The other issue we've had is that, what happens when you roll 11 seconds to find something but you'd never spend that time?

It's artificial and game-y to roll 11 seconds on 2d, and say "I know it'll take 11 seconds, so I'll stop looking immediately."

Worse, it

*feels*like cheating to stop, so people feel compelled to keep looking.

What my players suggested, and do, is announce ahead of the roll the maximum time they'll spend looking. "If I don't find it in 5 seconds, I stop." So they have a predecided out, and it doesn't feel like a prescient decision since they don't know how long it'll take, just how long they'll spend.

But I had another idea this past weekend, and I'm trying to think of how to expand it.

**Another Way To Roll**

In Sunday's game I tried something else - don't roll for time,

*roll to find it*. Instead of 1d6 seconds to pull something out of your pocket, you've got a 1 in 6 chance of finding it. After 6 turns, you just find it. It added a lot of fun, because it was "do I keep trying, or just give the hell up?"

That second part is generous - it never takes more than 6 seconds to find something, which anyone who's looking in their backpack only to eventually need to dump it over to find something will attest.

For a loose item in a bag, it's 2d seconds. This is trickier - with 1d it's clear you have a 1 in 6 chance of any given result. For 2-12, it's not so clear. But on average you'll find it in 7 seconds, and your maximum time is 12 seconds. I'm still mulling this over - it's a question of what the right chances would be. If I wanted the odds to be the same as the 1 in 6, I'd just say roll a 4 or less - you've got basically the same odds as rolling a 1 on a d6. That seems a bit generous, though - you shouldn't have a 16.66% chance to find it in 2 seconds, then 16.66% chance to find it in 3, etc., should it?

So I'd like to come up with a number for it, but math isn't my strong suit. Anyone? Doug? Bueller? What is a "good" roll-under number to match the odds that, yes, sometimes it'll take 12 seconds to find things but most of the time it won't.

**Is there another way to do this?**- Yes, you can simply have the GM roll in secret and announce when they find the object. This both get rid of the "meta" question (they really don't know how long it will take) and leaves the normal system in place. But I'm not a big fan of things that make me roll in secret, track it in secret, and otherwise take rolls out of the hands of players. If it's possible just stick odds on a roll and let the players roll (and debate using Luck to get it out faster, or bitch about how they rolled all the 1s out when they did damage, and complain about people rolling their dice), I like to do that.

Rolling is fun, so I like to hand as much fun to the players as I can.

If you want to keep a steady probability that'll add up to 1 over 12 seconds, then you want 1/12. That's a roll of 2 or 3 on 2d6.

ReplyDeleteThe advantage of the secret roll, it seems to me, is that the PC can keep looking but change his mind when the fight changes around him. The maximum search time declaration doesn't allow for that.

Nice, so 3 or less? I could live with that.

DeleteThe secret roll is just a pain for me and less fun for them, which is a whole lot of disadvantage to try and overcome. :)

Or roll 1d, which is 1/6, and if you get a 6, roll another. 4-6 finds it. That way it's still a simple roll with a flat probability, but the time is right.

ReplyDeleteI'd rather not make it two rolls (or even one roll of two specifically separate dice) just avoid the let-down feeling, like when you roll a critical and get "normal damage only." Oh, why did I roll, it was bypass defenses and normal damage anyway. That sort of thing. I'll try 2 or 3 on 2d, that seems like it could be fun.

DeleteThe cumulative probability doesn't add up to 1 over 12 seconds because of joint probability. Look at http://www.wolframalpha.com/input/?i=binomial+distribution%2C+probability+of+success+1%2F12%2C+8+trials at the bottom you can see the chance of at least one success in 8 trials is where the probability goes over 50% and you should expect a success to have occurred.

ReplyDeleteWhat's a better approximation of it, then? 2 on 2d seems a bit too harsh, at least at a glance at the odds (2.77% each time - that's less often than my PCs critically hit)

DeleteDepends on what you aim to achieve, really. If you want the cumulative probability to go up in a way consistent with the original roll mechanic, you have to change the required roll every time, and I was assuming that was too much mechanical faff for Peter's play style. (This applies to both the 1d and the 2d versions.)

Delete"Simple and easy to remember" + "roughly accurate" is better than "very accurate but harder." Roger knows me by now, I think.

DeleteI am not really advocating for a more complicated play style just against ignorance of what actual probabilities you are really having the players roll. I also just wanted all those statistics classes to have been for something :)

DeleteThis also opens up some ideas:

ReplyDeletePerk (well organized) your search rolls are at +1

Quirk (messy pockets) ... somehow add another die or penalty...

That's what I am thinking.

DeleteTrying to roll 1 on 1d gives a 52% chance at four times, but only goes to 67% at six. Trying to roll a 2-3 on 2d gives a 50% chance at eight tries and 65% after twelve.

ReplyDeleteThat means that the median numbers are close enough but there is a tail that wasn't in the original method so the mean is higher. A cap helps solve that but still leaves a higher chance of maxing the time: 33% for 6 seconds with 1d, 35% for 12s with 2d.

Using d4 or d5 would pull the mean in but also reduce the median, which may be a worse poison, and, besides, you're playing GURPS.

So your system hurts the PCs a little bit. I'm pretty sure it's the best reasonable way. It might even be more realistic (especially without a cap!).

How about:

ReplyDeleteRoll two on 2d6, *but keep ones*? So if you roll snake eyes you get it, if you roll 1,(something else) you only need to roll 1 on 1d6 on subsequent turns to get it. "I remember where it is now!"

Relatively easy to remember, gives the illusion of progress, and the median time is 7 turns. (mean time is 8.7, or ~7 if you max the time to 12 turns.)