I find throwing OP


Years-old doodle I made, but illustrates my feedback on the situation. Without training at game-start, I would expect throwing to not be significantly more accurate than melee. Maybe manhacks are a special case due to size, uncertain. Of course, this is based on an old-as-balls game version and may have since been rectified. Less precise than melee and guns both, would feel a lot more natural.

At the least I’d reduce the basic experience multiplier a bit since you shouldn’t be able to reach 10 skill in single season.
This multiplier is:

Also as far as I can tell all thrown attacks start with 0 dispersion and then get modified, I’d probably add a base deviation which can then be reduced by sufficient skill:

For reference, each 1 added to deviation is roughly 45 dispersion.

Not a perfect fix, but I can’t think of any good arguments against it, thoughts?

Taking a stab at some real-world evidence for this thread:

The best MLB pitchers can reliably hit the ‘strike zone’, a non-moving target about the size of a human torso, at a distance of 20 meters using a 5 oz baseball traveling at 95 mph. That likely represents the upper end of possible human throwing skill, as it comes from professional baseball players who spend years practicing throwing this particular object in this particular circumstance.

For larger objects and weapons, some throwing competitions which I can’t post external links to (search for Epic Blade Time on youtube) seem to indicate that the winners of these knife / tomahawk distance throwing competitions hit a target was at a distance of about 22 meters - however especially from the second video it does not look like they are able to do so reliably and hits at this distance or beyond represents in part a lucky throw. I also have no way to estimate the skill of the participants but they are likely very skilled at throwing if they are distance competition winners.

A modern throwing tomahawk weighs 12-15 ounces (not necessarily the ones in the video however). I was not able to find speeds but it may be possible to roughly calculate those based on estimating the throw angle in the video and the distance traveled.

This XKCD what if article 44 (High Throw) compares the throwing distance of the average person to that of an MLB pitcher, which could be great for skill benchmarks. The article is for a baseball thrown upwards but the comparison is likely based on horizontal throwing data. Note that this is distance only (and distance upward at that) and has no details on accuracy.

From the article:
Average guy: 15 meters.
Good arm: 25 meters.
MLB 80 MPH fastball: 50 meters.
World record 105 MPH fastball: 70 meters.

I think all is said, if the balance is done using real life statistics. I find that so strange, given that rifles clearly had their range nerfed for gameplay interest reason.

gameplay should be prevalent over realism…

[quote=“Comassion, post:23, topic:13106”]The best MLB pitchers can reliably hit the ‘strike zone’, a non-moving target about the size of a human torso, at a distance of 20 meters using a 5 oz baseball traveling at 95 mph. That likely represents the upper end of possible human throwing skill, as it comes from professional baseball players who spend years practicing throwing this particular object in this particular circumstance.[/quote]I’m going to strongly disagree with this one.
Let’s start with “best MLB pitchers”… no, just no. Your average highschool pitcher can pull that accuracy reliably too, the speed would vary, but the accuracy would not do so much.
Similarly, these same baseball players can throw a pass from the fields to the bases reliably, which is a far greater distance.

After that, you get the trouble that these MLB players would be STR 14, not, say, steroid-abusing base STR 18 monsters that later took bear mutations and installed the strength CBMs and railgun.

That’s what I said…

[quote=“Aabbcc, post:25, topic:13106”][quote=“Comassion, post:23, topic:13106”]The best MLB pitchers can reliably hit the ‘strike zone’, a non-moving target about the size of a human torso, at a distance of 20 meters using a 5 oz baseball traveling at 95 mph. That likely represents the upper end of possible human throwing skill, as it comes from professional baseball players who spend years practicing throwing this particular object in this particular circumstance.[/quote]I’m going to strongly disagree with this one.
Let’s start with “best MLB pitchers”… no, just no. Your average highschool pitcher can pull that accuracy reliably too, the speed would vary, but the accuracy would not do so much.
Similarly, these same baseball players can throw a pass from the fields to the bases reliably, which is a far greater distance.

After that, you get the trouble that these MLB players would be STR 14, not, say, steroid-abusing base STR 18 monsters that later took bear mutations and installed the strength CBMs and railgun.[/quote]
I would say that MLB pitchers can reliably hit a target zone much smaller than that, they’re not aiming at the strike zone, they’re aiming at a point in the strike zone they think the batter will have trouble hitting. The observation that they can throw a 5oz ball at ~95mph is pertinent though, faster balls means more strikes, so that’s something they are trying to improve.
A pass from the outfield is totally different, the “target” is “close enough to the catching player that they can take a step and then catch it”. I’m not saying no one can make throws with meaningfull accuracy from this distance, but that’s certainly not the norm in baseball.
I’d argue that MLB pitchers are effectively str 18 for the purposes of throwing a baseball, we don’t have this kind of specialization in DDA, but we should recognize it in reality. That having been said, baseball’s aren’t the only or even dominant thrown item, which is why we need a number of examples.

Sweet, this is exactly the kind of thing I’m looking for, I look forward to the Epic Blade Time videos in particular, I couldn’t find any good info on throwing weapons as opposed to things like baseball and track&field stuff previously.

At the moment, throwing has a problem of not having properly defined accuracy per item.

That is, to-hit is ignored totally and has no replacement (except some values calculated from volume and weight).

It isn’t necessary to balance it, but without it thrown katanas will cut well, which pretty hilarious.

I have an uncle that was fond of hunting things the “natural” way. I can provide effective ranges.

Note: Strength does not usually play into throwing effectiveness at all once you get past extremes. Pitchers need a strong arm, knife hunters do not. The principles between the two are completely different. (I will elaborate on this further in a later post if you wish.)

A rough guide to the real life applications of throwing weapons:

Sport Knives: Used to hunt very small game such as birds and rabbits. Max range around 25 feet if the person throwing is VERY skilled. Not at all suitable for killing people, they are easily stopped by clothes and it is impossible to throw them accurately with enough force to pierce a human skull or sternum.

Slings: Holy sweet mother of god these things can be lethal! Good against slightly larger game than knives (I’ve seen someone kill a coyote with one) have an effective killing range of around 60 feet for the average slinger more around 100 if you’re skilled. A lead bullet (correct term for a sling projectile) can punch straight through one side of a human being and go right out the other. There are techniques for getting your bullet to fly for several hundred meters, but they aren’t practical outside of Roman Era mass combat.

Tomohawks: A myth. Have never been used as throwing weapons historically. Misconception originated from popular “tomohawk throwing” competitions. Useless against live targets because they require pre- planned positioning and spacing to account for the fact that they spin.

Javelins: These are lethal and have a decent range. An olympic thrower can chuck one about the length of a football field, but not with any degree of accuracy. Practical effective range of around 50 meters for a skilled thrower. Were historically used against people. Range and velocity can be extended with an atlatl, but you’re not going to hit anything smaller than a wooly mammoth or a lumped mass of germanic barbarians. (Note: These require some strength)

Hunting Spears: This is a wierd one. They’re not actually “thrown” so much as “dropped”. You typically sit in a tree or stand and then drop the spear down on an animal (usually a boar from my experience) that passes underneath. They don’t make very practical weapons to throw from ground level as they are heavy in the front.

Subsection A: Weeb Shit

Shruiken: Generally thrown in the face as a distraction. Not an effective killing tool.

Kunai: Weren’t thrown and aren’t designed to be. Naruto lied to you.

Subsection B: Historical shit that I’m only passingly familiar with

Throwing Axes: Used to embed in shields and disrupt the front line in mass combat. I doubt that they’re very accurate, but if they hit someone it would probably break bone. Were large and impractical to carry in significant numbers.

Chakrams: Historical accounts say that a thrown chakram could cleave an arm or a leg clean off. I can’t corroborate this. The Sikh tribe still uses them. I’ll look into it.

Darts (2.5 foot long historical hunting tool, not the bar game): Seems like a less effective javelin. Probably only good for small game.

Unfortunately I can’t provide any numbers right now, but as someone who has a big interest in history, I can attest that slings are woefully underperforming in Cata. A skilled slinger is a force to be reckoned with, not only in mass combat, but also in a huntig situation (which is probably much more relevant to Cata). At the very least, the more elaborate sling ammo like lead bullets should get some hefty bonus to armour pierce.

Let me tell from my own exp.

Throwing Knives = Shit for killing, maybe effective against cats.

Throwing Axe = I dont think they are reliable enough to be usefull as a ranged weapon (then again I was using a hatchet)

Javelin and such = I would say they are the best throwing weapons in terms of deadliness and considering that they where used to make shields useless you can see why.

Rocks, etc = Lets see…You can kill someone with a rock and the accuracy on long range is not too bad either (something like 15~18 tiles should be the max usefulness).

Slings = Damn those things are too scary, slingshots are still in use when hunting (especially rats and geckos) but the effectiveness are most ammo dependent, dont know what would take to kill someone though. Slings…jeez I would not be surprised if someone could take a mammoth with those, nah just joking, I dont have much exp with those IRL it looks hard master (4 hits out of 10) but iam sure it was only a question of time.

So IMO throwing should be affected by what weapons you are throwing, but it looks like the problem is more with the code. Now about exp growing too fast, I do think it is fast but throwing itself is not something that hard to master soo there has to be a balance on it that doesnt make it super hard to gain exp but neither super easy as it is.

There is another major oversight in the way ranged attacks are treated in CDDA currently, which is to say that they are treated as instant projectiles which the target has no opportunity to evade. This was an assumption that was presumably made in regard to guns, for which this statement is effectively true unless you have vastly superhuman perception and reflexes.

However, this assumption is blatantly false for most throw projectiles, particularly massive ones like a tomahawk, spear, or even a baseball. These weapons rely on their mass much more than velocity to impart momentum on their target, as such their travel speed is quite slow - well within human perception and reaction times to potentially evade at any significant range, if the defender is aware of them.

Bows are a middle ground case. When fired on a reasonably flat arc, they are far faster than any thrown weapon, and would be quite difficult to evade after firing, though not impossible at extended range. When fired on a parabolic arc however, as one might see at even longer ranges or against massed opponents, they would be quite easy to individually evade due to their extended flight time and their much slower velocity at the top of their flight arc.

In any case, while it is fine to represent all missiles as impacting immediately for the sake of representation, missile velocity could be added, with an agility/dodge roll allowing a modifier to evade said missiles, with the slower missiles suffering much more against agile opponents, and the effective velocity dropping with range, so that hitting an evasive opponent at long range becomes unlikely even if your aim and lead is perfect - they simply aren’t going to wait for it to hit them, any more than they would allow you to calmly hack them to pieces with a machete without evading.

The fact is that even guns can be evaded - they simply cannot be evaded ‘after the fact’. An agile opponent is still harder to hit, as long as they are actively evading BEFORE the bullet is fired, effectively making aiming harder, and an unaware or still opponent would have no chance to evade whatsoever. Hitting something like a feral zombie at range should be a much more difficult undertaking than shooting a fat zombie with no dexterity to speak of.

There is another major oversight in the way ranged attacks are treated in CDDA currently, which is to say that they are treated as instant projectiles which the target has no opportunity to evade. This was an assumption that was presumably made in regard to guns, for which this statement is effectively true unless you have vastly superhuman perception and reflexes.

However, this assumption is blatantly false for most throw projectiles, particularly massive ones like a tomahawk, spear, or even a baseball. These weapons rely on their mass much more than velocity to impart momentum on their target, as such their travel speed is quite slow - well within human perception and reaction times to potentially evade at any significant range, if the defender is aware of them.

Bows are a middle ground case. When fired on a reasonably flat arc, they are far faster than any thrown weapon, and would be quite difficult to evade after firing, though not impossible at extended range. When fired on a parabolic arc however, as one might see at even longer ranges or against massed opponents, they would be quite easy to individually evade due to their extended flight time and their much slower velocity at the top of their flight arc.

In any case, while it is fine to represent all missiles as impacting immediately for the sake of representation, missile velocity could be added, with an agility/dodge roll allowing a modifier to evade said missiles, with the slower missiles suffering much more against agile opponents, and the effective velocity dropping with range, so that hitting an evasive opponent at long range becomes unlikely even if your aim and lead is perfect - they simply aren’t going to wait for it to hit them, any more than they would allow you to calmly hack them to pieces with a machete without evading.

The fact is that even guns can be evaded - they simply cannot be evaded ‘after the fact’. An agile opponent is still harder to hit, as long as they are actively evading BEFORE the bullet is fired, effectively making aiming harder, and an unaware or still opponent would have no chance to evade whatsoever. Hitting something like a feral zombie at range should be a much more difficult undertaking than shooting a fat zombie with no dexterity to speak of.

This is only aplied to npcs, as far as i can tell the majority of the cdda enemies are not the agile type.

Well, the base zombies certainly don’t have much dexterity - but when your melee skill is low, there are quite a few enemies in the game that have quite good evasive abilities. There are also classes of zombie that are MUCH more nimble than the player, with the ability to leap several meters at a bound or outrun you handily. Then there are all the non-zombie types, such as triffids, giant insects, and so forth, some of which are also quite evasive in melee, and could presumably also evade missile weapons.

Well, the base zombies certainly don't have much dexterity - but when your melee skill is low, there are quite a few enemies in the game that have quite good evasive abilities. There are also classes of zombie that are MUCH more nimble than the player, with the ability to leap several meters at a bound or outrun you handily. Then there are all the non-zombie types, such as triffids, giant insects, and so forth, some of which are also quite evasive in melee, and could presumably also evade missile weapons.

When I said agile I meant they are not going to try to dodge ranged weapons, because they neither have the smarts or give a fuck about dodging.

Fear my glass shard !!

When I was a kid we used to hunt rabbits by throwing rocks. Not surprisingly this is not very effective and I would venture that when we started out we were at skill 0 and moved up to probably about skill 8. With a average sized rock 2 1/2 inches around at skill 0 we hit maybe 1% of the time, and over time when we got to about skill 8 we were hitting them about 30% of the time. This of course is a small animal sitting / running sometimes at high speed. Objects that are small are much harder to hit when moving. As for zombies and human sized creatures, hitting them close up should be 100% ( its not hard to hit something only 2 or 3 meters away ) Head shots however should not be this accurate as thats a smaller target. Again tho I dont have real world data only my personal experience and its been a long time. Here in AZ tho we have cacti that are about the size and shape of a human so i had experience with trying to hit them. Its not as easy as it might seam esp at lower levels. I think someone with 0 skill in throwing such an object at 5 meters should be able to hit it maybe 50% of the time with head shots being down near the 1% mark.

100% Hit rate in relaxed non-lethal environment from national competitors using spin method.

I’m sad to be a part of my species right now, however awesome this is:

There’s a TON of knife videos on YT and while the internet may be a completely reliable source I thought I’d dig up some figures:


http://www.knifethrowing.info/thrower-meeting_world_championship_huthwaite_2015_results_ergebnisse.html#results

Olympics:

A Javelin can go through shield on occasion:

Force, mass, surface area of impact, friction, etc all heavily influenced by how well a person understands their body’s motions.

And yes, novices are as likely to injure themselves as the target. As I know from personal experience.

Throwing blunt objects like rocks? Go to the source lol.

P.S. If you think throwing is OP now activate hydraulic muscles, railgun and throw a live grenade or a steel spear.

Can attest that throwing is both extremely deadly and too fast in terms of experience gain. It also takes much less endurance compared to melee to use.

Also of note, thrown weapons never break. So, even budget steel broadswords are a weapons worthy of Zeus-grade throwing, obliterating (sometimes literally) unarmoured targets with 70~90 damage (up to ~8 tiles with training). One limiting factor is weight and volume of most weapons. But then we have a gift of dive knife with 0.01 volume and negligible weight (~30-40 damage with 5 throwing at 20+ tiles distance).

Seems that quickly nerfing throwing is not to be done? Devs should be more pragmatic, if nothing is done until statistical data on knives throwing IRL are collated, then we won’t see an update on throwing this year!