroes with 0 dice

Ranged:
- 4 heroes with 3 dice
- 6 heroes with 2 dice
- 11 heroes with 1 die
- 21 heroes with 0 dice

Magic:
- 7 heroes with 3 dice
- 4 heroes with 2 dice
- 11 heroes with 1 die
- 20 heroes with 0 dice

Less than half of the heroes have 0 melee dice, while about half of them have either no magic or no ranged dice.

It looks to me like they were trying to make it so that every part has a good chance of being half melee, with the other two being ranged or magic (with a very slight emphasis on magic). But that's probably way too simplistic of an analysis, since it doesn't take into account the power differential between characters, conquest values, speeds, etc. And it assumes that FFG had an agenda when making characters and doesn't just make what they feel will be cool/fun.

YellowPebble said:

Well, Melee heroes are often slower than the others and they can only hit adjacent enemies. Yes they do the most damage per roll, but their ability to get to the target is limited, and they can generally only hit one target at a time. They can Battle, but then their effective range is even shorter. They will usually kill the thing they hit, but they only kill one thing at a time. The excess damage they do over and above what they needed to kill is impressive and all, but it's ultimately wasted.

Before you go poking holes in that, I'm aware that there are some special circumstances that can change what I'm talking about above. Reach can get you out to two spaces (oooh), Sweep can hit more than one target per attack (aaaah) and overkill is certainly handy against Undying enemies. This is all true, and I'm sure there are other special abilities besides. I'm talking in general here, on average. I'm not considering special abilities, at least not for the moment.

Ranged and Magic heroes can go for anything in line of sight (within reasonable range.) They won't do as much damage per roll, but they can put it almost anywhere they might need it. Also, abilities like Blast are a lot more common on Magic attacks than Sweep is on Melee, so hitting multiple targets with one roll is a lot more common with Magic. Ranged still usually only hits one target per attack, but a range attacker has more targets to choose from than a Melee attacker (everything in LoS rather than just everything adjacent.)

You've no doubt heard people talking about how an efficient group of heroes can often clear a room of its initial monsters before they even get a chance to act. That happens primarily because of ranged and area effect attacks, at least in my experience.