The other major consideration, is if you are summoning units that get blasted off the table, unless you win via major victory, you can easily end up with over 100% of your starting force killed, preventing you from attaining a minor victory, so summoning really is a gamble, you either summon very conservatively, allowing you the option to gain a minor victory, or you go balls to the wall knowing your only hope is via a major victory.
At the same time, though, you can prevent your opponent from having much hope of achieving a major victory themselves, and you can completely subvert the sudden death victory game conditions by starting outnumbered, but outnumbering your opponent by the end of the first turn.
In my handful of games so far, I haven't played against anyone else pushing summoning, so I haven't had the experience of facing it down.
That said, I have played my own games without summoning, with some summoning, and with heavy summoning.
The 'some summoning' games were with one or two casters, no special characters - so if you try a summon you aren't casting anything else with that wizard - and with few casting bonuses - just a corpse cart in one, and just a randomly rolled terrain effect in the other, and thee seemed to work pretty ok for the most part, not too much different from non-summoning games (no wizards - a skeleton force led by Krell and a wight king BSB), or wizards that just don't summon (a ghoul list led by a ghoul king who spent his casts on his signature spell or armor buffing himself).
The heavy summoning games included either multi-cast special characters (Arkhan and Kemmler in particular, I haven't broken out Nagash yet) or 3+ casters and multiple stacking casting bonuses (Arkhan + morghasts; terrain + corpse cart + summoned morghasts), and those games all got out of hand quickly. Things that seemed like aggrevating factors:
* Multiple summoning attempts in the same round. Adding a unit every other turn is not the hugest deal. Adding two or more units every turn is. Likewise, some of the buffs are big deals (vanhels from our necros in particular). When summoning comes instead of buffing a key unit, that's a major trade off, but when you can do both? not so much.
* Casting bonuses making summoning too easy. Once success is automatic for regular units and monsters or double sized summons start appearing on average or even sub average rolls, things get out of hand quickly. Without +2 or more to cast, trying to summon a terrorgheist feels like a hail mary play that deserves to swing the game if you pull it off, because trying for it is far more likely to result in you simply wasting your wizard for a turn, and again, with spells like vanhels that you could have more easily cast instead, is no small cost.
* It is too easy to avoid dispelling altogether when summoning. Most spells, you need to be in range of enemy units, or of friendly units about to engage enemy units, in order to be relevant, bringing you into range for dispelling. Likewise, most non-summoning spells have line of sight requirements likely to bring you into line of sight of the enemy. Summoning breaks both of these, and summoning wizards can have huge impacts on the game from the back of their deployment zone and / or behind line of sight blocking terrain. Granted, if you're stacking huge casting bonuses dispel attempts aren't likely to be successful anyway, but still, it's a check on casting that summoning bypasses in a way that maybe it shouldn't.
* Summoned units that act on the turn they are summoned give opponents no way to respond, and this can result in games that are more frustrating than fun, and feel non-interactive. Harbingers are very likely to succeed on charges from 9.1 inches away, archers and terrorgheists that can fire the moment they are summoned can also just feel unfair. All the more so for summoned casters who can in turn summon more of themselves in the same hero phase, from what I've heard. For the record, this is also an issue for 'deep striking' sigmarites and ambushing units, so it isn't a problem exclusive to summoning.
* Victory conditions. As mentioned, summoning can interact poorly with sudden death, and it throws off the regular victory conditions as well, skewing the odds of total victory towards the summoner and minor victory towards their opponent.
There are some house rules I've been considering proposing:
- limits on number of summoned units a caster can control at a time, maybe restrict it to the number of spells they can cast in a single round. Once a necromancer has summoned a unit, or arkhan two, or Nagash eight, they can summon no further units until one of the units already summoned has been destroyed. If for some reason the number of spells a character can cast is reduced (relevant for nagash), then either more units must be destroyed before they can summon again or, alternatively, the summoner chooses units already in play to be destroyed until the summoned unit count is back down to the amount the character can now control (this in particular would have some interesting narrative / scenario type implications for games played against Nagash - who builds up an army quickly but as you damage him that army crumbles - but as I haven't actually played Nagash personally this is a pretty idle suggestion without any experience behind it).
- Similarly, if a summoner is destroyed, any units they've summoned are either destroyed, or become much easier to destroy (reduced to bravery zero and have to test bravery every round regardless of casualties, in emulation of our old crumbling rules, perhaps). This takes some of the skew off the major victory thing, as the opposition can assassinate the casters to get rid of excess summoned units, keeping major victory as a possibility. I like crumbling in particular as it has a cool narrative feel to it that seems fitting with the tone of Age of Sigmar.
- To attempt to unbind a spell, a wizard must be within range and line of sight of either the caster OR the effect (ie, the unit targeted, or the first model to be summoned, which would be placed as a marker before resolving unbinding attempts, and simply removed if the unbinding attempt is successful. Alternatively, range and line of sight are simply not required when attempting to unbind summoning spells. This, however, proves problematic when casters with bonuses to unbinding take the field, which could then hard counter any summoning attempted by caster that aren't cheezing out with stacking casting bonuses.
- Speaking of, maybe put a limit on stacking casting bonuses. Like, only from two sources for instance. Ie, if arkhan is next to some morghasts then he can stack their bonus with his staff, but if he's next to a corpse cart as well then he doesn't get that bonus on top of the two he already has. Or maybe limit it to a single bonus with no stacking at all, but while that might be more 'balanced', it seems counter to the narrative to prevent, say, Nagash from benefitting from morghasts, for instance. Or maybe simply cap total casting bonuses at +2, unless a single bonus is greater? Even that seems like a bit much when most summoning spells resolve on a 5 and thus could be nearly auto-cast at that value, plus again it seems counter-narrative to block Nagash from benefiting from morghasts.
- Perhaps outright bonuses to casting are a bit much to be thrown out willy-nilly. Maybe instead some bonuses should be 're-roll ones' or 'increased range' or 'd6 extra wounds worth of models' or the like. Maybe some of the current casting bonuses should do something else entirely - maybe morghasts could provide a bonus to bravery checks (analogous to their old crumble reduction), bolstering existing units instead of making it easier to bring new ones to the field; or the corpse cart's lodestone could add d6 extra zombies to any new unit of zombies summoned in its range instead of making all spells easier to summon, etc.
- Maybe re-institute auto-failure for casting rolls of 2 or 3, so at the very least these spells are never automatically successful, and/or an auto-pass for unbinding on double sixes. On that subject, instituting a 'one always fails' policy on to-hit, to-wound, and saves might also be a good call, but that's off topic.
- Maybe restrict summoned units from acting at all the turn their summoned, giving opponents the opportunity (or at least the chance of an opportunity, dependent on initiative roles) to respond in some way to the appearance of new units before they start inflicting damage. A terrorgheist appearing on the table is a momentum shift enough, a terrorgheist appearing and removing one of the enemy's key heroes at the same time before they can react in any way is maybe too much. Similar restrictions might be worth considering for other 'reserve' units. Or maybe not.
