This actual play post is to share my recent discovery regarding stakes in Shadow of Yesterday.

In the past I've done stakes pretty traditionally.  I found the following type of stakes:

Win/Lose - binary, boring, but sometimes works.

Win/Complicated Win -  awesome.  If the conflict is "These nasty orcs want to chop us up", a win gets you "The orcs are all chopped up,  no problem."  If you lose, you get "The nasty orcs are dead, but you're wounded and bleeding, and your axe broke on that last orc.  Oh and you hear an orcish horn.  Sounds like there's more on the way."  Or if the conflict is "I want to pick this lock and get to the jewels", a win gets you a picked lock, while a lose gets you a picked lock, along with two guards who come snooping around to see what that scratching noise is.

I can sum up these stakes by saying that while we're playing with the facts of the game, the GM is still the master narrator, in complete charge of the story and its direction.

Something happened at my last TSOY game though that changed that.  I'll share the stakes that we used, and then discuss a little more.

The party was at a gambling house, trying to track someone down. The trail had grown somewhat cold.  One of the party had Secret of Contacts, and declared that one of the bouncers was a good chum of hers.  The player asked me if the bouncer went and hung out with other bouncers from other places in the city.  She wanted to know if he could be convinced to talk to them and find some leads for them about the guy they were tracking.  I said "Sure, lets go to dice.  If you win the roll, the bouncer leaves work in a few hours, goes out to his favorite watering hole with his buddies, talks shop, and then does a little digging for you, and finds out some information, which he will come back and share.  However if you lose the roll, he goes out with his buddies, asks a little, and gets nothing."    I could have improved on this further perhaps by saying that if the conflict was a loss for the player that he would return with incorrect information, or even that all of the questions would complicate the situation by alerting their quarry.

Later, the party was looking for someone else, in this case, the brother of one of the characters.  The brother's name was Stephan, the player's character was named Josev.  They went to the home of an Uncle, a man named Thome, to see if he'd seen Stephan.  Josev went in to speak with him, said "Hi Uncle, have you seen Stephan?".  I replied, "Your uncle says no, he has not seen him.  Now get your dice out.  He is lying to you.  Lets see if you realize that or not."  So we went to dice, the stakes were that if the player won, he was aware that his uncle was lying to him.  If he lost, he was oblivious.  The player lost the roll, and even did BDtP, but his Uncle was apparently an accomplished liar, and so Josev left, convinced that his Uncle had not seen Stephan.

Then, the players wanted to start asking the usual suspects if they'd seen Stephan.  We decided to go to dice and do a streetwise check, and one of the players immediately upped the stakes by asking for these stakes: "If we win the roll, Stephan walks right into us."  I went for it, and in order to meet the scale and scope of the stakes, said "Fine, if you lose, Stephan has gone into hiding under the protection of your Arch Rivals across town."

I'm sharing this because this was a total "light bulb" moment.  I suddenly saw a way to use stakes that I had not seen before.

In the first set of stakes with the bouncer, I was able to use them in a way that would either help or possibly hinder - or at least complicate, the investigation.    Though I was and am happy with the way we did stakes for the Uncle, if I'd wanted to use them differently, instead of saying that the uncle was lying, and using dice and stakes to determine how good a liar he was, I could have used the stakes and dice to determine the facts.  I could have said that if the player won the roll, then they'd discover that Stephan was staying with the Uncle.  while if they lost,  he was not, and the uncle had not seen him.

As I said, this realization has changed how I look at stakes, and might change how I play TSOY entirely.

Cheers!

Let me rephrase your realization to see if I understood you correctly:

You realized that the conflict might, in addition to controlling character success and failure, also control setting facts such as where the prey of a player character is hiding. Effectively, the dice determine whether the SG will complicate a situation with further steps; a success allows characters to cut directly to what they're interested in.

That's a quite valid way of looking at the matter. To illustrate a different viewpoint, though, the way I'd have done it would have been to set the player character directly against Stephan - Stephan, after all, was the one who was trying to keep out of sight, so it'd make sense to me if he had to defend himself with his own ability at hiding. He would also be the right guy to go into BDtP with, if somebody needs to suffer the consequences of that. This is not to say that Thome couldn't get involved - he could simply support his nephew, for example, if he wanted to help with the hiding.

The reason for why I like to do things this way is that I want to preserve the rights to conflict procedures (and other protagonizing mechanics) for important secondary characters, as well as player characters. The sort of play you suggest here essentially handles the whole setting outside the player characters as a single antagonist, where winning conflicts against any individual part of it (like uncle Thome) suffices to win against all of it. In this case, for example, you're proposing to resolve Stephan's efforts at hiding by having Thome roll in his stead. That would only make sense to me if I'd decided that Stephan was putting his trust on Thome to hide him.

I can feel the pain related to useless conflicts as well, though; the job of the Story Guide is largely about recognizing and proffering significant and interesting conflict situations while framing through irrelevant stuff, so when you introduce uncle Thome, you certainly should have some real function for him in the story, too. Knowing where Stephan was hiding is just that sort of significance, so in practice I'd probably play it almost the way you suggest - if Thome buckles, he betrays Stephan and helps Josev find him. I wouldn't think of this in terms of shuffling setting facts, though; for me it's just plain dramatic coordination to tie strings of story together and give NPCs who get the spotlight some real relevance in the situation, too. So while I wouldn't outright have Stephan's hiding place change based on the results of the check, I'd certainly decide when Thome comes to the scene whether he has anything relevant to offer to the situation (and, consequently, whether there is a basis for conflict here or not).

But I can easily see playing it the way you suggest, too - the old fogies call that No Myth play, where the GM's dramatic coordination reaches into backstory elements like where Stephan is hiding with impunity.