Nerfed crossbows?

[Silas Dunsmore]

I have noticed that crossbows are severely underpowered compared to
Nethack.  Is this intentional?

A Slash'EM Gnomish Ranger starts with a +1 crossbow and has the natural
Ranger ability to fire one more missile per turn than other classes, yet
he has no chance of firing two bolts in one turn.

To gain even a 25% chance of firing two in one turn, he must be:
        a Ranger, Expert in crossbow, or
        a Ranger, Skilled in crossbow, and using at least a +2 crossbow, or
        a Ranger, Basic in crossbow and using at least a +5 crossbow,
        non-Ranger, Expert in crossbow, and using at least a +2 crossbow, or
        non-Ranger, Skilled in crossbow, and using at least a +5 crossbow.

This is because crossbows are GUNs in Slash'EM, and their rate-of-fire
is -1, which subtracts 2 (not 1) from the multishot calculation, if the
game is compiled with FIREARMS.  No such subtraction is made if FIREARMS
is not used.

The Ranger's flurry technique does not help because crossbows do not use
the P_BOW skill.

Crossbow bolts are already inferior to arrows.  This just doesn't seem
right.

If there is agreement on this, I am willing to make a patch.


        Silas Dunsmore



Here's my analysis:

In objects.c, GUNs have these parameters:
        (name,app,kn,bi,prob,wt,cost,range,rof,hitbon,ammotyp,metal,sub,color)
And crossbows are defined as:
        GUN("crossbow", (char *)0, 0, 1, 45, 50, 40, 12, -1, 0,
        WP_GENERIC, WOOD, P_CROSSBOW, HI_WOOD)
Note the -1 for the rof field.

In dothrow.c, in the throw_obj() routine, there are some multishot
calculations that are relevant for the case in question:

            switch (P_SKILL(weapon_type(obj))) {
            default: break; /* No bonus */
            case P_SKILLED: multishot++; break;
            case P_EXPERT: multishot += 2; break;
            }
and
            /* 1/3 of object enchantment */
            if (launcher && launcher->spe > 1)
                multishot += (long) rounddiv(launcher->spe,3);
and
            switch (Role_switch) {
            case PM_RANGER:
                multishot++;
                break;

The various tests for elven bow, elf/orc race, etc. don't apply.

We will see later that a multishot bonus of 4 must be reached to have a
25% chance of firing just 2 missiles.
Similarly, a multishot bonus of 5 must be reached to have a 20% chance
of firing 3 missiles and another 20% chance of firing 2 missiles.

Let's say that we have a bonus for being Skilled, and a bonus for being
a Ranger, but no bonus from weapon enchantment.  That gives us a
multishot score of 3.

Then we randomize:
            if (multishot > 0)
            multishot = rnd(multishot); /* Some randomness */
            else multishot = 1;
now our hypothetical multishot might be 1, 2, or 3.

Then we hit the FIREARMS-specific code:

#ifdef FIREARMS
            if (launcher && is_launcher(launcher))
            {
                if (objects[(launcher->otyp)].oc_rof)
                    multishot += (objects[(launcher->otyp)].oc_rof - 1);

.oc_rof is -1 for crossbows.  So this subtracts 2 from multishot.  That's
the critical change that makes crossbows so unattractive.

This puts our hypothetical multishot at -1, 0, or 1.

Then there's a bounds check:
        if (multishot < 1) multishot = 1;

So our Skilled (+1) Ranger (+1) with the starting crossbow (no mod) has
no chance of shooting more than one bolt.  You need an intitial multishot
score of 4 to have the 1-in-4 chance of getting a final multishot score
of 2, and thus shoot two bolts.


-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
_______________________________________________
Slashem-devel mailing list
Slashem-devel@...
https://lists.sourceforge.net/lists/listinfo/slashem-devel

[J. Ali Harlow-2]

On 2007-08-14 04:45:42 AM, Silas Dunsmore wrote:

> I have noticed that crossbows are severely underpowered compared to
> Nethack.  Is this intentional?
>
> A Slash'EM Gnomish Ranger starts with a +1 crossbow and has the
> natural
> Ranger ability to fire one more missile per turn than other classes,
> yet
> he has no chance of firing two bolts in one turn.
>
> To gain even a 25% chance of firing two in one turn, he must be:
> a Ranger, Expert in crossbow, or
> a Ranger, Skilled in crossbow, and using at least a +2
> crossbow, or
> a Ranger, Basic in crossbow and using at least a +5 crossbow,
> non-Ranger, Expert in crossbow, and using at least a +2
> crossbow, or
> non-Ranger, Skilled in crossbow, and using at least a +5
> crossbow.
>
> This is because crossbows are GUNs in Slash'EM, and their
> rate-of-fire
> is -1, which subtracts 2 (not 1) from the multishot calculation, if
> the
> game is compiled with FIREARMS.  No such subtraction is made if
> FIREARMS
> is not used.
>
> The Ranger's flurry technique does not help because crossbows do not
> use
> the P_BOW skill.
>
> Crossbow bolts are already inferior to arrows.  This just doesn't
> seem
> right.
>
> If there is agreement on this, I am willing to make a patch.

Just a quick response to this since I'm feeling guilty about the number  
of messages from you that I still haven't replied to!

It seems right to me that an expert archer should be able to use a bow  
and arrow more effectively than a crossbow (even though this may not be  
historically accurate). For an untrained archer the crossbow should  
certainly be more effective. It should also be the case that training  
with crossbows has less benefit than bows since it involves less skill.

I'd want to check whether there is any justification for the different  
effects under FIREARM. It does seem odd to me at first sight.

Cheers,

Ali.

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
_______________________________________________
Slashem-devel mailing list
Slashem-devel@...
https://lists.sourceforge.net/lists/listinfo/slashem-devel

[Silas Dunsmore]

On Tue, Aug 21, 2007 at 08:43:04AM +0100, J. Ali Harlow wrote:

> On 2007-08-14 04:45:42 AM, Silas Dunsmore wrote:
> >I have noticed that crossbows are severely underpowered compared to
> >Nethack.  Is this intentional?
>
> Just a quick response to this since I'm feeling guilty about the number  
> of messages from you that I still haven't replied to!
>
> It seems right to me that an expert archer should be able to use a bow  
> and arrow more effectively than a crossbow (even though this may not be  
> historically accurate). For an untrained archer the crossbow should  
> certainly be more effective. It should also be the case that training  
> with crossbows has less benefit than bows since it involves less skill.

Mm hmm, but remember that Rangers get multishot on all projectiles,
including daggers, darts, shuriken, sling stones, and firearm ammo.  So
crippling only the crossbow's multishot seems odd.


If we're arguing from reality, crossbows should be easy to aim and fire
once, but take a long time to reload, possibly with a strength modifier.

I could definately see using a loaded crossbow as a battle opener, then
swapping over to a hand-to-hand weapon.

But if we go that route, the bolts should do a _lot_ more than d4 small/
d6 large.


> I'd want to check whether there is any justification for the different  
> effects under FIREARM. It does seem odd to me at first sight.

I think one major gripe I have with the firearm code is this:

                if (objects[(launcher->otyp)].oc_rof)
                    multishot += (objects[(launcher->otyp)].oc_rof - 1);

This means that a rof of 1 is the same as a rof of 0.  And an rof of -1
has an actual penalty of -2.  You cannot specify a rof that will have a
penalty of -1.  Inconsistent.


I'm glad to get some dialog going; let's hear from the rest of you!

--
Later, eh?    Silas

Narrator:      There is no camera trickery in this shot.  The human eye would
               have seen just what the camera saw.  The room was designed by a
               master psychologist, Albert Ames, to illustrate how the brain
               interprets what the eyes see.

NARRATOR and MIDGET walk toward each other and become their "normal" sizes.

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
_______________________________________________
Slashem-devel mailing list
Slashem-devel@...
https://lists.sourceforge.net/lists/listinfo/slashem-devel