Standard Deviation in Statistics

[blackjack]

As a math graduate, I think it is extremely inelegant for the Statistics section to contain the expected values and the actual values for damage dealt and received... without also including the standard deviation.

I've thought about it, and it isn't difficult to do it... the question is, would players and devs find it useful? (personally, I think that the damage stats are pointless, but since it is there...)

If there is interest, I could take a look at the Wesnoth code over the weekend and suggest a patch. Unfortunately I don't know how to use the standard patching system but I think the code wouldn't be too long, perhaps I could just paste it here.

(I'm not sure how the interface aspects would work out, since the current display can get cluttered when you get into 5 digits)

[silene]

personally, I think that the damage stats are pointless, but since it is there...

The reason it was added was: some players were thinking Wesnoth is cheating. They believed their units never hit, and AI units always hit. There has been far less complaints since these statistics are here.

I don't know how to use the standard patching system

If you are able to write the code necessary for the game to compute and display standard deviation, generating a patch will be piece of cake in comparison.

As for the usefulness I don't see any; but maybe you will convince other developers of the interest of such a statistic.

[Circon]

Deviations would be vanishingly small after a short period of time. When 1000 damage is dealt (and the numbers mount up fast!), that's somewhere around 300 strokes of 6-7 damage on average. I passed my maths test with flying colors, but that was a year ago, so I don't have the rulebook in my head any more. Still, it looks like standards would be very close. Personally, I have never seen damage dealt differ from damage expected by anything over 1%, or 0.5% after 1000 damage.

[grochti]

I´ve seen more than 10% difference more than once. My hardest game started with damage done : 50 ; damage (exp.) : 320 . Damage Taken 350 ; damage (exp) 280.
Thats when i resigned

and my opponent told me something like "Dont be sad because i allways beat you"

[Popito]

I would find it nice!.
And worst is i find devs should find it even nicer! .

And its nice to be lucky sometimes, but when you lose a game only because luck, well you feel quite pissed. There are some critical moments in every game where armies meet and if one side gets all the luck the game is over as the loser survivors are too few or too weak to stand for the economic or stright battle.

I have seen lots of games where opponents got same luck but one having it at silly unimportant combats and the other having his in the main battle.

And damage its not so important as unit loss, you can stand damage as well as is evenly distributed among your guys, but if all your bad luck is owned by 3 unlucky dead guys then you are more in trouble. And its something we wont see in damage statistics .

Its nice you can say i won or i lost, and luck made most of it. Or find i was lucky and i lost and try to learn from the winer how has he done it .

I almost enjoy the same watching a game than playing it myself and getting more data will be welcomed .

Popi