TOPIC: Last Second Changes Made - Sorry! FINAL Contact Sheets

Endgame wrote: Analysis should likely never average in crit damage based on crit percent , as the result is binary- a slide either crits or it doesn’t - it should be expressed as 2 values. First the more likely result (generally the non crit), then likely result.

That's not how statictics works

Arcanist Kolixela wrote:

Justice wrote:

You mean a fighter ability that could be restored by the Cleric with Restore Power?

What fighter power are you referring to? Weapon Specialization?

Just because you can use a Restore Power on an ability doesn't mean that you should LOL

Not arguing that you SHOULD just saying that it's not valid comparison because one uses a 1/game effect and the other (uses 2 one/game effects) does not is a bad thing to use as a claim

I hear ya.

Also weapon specialization is a 1/combat. Unless you were referring to a different mechanic that is 1/game.

Arcanist Kolixela wrote:

Endgame wrote: Analysis should likely never average in crit damage based on crit percent , as the result is binary- a slide either crits or it doesn’t - it should be expressed as 2 values. First the more likely result (generally the non crit), then likely result.

That's not how statictics works

There is no in game condition where a fighter slides for 48% crit damage, thus it is a meaningless calculation.

Endgame wrote:

Arcanist Kolixela wrote:

Endgame wrote: Analysis should likely never average in crit damage based on crit percent , as the result is binary- a slide either crits or it doesn’t - it should be expressed as 2 values. First the more likely result (generally the non crit), then likely result.

That's not how statictics works

There is no in game condition where a fighter slides for 48% crit damage, thus it is a meaningless calculation.

There’s no situation where a drunk driver kills 48% of a person in an accident. It is binary, either they kill someone driving drunk or they don’t, we shouldn’t create these statistics that assign any percentage other than 100% or 0% to things related to drunk driving. That is how statistics work.

Can we PLEASE stop the conversation? It has nothing to do with the tokens anymore and is not helping anything.

JEFF has to READ this whole chain for any USEFUL information. So let's stop the bickering?

Make a new thread if you want to bicker, I'm all for it

Endgame wrote: Analysis should likely never average in crit damage based on crit percent , as the result is binary- a slide either crits or it doesn’t - it should be expressed as 2 values. First the more likely result (generally the non crit), then likely result.

This has to be one of the stranger criticism I've seen.

I'll just say the calculation of average result of a random process as a weighted average of the likelihood of each possibility and the value each of those possibilities takes on is very common statistical procedure.

For example, the average result of a d6 is 3.5.

It would be strange to insist one should never report an average result of a d6, because either the result is always only one of: 1, 2, 3, ... 6 - and it should only be expressed as 6 values:

1 with probability 1/6
2 with probability 1/6
3 with probability 1/6
4 with probability 1/6
5 with probability 1/6
6 with probability 1/6

If you would prefer to see:

Crit but not hit chance: crit damage value
Hit chance: hit damage value

I welcome you to reverse engineer those numbers and post them yourself - they are easily derivable from the formulas I provide.

I'll continue to report single numbers that characterize the average result, and stating my assumptions as to the underlying probabilities.

Arcanist makes a good point. For a Wizard to do that much damage means a Healer is doing zero, so it’s fair to say the cleric and Wizard should each share the credit for the damage dealt.

Arcanist Kolixela wrote: Can we PLEASE stop the conversation? It has nothing to do with the tokens anymore and is not helping anything.

JEFF has to READ this whole chain for any USEFUL information. So let's stop the bickering?

Make a new thread if you want to bicker, I'm all for it

Yeah fair enough - I'm done posting here till we get new data from TPTB.

Endgame wrote: Analysis should likely never average in crit damage based on crit percent , as the result is binary- a slide either crits or it doesn’t - it should be expressed as 2 values. First the more likely result (generally the non crit), then likely result.

I alway show three values. Min, Max and Average.

Average can be lower than Min, because Min and Max are based on actually hitting. There always is a 10% Of missing in VTD. I haven’t factored the reslide into that. I’ll do that next.

Matthew Hayward wrote:

Endgame wrote: Analysis should likely never average in crit damage based on crit percent , as the result is binary- a slide either crits or it doesn’t - it should be expressed as 2 values. First the more likely result (generally the non crit), then likely result.

This has to be one of the stranger criticism I've seen.

I'll just say the calculation of average result of a random process as a weighted average of the likelihood of each possibility and the value of each of those possibilities is very common statistical procedure.

For example, the average result of a d6 is 3.5.

It would be strange to insist one should never report an average result of a d6, because either the result is always only one of: 1, 2, 3, ... 6 - and it should only be expressed as 6 values:

1 with probability 1/6
2 with probability 1/6
3 with probability 1/6
4 with probability 1/6
5 with probability 1/6
6 with probability 1/6

If you would prefer to see:

Crit but not hit chance: crit damage value
Hit chance: hit damage value

I welcome you to reverse engineer those numbers and post them yourself - they are easily derivable from the formulas I provide.

I'll continue to report single numbers that characterize the average result, and stating my assumptions as to the underlying probabilities.

Let me introduce you to a great miniatures game called Warmachine. At a high level, it is a 2d6 resource management game where at the start of your turn, you allocate resources, and you generally can not change it mid turn.

There is a mechanic in the game called chain attack - in certain conditions, at roughly 25% to 33% of the time, chain attack will dramatically increasing damage output.

Many people modeled the average chance of the increase damage into how much resources you would need to do X damage (for this situation, we shall use 3 resources). Many would use this modeling in game to allocate 3 resources to destroy target Y (which could absorb X damage.)

In reality, what you get is 2 different game states. In state 1, which occurs 25-33% of the time, you will deal X damage using only 1 resource. In state 2, which occurs 67 to 75 % of the time, you will need 5 resources to deal X damage.

If a player follows the average, they will either over allocate by 2, or under allocate by 2, and both of those results can lose you the game.

While you can use your crit average to calculate the average damage done over all runs, maybe to predict that A combat will, on average, run 3 rounds, it may never be accurate. In some percentage of runs, the crits will occur and the combat will finish in 2 rounds. In the remainder, combat runs 4 rounds.

Arcanist Kolixela wrote: Can we PLEASE stop the conversation? It has nothing to do with the tokens anymore and is not helping anything.

JEFF has to READ this whole chain for any USEFUL information. So let's stop the bickering?

Make a new thread if you want to bicker, I'm all for it

I checked the rare tokens this morning, but there wasn't anything to talk about - the rare sheet appears to be unchanged from the previous "final" sheet that didn't really generate any comments either.

Iross wrote: Firstly, thank you Jeff for all your hard work.

I see that the Amulet of Flawsight has been downpowered again since the last version. I really do think it needs to be improved, I'm honestly not sure that Rogue builds would want to equip it when there are existing R necklaces which increase STR or damage.

At the moment, the Amulet offers, once per game, a possibility of adding the usual 15 Sneak Attack damage to an attack that you would normally be able to add anyway. That's only if you use the ability before seeing if the attack hits, and only if a monster is encountered which is immune to Sneak Attack. It does also affect Critical Hits, but given that you have to use the Amulet before sliding, and then slide a natural 20 on that exact turn, it's so unlikely that it's not really a factor in assessing the Amulet's usefulness (other posts have estimated that in general maybe 5% of attacks are a natural 20, but that's without having to do it on the same turn you use a 1/game ability).

So a conditional once per game 15 damage, which still uses up your class ability to activate. Compare that to Amulet of Vigor, granting a permanent +1 to hit and +1 to damage on all attacks, along with +3HP. If an adventure involves 12-15 rounds of combat, the Amulet seems far more useful and is a generic item equippable by all classes.

Consider also Amulet of Shock. A static +2 damage on all attacks, again equippable by all classes. So +24-30 damage during an adventure.

And then consider the other class-specific R necklaces in this set, which have abilities including a static +4 damage on all (2h) attacks, expanded crit range, a static +2/+3 damage on all attacks, and non damage-based class abilities.

It seems to me that the Rogue necklace is weaker than both the generic necklaces equippable by all classes and the other class-specific necklaces in this set. Thematically it meets the aim of having class-defining necklaces, but it just doesn't seem useful enough for a Rogue to want to use it.

Bumping this one more time for visablilty, as this is an significant issue for new Rogue players and would be nice to ensure proper balance. Adding +1 or 2 to hit or +2 damage would be completely reasonable for the Rogue necklace, more so now that this was limited to once per game.