I posted this over at the atari forums, and it was suggested there that here would be a better place to have it seen and discussed. I'm new to the game and the forum. I've been reading up on strategies and enjoyed the debate between two weapon use, and two handed weapons. I was bored at work, so i did an analysis of the two styles with excel. For reference i used a third level fighter, with 18 strength. On the two hander side, i used Greataxe and Greatsword. For two weapons, i used Long sword and short sword. (Incendentally i tried Rapier/Rapier and it was inferior to LS/SS. LS/Rapier was only marginally better than LS/SS.) I chose an AC 18 opponent because that is where the 2H wielder has a 50-50 chance to hit. A Greataxe wielder will average 6.875 damage/strike. Greatsword is slightly better at 7.15 LS/SS without two weapon feat is 3.41 damage per full round. With the feat it increases to 6.16 These numbers are averages, which DO include critical hits. It is probably better to think of it in terms of damage per 100 full rounds. Over 100 full rounds, a Greataxe = 687.5 Total damage, Greatsword = 715, LS/SS no feat=341, LS/SS with feat= 616. In percentage terms, the greatsword does 16% more damage than LS/SS with feat. As the enemy gets harder to hit, the 2H weapon gains an increasing advantage. Against a 23 AC opponent, Greatsword Averages 3.575 per round (357.5 /100 rounds) and the LS/SS averages 2.31 per round (231 /100 rounds). 55% more damage with the Greatsword. Margin slims as enemies get easier to hit. Against AC 13 opponent, Greatsword = 10.725 vs the LS/SS's 10.01 Only a 7% difference. It must be noted that without two weapon feat, AND a light weapon in the off hand, the numbers aren't even close. Thats the straight numbers. IMO the 2 hander has an even greater advantage, because a 2 weapon user loses the second attack when they have to move to attack, AND have to keep the primary penalty. This is significant in my experience with my dual weapon rogue. For extra kicks, and to be fair, i ran it considering that the 2 hander spent a Feat on weapon focus where the other fighter spent it on 2 weapon. Vs AC 18 his damage goes up to 7.865 for a 19% increase over LS/SS Vs AC 23 his average goes up to 4.29 for a 73% increase over LS/SS Vs AC 13 his average goes up to 11.44 for a 7% increase over LS/SS The difference is greatest against the toughest opponents, where it counts the most. The 2 weapon combatant fares the best against the weaker opponents, where the difference probably means less. If anyone wants me to run specific weapon/bonus combos, or wants to see my numbers, let me know.
uhm - If someone rolls a d20, theoretically, the average is 10.5. * Each number from 1 to 20 would be rolled one time in 20 consecutive rolls. Each number is equally weighted. Its set of numbers would consist of each number being rolled just one time, and, when averaged, would be 10.5. avg d20 roll = 10.5; avg 2d20 roll = 21 avg d12 roll = 6.5; avg 2d12 roll = 13 avg d10 roll = 5.5; avg 2d10 roll = 11 avg d8 roll = 4.5; avg 2d8 roll = 9 avg d6 roll = 3.5; avg 2d6 roll = 7 avg d4 roll = 2.5; avg 2d4 roll = 5 Having said that [if you are really in to this kind of information], what would be the final result of using each number from 1 to 20 for a d20? * Try using (1st) AC15 and (2nd) AC20 and (3rd) AC25 comparing a "to hit" number sequence from 1 to 20 for each of the three AC values. Do this for the same character [has the same stats] given a greatsword and, then, repeat the process for the long sword and short sword combination. For weapon damage, use the averages listed above for every hit, except for citicals which is 2x damage. Let us know what you determine.
I'm not sure if you're refuting my numbers or not. Let me explain in the terms that you've used. There's no need to do a study on each attack roll. It's success/fail, not incremental. My 18 str Level 3 fighter needs to roll an 11 to hit. If he rolls a 1 through ten, he misses. If he rolls 11 though 20, he hits. 50% of his rolls will be hits. Average damage for a greatword is 7, + 4 for strength for an average damage of 11. If he swung his sword 100 times, he would hit 50 times doing an average damage of 11 each hit for a total of 550 damage. Divide by his 100 swings is an average of 5.5 damage per swing. You can get the same result by multiplying his chance to hit (.5) by average damage (11) I added the critical hit average damage in the same manner, only it is a little more complicated and hard to explain. The only thing i'm not sure of is if it is just weapon damage that gets doubled (or tripled) on a crit, or if it is weapon+strength bonus x 2.
It's both weapon damage and strength that gets doubled/tripled. It seems strange to me that the numbers would be this much biased even at this low a level, but i haven't done the calculations so I don't know. If I am not too lazy i'll write a little program to run the numbers through. Even if your numbers are quite accurate though, there are a few other things to consider. First you get a greater range of options with two weapons. I.e. in PnP D&D where you can't have all weapons holy,flaming,shocking,frosty,axiomatic and whatnot at once with two weapons you might have one a holy cold iron LS and the other a Flaming mithril SS or something. This gets even more important once the weapons have some special bonuses i.e. weapons of luck or some such. There is also the higher levels to consider where the chances to hit become much better if your BAB is +10 you have +4 for strength +2 from enhancment and +1 from bless for a grand total of +17 to hit anything under an AC of 17 is an auto hit (barring a roll of 1) and even for AC 23 you still have only a 1 in 4 chance to miss so i think the numbers will come out a little differently there. Last bot not least even if TWF is not the most effective strategy for straight high Str fighter it might be quite viable for the weak rogue that has only a +1 Str bonus but a +4 dex one and if he's flaking and doing an additional 4d6 sneak attack damage per hit the str bonus is really not the important part. All that said I do think that TWF is in the numbers probbably in many cases suboptimal to a twohander but there is allways the roleplaying aspect of being the whirling dervish with two weapons flashing around you to consider. Just my 2 cp Rawel
I prefer my characters to dual wield Katana (bastard sword) and Wakizashi (shortsword) the bigger damage dice for the katana over the longsword example you used helps to close the damage gap you exposed, and like Rawel said, once both weapons are enchanted to do more damage dice per hit, the damage potential for 2 weapons is HUGE!
Also remember that when fighting multiple foes, TWFers shine because they have less "waste" damage. If the 2-hander can do 15 points of damage on average and the TWF averages 6 with one weapon and 4 with the other (note these numbers are completely arbitrary), it doesn't matter if the guy they're fighting only has 5 hit points left. They both drop the guy with one blow, but the TWFer can then turn around and hit the guy's friend with his off-hand.
It really depends on the build you're going for... The system rewards you for extra strikes if you have some levels of rogue for Sneak Attack damage, since Sneak Attack is so easy to achieve. If you're going for a straight up fighter though, I think then I think it's hard to beat a Two Handed Reach Weapon and if you want extra strikes, you can invest in the feats for Whirl Wind Attack (although I've tried that out and honestly it's really not worth it). Other thing to consider is that with a Two Hander, you only have to craft one weapon, whereas with TWF you need to craft two weapons, meaning double the experience and gold. Another point for Two Handers is that it takes a Full Action to get the benefits of TWF. Two Handers are much better when you plan on using Improved Trip a lot. And a negative point about TWF is that you'd want a light weapon for off hand, so either you need two different weapon types or you have to go with two light weapons. The problem with two weapon types is your off hand wont share your Weapon Specialization and Weapon Focus with your primary weapon. I had a party with both: one was a Fighter 8/ Monk 2 with Longspear/ Whirlwind Attack, and the other was a Ranger 4 (I wanted a doggie)/ Monk 2/ Fighter 2/ Rogue 3 TWF Katanas. The Fighter was awesome on criticals (around 90 damage!), and also the Reach was very, very cool since he could reach over people in crowded situations and he was able to use 5 foot steps a lot more, meaning he got a lot of Full Attacks that he wouldn't without a Reach weapon. The Ranger, meanwhile, was pure slaughter with Greater Invisibility. So she did greater damage more consistently, although she needed more buff spells than the fighter. In conclusion, I'd take a Two Hander for most builds, including Fighters and Clerics. And I'd only go for TWF only with Rangers, Rogues and Bards (since they incur Spell failure if they use a shield, and Ranged Attacks suck). Everything about Two Handers are better, the only thing that makes it close to balanced is TWF gets a little more benefit out of buff spells. So at low levels, Two Hander is better (more accurate) and then for a while TWF is a little better, but after you get Keen and/ or Improved Criticals, Two Handers pull ahead again. Cheers, Dai
Good discussion guys. I'm not saying that TWF is inferior in all cases. That wasn't my goal. I ran the numbers for my own enlightenment, and decided to share with the community. The numbers do get really close when the chance to hit is almost automatic, but it was an order of magnitude more difficult to figure it out at that end of the scale. (Does rolling a 1 still always result in a miss in 3.5? I haven't played PNP for at least 10 years, so my knowledge of 3.5 isn't up to snuff) I think though, in most cases, TWF is inferior to a two hander, since with most specialized builds to give an advantage to TWF, there is an alternate two hander build that will likely be superior. A good sneak attack, or special magic weapons could be a real exception though. Sorta off topic question, does Toee give the double damage bonus to 2 handers using power attack? (+2 damage instead of +1 for -1 to hit)
Well in the end I turned out not to be too lazy and wrote the program to simulate the actuall attacks. Here are my results for a few different situations: who dam/round a 3rd level fighter with 18 str no feats greataxe vs. AC 18 - 6.859 same vs. AC 13 - 10.264 same vs. AC 23 - 3.436 same with greatsword vs. AC 18 - 7.144 same with greatsword vs. AC 13 - 10.793 same with greatsword vs. AC 23 - 3.569 a 3d level fighter with 18 str TWF long sword/short sword vs. AC 18 - 6.188 a 3d level fighter with 18 str TWF long sword/short sword vs. AC 13 - 10.062 a 3d level fighter with 18 str TWF long sword/short sword vs. AC 23 - 2.313 now let's move them up to level 6 I'll only go with GS from now on for the THF and I will suppose a specialization and weapon focus in the primary weapon for both cases. I also choose to have a new AC of 26 since frankly as a DM i've never sent opponents with higher AC's at my 6Th level chars. in this case i have a BAB 6, abi +4, Strength bonus to dam +6 +1 bonus to attack for weapon focus +2 bonus to damage for WS GS vs. AC 18 - 19.296 GS vs. AC 13 - 27.795 GS vs. AC 23 - 10.766 GS vs. AC 26 - 5.747 now we have a primary BAB 4, abi +4 strength bonus to dam +4 +1 bonus to attack for weapon focus +2 bonus for to damage for WS second attack at BAB -1 and SS attack at BAB 4 but no weapon focus and no WS LS/SS vs. AC 18 - 13.565 LS/SS vs. AC 13 - 20.279 LS/SS vs. AC 23 - 6.814 LS/SS vs. AC 26 - 3.669 now suppose all of the weapons are shocking and see what the results are. note that offcourse given the higher amounts of random numbers used the accuracy is slightly smaller but i think that Sigma is still much smaller then 0.02 GS vs. AC 18 - 22.434 LS/SS vs AC 18 - 18.902 GS vs. AC 13 - 32.409 LS/SS vs AC 13 - 3.409 at the higher AC the gap widens again. I was very much surprised to find out that even at sixth level with shocking weapons the THF wins. Probbably at higher levels the gap might shrink a little more given that the top attacks will be most important but I think this post is long enough if anyone feels they would like to see more results or results for some specific situations I shall be happy to provide them. Rawel PS.: Note I didn't count the probabilities i just simulated the attack using rand()%X for dX rolling in C++ and about 100 mil. tries. I'm aware that the results might be skewed due to many things not the least the fact that rand is not really random and using modulo for the dice roll is even less inteligent. I was too lazy though too work out either the math or a way to get a smarter dice rolling mechanism if anyone knows a smarter one off the top of the head i'll be happy to implement it. PPS.: Gerwen I would actually be quite interested in seeing your numbers since it seems strange to me that they differ from the ones i got even if only very slightly i think I might be doing something wrong but i would like know what.
I've attached my spreadsheet. I marked it up with notes on cells to make it a little more understandable what i've done. Hopefully you can figure it out.
