Data science investigations into the mechanics of the world's greatest role playing game

[5ekyu]

Strongly agree!

Though I don't doubt that a machine learning algorithm could identify the optimal moves of any permutation of creatures and PCs if it included complete information. Things like AlphaGo, that do this sort of thing, already exist.

The d20datascience article clearly lacks a lot of relevant information.

Actually, based on various advances over the years with other games use of elements intelligently can be "learned" by machine operations and sophisticated processes... in the hands of experts from both the subject matter *and* the programming side and with a lot of computer power.

however, for any such model to be accurate and reflective of anything relevant to the subject matter, it must include *all* the relevant subject matters that relates to performance.

Cutting out entire swaths of the gameplay and tactical puzzles and then arriving at a conclusion...

Well, its the equivalent of running sophisticated analysis of chess and determining that the knight is the most powerful piece - after you cut out queens, rooks and bishops from your "chess board analysis" because they were more difficult to model than you could handle.

Its humorous because of how many rush to it because it reinforces their biases.

 

[GlassJaw]

For those familiar with the mathematician Lanchester (and Lanchester's Square Law), this makes a lot of sense.

https://en.wikipedia.org/wiki/Lanchester's_laws

 

[smbakeresq]

I think they are more useful for what they discredit.

For example, some say that initiative bonuses are very powerful. I think that's silly.

It is and it isn’t. Initiative when you are just going to trade attacks and HP isn’t that important. Initiative to cast a Slow spell before the enemies go and thus increase the damage they take and reduce their damage output is very important.

As far as -5/+10 feats they are powerful, however part of their power is they are one-sided, in most cases only the PC get to use them. I much prefer the old power attack, which I use in my games based of your proficiency bonus. This way they scale with the PC.

 

[Oofta]

It is and it isn’t. Initiative when you are just going to trade attacks and HP isn’t that important. Initiative to cast a Slow spell before the enemies go and thus increase the damage they take and reduce their damage output is very important.

As far as -5/+10 feats they are powerful, however part of their power is they are one-sided, in most cases only the PC get to use them. I much prefer the old power attack, which I use in my games based of your proficiency bonus. This way they scale with the PC.

How do you implement that? If I have a +3 proficiency can I take up to a -3 to get +3 damage?

I don't particularly care for the -5/+10 myself, it seems overpowered at high levels depending on the campaign and group but I'm always interested in alternatives.

 

[smbakeresq]

How do you implement that? If I have a +3 proficiency can I take up to a -3 to get +3 damage?

I don't particularly care for the -5/+10 myself, it seems overpowered at high levels depending on the campaign and group but I'm always interested in alternatives.

One handed weapon its up to proficiency and get same as bonus damage.

one handed in two hands its up to proficiency and 1.5x as bonus damage

two handed its its up to proficiency and 2x as bonus damage.


I also allow one handed weapon users to take it. The second part of the feat doesn't require a two handed weapon.



An interesting variant that I have not played yet but heard about was the -5 was to AC and not to hit, your attempt at a big swing threw you off balance. I think this is from someone who played Dragon Age, where powerful swings did just that.

 

[Satyrn]

The more interesting conclusion has that high level fighters can take on crs 150% above their level. That follows my intuition based on experience that the Cr system starts to break down at higher levels

Of course high level fighters can take on CRs above their level. A monster's CR primarily tells you what level the "typical PC" has to be before it won't drop to 0 hit points if the monster deals its "average damage" against the PC in a single round.

It's very important to undrstand what those phrases I put in scare quotes really mean, too:

1) The "typical PC" seems to have a d8 hit die and +2 Con mod. Fighters are likely touger than that.
2) A monster's "average damage" is calculated by assuming all its attacks will hit, and the PC fails all its saves. Since fighters tend to have among the highest ACs in the party, this assumption won't often hold true against a fighter.

The monster manual needs to spell out what CR means a lot better than it actually does, because the DMG only hints at all this.

 

[Gadget]

Not having read and done a full analysis, but on first blush this seems to confirm the design pattern 'bounded accuracy' wanted to enforce: Hit points & damage are the two major factors that differentiate the high and low level characters, though of course other things like attack bonus and AC do tend to increase somewhat over the levels as well.

 

[smbakeresq]

Of course high level fighters can take on CRs above their level. A monster's CR primarily tells you what level the "typical PC" has to be before it won't drop to 0 hit points if the monster deals its "average damage" against the PC in a single round.

It's very important to undrstand what those phrases I put in scare quotes really mean, too:

1) The "typical PC" seems to have a d8 hit die and +2 Con mod. Fighters are likely touger than that.
2) A monster's "average damage" is calculated by assuming all its attacks will hit, and the PC fails all its saves. Since fighters tend to have among the highest ACs in the party, this assumption won't often hold true against a fighter.

The monster manual needs to spell out what CR means a lot better than it actually does, because the DMG only hints at all this.

The descriptions are poor and the DMG isn’t any better.

 

[Krachek]

By analyzing DnD close enough we will answer the universe mysteries.

 

[Satyrn]

The descriptions are poor and the DMG isn’t any better.

That's what I'm saying.

It took me a while to figure out what CR was actually trying to measure - and how well it does so.

It doesn't measure a monster's defensive staying power - its hit points - in any way that reflects the "growth spurt" that the PCs get at 5th level for example - where the fighters are essentially doubling their damage output and spellcasters are getting fireball and the like.

I come to see CR as nothing but the roughest measure of a monster's power, and not in monster vs PC way. Rather, it's really just telling you which monster is stronger than another, so that if a CR 4 creature didn't provide quite the oomph you expected it to, you should try a CR 5 next.