Jump to content
Fanfan

XwingV2 jousting values

Recommended Posts

On 7/29/2018 at 1:32 PM, Brunas said:

So I'm always a big fan of efforts like this, but I wanted to cover this specifically because it's false, and very easy to demonstrate that it's false.  It feels true, because adding results feels better than changing them, but I'll show my work.  I'm just going to assume you have a target lock, but if you have a focus as well ATC 2.0 jumps past 1.0 very quickly.

 

ATC 2.0, just target lock at range 2: http://xwing.gateofstorms.net/2/multi/?d=AAAAAAAAAAA&a1=MwAAAAAABAA (2.25 damage)

ATC 2.0, "i'm not willing to spend my target lock but didn't bring fire-control system because I'm bad" at range 2: http://xwing.gateofstorms.net/2/multi/?d=AAAAAAAAAAA&a1=MQAAAAAQAAA (1.5 damage)

ATC 2.0, FCS, unwilling to spend target lock: http://xwing.gateofstorms.net/2/multi/?d=AAAAAAAAAAA&a1=MQAAAAAUAAA (1.98 damage)

ATC 1.0, just target lock range 2: http://xwing.gateofstorms.net/1/basic/?q=IAAAAAAAAAAAAQAAAAAAAA (2 damage).

Obviously, when you add force, focus, or range 1 2.0 ATC outshines 1.0 quickly.  These are the edge cases where it's even close.

 

EDIT: I guess I should add - the tie advanced is the same price as an xwing, and strictly better offensively.

 

Second, Fang Fighters having a low "defensive efficiency" leads me to believe that you've ignored concordia faceoff.  While I get this is an early revision, this is incredibly misleading.  They have the best defensive ability it the game.

 

I have been debating how exactly to model generic TIE Advanced in 2.0. I'm modelling a full action economy, so I have to make some assumptions about how often the TIE Advanced will be taking a TL action vs a focus action during a round. For reference a vanilla ship's action economy going into the round is pretty basic... either take a focus action, or not. I haven't done it yet so this is a good excuse to put some thoughts onto paper. It seems like there are 5 different scenarios:

#1-2: already has a target lock on what will be the eventual defender later this round.

  • No new token action this round. (bump or barrel roll)
  • Takes the focus action

#3-5: does not have a target lock on what will be the eventual defender later this round.

  • No new token action this round (bump or barrel roll, or take a target lock on someone that isn't the eventual target)
  • take the focus action
  • take the target lock action on what will be the eventual target

 

So in net the possibilities are:

  • ATT3 + TL, no focus
  • ATT3 + TL + focus
  • ATT2, no tokens
  • ATT2 + focus
  • ATT3 + TL, no focus

 

The 1st and 5th scenario lump together into the same result. If I were in charge of playtesting at FFG, I would instruct the playtesters to go collect analytical data about the % occurrence rate of each of these scenarios, and then enter those values into my scripts to calculate the resulting jousting value. Or, more specifically, I would calculate across a range of values to show the relative ship performance as a function of player decisions impacting the action economy.

 

OK, now that it's done, guess I'll go code this up tonight. Vader is already done. :-)

 

 I would love to pick your brain about what you think the % occurrence rates would be for each of these scenarios.

 

P.S. this assumes the scenario probabilities are range-independent, which is not entirely true, but it's a solid starting point.

Edited by MajorJuggler

Share this post


Link to post
Share on other sites
3 hours ago, FTS Gecko said:

@Commander Kaine, the quote button is a privilege, not a right.

For the love of Dog, if you're just going to write one line, don't quote fifty. 

 

3 hours ago, FTS Gecko said:

@Commander Kaine, the quote button is a privilege, not a right.

For the love of Dog, if you're just going to write one line, don't quote fifty. 

 

3 hours ago, FTS Gecko said:

@Commander Kaine, the quote button is a privilege, not a right.

For the love of Dog, if you're just going to write one line, don't quote fifty. 

 

3 hours ago, FTS Gecko said:

@Commander Kaine, the quote button is a privilege, not a right.

For the love of Dog, if you're just going to write one line, don't quote fifty. 

 

3 hours ago, FTS Gecko said:

@Commander Kaine, the quote button is a privilege, not a right.

For the love of Dog, if you're just going to write one line, don't quote fifty. 

39200165_210689496466701_845252232315981

Share this post


Link to post
Share on other sites
4 hours ago, MajorJuggler said:

 

I have been debating how exactly to model generic TIE Advanced in 2.0. I'm modelling a full action economy, so I have to make some assumptions about how often the TIE Advanced will be taking a TL action vs a focus action during a round. For reference a vanilla ship's action economy going into the round is pretty basic... either take a focus action, or not. I haven't done it yet so this is a good excuse to put some thoughts onto paper. It seems like there are 5 different scenarios:

#1-2: already has a target lock on what will be the eventual defender later this round.

  • No new token action this round. (bump or barrel roll)
  • Takes the focus action

#3-5: does not have a target lock on what will be the eventual defender later this round.

  • No new token action this round (bump or barrel roll, or take a target lock on someone that isn't the eventual target)
  • take the focus action
  • take the target lock action on what will be the eventual target

 

So in net the possibilities are:

  • ATT3 + TL, no focus
  • ATT3 + TL + focus
  • ATT2, no tokens
  • ATT2 + focus
  • ATT3 + TL, no focus

 

The 1st and 5th scenario lump together into the same result. If I were in charge of playtesting at FFG, I would instruct the playtesters to go collect analytical data about the % occurrence rate of each of these scenarios, and then enter those values into my scripts to calculate the resulting jousting value. Or, more specifically, I would calculate across a range of values to show the relative ship performance as a function of player decisions impacting the action economy.

 

OK, now that it's done, guess I'll go code this up tonight. Vader is already done. ?

 

 I would love to pick your brain about what you think the % occurrence rates would be for each of these scenarios.

 

P.S. this assumes the scenario probabilities are range-independent, which is not entirely true, but it's a solid starting point.

Frankly, I don't think any estimate I'd have for what you're looking for that's any better than a blind guess.

 

It's all so dependent on the pilot skill of every ship involved, which is a question of the meta which doesn't exist yet.  There's even more mess to the confusion.  Even if I line up the target lock + focus, how often does that advanced still have a focus for offense by the time it attacks?  How willing am I to spend that focus on defense?  How often does the locked target die before the Advanced ever shoots?

 

I'll note from the (very) few games I've played with a tempest, they had a lock for every shot but one (locked target died before tempest attacked).  Mostly because the re-positioning has been toned down so much, but we're all bad just throwing random stuff on the table now regardless :)


Vader is of course really easy - if he's shooting, he has a target lock.  Not like anyone can block him!

Edited by Brunas

Share this post


Link to post
Share on other sites
15 minutes ago, Brunas said:

Frankly, I don't think any estimate I'd have for what you're looking for that's any better than a blind guess.

 

It's all so dependent on the pilot skill of every ship involved, which is a question of the meta which doesn't exist yet.  There's even more mess to the confusion.  Even if I line up the target lock + focus, how often does that advanced still have a focus for offense by the time it attacks?  How willing am I to spend that focus on defense?  How often does the locked target die before the Advanced ever shoots?

 

I'll note from the (very) few games I've played with a tempest, they had a lock for every shot but one (locked target died before tempest attacked).  Mostly because the re-positioning has been toned down so much, but we're all bad just throwing random stuff on the table now regardless :)

 

I can approximate pretty well the likelihood of a ship having focus for either attack or defense with my action economy modelling, given that you know what tokens it starts with in the beginning of the combat phase. I have had that functionality for a couple years now, and is the missing link that makes the jousting numbers actually useful for a wide variety of pilots, which have a wide variety of action economies. In this case, the main problem is figuring out what useful tokens the ship has going into the combat phase.

 

If you can distill it down to the following, then you can get the net action economy going into a combat round:

  • Chance of having a TL at the start of the round leftover from the previous round, AND you'll be able to use this TL on your target
  • equivalently, the chance of not having a useful TL going into the start of the round. (1-the above)

Then:

  • The chance that the ship can take a token action, i.e. no bump, stressed, or single action reposition (I'm using the same default global % chance that I'm using as a default for all ships)

Then: 

  • The chance that the ship takes a target lock but then can't spend it this round to attack that target (same net effect for this round as no action this round + no useful TL entering the round)
  • The chance that the ship takes a target lock and can use it this round
  • The chance the ship intentionally takes a focus this round, without having a target lock on anyone.

That yields the probabilities for a 4-state outcome. (technically 5, which consolidate to 4). Then just use that in the jousting calculator blender, and you get... an answer. :-)

Edited by MajorJuggler

Share this post


Link to post
Share on other sites
8 hours ago, MajorJuggler said:

 

The 1st and 5th scenario lump together into the same result. If I were in charge of playtesting at FFG, I would instruct the playtesters to go collect analytical data about the % occurrence rate of each of these scenarios, and then enter those values into my scripts to calculate the resulting jousting value. Or, more specifically, I would calculate across a range of values to show the relative ship performance as a function of player decisions impacting the action economy.

 

I don't think there's any evidence FFG care about this at all.  They use the Force.

Share this post


Link to post
Share on other sites
9 hours ago, MajorJuggler said:

 

I can approximate pretty well the likelihood of a ship having focus for either attack or defense with my action economy modelling, given that you know what tokens it starts with in the beginning of the combat phase. I have had that functionality for a couple years now, and is the missing link that makes the jousting numbers actually useful for a wide variety of pilots, which have a wide variety of action economies. In this case, the main problem is figuring out what useful tokens the ship has going into the combat phase.

 

If you can distill it down to the following, then you can get the net action economy going into a combat round:

  • Chance of having a TL at the start of the round leftover from the previous round, AND you'll be able to use this TL on your target
  • equivalently, the chance of not having a useful TL going into the start of the round. (1-the above)

Then:

  • The chance that the ship can take a token action, i.e. no bump, stressed, or single action reposition (I'm using the same default global % chance that I'm using as a default for all ships)

Then: 

  • The chance that the ship takes a target lock but then can't spend it this round to attack that target (same net effect for this round as no action this round + no useful TL entering the round)
  • The chance that the ship takes a target lock and can use it this round
  • The chance the ship intentionally takes a focus this round, without having a target lock on anyone.

That yields the probabilities for a 4-state outcome. (technically 5, which consolidate to 4). Then just use that in the jousting calculator blender, and you get... an answer. :-)

Right, I'm still not sure I (or anyone?) can accurately predict the values you're looking for, at the moment at least.

Maybe in this case it would be more useful to work backwards? e.g., according to the model, the tie advanced needs to have a lock X%, and a focus Y% offensively to be worth Z points, then measure individual games against those metrics.

Share this post


Link to post
Share on other sites
49 minutes ago, Brunas said:

Right, I'm still not sure I (or anyone?) can accurately predict the values you're looking for, at the moment at least.

Maybe in this case it would be more useful to work backwards? e.g., according to the model, the tie advanced needs to have a lock X%, and a focus Y% offensively to be worth Z points, then measure individual games against those metrics.

 

For ships that have uncertain real-world triggering rates that's exactly what I do: sweep the trigger rates over a variety of ranges (usually just 0% to 100% in 10% increments) and calculate the resulting efficiency at each point. If there's just one input variable (like x7 triggering) then it's a pretty simple 1D plot. If there are 2 input variables then it's a more complicated 2D plot, or equivalently a larger 2D table of results. If there are more degrees of freedom then it gets annoying to look at (the technical term for higher dimensional plots is actually a hyperspace). In any event, based on the resulting efficiencies, if you have 1-2 input variables then you can trivially look at the table to determine what trigger rates result in the ship being underpowered / overpowered / in the goldilocks zone.

 

In the case of the TIE Advanced I can distill it down to 3 representative variables:

  1. how often it has a useful target lock going into the start of the round
  2. how often it takes a new target lock action that becomes useless
  3. how often it takes a focus action without having a target lock

 

#1 is a multiplication of 2 factors:

[probability of not spending the target lock on the previous round]

x

[probability of target still being around to shoot at next round]

 

1a can be hard-coded to something like (1/2)^3: i.e. always spend the lock to modify dice. It's a simplistic approach but not totally unrealistic. A tactical player might also save the locks from range 3, hoping to use them next round at range 1-2, so you can nudge this up a little.

1b can be varied across a range, like 20% to 80%.

 

#2 can also be varied across a range: say 0% to 50%

#3 I can slave to be relative to #2, lets call it half the rate of #2, so that the probability of taking a useful target lock action is 1-(1.5x #2). Again, an approximation, but a good start without having any analytical data yet.

 

Using those ranges gets you in the neighborhood, which says that the IN2 TIE Advanced is probably slightly less terrible than most people would assume. But it's almost certainly not as good an an X-wing. The crit effect will help it (I haven't modeled this for the v2 TIE Advanced yet), but it still won't be anywhere close to the cost efficiency of an ATT2 TIE Fighter buffed by Howlrunner. I'm not publishing results for 2nd edition so don't want to put specific numbers out there, but it's also a fairly obvious conclusion, that people will naturally figure out about half a game into playing with 2E.

 

 

The validation step is getting analytical playtesting data to see what real-world trigger rates can be, but as @SOTL has pointed out this is not something that FFG does, in part because even if they had that data they wouldn't know what to do with it.

Share this post


Link to post
Share on other sites
1 minute ago, MajorJuggler said:

The validation step is getting analytical playtesting data to see what real-world trigger rates can be, but as @SOTL has pointed out this is not something that FFG does, in part because even if they had that data they wouldn't know what to do with it.*

That might be true, but I think more importantly they just don't want it even if they did know what to do with it.  I wasn't being flippant in saying that they use the Force, and I've seen the same approach across many of their games.  They simply aren't designing with mathematical efficiency in mind and my read of it, across all their games I've been involved in, is that they see such things as a dark art rather than a helpful tool

Share this post


Link to post
Share on other sites
47 minutes ago, SOTL said:

That might be true, but I think more importantly they just don't want it even if they did know what to do with it.  I wasn't being flippant in saying that they use the Force, and I've seen the same approach across many of their games.  They simply aren't designing with mathematical efficiency in mind and my read of it, across all their games I've been involved in, is that they see such things as a dark art rather than a helpful tool

 

Yeah, it depends somewhat on the designer, but that certainly looks like the [edit: strong] trend. Philosophy of design is also different than understanding of design. Alex seems to appreciate that math is a useful predictive tool, vs Frank who has said in one of our interviews that XYZ specifically can't be done with math (ironically as I was implementing it during the interview). Max I don't know at all where he stands -- I have only heard him say that his primary concern is to make cards that is "fun".

Edited by MajorJuggler

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...