Search results

That’s the part I live for!!!

I have a sneaking suspicion that since he’s not simulating (his words) that he’s doing something closer to allocating average DPR per turn based on initiative. If so that methodology will tend to produce very skewed results, often amplifying small differences based on discrete breakpoints.

Got it. The correct test shows ‘pretty close’ so we must use an incorrect test to get the results you want. I do agree that looking at the fully paralyzed case can give some context as an upper bound. If paralyzed is 25% reduction in party effectiveness then showing a 20% reduction in party...

The effectiveness of the player is how much he decreases the party effectiveness with only him having 1 less mod…. Changing the rest of the party nullifies that test. You are no longer testing player effectiveness but party effectiveness. That’s what you are doing with your tests.

Missing 1 mod is now equivalent to paralyzed? I still don’t understand why you don’t just run the proper test through your simulator.

That’s everything. That’s what your entire claim was and what the whole disagreement is all about. In actual play having your PC not be fully optimized (by taking 1 lower mod) will barely register anywhere. You don’t get to control the other PCs. So they need to be treated as constants...

Strictly speaking my claim wasn’t that you’ll win 100% of the time either way so it doesn’t matter. There’s obviously some effect. The question is how much. In a typical encounter where you already have a very high win rate (whatever the number) it matters very little to your win rate. In the...

IMO totally unrealistic examples don’t show anything except that in totally non-realistic circumstances it could do something.

Why you wasting time on an example that cant actually show what you want to show?

Explain this? Why is pointing out that players are overwhelmingly favored to win 5e combat mean that rules and math don’t matter?

Single player isn’t how d&d is played either. It must be 1 PC in the group context to match with your point. If it’s really that impactful, why the hesitation to show it in that context?

Not when most combats have you winning 99% of the time. Having a better stat mod might up that to 99.5%.

Which won’t prove your claim that 1 mod is that important… It’s more realistic perhaps, but doesn’t align with your claim.

The test needs to be set up fairly and properly to demonstrate your point. I think that kind of proves my point…

I think you are using two different measures of effectiveness.

Oh god. That’s not how that works at all. Statistical significance is about testing a hypothesis and finding that there’s only a 5% chance or 1% chance for your actual data results to have happened by random chance (or whatever you’ve chosen as your critical value).

I think you only should be replacing 1 of the original 4 fighters to see the true effect of a single pc taking 1 lower str and con mod.

@FitzTheRuke How do you forecast how many books you initially try to obtain?

+1 mod is usually about a +20% increase after damage and accuracy are factored in. But I still agree that it’s mostly inconsequential. It’s only maybe 5% of party output. And combats aren’t calibrated that closely and so 5% more output is basically never going to be the difference in winning...

I asked Chat GPT to categorize the books. I did a spot check and it appeared to be correct. Starter sets were listed as other (not shown). Age is in Months. A few interesting points: He's sold more non-core books than core books. That's surprising. I wasn't expecting that result. Non-core...