Russia 122 for 2 at Lunch

[dhump]

it's not as good as you think, because it's far too simplistic. skills must be relative to the conditions and opponent.

associate cricket for example, isn't all scores of 75 all out, which the low batting skilling would imply. the fact is a player can score highly against players of similar skill, but struggle against higher skilled players. similarly a bowler might be prolific in county cricket, but lack the yard of pace or extra movement/variety to succeed in international cricket. the game should be able to model that.

This is very true that not all associate cricket teams will fall for 75 all out, but chances are much higher that they would fall for something cheap. Visible difference is important and will make the game much more interesting (end of the day we have to understand its a computer game and A.I. can be limited so implementations POV should be kept in mind). I can understand what you mean and would love to play something so intelligent but sounds very difficult to get it anytime soon.

I think I always think more about career mode and that can be made into an amazing experience overall.

 

[cricket_online]

it's not as good as you think, because it's far too simplistic. skills must be relative to the conditions and opponent.

associate cricket for example, isn't all scores of 75 all out, which the low batting skilling would imply. the fact is a player can score highly against players of similar skill, but struggle against higher skilled players. similarly a bowler might be prolific in county cricket, but lack the yard of pace or extra movement/variety to succeed in international cricket. the game should be able to model that.

That's the yardstick for averages and not one-off scores, i.e. if a player with certain skill level plays for a long time against varied bowling across different conditions. I didn't say a batsman with 50% will always score 40. What I'm saying is that you would expect a batsman with 50% skill level to average 40 over a prolonged period of time. of course the batsman would likely score less in seaming conditions or against better bowlers but at the same time might score more on flat pitches and against lesser skilled bowlers.

 

[blockerdave]

This is very true that not all associate cricket teams will fall for 75 all out, but chances are much higher that they would fall for something cheap. Visible difference is important and will make the game much more interesting (end of the day we have to understand its a computer game and A.I. can be limited so implementations POV should be kept in mind). I can understand what you mean and would love to play something so intelligent but sounds very difficult to get it anytime soon.

I think I always think more about career mode and that can be made into an amazing experience overall.

no you've misunderstood, i'm saying that when associate plays associate...

conditions being equal, an associate batting lineup will fare better against an associate bowling attack than against a top-class test attack so it isn't as simple as saying "low rating = crap player"... it's rating in the context of the opponent.

It's also irrelevant in the context of AI or any form of intelligence: it's just maths and modelling the timing windows, edge probability etc. the game should be able to do that.

The simplest example I can give would be football manager - if I'm managing a conference side and have a winger with pace 11, he's likely to be quicker than most fullbacks he's gonna come up against. put him in the premiership, and he's slow. similarly, even Chris Martin could probably smash my bowling around, the problem is when he's facing a test player.

so my point was that the model you and @cricket_online were suggesting - and any model that takes the absolute skill level, rather than the skill level relative to opponent - was over simplified and wouldn't improve the game.

 

[cricket_online]

no you've misunderstood, i'm saying that when associate plays associate...

conditions being equal, an associate batting lineup will fare better against an associate bowling attack than against a top-class test attack so it isn't as simple as saying "low rating = crap player"... it's rating in the context of the opponent.

It's also irrelevant in the context of AI or any form of intelligence: it's just maths and modelling the timing windows, edge probability etc. the game should be able to do that.

The simplest example I can give would be football manager - if I'm managing a conference side and have a winger with pace 11, he's likely to be quicker than most fullbacks he's gonna come up against. put him in the premiership, and he's slow. similarly, even Chris Martin could probably smash my bowling around, the problem is when he's facing a test player.

so my point was that the model you and @cricket_online were suggesting - and any model that takes the absolute skill level, rather than the skill level relative to opponent - was over simplified and wouldn't improve the game.

I didn't suggest that. Read my previous post above.

 

[blockerdave]

That's the yardstick for averages and not one-off scores, i.e. if a player with certain skill level plays for a long time against varied bowling across different conditions. I didn't say a batsman with 50% will always score 40. What I'm saying is that you would expect a batsman with 50% skill level to average 40 over a prolonged period of time. of course the batsman would likely score less in seaming conditions or against better bowlers but at the same time might score more on flat pitches and against lesser skilled bowlers.

averages are irrelevant. my point is that the model cannot take into account the absolute skill level of the batsman, it needs to take into account the skill level of the batsman relative to the bowler. only a game engine that models that can get it right.[DOUBLEPOST=1444053770][/DOUBLEPOST]

I didn't suggest that. Read my previous post above.

well i disagree with the implications of your post and the model. only modelling outcome based on both skill levels works.

 

[blockerdave]

I didn't suggest that. Read my previous post above.

it's still wrong, utterly wrong. look at the "first class" averages of some associates, who play the majority of their first class cricket in the intercontinental cup, they would be implied to have a higher skill level on your model than is correct. whereas if they played more first class domestic cricket in the main nations, or played test matches, their average would be likely lower.

you cannot just look at the skill levels of the batsman, or bowler, in isolation; only plotting them relative to each other can give a proper result.

 

[cricket_online]

averages are irrelevant. my point is that the model cannot take into account the absolute skill level of the batsman, it needs to take into account the skill level of the batsman relative to the bowler. only a game engine that models that can get it right.

The model needs to be constrained within certain limits, i.e. averages. Yes any batsman's performance or results will depend on type of opposition (bowler's skill) and conditions (seaming, flat etc.) but over a period of time a certain type of batsman (say 50% skill level) should average around a particular range (say 40). Otherwise how do you even come up with the model if there's no benchmark (average) to reference?

 

[blockerdave]

The model needs to be constrained within certain limits, i.e. averages. Yes any batsman's performance or results will depend on type of opposition (bowler's skill) and conditions (seaming, flat etc.) but over a period of time a certain type of batsman (say 50% skill level) should average around a particular range (say 40). Otherwise how do you even come up with the model if there's no benchmark (average) to reference?

one more time, averages are not the answer. Mohammad Shazhad of Afghanistan averages 49 in first class cricket. under your model that gives him what was it, 70% skill?

averages are a function of plotting the relative skill of the batsman and the bowler.

the game should be able to model a situation in which a player may perform well at one level, and not at a higher level. it's not rocket science.[DOUBLEPOST=1444054458][/DOUBLEPOST]how would your model cope with someone like Mark Ramprakash, with a 26 run differential between his First class average of 53, and his test average of 27? Answer, it couldn't.

but a model that plotted skill relative to opponent, whilst it might struggle to encompass SUCH a large gap, can at least cope with the concept of different performance according to level of opponent.

 

[T.J.Hooker]

On this note i would love to see a twist mechanism that determines the angle of the bat blade...not sure how it could work but would add a new dimension to intuitive play...

Well, the way I'm imagining a target system working is that you left click to set the space in the air you're targeting - so you're setting a point through which you're aiming to swing the middle of the bat - and then you swing the mouse, with the shape of your swing determining certain characteristics of the shot.

So I guess you could control the face of the bat like that. So you bend your mouse stroke towards cover or cover point just before the moment of contact, for instance, to open the face.

 

[blockerdave]

Well, the way I'm imagining a target system working is that you left click to set the space in the air you're targeting - so you're setting a point through which you're aiming to swing the middle of the bat - and then you swing the mouse, with the shape of your swing determining certain characteristics of the shot.

So I guess you could control the face of the bat like that. So you bend your mouse stroke towards cover or cover point just before the moment of contact, for instance, to open the face.

Mouse control? That's it - release date confirmed for start of the cricket season, 2008.

 

[T.J.Hooker]

Well, the reason I'd go for mouse is that it's the analogue device with the most precision. I don't think an analogue stick would give you enough control to actually aim at the ball properly, and clearly the larger area of a mouse swing gives you potentially a much more precise and controllable swing.

Analogue sticks are too fiddly to map shot power to the swing, for instance, but a mouse swing might be able to manage it.

 

[cricket_online]

one more time, averages are not the answer. Mohammad Shazhad of Afghanistan averages 49 in first class cricket. under your model that gives him what was it, 70% skill?

averages are a function of plotting the relative skill of the batsman and the bowler.

the game should be able to model a situation in which a player may perform well at one level, and not at a higher level. it's not rocket science.[DOUBLEPOST=1444054458][/DOUBLEPOST]how would your model cope with someone like Mark Ramprakash, with a 26 run differential between his First class average of 53, and his test average of 27? Answer, it couldn't.

but a model that plotted skill relative to opponent, whilst it might struggle to encompass SUCH a large gap, can at least cope with the concept of different performance according to level of opponent.

The averages I mentioned in relation to skill bars was to model the players appropriately. I wouldn't give Mohammed Shahzad a rating of 70% of skill bar and give him a rating of more like 40% skill bar. The skill bar is a way to create players across varying skill sets - right from Chris Martin to Don Bradman - and that's about it.

To put it another way, the skill bar depicts what a batsman should average over a lengthy period of time if the batsman faces a 3 Star bowling on a particular kind of pitch (say regular pitch with just a tinge of green and no cracks). Thus Bradman will have the bar full, Viv Richards will have the bar 80% full, Shahzad in your example will have the bar 30-40% full and chris Martin will have it maybe 5% full or lower. Thus Shahzad can continue averaging 49 against 1 star bowling (associates) but will struggle at intl level.

And I reiterate you need to have constraints and benchmarks (averages) to come up with a model else how exactly will you come up with a model? How exactly will you specify how good or bad a batsman is without the same?

 

[grkrama]

How would you showcase a batsman like say garry ballance or pujara who have healthy averages but would struggle against pacy attacks on seaming conditions ?


I dont think averages directly related to skills is the answer. Cause that will lead to tailender kind of phenomenon we have with DBC14 now were the team averages a score.

I would rather say relative averages to particular bowler type and/or say chance of getting out / scoring against different type of bowlers along with comparison to their skill level should be constructed. Rather than by going say he has 50 % skill let him average 40 for the series.

 

[Markkkkk]

All we would need is a modifier like in Cricket Coach, where you can increase a persons abilities in certain formats and in that case against certain bowling, also having a weakness like pace bowling would help aswell...

 

[Gamer Pradosh]

How would you showcase a batsman like say garry ballance or pujara who have healthy averages but would struggle against pacy attacks on seaming conditions ?


I dont think averages directly related to skills is the answer. Cause that will lead to tailender kind of phenomenon we have with DBC14 now were the team averages a score.

I would rather say relative averages to particular bowler type and/or say chance of getting out / scoring against different type of bowlers along with comparison to their skill level should be constructed. Rather than by going say he has 50 % skill let him average 40 for the series.

These discussions are great for the game but will these factors be accepted by the gamer point of view...If one batsman is batting really well say in his 80s at a great strike rate and the bowler against who he is weak comes to bowl and takes his wicket, what are the factors come in for his dismissal??

If the weakness factor dominates in that position it is just not acceptable even for a gamer like me when the scoring is smooth..We all need that "why did I play that" kind of a feel even on getting dismissed when playing well...
We need a setup where the gamer feels "that was bad" "Beauty of a delivery" kind of stuff because on the whole its a game that reaches every type of gamer...The discussions on the way the game should be is really good for hardcore gamers like us but that is very low percentage when compared to that casual ones...So my view on this is that we should take the game with an approach where even the casual gamers or gamers who have not played cricket to understand the concept ,the strategies ,the plans, the execution and even rotation of bowlers for effectiveness , if the game is able to exhibit that then we get a group or fans of the game who would get better with the concept of cricket...If next version does that then we can aim for these deep analysis put into the game when many are familiar with what the game actually wants us to do...

Just my view...But a game with the discussions above being made would be awesome for hardcore cricket fans like us...