Pages

Wednesday, March 27, 2013

GURPS Combat Choices: Deceptive Attack and Telegraphic Attack

Update: Martin Leuschen pointed out an error in the table, since fixed. My thanks to Martin for keeping my honest.

Patric Halter over at Renovating the Temple beat me to it with this post, but here is my take on Deceptive Attack and Telegraphic Attack.

In GURPS combat, Deceptive Attack is a combat option that lets you penalize your opponent's defense by -1 for every -2 you take to your attack roll. The only limits are your GM's indulgence, and a minimum final skill level of 10.

Telegraphic Attack, on the other hand, is a combat option that goes in the other direction: you can add +4 to your attack skill, but your opponent's defense increases by 2. Unlike DA, you don't get to pick how much Telegraphic Attack you want to use: it's +4/+2 or nothing. Also, TA doesn't affect your chances of scoring a critical hit. To determine if you got a critical, you use the skill you had before adding the TA bonus.

So, when do you want to use Telegraphic Attack? When do you want to use Deceptive Attack, and how much? Here's a handy table I made.

Defense
Attack 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
3 TA TA TA TA TA TA TA TA TA TA Flee Flee Flee Flee Flee Flee Flee Flee
4 TA TA TA TA TA TA TA TA TA TA TA Flee Flee Flee Flee Flee Flee Flee
5 TA TA TA TA TA TA TA TA TA TA TA Flee Flee Flee Flee Flee Flee Flee
6 TA TA TA TA TA TA TA TA TA TA TA Flee Flee Flee Flee Flee Flee Flee
7 TA TA TA TA TA TA TA TA TA TA Hold Hold Flee Flee Flee Flee Flee Flee
8 TA TA TA TA TA TA TA TA TA TA Hold Hold Hold Flee Flee Flee Flee Flee
9 TA TA TA TA TA TA TA TA Hold Hold Hold Hold Hold Flee Flee Flee Flee Flee
10 TA TA TA TA TA TA TA Hold Hold Hold Hold Hold Hold Flee Flee Flee Flee Flee
11 TA TA TA TA TA TA Hold Hold Hold Hold Hold Hold Hold Flee Flee Flee Flee Flee
12 TA TA TA TA Hold Hold Hold Hold Hold Hold DA 1 DA 1 DA 1 DA 1 Flee Flee Flee Flee
13 TA TA Hold Hold Hold Hold Hold Hold Hold DA 1 DA 1 DA 1 DA 1 DA 1 Flee Flee Flee Flee
14 Hold Hold Hold Hold Hold Hold Hold DA 1 DA 1 DA 1 DA 1 DA 2 DA 2 DA 2 DA 2 Flee Flee Flee
15 Hold Hold Hold Hold Hold Hold Hold DA 1 DA 1 DA 1 DA 2 DA 2 DA 2 TA TA TA Hold Hold
16 Hold Hold Hold Hold Hold Hold Hold Hold DA 2 DA 2 DA 2 DA 2 Hold Hold Hold Hold Hold Hold
17 Hold Hold Hold Hold Hold DA 1 DA 1 DA 1 DA 2 DA 2 DA 3 DA 3 Hold Hold Hold Hold Hold Hold
18 Hold Hold DA 1 DA 1 DA 1 DA 1 DA 1 DA 1 DA 1 DA 3 DA 3 DA 3 DA 3 DA 1 DA 1 Hold Hold Hold
19 Hold Hold DA 1 DA 1 DA 1 DA 1 DA 2 DA 2 DA 2 DA 3 DA 3 DA 3 DA 4 DA 1 DA 1 Hold Hold Hold
20 Hold Hold DA 1 DA 2 DA 2 DA 2 DA 2 DA 2 DA 2 DA 2 DA 4 DA 4 DA 4 DA 4 DA 2 DA 2 Hold Hold
21 Hold Hold DA 1 DA 2 DA 2 DA 2 DA 2 DA 3 DA 3 DA 3 DA 4 DA 4 DA 4 DA 5 DA 2 DA 2 Hold Hold
22 Hold Hold DA 1 DA 2 DA 3 DA 3 DA 3 DA 3 DA 3 DA 3 DA 3 DA 5 DA 5 DA 5 DA 5 DA 3 DA 3 Hold
23 Hold Hold DA 1 DA 2 DA 3 DA 3 DA 3 DA 3 DA 4 DA 4 DA 4 DA 5 DA 5 DA 5 DA 6 DA 3 DA 3 Hold
24 Hold Hold DA 1 DA 2 DA 3 DA 4 DA 4 DA 4 DA 4 DA 4 DA 4 DA 4 DA 6 DA 6 DA 6 DA 6 DA 4 DA 4
25 Hold Hold DA 1 DA 2 DA 3 DA 4 DA 4 DA 4 DA 4 DA 5 DA 5 DA 5 DA 6 DA 6 DA 6 DA 7 DA 4 DA 4
26 Hold Hold DA 1 DA 2 DA 3 DA 4 DA 5 DA 5 DA 5 DA 5 DA 5 DA 5 DA 5 DA 7 DA 7 DA 7 DA 7 DA 5
27 Hold Hold DA 1 DA 2 DA 3 DA 4 DA 5 DA 5 DA 5 DA 5 DA 6 DA 6 DA 6 DA 7 DA 7 DA 7 DA 8 DA 5
28 Hold Hold DA 1 DA 2 DA 3 DA 4 DA 5 DA 6 DA 6 DA 6 DA 6 DA 6 DA 6 DA 6 DA 8 DA 8 DA 8 DA 8
29 Hold Hold DA 1 DA 2 DA 3 DA 4 DA 5 DA 6 DA 6 DA 6 DA 6 DA 7 DA 7 DA 7 DA 8 DA 8 DA 8 DA 9
30 Hold Hold DA 1 DA 2 DA 3 DA 4 DA 5 DA 6 DA 7 DA 7 DA 7 DA 7 DA 7 DA 7 DA 7 DA 9 DA 9 DA 9

Caveat: this table only considers Deceptive Attack and Telegraphic Attack. If you also considered Rapid Strike, Dual Weapon Attack, hit locations, or other combat options, things would change a lot. But that will have to wait for later.
The attacker's skill runs along the side, and the defense runs across the top. The entry for a given skill and defense shows your best choice. "Hold" means that you can't improve your chances with either a Deceptive Attack or a Telegraphic Attack; you're already at the best level. "Flee" means that no matter what you do, your attack is going to be worth less than zero. Yes, I mean that on your own turn, you are more likely to get hurt than to hurt your opponent. Might as well run away.

You probably don't want to waste time looking things up on this table all the time, so let me try to reduce this information to a few rules of thumb.

  1. Don't use Deceptive Attack if your opponent's defense is already at 3 or 4. 
  2. Don't use Deceptive Attack unless you can get your opponent's defense below 13.
  3. If the sum of your skill and your opponent's defense is above 24, try to bring it down to the range of 22 to 24. Each level of DA will reduce the sum by 3.
  4. If the sum of your skill and your opponent's defense is 19 or below, use Telegraphic Attack.
  5. But don't use Telegraphic Attack if your skill is above 13.
  6. If your opponent's defense beats your skill by 8 or more, don't even try making an attack this turn. 

These rules don't get you optimal results in all cases, but they cover most of the chart pretty well.

For those of you interested in all the math behind the scenes, for now I'll just say that in this table, I considered the value of a normal hit to be 1, and discounted it by the probability that it gets past the defense. Then I valued a critical hit on the attack at 1.3, a critical failure on attack at -1.5, and a critical failure on defense as worth an extra 0.1. A critical success on defense imposes a critical failure on the attacker, and so is also worth -1.5.

These numbers are somewhat arbitrary, and anyone could reasonably question them, but after fiddling around with various values, I find that they don't alter the results very much.

A more detailed explanation of the number-crunching behind the chart will have to wait for another day.


15 comments:

  1. Very interesting in general, and I like it (I'm the guy who did the chart on Patrick's blog). Here's some comments:

    1. You have TA as the best option for 16 v 16. Since this turns the attack into 20 v 18, the odds shouldn't change at all. I suspect one of you calculations in a row off, since TA *does* make sense for 14 and 15 v 16 and you have it for 15 and 16 v 16.

    2. I know you mentioned this in passing, but my own further efforts really really show that rapid strike is the got to vs lower defenses, completely swamping DA most of the time. DA adds a fraction of an effective hit, RA can add multiple effective hits. Consider your 30 v 8 row - DA 4 is nice, you go from around 0.8 effective hits to around 1.0, or a 25% increase. If you RS2 with DA1 (still skill 16) you are looking at around 3 effective hits.

    With WM or EE it gets obscenely good, and if you have 30 attack skill, you should have an advantage to go with it that halves RS to -3. RS5DA1 should probably be your "go to" in many situations. (Although RS5 to the legs is a good trademark move - "Let's just call you Stumpy."

    This isn't even considering the side issue of someone with low defenses being hit multiple times suffering iterative defense penalties...

    3. It's probably worth throwing in an eye shot, a leg shot, and a vitals shot as well, although there's severe abstraction issues. Still working on that myself.

    4. I'm still mulling how to show it, but I think it is important to determine how much better each approach is than "Hold." Some of the recommendations on your chart increase your odds by less than 10%, others by more than 50%.

    5. As Peter points out, you rarely know your opponents defenses exactly. Rephrasing everything in terms of "none/poor/med/good/impenetrable" for the rules of thumb probably makes more sense.

    6. Sigh. This whole thing is too multidimensional to fit nicely on one spreadsheet or graph...

    ReplyDelete
    Replies
    1. One quick note: "hit multiple times suffering multiple defense penalties."

      That's a common and sensible house rule, but by RAW, defenses are not lowered by shock.

      Delete
    2. I think that what targeted attacks do for you is sink your skill to lower effective levels before you look on the chart. Sure, it makes the results of the hit more badass, but I think the way to interpret these charts is in terms of how you spend "leftover skill" after you account for hit location, lighting/terrain penalties, and other things.

      Delete
    3. Wow, thanks for the detailed reply.

      1. You're right. I had the critical fail chance falling to 1/216 at 17, rather than at 16. I'm going to have to update the table.

      2. Yes, RS is huge, and it's coming. I expect a combination of DA and RS will usually be optimal. How to handle defense degradation with multiple hits? I don't know yet.

      3. I'm waiting on hit locations for now, because that option is heavily influenced by DR, and I haven't even started on looking at the effects of DR on a plain old torso hit.

      4. I did consider posting a table that explicitly included all the probabilities, but I'm going to hold that for later when I can go into it in more detail. If you've been looking at the same calculations, then what an overwhelming mass of numbers it can be. I'm trying to find simpler ways of displaying it.

      5. True. This might be of more benefit to GMs than to players. Still, in many cases, if you can say your opponent's defense is somewhere in a range of 4 numbers, you get the same answer.

      6. What do you mean? You only have to consider every weapon and all of its possible wounding types, the attacker's skill, the defense level, Deceptive Attack at all its various levels, Telegraphic Attack, Rapid Strike at -6, -3, or -1, Dual Weapon Attack with all the possible combinations of weapons, hit locations (including different DR on different locations), Extra Effort, the varying value of critical success and failure in different situations, and everything else I haven't thought of. See, it's not so bad....

      Delete
    4. When I challenged/asked Bruno to look at my Setup Attack article and the math that comes from it, she tackled it, as I would have, with a Monte Carlo simulation with a half-million iterations each.

      This is likely how I'd approach the "what works" question, though it's quite brute-force. I'd randomly pile on yes/no options for certain skill level pairings - or even random ones - and look at trends generated by the data. Going through all the GURPS combat options is insane by design! :-)

      Delete
    5. Douglas, I think "hit multiple times suffering multiple defense penalties" refers to penalties to Parry and Block for multiple uses. That would have to be accounted for in an evaluation of RS.

      Delete
    6. And then of course you have to have separate simulations for Dodge versus Parry (Fencing and Otherwise) versus Block.

      It really is nearly impossible to "game the system" with GURPS - there are just too many variables to figure it all out!

      Delete
    7. Grouchy Chris is right about the iterative defense penalties. Currently, I'm quarter-assing it by just assigning a blanket -1 to effective defenses per extra attack, but that's a bigger SWAG than most of the approximations here.

      With respect to "Rapid Strike at -6, -3, or -1," isn't it actually "Rapid Strike at -6, -3, or ROUNDUP(-1.5*n)?" Or did I bork the OoO?

      With respect to locations, yes, there's massive SWAG there too, but neglecting "stab him in the eye" as an attack option because it is too confusing makes the table much less useful, IMHO.

      At some point, we also have to account for DetA, DefA, and AoA.

      Also, the date on the Babylonian calandar...

      Delete
    8. I can't remember which way the rounding goes. I have a hard time remembering rounding rules in general. Someone want to write a post on all the different things you need to round in GURPS, and which way they go?

      Delete
    9. Except for my own work, where I muck this up, I THINK I've been told you round everything down in GURPS except for one thing - maybe DR halving?

      Delete
  2. Another way to look at, say, the first few columns is to pick ever-escalating high-reward targets. So While 14-16 vs. Defense 3 or 4 is indeed "Hold," I'd say that either 17 or 18 should be "Arm/Leg." 19 might be "Vitals," 21 is "Head/Neck," 23 is "Brain," and 25 or 26 might be "Eyes."

    To my previous point, that might be if you choose "get a hit" over "hit something important." This would be inverted if even if you DO hit, it's useless. So if you can do an awesome 1d+3 attack vs. DR 10 to the torso, who cares what the Deceptive Attack penalty is? You need to first absorb the location penalty for something you CAN strike, and THEN think about Deceptive or Telegraphic attacks.

    ReplyDelete
    Replies
    1. Hm, maybe I should just change attack 14+ vs defense 3-4 to "demand a surrender."

      I hope to eventually get around to writing about hit locations and how much benefit they bring, but compared to the swing vs. thrust decision, or DA and TA, they are hard to quantify. I'm going after the low-hanging fruit first.

      Delete
    2. GC: Hm, maybe I should just change attack 14+ vs defense 3-4 to "demand a surrender."

      That is also classic "Telegraphic attack to the eye" territory. The so called "Berserker Off Switch."

      Delete
  3. Hey GChris, I just noticed another thing:

    Why does the horizontal line at skill 15 change from TA to Hold at defender skill 19? It seems to me that all the values for 15At vs. 16+Df should have the same effective hits for those two strategies.

    Also, what I was pondering when I saw that:
    If a crit hit is worth 1.3 effective hits (EHs)and a crit miss is worth -1.5 EHs, I'd think most of the 16+ def range would be DA or flee. Even with 16+ attack, a crit hit happens on 3-6 (~10%), so ~.13EHs, a normal hit on 7-16 and 18 Def, so ~.004EHs, and a crit miss on an 18 (~.005 EHs) or a 7-16 and a 3-6 Def (~9%, so ~.135 EHs), for a roughly dead wash (it's actually slightly in favor of the defense, if you give the WAGs too many sig figs).

    The 15 row, TA or no, has half the crit hits calculated above (TAs do not affect positive crit chances) so they should not calculate as positive...

    .
    .
    .
    This Yak may end up flensed.

    ReplyDelete