Strain: alternative effects

Some thoughts:

1. It largely depends on the GM's style. How often do you roll for tasks in your group? If rarely Strain is powerful, if very often irrelevant.

2. I do not think +1 is that small. 10% is quite a bit. And you can spend multiple strain points! And you can use multiple Strain Points to improve your chances AND reroll.

3. I am not a big friend of adding dice to rolls (did not like it in 2nd editione either), as it destroys the nice liner probability curve I prefer. But that is a personal thing of mine. At least you should look in the math, how much one dice changes the chances (2nd edition has a table in the back of the GM's book I think).

I was going to start a new thread but I found this old one and thought, hmmm, relevant....

I was playing with some game mechanic stuff the other evening and similar to this topic the economy of strain came to mind, and I was pondering it. Now rather than do anything logical like playtest, I decided to throw math at the problem.  

So here's a little probability chart, quick and dirty. Presuming my math is right....



So unsurprisingly you have a progressive 10% chance per TN increase when rolling 1d10, so if you spend a strain, 1 strain = 10%. Not bad. At a 2d10 level though, now we set a curve, where you get the most return at lower TN levels. At TN 3, a point of strain will give you a +13% increase whereas at TN 6 that same point of strain will only grant you 7%. The curve gets much more steep at 3d10 as you can see, where our same TN 3 example nets 12.7% and TN 6 gets only 3.7%.

What I find interesting about this, is that strain really helps you when you're less good at something, than when you're already good at something, like a sort of diminishing return.  

So naturally I look for an example, so I went for the Pregens in the back of the quickstart guide to play around with. Let's take a peek at Drew Latimer. Let's say we have Drew piloting a jumpcraft into a landing site, and because of conditions we called for a role - there's a heavy crosswind (-2) but Drew checked the weather satellites before departure and pre-plotted a landing vector (+2) so we're at his base TN of 6 (Skill 4 plus Coordination 2, saying his 'Dextrous' nature makes him better at the controls!). Rolling 3d10, he has a 93.6% chance of success. Solid, but were he to put a little stress into it, grit his teeth, he could get a 3.7% bump up to 97.3% success chance. While good, that hardly seems worth the investment.
But what if it's a smaller chance? Perhaps our friend Drew suffered a much harder landing than expected and now, waiting for rescue, he's decided to set up a makeshift filter for drinkable water. I'll say this falls under his Survival Training, which he has a TN of 4 (Skill 1 plus Cognition 3 since Resourceful would apply). With 1d10 to roll, that's a 40% chance. Spend 1 Strain and it goes to 50%. Well that's better.

But hold on. Let's throw a wrench into things!

Here's my thinking: 1 strain can also buy you a reroll. Okay. Lemme put on my statistics hat here. A re-roll is a complete reroll, ignores previous results and just goes for new ones. Since these are independent probabilities, because the first roll does not influence the second roll, and a re-roll would only occur upon failure of the first roll, we can reevaluate as follows. In Example 1, if he chooses to not spend the strain, and decides to save the strain to do a reroll only in the event he fails the roll, the recalculated probability of success is 99.6% [93.6% + (6.4% * 93.6%)], or a gain of 6% per 1 Strain. With his low skill example of TN 4, then the recalculated reroll-on-failure success percentage is 64%, or a gain of 24% per 1 Strain. 

So it seems unless my math is warped....that spending a Strain on a reroll is always the better deal. 

Oh you're right! I forgot about the table, geez, why did I go and redo the math again??? Oh well I guess it was a good exercise.

You encouraged me to do further testing, and you're correct to a point - it seems that for all values above TN 1, reroll is better, but for TN 1 values indeed, adding +1 TN is the better way to go. 

Yeah, I definitely understand your opinion, though I disagree with it personally. The whole point of multiple dice and rerolls is to turn a linear scale into a probability curve. The old game design thought always was that, when you are very good at something, a linear scale fails to represent this. For instance in my little example of landing a jumpcraft in sheer winds, let's look it over in a little detail.

So on a linear scale, the 1d10 scale, a -2 TN is a flat 20% penalty to success. Which means that no matter how amazing a pilot you are, sheer winds would -always- add a 20% chance of failure to land, even if your target number was TN 9. 
If instead we use the multiple dice, or 'dice pool' scaling, at 3d10 the -2 penalty on Drew's TN 6 would account for a 15.2% failure chance. If he only had 2d10, it would be back to 20%. This keeps decreasing as TN goes up, so at TN 7 on 3d10, a -2 is only 9.8%, TN 8 it's 1.9%, etc. As in, the better the characters' skill, the less minuses affect their success chance. Which makes sense! And it makes sense in a non linear way. At least I think so!