[interface] we need better statistic (my propose)

Brainstorm ideas of possible additions to the game. Read this before posting!

Moderators: Forum Moderators, Developers

Forum rules
Before posting a new idea, you must read the following:
Post Reply
igor
Posts: 11
Joined: September 14th, 2010, 5:46 pm

[interface] we need better statistic (my propose)

Post by igor »

Hi

In the first place, sorry my English is bad but I play wesnoth since 1.6., and play often, so I’m experienced player.

Few SIMPLE things we very need in wesnoth...

1. Problem: Players during or after the game say I lose because bad luck. Enemy’s usually think differently :). Luck is important part of wesnoth, so we need to count luck.
Proposition: It must work like damage calculator (button "s"). For example, adept/adepts do 10 attacks and hit 7 times. So, that is normal luck overall. I propose to count luck per turn and per game at hit points. For example, if 2 fighters (5-4) attack archer (do 4*2=8 hits) in a watervillage (20%) and hit 6 time luck is… (normal luck is: 8*5*80%=32hp, in game was: 6*5=30hp, so that is a little unluck, this two attacks gives (30-32=-2) point to overall per turn luck statistic. So we may see how luck do a game :). A little difficultly to count a luck per unit, but this will be cool too.

2. Problem: We don’t know who play fast who slow. How fast? How slow? Players during or after the game may say you win because you think long. I have no statistic to say – not, time was the same :)
Proposition: count a time which players use per turn/per game. It can be add to statistic (button “s”)

I suppose, developers will see my text and do some comments (please… and thanks :) ) to my propositions.



Best regards, igor
Last edited by igor on January 13th, 2012, 2:50 pm, edited 3 times in total.

Caphriel
Posts: 994
Joined: April 21st, 2008, 4:10 pm

Re: we need better statistic (my propose)

Post by Caphriel »

Tracking player statistics is Frequently Proposed Idea number 13, and as such, isn't going to be implemented. Read the "About the Ideas Forum" link at the top of this page, in the forum rules box.

User avatar
Gambit
Loose Screw
Posts: 3266
Joined: August 13th, 2008, 3:00 pm
Location: Dynamica
Contact:

Re: we need better statistic (my propose)

Post by Gambit »

Yeah, we don't do that sort of statistics stuff in MP. Especially not statistics that exist only so that you can discriminate against certain players. Locked.

edit: The OP has clarified that he wants these stats added to the statistics dialog on a per-game basis. He doesn't want server-side statistics tracking. Unlocked.

AlexanderK
Posts: 24
Joined: August 4th, 2011, 3:41 am

Re: [interface] we need better statistic (my propose)

Post by AlexanderK »

Very interesting idea, it would be nice to have per percentage statistic too for 20-70% defense, to see how much hit and misses you have for them, not per unit but per hit chance with option of excluding berserker frenzy attacks.

There is a evil laugh when berserker attacks dark adept or mud crawler, how about to add some sort of crazy stupid laugh from your opponent when your thunderers miss miss miss 20 times in a row. :mrgreen:

Caphriel
Posts: 994
Joined: April 21st, 2008, 4:10 pm

Re: [interface] we need better statistic (my propose)

Post by Caphriel »

Given the clarification to per-game basis, disregard my previous post. I for one would find per-game time-use tracking to be interesting.

User avatar
Cackfiend
Posts: 436
Joined: January 28th, 2007, 7:36 am
Location: Florida, USA
Contact:

Re: [interface] we need better statistic (my propose)

Post by Cackfiend »

AlexanderK wrote:Very interesting idea, it would be nice to have per percentage statistic too for 20-70% defense, to see how much hit and misses you have for them, not per unit but per hit chance with option of excluding berserker frenzy attacks.

There is a evil laugh when berserker attacks dark adept or mud crawler, how about to add some sort of crazy stupid laugh from your opponent when your thunderers miss miss miss 20 times in a row. :mrgreen:
I literally ask for the exact same thing in my post 3 posts below this one...
http://forums.wesnoth.org/viewtopic.php?f=12&t=35843

=x


also, im pretty sure no one thinks a player won because they took more time for their turns... that being said, having your average turn time in the statistics menu might be handy sometimes so it wouldnt hurt to have it in there
"There's no love in fear." - Maynard James Keenan

I'm the guy who's responsible for 40% Gliders in all hexes... I can now die a happy man. =D

Caphriel
Posts: 994
Joined: April 21st, 2008, 4:10 pm

Re: [interface] we need better statistic (my propose)

Post by Caphriel »

Well, if nothing else, time statistics could be used to support an argument for more flexible time settings, perhaps :)

igor
Posts: 11
Joined: September 14th, 2010, 5:46 pm

Re: [interface] we need better statistic (my propose)

Post by igor »

OK.

So, luck statistic calculates AFTER attack always.

I realized that I should better explain the proposed method of collecting statistics, the influence of luck on the game. Interest is not the luck in the pure state. Its impact on the game more interesting. And this statistic is interesting to analyze with “inflicted and received damage statistic” that already exists in the game (button "s"). Thats why I offer these same units of measurement – health points. Units of measure of luck in its pure form - it's certainly percentages, but thats not usefull to analyze. Percentages may be counted as additional statistics.

When preparing to attack, we know in advance the conditions under which it is. We know the number of collisions and the possibility to hit the target. Accordingly, we are counting on something quite definite. In fact, its a a factor of location, type and strength of attack. Baseline values of that ​​are known, so we may to calculate “baseline” damage. That damage I will call reference (normal) luck. Other damage is luck or unluck. Statistics for the proposed method shows the influence of such factor as luck game on coating and obtained damage of units. It's like “how many health point you obtaine because (by) luck (or unluck)” in a specific conditions (tactic actions, type of attacks, time of the day and others) . Together with the statistics of the damage that can be analyzed, we may see, for example, we give X hp damage and Y of them was bacause of luck.

Thus, we can to make conclusions on the feasibility of our tactical actions, because the factor of luck takes into account are influence in our statistic.

For example, at night 2 archers (player A, 15-2) attack wose in forest (player B, hp 52) and two Orcs (player A, 12-2, hp 40 and hp 7) attack elvish archer in forest (player C, hp 9, 4-3) during players A turn (see attachment). See per turn luck statistic there.

We may see in table 5 (per turn luck points stat) player C had best luck, its because orcs missed archer every time, and thats why player A had unluck in per turn statistic despite lucky fire attack on wose (player B).


In attachment you may see my calculations, if I can do that in Excel it's no problem to do that in the game :), question is only in approve my propose. You may edit column O (table 3) to change results of attacks and see how luck will be changed.

How to count slows and berserkers ability I will explane later if this propose will be approved.
Attachments
luck_stat.zip
(9.13 KiB) Downloaded 151 times

AlexanderK
Posts: 24
Joined: August 4th, 2011, 3:41 am

Re: [interface] we need better statistic (my propose)

Post by AlexanderK »

well, igor, your table makes sense, BUT sometimes taken damage and suicide is also can be considered as luck:

1. when you need to deal some damage and die to let this hex free to let another unit to attack to kill enemy leader for example and when your unit will survive with 1-3 hp it's not good. sacrifice pawn to slay the king.
2.in "no where to run" situation when enemy unit is about to level up and you need to suicide your heavily injured troop on another enemy unit with low xp bar. and if you fail to suicide enemy will level up.

How will you count "luck" points in such situations?

igor
Posts: 11
Joined: September 14th, 2010, 5:46 pm

Re: [interface] we need better statistic (my propose)

Post by igor »

AlexanderK wrote:well, igor, your table makes sense, BUT sometimes taken damage and suicide is also can be considered as luck:

1. when you need to deal some damage and die to let this hex free to let another unit to attack to kill enemy leader for example and when your unit will survive with 1-3 hp it's not good. sacrifice pawn to slay the king.
2.in "no where to run" situation when enemy unit is about to level up and you need to suicide your heavily injured troop on another enemy unit with low xp bar. and if you fail to suicide enemy will level up.

How will you count "luck" points in such situations?
I think, no. That may be considered as tactical actions. I give you same example - block recruting in the enemy castle (with big chance to die) :). It will be good for tactic statistic, but we must have artificial intelligence to analyze all possible options. But in any case its not a luck.

I made one assumption, the statistics of luck depends on the force of impact. If you start entering additional dependencies proportion of luck in its pure form is reduced even more. Compliance with these statistics, to its name will become questionable.

You must to understand luck is not related with units die. Luck is related with random number generator (will you or your enemy hit or miss). And this is the meaning of proposed statistic. So answer on your question is... I will count luck points like all times, when hit - its luck, when miss - unluck of depending on the force of impact and chances to hit. So, if your wose in forest with 1hp was attacked by archer and die after 1st shot it gives minus in luck statistic, if them die from 2nd and after shots its gives plus to your luck statistic. Its because chance to hit wose is 60%.

You may see formuls in my Excel file (see bellow message).

So for simply understanding:

luck in pure form - shows which side takes "random number generator" :) (probably of hits are accounted). And in which proportion of...
proposed here method to count luck statistic - shows how luck in pure form appear in the applied and received units damages (health points).

OK, I propose to show both form of luck statistic and will prepare the sample of calculations little later (for 2nd method was before)

AlexanderK
Posts: 24
Joined: August 4th, 2011, 3:41 am

Re: [interface] we need better statistic (my propose)

Post by AlexanderK »

igor wrote:
AlexanderK wrote:well, igor, your table makes sense, BUT sometimes taken damage and suicide is also can be considered as luck:

1. when you need to deal some damage and die to let this hex free to let another unit to attack to kill enemy leader for example and when your unit will survive with 1-3 hp it's not good. sacrifice pawn to slay the king.
2.in "no where to run" situation when enemy unit is about to level up and you need to suicide your heavily injured troop on another enemy unit with low xp bar. and if you fail to suicide enemy will level up.

How will you count "luck" points in such situations?
I think, no. That may be considered as tactical actions. I give you same example - block recruting in the enemy castle (with big chance to die) :). It will be good for tactic statistic, but we must have artificial intelligence to analyze all possible options. But in any case its not a luck.
In my example you want your unit to be dead and if doesn't die on 30-40% terrain you tactical manoeuvre with 30*30*30...n of hits% probability fails that's sucks when you fail with such a low chance. In your example you just move your unit in key position with 100% probability.
igor wrote:I made one assumption, the statistics of luck depends on the force of impact. If you start entering additional dependencies proportion of luck in its pure form is reduced even more. Compliance with these statistics, to its name will become questionable.
so how do I count luck for dwarvish thunderer and loyalists spearman ranged attack if they try to shoot at fully healed/18hp/1hp orcish grunt at night standing on hills (50% defense)?

igor
Posts: 11
Joined: September 14th, 2010, 5:46 pm

Re: [interface] we need better statistic (my propose)

Post by igor »

igor wrote:
OK, I propose to show both form of luck statistic and will prepare the sample of calculations little later (for 2nd method was before)
So, "luck" in its purest form in attachment (tables 4.1 and 5.1) Method to know how often random number generator takes players sides (per turn).
Attachments
luck_stat2.xls.zip
(10.69 KiB) Downloaded 187 times

Post Reply