First off, we need to ask “Why should we roll often?” The main reason is that there are parts of the game that rely on rolling a lot. If you only roll a few times per session, then players will not be rolling Unskilled often, will almost never be learning new Skills, won’t be using Action Points and thus will have little reason to generate them, won’t be making the most of their Kits if you are using those. A bigger problem is that is Fallbacks become way too strong - if each player only makes two or three Skill Rolls in a given session, then that Fallback could easily mean that they get nothing but Successes all session. If you find yourself in this situation, I think it’s best to restrict the use of Fallbacks (perhaps to once per adventure if your adventures last a few sessions) until you find yourself making more rolls. As an aside, this can also happen when you play short sessions - you may be calling for plenty of rolls, but if you only play for 90 minutes and 40 of those minutes are combat, then you still end up with too many Fallbacks per session. If rolling dice is rare, then giving characters Minor Conditions can almost feel like cheating because you know that they’ll be able to recover before their next roll, but when you call for plenty of rolls, including multiple rolls in succession, then Minor Conditions will work well, too, providing a short-term penalty and lasting only until the end of the scene.
Okay, so whose responsibility is it? Well, it is everyone’s responsibility, but the GM plays a pivotal role here. Players should be asking for rolls. Their characters improve when they roll more - especially if you are using the rules for Natural Advancement (p. ??X). If you’re not using Natural Advancement and you want to encourage the players to roll more, you should switch to those rules and tell the players the reason. The GM should generally allow players to make the rolls they want to as long as there is something at stake, even if it is minor - just keep in mind that if the stakes are small then the consequences should generally be small. Giving a player a Major Condition on a roll they didn’t have to make has the effect of discouraging them from asking to make rolls. Use Favors and Flaws more, or Minor Conditions that they can easily resolve if they roll a Cost. As the GM, if you’re already happy with how the action point economy and fallbacks are working, then you probably don’t need to change anything, but if you feel like the mechanics aren’t being used, you should try to call for more rolls. You can call for rolls whenever the players attempt to do something where there is something at stake - the only restriction is that you can think of a good Twist, and there is no rule that says you need to think of one in five seconds or less; you’re allowed to take a little time to come up with something good. You can call for multiple rolls in a row if you do it using Linked Rolls. Remember, it’s a very bad thing to require that a player succeeds on three consecutive rolls to perform their intended action, since the odds are very good that they will fail at least one of those rolls, but Linked Rolls are designed to solve that exact problem: making a Linked Roll is (on average) a slight benefit to the player, so they may be glad that you are calling for one, and linking a few rolls consecutively is a good way to give players a chance to learn new Skills and emphasize the interrelationships between Skills - being good at three Skills related to burglary is better than being good at just one, but only if you call for more than just a single Stealth roll to resolve the whole thing.
Why doesn’t it happen naturally for some groups? It’s all about habit, and it comes down to your experience with different games. Some games don’t have an economy that cares about how often you roll. Some games encourage you to wait for a really significant moment and then make one roll to resolve the whole scene. Some games want you to roll only for certain actions. If your group is used to another game with a different philosophy about rolling, then the adjustment can take a little time. The game Strike! is most like in this respect is Burning Wheel - “Advancement is Lifeblood” is what Burning Wheel claims, and that means that making rolls drives the action and drives character advancement, so the game stalls out if you’re not rolling the dice.
How can we make it happen? Since it’s about habit, all it takes is a bit of conscious practice. Tell the players that you want to call for more rolls this session, and ask them to remind you about it. If you are being conscious of it and looking for opportunities, you will find them, especially if you have your friends helping out. They should be glad of it, because they will get more character advancement out of it as they learn new Skills. As noted above, the Natural Advancement rules act as an incentive to players to want to roll more, even making Conditions desirable as they help players earn new Complications.
One other quick aside about Fallbacks: a fan said that his players were simply saving their Fallbacks for the final confrontation which they knew was going to occur and then never having to worry about their assured success. There are a couple of simple tricks to play with that - drain their fallbacks early by having other impactful rolls earlier in the session. What are they losing by accepting the Twist and hanging onto their Fallback? Apparently, not enough. So Twist harder and make it worth using their Fallback early. Then, once you get to the climactic scene, take the fight to the players - make them make rolls with defeat as a possible outcome but where success won’t mean their final victory. Then - if they roll Twists - they will have to use their Fallbacks to avoid defeat on rolls without guaranteeing victory. This is essentially the same advice as before about draining their Fallbacks, just within a scene instead of across scenes. If this is insufficient, there is a simple mechanical fix: one scene per session, at the start of the scene, the GM will declare that Fallbacks are off limits. With that rule, players will want to use their fallbacks earlier since they won’t be able to use them in the climactic showdown.
Examples are always good. Here is an example to demonstrate how to have more rolls in a scene and inspire good play.
The Bad* Way: Zhen wants to poison the Duke. She must make one Wealth Roll to buy the poison, and then Rosen makes one roll to sneak the poison into the Duke’s drink. That’s only two rolls, but they resolve everything.
*Bad for Strike! specifically - the same may not hold true for every game. A game without rules that allow you to “fail forward” might suffer badly from having lots of rolls and thus too many chances to tank the whole thing with one bad roll.
A Better Way: Zhen first must seek out the ingredients for the poison. The stakes here could be whether the ingredients will be readily available or whether more effort will be needed to source them (a side-quest or having to trade away something valuable). Having found a source, Zhen and his confederates must make a Wealth Roll to actually buy them, where the stakes could be whether the ingredients found are of high quality or whether they are subpar, making this a Linked Roll. Next, Zhen must mix the poison correctly - the stakes here are whether or not the poison has all the required qualities (undetectable, fast-acting, irreversible, deadly). A Twist here might give the mixture an unfortunate strong smell, making it harder to cover up. The one roll to acquire poison has been turned into three, and each one carries with it new details about the world: Where does one go to get ingredients for poison in this city? What type of ingredients are necessary? What could be used in a pinch if the ideal ingredient is absent?
The next stage - Rosen’s infiltration - could be spread out into rolls like that as well. A sequence might look like: extortion to alter the patrol schedule, Linked into climbing to get over the wall, Linked into stealth to sneak into the kitchens; from there, test your disguise to pass for a servant, Linked into sleight of hand to slip the poison into the King’s chalice unnoticed. The key thing is that getting a Twist on most of these rolls won’t scuttle the plan entirely, since most of them will be Linked Rolls. Note also that the crucial make-or-break roll needn’t always be the final roll, either. It might be that the key roll here is sneaking into the kitchens, with a Twist there representing capture and imprisonment, while a Twist on the final sleight of hand roll might simply indicate that someone saw you do it, though not the Duke, and though you are successful in poisoning him, there is now someone who has some very strong leverage over you.
You can hopefully see how adding these extra rolls helps you put the players’ actions into context and leads to a more satisfying result. Once the players outline their plan, it won’t go amiss for you to take a minute’s break just to think about the rolls you anticipate and how they relate to one another, and to decide which should link into which. The players might still change their plans partway through and surprise you, but that is always a risk.
Oh, and one more thing before I go… the reason I wrote this post is because I struggle with this very thing when I am the GM. I have a tendency when a player has a good idea to let them make one roll to resolve the whole thing, and my scenes can wind up unsatisfying when I do that. So I decided to write about it to improve my own GMing and help you all improve yours. I hope you find it useful!