Designer Kevin Wilson recently weighed in way at the end of a thread on the Sorceror King about potential upcoming changes to the Avatar battles (in the form of allowing the Avatar to place orders). This reminded me that perhaps I should post a set of Avatar battle changes that I have been experimenting with. Note that these are not just 'off the cuff' suggestions - I have been, and continue to, test them experimentally through extensive self-play.

1) The Avatar drops all threat and cards when the battle begins, but then begins to collect threat again as normal. This threat can be used as in Lieutenant battles (ie. 2 threat for 1 movement, or 1 boost). This change brings back the time pressure that normally prevents the 'snipe around a corner' tactics in Descent.

2) The Avatar has permanent access to the following abilities for the listed Threat cost:

Aim: 4 threat

Dodge: 4 threat

Dark Balm: 6 threat

These are to add strategic interest to the fight, reduce variance (the outcome really shouldn't depend on whether or not the Avatar misses his first attack) and prevent cheesiness like infinite-webs-stopping-the-Titan-from-ever-attacking.

3) Hero orders are removed when the Avatar fight begins. The Avatar fight should begin with a clean slate for both sides. Otherwise, the Heroes would always have guards and dodges on for the Avatar's initial attack

4) The Avatar's should receive more than 2 HP per conquest earned in the Final Keep. I haven't settled for sure on what this number should be, right now I'm thinking around 5 HP per conquest is appropriate. Gold-equipped Heroes simply don't take many casualties, and they can also do a ridiculous amount of damage, making 2HP per conquest completely insufficient.

I have been testing these changes using the Spider Queen and the Hero party from my group's recently finished game, where they soundly defeated her. I re-ran the battle at increasing levels of hero 'upgrade,' where 1 upgrade is defined as one extra skill/2 dice/secret master training for every Hero. As to which upgrades were selected, I used the progression that the Hero party had used in our game, which I make no claim to being optimal, but which was certainly sufficient to easily crush me. In our actual game, the total upgrade level of the party was 9.5 by the end (.5 = 1 power die instead of 2). For the tests, I also assumed the Spider Queen had every relevant Avatar upgrade. For each level of Hero upgrade, I then recorded the amount of bonus HP (due to Conquest in the Final Dungeon) that the Avatar WOULD have needed to win the fight. The results look something like this:

# Hero Upgrades Bonus HP Needed

0-3 0

4-6 60-70

7-8 150-170

9 300

Finally, I also had played out the fight using the original rules and the final upgraded strength of the Heroes, and found that I would have needed 1100 bonus HP to win the fight. So the rules I've suggested definitely make a major difference, but in a very reasonable direction I think.

Experimentally sound, as always. How often did your heroes die during the final 4 dungeon levels + the keep level? Seems like 60 conquest is a high number to shoot for when facing heroes with 9 upgrades (if we're using 5 hp per ct).

I think the avatar would need something else to spend his points on. I was thinking about allowing him to purchase gold dices at 150% of the normal rate (or maybe any generic upgrade at 150% cost), but I didn't have time to test this.

Allowing the OL to spend threats to upgrade movement and attacks is needed indeed to discourage hero camping in the final battle.

Galdred said:

I thought there were three generic Avatar upgrades that did this already?

Big Remy said:

I was talking about allowing the OL to purchase these with the CT earned in the final dungeon, but at an increased rate, then about thundercles suggestion to allow the OL to spend threat to increase attacks and movements in the final battle.

Ah okay, that makes sense now.

Galdred said:

My what? *looks up*

Yeah, pretty sure that was Immortal suggestin' that stuff.

you're right, and I cannot edit it

First, as to OL conquest obtained in the final Keep:

In my original Spider Queen game, I obtained only 12 CT in the ENTIRE 4-dungeon-levels + Overlord's keep level. However, in that particular game, I was definitely outplayed throughout the campaign and deserved to lose. So I wouldn't want any changes I suggest to change that outcome - but it was clear to me the Spider Queen would, under the original rules, have lost even to a considerably less powerful Hero party.

Second, an update:

I've now tried the same experiment against the Sorcerer King, with the addition of Kevin Wilson's post allowing the Avatar to guard making the 'trick' of that fight less of a non-issue and modifying the Hero party slightly to take into account their new opponent (for instance, the Scarab of Death was by far the best Ranged weapon against the Spider Queen, but against Sorcerer King, the Sorceror's Orb is far, far better). I'm definitely finding that due to his vastly superior defense, a few extra HP on the Sorcerer King go MUCH farther than they do on the Spider Queen. This is not surprising of course, since it's clear that the Sorcerer King has the best defense by far out of any Avatar just by reading the cards. However, my initial thoughts of 5 HP per CT may have been excessive; although the Spider Queen and the Sorcerer King represent the two extremes in terms of Avatar defense, so the others are probably somewhere in the middle. So I'd split the difference and say either 3 or 4 HP per CT would be more approrpriate.

That still leaves the Spider Queen hopelessly weak against decently strong Heroes, but this is most likely an issue with the Spider's Nest upgrade (the one that puts a Web on anyone who Battles during the final level of her Keep). Most likely in the designers' minds, this upgrade is supposed to be reasonably good, whereas in actual fact it is entirely worthless. This is because the Heroes can easily still get two attacks each by declaring a Ready with Attack + Guard, and that getting immunity to Web isn't even all that hard anyway.