I've been playing Mordheim with Leah over the past two days.
The games have been just the way I remember them. You get the fiddly start up where you try to generate an optimal team without the full resources that you need to do it (so you end up making a team that generally fits what you're aiming for, but you hope that a few games will generate the funds needed to truly create the team you want). Then you get the actual game play where strategy, team synergies, and quirky skills can give you an advantage, but a string of good (or bad) rolls can easily overcome player skill or strategy.
We've both decided to play four teams on a semi-random play selection roster.
Each players uses a team, and once the game is over each player lines up the leaders of their three remaining teams then rolls a d6 to determine which team to play next. 1-2: Team A, 3-4: Team B, 5-6: Team C. It means you don't get the one team played constantly, there's always a change-up. Eventually one team might get played more than the others, but that's just random; and even in the worst case scenario where one team gets played every second game, it still doesn't get too far ahead. Mordheim has some built in rules to accommodate some disparity in teams (such as heroic "dramatis personae" who might join a lower ranking team with a better chance of joining if there is a larger gap in relative power level).
But the notion of semi-random die team selection and self-regulating game mechanisms aren't really the focus of this post. What has really stood out in these games, and the factor that has been a make-or-break element in many of the scenarios has been the existence of saving throws. In many cases a character has two or more levels of saving throws (roll to dodge an incoming blow entirely, roll to parry a successful strike, roll to absorb the incoming damage with your armour, roll with a magical amulet or spell to avoid the worst, etc.)
(I looked at Saving Throws in the original Game Mechani(sm) series back in 2009, I don't think my opinion has changed all that much.)
I've always seen saving throws as the hallmark of old-school gaming. They definitely seem to be epitome of competitive play and the antithesis of "modern" narrative game forms.
Different games handle the mechanism in different ways, but have the same basic pair (or trio) of outcomes. A saving throw is a character's last chance to resist the effects of injury or story. A successful use of the mechanism may cancel (or minimise) the effects currently being resolved, while an unsuccessful use of the mechanism allows the effect to continue unimpeded.
You can often tell the type of player based on their reaction to saving throws. Those who favour their characters as "Mary Sues" often maximise their saving throws to prevent the chance of storylines happening to them (they prefer to be pretty princesses insulated from their world). On the other hand, those who want to immerse themselves in the world often don't care what the saving throw figures may be, such players may even ignore the fact they are able to call saving throws into play.
Saving throws add a level of stability. Without a saving throw, a single bad roll could cause a player to lose their character in a situation. Narrative styles of play may provide the player with some kind of option to perform a dramatic "last deed", or perhaps end up severely injured but able to return in a later story. Games focusing on the mechanisms (such as D&D) don't have as much of an input from the story into the mechanisms, instead the mechanisms completely drive the story, so a bad die roll simply means the end of things for that particular character unless saving throws are permitted.
In a competitive game like Mordheim, saving throws might slow down combat, but they turn the sequence of events into a back-and-forth scuffle, where two characters might become so focused on breaking a deadlock that they ignore the actual objective of the scenario (which might be to capture a building). The saving throw element built into the combat makes the conflict feel more important, while stabilising it's outcome compared to the true strategies elsewhere that might win or lose the game. Die rolling is apart of the paradigm for miniature battlegames, you expect it, and you expect the back and forth. This is especially true with d6 based games where the chunky granularity of results often needs something nuanced like saving throws to provide finer differentiation between opposing forces.
Saving throws are a way of preventing story from happening in a number of ways. Firstly, they cause actions to take longer to resolve. An action that can be resolved with a single die roll or card draw allows the story to flow more easily...an action that requires a mechanism for possible success, a further mechanisms to see if the possible success is countered, then a final process to determine the output of those mechanisms into\the story...that second option clearly doesn't allow good story flow.
Secondly, if you want things to happen in a game, saving throws are purely designed to prevent that happening.
I can see the reasons why some gamers embrace the notion of saving throws. They are a way to determine how tough a character is in the face of adversity or potential injury.
But the flip side is just as easy to see. I especially saw this in an ongoing LARP campaign, where players were able to create effects that basically caused saving throws against storylines that we were trying to use to keep new players interested. This reinforce stability in the game, which was great for these long term players who were trying to show that they had established a solid power base, but didn't make for much fun among those players who wanted a bit of action.
Just something to keep in mind when you consider adding saving throws to your games.