AC/EV/SH statistics


Although the central place for design discussion is ##crawl-dev on freenode, some may find it helpful to discuss requests and suggestions here first.

Ziggurat Zagger

Posts: 11111

Joined: Friday, 8th February 2013, 12:00

Post Wednesday, 11th March 2015, 22:43

AC/EV/SH statistics

Is it possible to add statistics about how many hits (count and percentage) were blocked by shield, how many hits were dodged and how many hits were reduced to 0 by AC? It could help to see how useful different lines of defense are.

For this message the author Sandman25 has received thanks:
mps

Tartarus Sorceror

Posts: 1774

Joined: Tuesday, 23rd December 2014, 23:39

Post Thursday, 12th March 2015, 15:36

Re: AC/EV/SH statistics

The numbers you really want are:
1. the proportion of hits that were blocked
2. the proportion of unblocked hits that were dodged
3. the amount of damage that was reduced by AC, divided by the total incoming damage that was unblocked and undodged
4. the proportion of damage that was reduced by resistances, divided by the total incoming damage after AC

Each of these (1-4) should be taken from data over some reasonable time window such as the last 100 turns during which enemies attacked you.
streaks: 5 fifteen rune octopodes. 15 diverse chars. 13 random chars. 24 NaWn^gozag.
251 total wins Berder hyperborean + misc
83/108 recent wins (76%)
guides: safe tactics value of ac/ev/sh forum toxicity

For this message the author Berder has received thanks:
Rast

Ziggurat Zagger

Posts: 11111

Joined: Friday, 8th February 2013, 12:00

Post Thursday, 12th March 2015, 16:39

Re: AC/EV/SH statistics

Yes, 1-3 is what I basically suggested (except for 3, reducing to 0 is identical to dodging/blocking). I am not really interested in watching those from last 100 turns because I can already press Ctrl+p.
I am not interested in 4 as it changes nothing (I will swap to rF+ for Fire Giant etc. with all characters).

Displaying the statistics can help to make decisions about removing shield, switching to lighter/heavier armour, also can help to compare different characters.

Tartarus Sorceror

Posts: 1774

Joined: Tuesday, 23rd December 2014, 23:39

Post Thursday, 12th March 2015, 19:44

Re: AC/EV/SH statistics

Why do you only care about whether the damage is reduced to 0? Most of the damage reduction from armor does not reduce it to 0 (usually).
streaks: 5 fifteen rune octopodes. 15 diverse chars. 13 random chars. 24 NaWn^gozag.
251 total wins Berder hyperborean + misc
83/108 recent wins (76%)
guides: safe tactics value of ac/ev/sh forum toxicity

Ziggurat Zagger

Posts: 11111

Joined: Friday, 8th February 2013, 12:00

Post Thursday, 12th March 2015, 19:45

Re: AC/EV/SH statistics

Berder wrote:Why do you only care about whether the damage is reduced to 0? Most of the damage reduction from armor does not reduce it to 0 (usually).


To compare it with EV/SH.

Tartarus Sorceror

Posts: 1774

Joined: Tuesday, 23rd December 2014, 23:39

Post Thursday, 12th March 2015, 19:51

Re: AC/EV/SH statistics

Sandman25 wrote:
Berder wrote:Why do you only care about whether the damage is reduced to 0? Most of the damage reduction from armor does not reduce it to 0 (usually).


To compare it with EV/SH.

This:
3. the amount of damage that was reduced by AC, divided by the total incoming damage that was unblocked and undodged
is what you want, if you want to compare it with EV/SH. That is:
1. block ratio tells you by what factor average incoming damage is reduced by your sh
2. dodge ratio tells you by what factor average incoming damage is reduced by your ev
3. armor ratio (as above) tells you by what factor average incoming damage is reduced by your ac

If you only looked at when the ac reduces incoming damage to exactly 0, you'd be undervaluing ac relative to ev/sh because ac does much more than that.
streaks: 5 fifteen rune octopodes. 15 diverse chars. 13 random chars. 24 NaWn^gozag.
251 total wins Berder hyperborean + misc
83/108 recent wins (76%)
guides: safe tactics value of ac/ev/sh forum toxicity

For this message the author Berder has received thanks:
Rast

Ziggurat Zagger

Posts: 11111

Joined: Friday, 8th February 2013, 12:00

Post Thursday, 12th March 2015, 20:01

Re: AC/EV/SH statistics

Berder wrote:If you only looked at when the ac reduces incoming damage to exactly 0, you'd be undervaluing ac relative to ev/sh because ac does much more than that.


Yes, I know. SH can block attacks which would be dodged by EV anyway, EV can dodge attacks which would be reduced to 0 by AC anyway and seriously there is a huge difference between dodging Crystal Spear and rat's attack.

Ziggurat Zagger

Posts: 11111

Joined: Friday, 8th February 2013, 12:00

Post Friday, 24th April 2015, 12:45

Re: AC/EV/SH statistics

Argonaut, thank you for the patch!
User avatar

Tomb Titivator

Posts: 778

Joined: Thursday, 13th March 2014, 20:15

Post Friday, 24th April 2015, 17:46

Re: AC/EV/SH statistics

Sandman25 wrote:Argonaut, thank you for the patch!
Oh, nice, btw how to use this in 0.16?
~online scoring~

Pig's in zen
Pig is nude
Unashamed

Ziggurat Zagger

Posts: 11111

Joined: Friday, 8th February 2013, 12:00

Post Friday, 24th April 2015, 18:16

Re: AC/EV/SH statistics

Eyesburn wrote:Oh, nice, btw how to use this in 0.16?


I believe the only possible way is to apply the patch to 0.16 branch manually.

Return to Game Design Discussion

Who is online

Users browsing this forum: No registered users and 64 guests

cron
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group.
Designed by ST Software for PTF.