Cricket Heroes - testing & feedback

[hedger_14]

I reckon rather than that, just add variation to the speed. E.g. Choose 130kph, there's a slight chance it will be just over 130kph and a chance it will be under and the less stamina you have, the higher the variation. I can try and come up with some complex sort of formula for it if you want later?

 

[kurtkz]

I'm going to agree with hedger on this - the player still needs some kind of input as to what speed he wants to bowl at.

@hedger: If you can draw up a formula (it can be fuzzy too, I don't mind) that'll be great. I'm still trying to implement the last batch of requests, so that'll be a big help. Currently I have variables like: stamina, min speed, max speed, spell #, ball condition, etc.

 

[hedger_14]

okay, working on it atm Anything else you want done while I'm at it?

----------

I'd use spell number as one of the variables but it would screw things up if you bowled players in one over spells.

 

[kurtkz]

Hmm nothing comes to mind. What about using average spell length instead of # of spells?

 

[hedger_14]

idk, can you just put that all into affecting stamina or something instead? I can try working out something for that later.

Also, how much would ball condition affect delivery speed? Like after 30 overs, would there be any difference in bowling speed for say a fresh bowler?

 

[kurtkz]

Hmm, I'm really not sure. I guess the condition only really affects speed off the pitch? Although, a scuffed up ball would have more air resistance...it's probably safe to just go with stamina as the affecting variable.

 

[hedger_14]

okay, my complex formula. I can edit it later maybe with other variables but this seems pretty decent for the time being.

Y = output bowling speed
X = input bowling speed
Z = stamina where 100% stamina = 10, 95% stamina = 9.5, 90% stamina = 9 etc.

Y = min{maxspeed, max{ [minspeed(* 1 when 5 < Z < 10),(* .92 when 2.5 < Z < 4.99),(* .88 when 1 < Z < 2.49(* .84 when Z < .99)], [ (ceil[Z]/10) / (min{1, (ceil[Z]/10 + (10-ceil[Z])/100)}) ] * X +/- (1.96 * 1.8) } }

sorry if that's too confusing with all the brackets. That should do a decent job. The C.I. is the +/- bit, so you will get the two extremes (a 95% C.I.) and pretty much a number has to be generated in between those numbers (Idk how to do that in a formula, lol)

----------

I forgot how quickly stamina depletes so you may need to make edits to your stamina formula

 

[kurtkz]

Lovely work I'll add that tonight.

 

[hedger_14]

and make sure you change the equality signs. Cause it was really going to be a pain in the ass to do it in the formula as it would have looked even more of a mess but yeah, it should be (5 <= Z <= 10), (2.5 <= Z < 5), (1 <= Z < 2.5), (Z < 1)

----------

Anything else you want done that won't take too long (it can take longer but I'll be finishing it in the morning if that's the case )

 

[kurtkz]

Nope, I think that's it for now, thanks - at least until I release the next update

 

[hedger_14]

tell me what your stamina model is atm as well, cause then I might see if I can improve that somewhat.

----------

Later though, going to bed now

 

[hedger_14]

Well it seems like England are playing for the draw. 0/2 after 5 overs...

----------

nvm, they score a four next ball... off an edge through the gap between third slip and gully.

----------

idk, but they're actually batting well now. They're 30/1, New Zealand referring an LBW decision originally given not out, which would have gone on to hit the top of middle, Compton the only dismissed batsman.

But they just look a lot better. Compton scored most of his runs through edges but apart from that, Cook has looked solid and Trott's looking alright. Big change from their first innings performance.

----------

okay, well Cook just played his first stupid shot of the innings, trying to work Boult to square leg, and got caught LBW, with the decision review allowing the original decision of out to stand, the ball clipping the top of leg-stump leaving the umpire's call decision.

 

[hedger_14]

Also, idk whether it would make putting in that formula any easier but for the part " (ceil[Z]/10) / (min{1, (ceil[Z]/10 + (10-ceil[Z])/100)}) ] * X +/- (1.96 * 1.8) "

it is a 95% C.I. where x-bar = (ceil[Z]/10) / (min{1, (ceil[Z]/10) / (min{1, (ceil[Z]/10 + (10-ceil[Z])/100)}) ] * X

alpha = a = 0.05

Z(a/2) = 1.96

And the SE mean = 1.8

Where the 95% C.I. = x-bar +/- Z(a/2) * SE mean
so you get the x-bar +/- 1.96 * 1.8

Don't know whether that makes anything easier for you.

 

[Ashu54]

Kurt,Waiting for your models eagerly......

 

[Ashu54]

OK the shots in which a combo of keys are req. dont work I mean the batsman plays forwrd def. stroke....

Also the square drive of ff is pretty useless.....