to_hit

[Tiblanc]

If you want to reduce the effect of high AC, you could use an opposed hit roll. Instead of rolling to_hit and comparing to AC, roll AC and compare both rolls.

Hit% is roughly :
If to_hit = AC : 50%
If to_hit > AC : (to_hit - AC) / to_hit + (1 - (to_hit - AC) / to_hit) / 2
If to_hit < AC : (AC - to_hit) / AC + (1 - (AC - to_hit) / AC) / 2

Having twice to_hit will give you about 75% chance to hit. To get 5% chance to hit, you would need to have AC about 10 times higher than to_hit.

That would most likely require a rework of everything, but it gives a smoother curve at both extremes. Also, there are a few things you can do to modify that curve to get whatever target values are required.

[Derakon]

So having both only increases the chances of hitting something by only 1-3% ? I'm glad I never bother using them since that seems way underpowered to be worth it.

This is in the late game. Earlier on they're much more worthwhile, as they make a relatively larger impact. And getting a 1% increased chance-to-hit against Morgoth may well be useful too...

[jens]

AC deflation, take two: a more balanced aproach

Well, no one has commented on ac and to-hit of late... So I guess deflating AC vs to-hit by 50% seems to be working OK.

After hearing that winners on the ladder had an average AC increase of about 20% I looked into this again. I realised that I deflated AC too much, so that AC becomes less interesting in the game. I have made some further experiments, calculations, tweaks and tests, and have as a result pushed a new commit with an overall better solution.

I have an Excel spread sheet on which I based my testing, but it's too large to post here. I can mail it if someone would like to take a look. It's not documented though, so understanding it might be a chore ;-)

[Magnate]

Well, no one has commented on ac and to-hit of late... So I guess deflating AC vs to-hit by 50% seems to be working OK.

After hearing that winners on the ladder had an average AC increase of about 20% I looked into this again. I realised that I deflated AC too much, so that AC becomes less interesting in the game. I have made some further experiments, calculations, tweaks and tests, and have as a result pushed a new commit with an overall better solution.

I have an Excel spread sheet on which I based my testing, but it's too large to post here. I can mail it if someone would like to take a look. It's not documented though, so understanding it might be a chore ;-)

I've seen your pull request and most of it looks ok - but why a minimum 12% chance to hit? That's 140% boost.

[jens]

I've seen your pull request and most of it looks ok - but why a minimum 12% chance to hit? That's 140% boost.

The 5% is just a historic number. The formula has been changed, probably several times since 5% was a good value, if it ever was. The reason for having such a number is because the formula breaks at the end, so the question is how broken we want hit chance to get before we stop it.

If we stop it at 5% the worst it gets is a monster hitting you 1 out of 20 times. The start of the discussion was because tough monsters were hitting players 1 out of 4 times. So setting the stop at hitting 1 out of 8 times still leaves a lot of leeway for breaking the scale.

On the other end of the scale we have the low level monsters with low powered effects that inflict status ailments or stat draining. I want them to still be somewhat relevant, and at 1 hit of 20 they are not, but at 1 in 8 you will get the occasional hit even with good ac. Another scenario is when raiding orc pits when you are far superior, now you would get hit on average once every monster turn, instead of twice in five monster turns...

But there is also the issue of balancing to-hit vs AC throughout the dungeon. When tweaking values in my spread sheet I learned that including a raised auto-hit number had positive effects. Basically it is easier to balance the breaking point with a higher auto-hit number. Which means I can make to-hit and AC balance quite well throughout the dungeon.

[jens]

Ah, maybe should mention the most relevant case as well, and not just the end cases. So you are fighting some monster that has a relevant to-hit compared to your AC. Say it's chance to hit you is 51%. The increase from 5% to 12% auto-hit means that something less than 7% will be added to that 51%, so in one scenario that monster will now hit you 54% of the time. Or another case where you have better AC and the monster hit chance will change from 20% to 26%. So in most cases the change has a very limited impact.

[Jazerus]

I'm liking the change - as a Paladin it makes some enemies, particularly melee uniques and other things with lots of blows, much more interesting to fight without making them unduly deadly.

[Magnate]

I'm liking the change - as a Paladin it makes some enemies, particularly melee uniques and other things with lots of blows, much more interesting to fight without making them unduly deadly.

Good. Jens - pls could you post a link to your spreadsheet? I'd like to see the to-hit-vs-ac for normal monster blows before and after your changes (and compared against your new proposals)

[jens]

Don't really have any way of posting it though... mail?

[jens]

There I made a smaller version that I could zip up here :-)

I just removed the tabs that worked my way up to the final tabs. Then I also had to remove one of the relevant tabs, so I suggest the first thing you do is make a copy of the first tab (to-hit vs AC 3.2), then in this new tab change the field 'AC value' (marked in red at the top) from 0,75 to 0,5. This will then be the tab for the comparisons in the current nightly. The last tab is for my newest set of changes. Here some of the formulas are different than in the previous tab.

What they do is take monster level and effect power, and compare these to 3 different columns of AC. I have low, mid and high AC. Low ac is 10 + lvl, mid ac is 20 + 2*lvl, high ac is 30 + 3*lvl. I have also inserted some jumps into the high AC table, i.e. at some points I don't use the linear model, but just put in a very high number.

Please also note that in my new suggestion I have changed values of power for the effects.

For ease of reading I also format the result so values are shown in red, yellow, green, or normal to indicate how 'dangerous' they are. What I want to accomplish is that there should be a clear progression from low ac to high ac, without getting any of the problems we had in 3.2

[jens]

Ah, yes, I modified the start of the high end AC after testing several characters trying to maximise AC at the start of the game. Those values are the highest AC values I achieved after running 5 characters, if they were higher than the formula for high ac I gave above.

[jens]

So I've been thinking a bit more about your question about 5% vs 12% automatic hit. I've been trying to find cases were @ gains a lot from that change. Most cases I found were totally ludicrus, but some come close. We only need to consider mages, all other classes have a high enough base to-hit that in order for this to be a significant change they need to meet monsters way harder than they can handle.

The most detrimental example I could find against the case of 12% automatic hit is a lvl 29 mage vs an Iron Golem.

lvl 29 mage (base to-hit 64) vs AC 90 Iron Golem (lvl 22)
at 5%
+0	+3	+7	+13	+20	
9%	21%	30%	43%	51%
at 12%
+0	+3	+7	+13	+20	
16%	27%	35%	47%	55%

To make it worse, lets say the mage in question has had good rolls for hp (180), and has found good AC (120). This mage would survive on avg 66 rounds versus the Iron Golem. He also started with bad stats, and has not found any good to-hit bonuses, so lets say he at level 29 only has +3 in to-hit. He would then during the 66 rounds he can stand and fight increase his successful hits from an average of 14 during the entire fight, to a wooping 18 during the entire fight. If you want to be extreme we can assume that he has a dex lower than 16, and is still using an unenchanted weapon, and has no other +to-hit. Then he would increase from hitting an average of 6 times to an average of 10.5 times. Oh, and the golem has 520 hp.

Yes, there are cases were your damage output increases radically, but you don't gain anything from it in those cases. But thinking about it I have come to realize I should have compensated for this in monster AC by increasing monster AC an additional 5%.

[jens]

Updated my commit so that monsters AC are now in total boosted with 20% instead of the previous 12.5%

Also updated the 'describe_monster_toughness' so the correct to-hit % is shown when looking at monsters.

Now I have to rush to the pre-release of M12!