Nirvanam Posted December 15, 2010 Posted December 15, 2010 Guys, pls help make it more accurate... I am trying to build a function map for batsmanship (to derive metrics for virtuoso analysis later in the season). I've tried to imagine it in the universal equation, Y -> F(x). So first level is: Virtuoso -> Function (Skill Mastery, Innovativeness, Reliability). At second level, Reliability -> Function(...) and so on. Pls help make it better... ps: how does one show the image in the message itself?
Nirvanam Posted December 15, 2010 Author Posted December 15, 2010 Thanks Tdigi, I used the image tag itself but still it showed only the link and not the image.
Nova Posted December 15, 2010 Posted December 15, 2010 Thanks Tdigi' date=' I used the image tag itself but still it showed only the link and not the image.[/quote'] This is how it is: wtf mate :hysterical: P.S. Nirvanam.. not laughing at you so don't cast a spell on my offsprings. :P
TumTum Posted December 15, 2010 Posted December 15, 2010 In this case it would be y = F(a,b,c,d,e,f,g etc) as you take into account many variables right? The toughest thing will be how to accurately weight every category.
Nirvanam Posted December 15, 2010 Author Posted December 15, 2010 This is how it is: :hysterical: P.S. Nirvanam.. not laughing at you so don't cast a spell on my offsprings. :PLOL! nah dont worry! Hey I did the same thing img-link-/img still it dint
Nirvanam Posted December 15, 2010 Author Posted December 15, 2010 In this case it would be y = F(a,b,c,d,e,f,g etc) as you take into account many variables right? The toughest thing will be how to accurately weight every category.Ya...at top level Virtuoso = F(Skill Mastery, Innovativeness, Reliability) next level Skill Mastery = F(sm.oppositions, sm.conditions, sm.situations), Innovativeness = F(performance in atypical contexts, performance vis-a-vis peers) and Reliability = f(Consistency, Contribution to team) So from every level you make the x1, x2, x3 as y1, y2, y3 and find out its new x1.1, x1.2, x1.3 until you can get to the granularity where we can figure out a metric. So, first step is to finish the map...then we can determine relevant metrics. As for the weighting...that'll come later...once we have determined the metrics. There are already 6 proper metrics. But trying to improve them...so I thought let's build the proper structured breakdown first....
Jersey #10 Posted December 15, 2010 Posted December 15, 2010 LOL! nah dont worry! Hey I did the same thing img-link-/img still it dint your link does not end with .gif or any other image file extension. What you posted is a link to the web page called picasa which has the image saved on it. The real image link looks like this http://lh4.ggpht.com/_m32z_4R1LBY/TQgG-OKDQkI/AAAAAAAAAMw/KLVA-23wyMk/s720/Virtuoso%20Function%20Map.jpg it will show up as an image in your post. Good luck next time.
Nirvanam Posted December 15, 2010 Author Posted December 15, 2010 your link does not end with .gif or any other image file extension. What you posted is a link to the web page called picasa which has the image saved on it. The real image link looks like this http://lh4.ggpht.com/_m32z_4R1LBY/TQgG-OKDQkI/AAAAAAAAAMw/KLVA-23wyMk/s720/Virtuoso%20Function%20Map.jpg <-- see how this ends with ".jpg"? now if you put this direct location of the link in it will show up as an image in your post. Good luck next time.Cool...will do that next time. Thanks!
Nova Posted December 15, 2010 Posted December 15, 2010 What is the purpose of this? Are you trying to build an analysis software, patent it and then sell it?
coffee_rules Posted December 15, 2010 Posted December 15, 2010 Why ? So that Tendulkar can be compared to Bradman?
Jadoo Posted December 15, 2010 Posted December 15, 2010 Nirvanam, what is this bro?? I saw the threads in the nerd section about how cats, dogs, and trees are illusions and all, put please don't bring this y=illusion onto the cricket section, some of us mere mortals work in finance, health etc, and don't want nor need to see this on the cricket section. Thanks very much, I know you are a very understanding person.
EnterTheVoid Posted December 15, 2010 Posted December 15, 2010 Nirvanam, what is this bro?? I saw the threads in the nerd section about how cats, dogs, and trees are illusions and all, put please don't bring this y=illusion onto the cricket section, some of us mere mortals work in finance, health etc, and don't want nor need to see this on the cricket section. Thanks very much, I know you are a very understanding person. :hysterical:
akshayxyz Posted December 15, 2010 Posted December 15, 2010 I suggest you to do a reading on Linear Algebra. As I commented on some other post of yours, yes you can have a Function with multiple variables, but the weights are subjected to perception. Best you can use this mapping of batsmen in multi-dimensional space is to find batsmen who are similar to others, or build a prediction engine that can predict to some certainty how a given batsmen is going to perform in given condition. For this usage, I suggest reading on machine learning and recommendation engines. To repeat, you will be able to define coordinates of each batsmen in your N-dimensional space, however which vector is better than other will still be subjected to perception - Visualize it like this - in a simpler 2-D space - lets say there are three points, (40,20) - (20,40) and (30,30) - the concept of "greater/better" point is highly subjective. Still, one way to achieve "universally accepted weights" would be to ask multiple "experts", but there lies even bigger problem, "experts" will change their opinion, based on the outcome - they would want to see their favorites on the top. So you will have second level problem, which experts' opinion should weigh more than others' :). As I said before - the best use,IMO, of this mapping is in field of "batsmen similarity"/ "recommendation" - finding similar batsmen, trying to predict "performance" of players in certain conditions and hence an Artifical -intelligent team selector!!! - but then, there is huge uncertainty factor (in other words "there are far too many variables than we might be able to put in a mathematical model and then be able to execute with today's computing power") - e.g. Based purely on stats Sehwag may be a bad choice against Bangladesh - but would you rule him out, if he scores @100 Avg against SA in a preceding series? Anyway - :thumbsup: for your enthusiasm towards maths - keep going, you will figure out the way to solve your problem(s) - or get a good job in field of AI/Machine learning. "Only mathematics is capable of any sort of magic" - me :).
akshayxyz Posted December 15, 2010 Posted December 15, 2010 BTW, IMO, having nesting variables is un-necessary, only usability is "modularity" of your implementation or parallel execution of independent sub-functions. Any F(g(a,b),h(c,d)) can be further simplified to - F(a,b,c,d). Listen to others' sentiments too, this is not the place to have quality discussion for your this endeavor of yours.
Nirvanam Posted December 15, 2010 Author Posted December 15, 2010 Well, the effort is to improve the Test Cricket's Virtuoso Batsman analysis that I did end of last season. Last time around the metrics/parameters were identified without any explicitly structured way of thinking although I feel the parameters address most important criteria to evaluate a player. But this time I wanna improve it and just put some structure to that thought so that there is enough reason to choose the parameters, and reduced duplication as much as possible. Nothing to do with making any software or anything. Just out of pure interest to verify my own hypothesis that the batsmanship of Sachin, Lara, and Bradman (although last time he was given default virtuoso status) is in a different league when compared to other ATG batsmen. Hence we can call them as virtuosos. My own analysis showed I was wrong...well not wholly though...it showed last season that Gavaskar also is in the group of virtuosos in addition to the three. Akshay, the idea is not build a mathematical model like Virtuoso = 1.9* Skill + 3.4*Innovation-square + 12*Reliability. It is only to build a conceptual equation...so it will remain it English -> function(English, english...) and not become mathematical here. From the conceptual equation I'll see what metrics address would address which x, etc. First want to be sure that the conceptual equation seems accurate. So, folks pls help improve it
Nirvanam Posted December 15, 2010 Author Posted December 15, 2010 Nirvanam, what is this bro?? I saw the threads in the nerd section about how cats, dogs, and trees are illusions and all, put please don't bring this y=illusion onto the cricket section, some of us mere mortals work in finance, health etc, and don't want nor need to see this on the cricket section. Thanks very much, I know you are a very understanding person.LOL! Sorry yaar, just interested in taking your inputs for that batsman analysis I did last time...want to improve it.
Nirvanam Posted December 16, 2010 Author Posted December 16, 2010 OK, here are the metrics/parameters for the function map -
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now