Hello all. I am a 3rd year software engineering student. For my final year project, I am proposing to develop a software that would help users to identify (or recommend) the most suitable racquet based on the build and playing styles of the user.

I have been following this forum for more than a year now ever since I had my own dilemma on choosing the most suitable racquet for myself. I have rough knowledge about a racquet's build; mainly about weight, balance point, flex, string tension and type, and grip size and type.

My question is, is it possible to recommend a racquet (or the best racquet build) based on numerical inputs from the user such as arm strength (probably in Newtons), wrist strength, swing speed (meters per second), etc.. Other non-numerical factors will also be considered such as playing style in the racquet determination.

I would really appreciate any insights received :)

I think its a pretty good idea.

Software wise it could be pretty good, but we would need some hardware as i couldn't tell you what my wrist strength was in newtons.

There would be a lot of factors you'ed have to consider, i know there is software out there to help you choose golf clubs.
Some variables i can think of right now are: Female/Male, weight, height, strength for sure(not sure how you would measure this) The length of your arms and legs. For rackets, there would be a lot of different things. Weight, grip thickness(+grip weight), strings(thickness, weight, and tension) Racket types, headheavy, balanced, and light. Racket flex also.

It sounds like a good idea, but there are problems.

A lot of the kind of numerical data you are talking about is very hard to measure.
For example, you need to quantify a player's technique (how "wristy" they are). I'm guessing you would compare the rate of forearm Vs shoulder rotation. I can't think of a way to calculate this without the use of sophisticated high-speed cameras, which very few people have access to.

I can't imagine recreational players going to such lengths in order to find a suitable racket anyway, and most more experienced players already have a pretty good idea of the kind of racket that suits them.

Also, I'm not sure you can accurately predict a player's preference based on numerical measurements. Take grip sizes - there's a thread here somewhere (search) telling you how to choose a grip size, based on measurements of your hand. However, a lot of people disagreed with these recommedations.

I think you really need to use far more qualititative data, of the kind a buyer can give off the top of their head (eg. Answers on a scale of 1-10).
Now the problem is you could easily construct a simple table to direct buyers to the correct combination of balance/weight/flex - making the software somewhat redundant.

You'd really need to make a database of actual racket models for it to be useful (and make it searchable by weight/balance/flex so it's useful to people who already know more or less what they want too).

It would be great if you could do that, but it would require a huge amount of research and data entry - especially if you take into account all the different combinations of weights and grips available in different regions.

Good luck though - I agree that the selection of rackets on the market is daunting.

I think it is a nice idea :) Wow sketchy , that is true.

I think finding the perfect racket would take a lot of research etc.

But it would be helpfull to badminton players.

What sketchy said was quite important to me.

Like how do you calculate someones technique :D.

You can measure wrist strenth with a test called the grip test , but I can get higher in the grip test than some people , but people with lower result in the grip test , with better technique could have more power.

So the effieciency of strokes etc.

Nice idea though!

Hello all. I am a 3rd year software engineering student. For my final year project, I am proposing to develop a software that would help users to identify (or recommend) the most suitable racquet based on the build and playing styles of the user.

I have been following this forum for more than a year now ever since I had my own dilemma on choosing the most suitable racquet for myself. I have rough knowledge about a racquet's build; mainly about weight, balance point, flex, string tension and type, and grip size and type.

My question is, is it possible to recommend a racquet (or the best racquet build) based on numerical inputs from the user such as arm strength (probably in Newtons), wrist strength, swing speed (meters per second), etc.. Other non-numerical factors will also be considered such as playing style in the racquet determination.

I would really appreciate any insights received :)just curious, are in in an engineering program or a computer science program? A real engineering student should know more than a 'rough' knowledge of racket physics and dynamics:D

To all those saying "how do you calculate technique" He specifically asked for "numerical inputs"

just curious, are in in an engineering program or a computer science program? A real engineering student should know more than a 'rough' knowledge of racket physics and dynamics:D

I think he clarified that when he said software engineer.

Great idea!!

Only one major flaw, a perfect racket today may not be the perfect racket tmrw..

non-quantifiable factors such as mood,desire and form will come into the picture. Not to mention newer racket models thats just been launched esp if it has received glowing reviews..

To all those saying "how do you calculate technique" He specifically asked for "numerical inputs"

Yes but we were trying to say a different point.

Say if one person gets 50 in wrist strength and he puts it into the system , although this person might have realy bad technique , causing the player to loose effiecnecy thus swing speed . So maybe the player wouldn't get the racket they wanted because the racket calculated would be too stiff.
If you get the idea.

Great idea!!

Only one major flaw, a perfect racket today may not be the perfect racket tmrw..

non-quantifiable factors such as mood,desire and form will come into the picture. Not to mention newer racket models thats just been launched esp if it has received glowing reviews..

This is only if you wanted to be majorly exact , but each individual racket would need to be measured ,to find the "ideal" racket. Because they vary between 2 grams :p:cool::eek:

Thank you all. Just as mentioned all your insights are much appreciated.

The System
To clarify things a bit, I shall share with you the scope of my project. I believe many would agree that choosing a racquet is really a subjective matter, and there is no right or wrong in any argument presented by each individual. The main goal for my system is to suggest/recommend racquets that best fits the needs of the user. The system should be able to narrow down to the 4 or less (just an estimation) racquets that best fits the user. It not a system that specifies one particular racquet that would definitely be the perfect racquet for the user. And it is not by any means a system to replace a job of a racquet experts. It’s rather a tool that can assist experts or staff in choosing a racquet for a customer (although it can be used by the customer him/herself). Think of it like a doctor’s diagnostic software for narrowing down the type of illness a patient is suffering based on a patient’s symptoms. It’s a recommendation system, not an assignment system.


Measuring Technique
However, I do want a solid method to accurately (or as effectively as possible) determine the best racquet for the user. My assumption is that this could be done by using quantitative readings such as maximum force that can be exerted by the user’s arm, speed generated by the wrist, etc. It is the reason of my first post in the first place: to enquire if it is possible to use this measuring technique. After reading through this tread, I believe it is now out of the scope of my project and my course discipline as this is related more towards the study of physics than the actual software development process itself. But, I do believe racquet manufacturer’s Research and Development department as such Yonex would have already done this (research and testing) in producing racquets we have today. So if there’s anyone that could help me to obtain the results of this research, it would definitely solve all my enquiries. It’s doubtful, but I’m hopeful.


Racquet Database
Regarding the database of actual racquets, that should not be a problem at all. My project requirements do specify the use of a database to be integrated into my system. The database will be used for the racquet determination and also for searching function for users to search racquets by weight/balance/flex.. etc. All I need to do is to “upload” all the racquet and its build details into the database. To narrow the scope a little further, my database would consist of racquets by Yonex and Yonex alone. Only racquets from the Armortec, NanoSpeed, and Arcsaber series will be used.


Users
Lastly, regarding the users of my system. The category of users I’m targeting my system for are novice users that have no or little knowledge about badminton or racquets. According to you Sketchy, you mention that it would be easier to construct a table direct buyers to the correct racquet build. Do you mean something like this Yonex selection chart?


http://farm4.static.flickr.com/3608/3548396749_df02172368.jpg?v=0


To me, the chart is fine as I already have background knowledge on how the build of the racquet is associated with its placing in the chart. But the problem is, most novice users would be left half guessing when selecting from this chart. If I’m not mistaken, there’s even a sticky thread devoted entirely to the discussion of this chart. I assume that even with this chart, it is still very ambiguous when trying to select a racquet.


Current problem
My problem still lies in the factors in determining a suitable racquet. I have now 3 methods in mind.

1. My initial proposal using quantitative data (generated force, speed, etc)

2. As mentioned by Sketchy, by using qualitative data (answers scale of 1-10)

3. Or using an even far more qualitative data (eg. Playing style – offensive, defensive, all-rounder; Playing time – Beginner, Intermediate, Advanced, etc)

Any ideas?

Sorry, I didn't really explain at all, but what I was thinking is really something more like the kind of keys you use to identify a species.

http://en.wikipedia.org/wiki/Dichotomous_key

They just happen to be presented in a kind of a table-like diagram quite often, so they look clearer / more appealling.

Eg. For ski bindings:
http://www.dinsetting.com/images/MarkerDinChart0708_sm.jpg

The first question might be something like "Do you play mainly singles or doubles?", and depending on your answer, it would rule out some rackets, and then following questions would further narrow down the list of possibilities. Eventually, you'd get left with one or two to choose from.

Anyway, I think you could come up with some kind of easy-to-follow, printable document that people could use.


As far as actual measurements go, I know my height and weight, and that's all (and all I have a means of finding out). I guess regular gym-goers know what they can bench etc, but probably not much that's going to be very relevant to badminton. And, like I said before, there are important factors such as technique, which are virtually impossible to quantify.
Personally, I think the ultra-vague system is probably a better way to go.


I'm not sure you can afford to restrict yourself to just novices / people who don't know much about badminton.
As far as I'm concerned, that would rule out all the high-end rackets, as they're just far too expensive to be worth a novice buying, and not even suitable, due to their stiffness. Actually, I personally would never recommend any Yonex racket to a novice, as you can do so much better for the money with other brands. That's why I think you should include at least a couple of other brands aswell. I guess if your system was hypothetically going to be on the Yonex website or something, then it would make sense to have only their rackets.


Cost is actually quite a big issue. I think a lot of people tend to choose a range (ARC / AT / NS) based on what that rackets are like, but the individual model based on price.
Eg. If they decide that they want a defensive racket, they will go with the Nanospeed series, and then pick the most expensive model within their budget.
You're always going to have a hard time convincing people that a cheaper racket is "better" than a more expensive one, and they'll always want the best they can afford.

Looking at that chart you posted, it's not really that ambiguous at all, once you rule out those rackets that are too expensive, and those that are too cheap (and therefore not as "good"). The only ambiguous parts are the mysterious "Repulsion" and "Hold" terms. I can only assume they don't translate well from Japanese.
If it were me, I'd have "Swingweight" (derived from weight and balance) on one axis, and "Stiffness" on the other. That basically tells people everything they need to know.

Cost is actually quite a big issue. I think a lot of people tend to choose a range (ARC / AT / NS) based on what that rackets are like, but the individual model based on price.
Eg. If they decide that they want a defensive racket, they will go with the Nanospeed series, and then pick the most expensive model within their budget.
You're always going to have a hard time convincing people that a cheaper racket is "better" than a more expensive one, and they'll always want the best they can afford.



Good point, perhaps the system should recomend rackets of similar characteristics of different price ranges.

As for calculating technique, you could formulate a system that approximates the force generated of a persons height/weight. Of course the system wouldn't be 100% accurate but it could serve as a guideline. Regardless no two people are going to be exactly alike, but it is possible to determine the average swingspeed for a person of X-height/weight. Whether these variables are relevant in determining what racket a person should use is another thing entirely.

Why not do some research with the Bio-Physics department, I'm sure you can find something there. I believe there is a formula which you can input a human's weight and height and estimate the force that he or she can generate.

Well, I could do research on Bio-Physics and/or perform physical tests on a group of badminton players to gather qualitative and quantitative data which could then be analyzed to build an effective means of determine the best racquet for a specific group of players. . But as I said, it goes outside the scope of my course discipline which is software engineering. I wouldn’t be credited for this finding even if I performed and documented this research it.



I would have to agree with Sketchy. Perhaps the best way to go is to develop a vague system for now. The system wouldn’t be as fancy as I initially planned it to be by using quantitative statistics (force, speed, etc) of an individual. And again, I stand corrected by Sketchy. I should widen the targeted audience of my system to not just novice players, but also to intermediate and semi pros. My proposed system would help them reduce the hassle of searching high and wide to determine racquets that suit their playing needs by offering vague recommendations. For professional players, perhaps the system would act as an information terminal rather than a recommendation system.



As for the racquet brands, I will still be sticking with just Yonex products. I have made my assumption that my system will be installed in an official Yonex store. Perhaps I will include other products such as strings, shoes, and shuttlecocks as well. But the main focus would still be on racquets.



So for now, I’m proposing a vague system. I’m going for qualitative data (inputted by user) as previously mentioned. The factors I’m considering are

1.Game discipline (Singles or Doubles)
2.Level of play (Beginner, Intermediate, Advanced)
3.Playing style (Offensive, Defensive, All-Rounder)
4.Budget


Are there any other factors I should consider? Is this practical? Will this system be useful to you if you were put in the shoes of my targeted audience? If no, what are some of the ways I can improve my system? Your honesty will help me a great lot.

Take a look at the Prince racket selector (http://princetennis.com/tennis/racquetselector/racquetSelector.aspx?cid=36&rid=1). It's for tennis but it might give you some more ideas. It factors in how much power you generate by asking you what your swing is like - maybe you could have a question asking the user whether they think they have a fast, moderate or slow swing. That would help you decide what stiffness they would be looking for.

Gender is obviously a *very* important factor you missed. In mixed doubles, the woman has completely different requirements from the man. Women generally prefer lighter, more flexible rackets, and men generally prefer rackets that aren't pink (eg. Arc9).

You may also want to take it into account if the player is particularly young (<15?) or old (>50?).

Also, I agree with everything Dan says. You do need to ask about swing type (speed, length, amount of wrist). Animations, like in the Prince racket selector, would be *really* helpful here.

That should be enough to make a fairly reliable recommendation. If you can make it visually appealing, like the Prince one, then so much the better.

Gender is obviously a *very* important factor you missed. In mixed doubles, the woman has completely different requirements from the man. Women generally prefer lighter, more flexible rackets, and men generally prefer rackets that aren't pink (eg. Arc9).

You may also want to take it into account if the player is particularly young (<15?) or old (>50?).

Also, I agree with everything Dan says. You do need to ask about swing type (speed, length, amount of wrist). Animations, like in the Prince racket selector, would be *really* helpful here.

That should be enough to make a fairly reliable recommendation. If you can make it visually appealing, like the Prince one, then so much the better.

The gender idea is a very good one that I missed as well as the OP. Age could be another important one but then I do know some people who are 11-13 and still hit very well, probably better and maybe even harder than beginner level adults. It's the same with older people, think of how many badminton coaches and other advanced players that are more senior (although I guess advanced players aren't really the target for this application). Perhaps if the user's age was below a particular age (maybe 15, as you suggested, Sketchy) or over another age, you could ask them another question about what level they played at. If that level was less than advanced, you'd have to look at rackets that were easier to use than the "normal adult" rackets of the same skill level. That wasn't particularly clear but hopefully you get the idea.

Being visually appealing and easy is so important in all software these days but especially software like this that may be running on a terminal in a big sports shop like Decathlon or something. Make it look good and also interesting so people actually want to use it. Also, make it easy and possibly suitable for touchscreens so it could be used in sports shops and by people who may not be too good with computers.

Another one you might want to consider is whether the person has any underlying injuries. For example, tennis elbow sufferers, even if they are advanced players will certainly not want an extra stiff racket.

Also, once you've decided how it's going to work, if you want help categorising the vast lists of rackets you might want to add in to it, I'd be happy to give you some assistance.

My parameter for choosing a racket are as follow and hope it might be helpful to you
A)cost
B)arm strength(an injury will reduce this overall affecting certain stroke only...)
C)swing speed( ...but can still produce a very fast swing if technique is correctly used with swing type)
D)swing type(long/moderate/short)
E)playing style(smash fanatics?)

Once the racket are shortlisted, additional info like F)overall stiffness, G)racket feel, H)double/single racket, I)racket characteristic such as thin frame for very fast maneuverability, high torsion, could be added in as comment for ease of choosing. Wow so many factors!:D:D:D

Hi all. I'm back. After dealing with my other course assignments and confirming with my project supervisor that this racquet recommendation project is feasible , I can finally focus attention back to this project.

I have a couple of things in mind today. Firstly, thanks to the ideas you guys have provided, I am able to determined the parameters of users to be inputted into this system to provide racquet recommendations. They are:

User details
1.Gender (Male or Female)
2.Game discipline (Singles or Doubles or Mixed Doubles)
3.Level of play (Casual, Intermediate, Advanced)
4.Playing style (Offensive, Defensive, All-Rounder)
5.Palm size (small, medium, large)
6.Budget

I understand many have said to include swing speed as one of the parameters for the recommendation process. I am unsure of this. Can I make an assumption that swing speed is proportional to the level of play of the user? I assume that a more experience player has better technique and physique to be able produce a higher swing speed. Is this practical? Someone please enlighten me on this.

Anyhow, based on these parameters, my system would will output the most suitable racquet. The racquet details that will be displayed:

Output of Recommended Racquet Details
1. Racquet series
2. Racquet model
3. Flex
4. Weight
5. Balance
6. Grip size
7. Design/colour
8. String type and tension
9. Price
10. Other available variant of the same racquet model (different weight, grip size, colour, or edition)

Also, I would allow the users to browse/search for other racquets with a specific build preference. This is to cater for users that do not wish to use the recommendation method as above to search for a racquet, but would rather browse for racquets or racquet builds that are already in mind. The search parameters would contain:

Racquet Preferences
1. Flex (Medium, Stiff, Extra Stiff)
2. Weight (2U, 3U , 4U)
5. Balance (Head heavy, head light, even balanced)
4. Grip size (G3, G4, G5)
5. String emphasis (durability, repulsion, hitting sound)
6. Design/colour
7. Cost

Also, please note that I will only be involved in producing recommendation for the Yonex Arcsaber, Armortec, and Nanospeed series to reduce my database scope.

Please do feel free to post any comments or constructive criticism that can assist in my project development :)

You have missed out on the most important thing in a badminton racquet, which are the materials used to make the frame and shaft and the aerodynamics of the "weapon". Otherwise any wood, aluminum, or even steel racquets of any dimensions, provided they are lightweight, will do instead of carbon graphite.
The two most important factors in developing or designing a badminton racquet are quality of the raw materials, with materials of high Youngs Modulus being the best, compatible with acceptable tensile strength, and the design of a racquet frame/shaft with the least drag or air-resistance in all the directions of the swing, which can start from the frame edge-wise and then face-wise. Both are to some extent somewhat contradictory to each other.

You have missed out on the most important thing in a badminton racquet, which are the materials used to make the frame and shaft and the aerodynamics of the "weapon". Otherwise any wood, aluminum, or even steel racquets of any dimensions, provided they are lightweight, will do instead of carbon graphite.
The two most important factors in developing or designing a badminton racquet are quality of the raw materials, with materials of high Youngs Modulus being the best, compatible with acceptable tensile strength, and the design of a racquet frame/shaft with the least drag or air-resistance in all the directions of the swing, which can start from the frame edge-wise and then face-wise. Both are to some extent somewhat contradictory to each other.

thanks taneepak. materials didnt occur to me as i was focusing too much on the attributes of the player than the racquet itself. I will add in materials for the frame and shaft as one of the search/browse parameters.

I have created a questionnaire to help gather requirements for my project as well as to justify the need of my system. Calling all badminton fans here, to help fill in this questionnaire for me. Your input is very much appreciated :o

http://FreeOnlineSurveys.com/rendersurvey.asp?sid=4fwaudsecopdcz9638089

I said gender first!!!! I demand being called awesome!!!

I said gender first!!!! I demand being called awesome!!!


Haha. And you're nothing less than that chewablemorphin. If you would like to be further credited for your contribution, pm me your name (real name preferably), your area of expertise (or job position) and the idea you have contributed and I shall include it in my acknowledgment section of my project documentation. This applies for everyone that has posted in this thread :)

And please do not forget to fill in my questionnaire if you guys haven't done so!

http://freeonlinesurveys.com/rendersurvey.asp?sid=4fwaudsecopdcz9638089

I am now moving towards the designing stage of my project. I have created a set of interview questions that was initially made for racquet store staffs / experts. But i guess it wouldn't hurt to get opinions from you guys here. It would be preferable if all questions are answered but I wouldn't refrain anyone from answering any questions individually. Just ensure that you copy the question on top of your answer so that I know which question you're answering. Do feel free to express your opinions ;)


1.Can you provide a brief description about yourself? Mainly about your job description (or area of knowledge).

2.A racquet build consist of flex, balance point, weight, grip, and its string. Are there any other build details that make a racquet model unique from one another?

3.From the build details mentioned in the previous question, how would you recommend a suitable racquet for a badminton player based on these racquet builds?

4.What are some of the problems that your customers face when selecting and purchasing a badminton racquet?

5.What are the most common criteria do customers look for when purchasing a badminton racquet? (e.g. price, design, technology)

6.An expert software system would provide an unbiased and objective recommendation to racquet customers as opposed to human recommendation. Is it feasible (practical) to have such a computerized software system for racquet recommendation in your racquet store? If no, what can be done to make it feasible?

7. Are there any personal comments you would like to add before closing this interview session?

I guess I will start.


1.Can you provide a brief description about yourself? Mainly about your job description (or area of knowledge).



I am a nineteen year old male student currently attending college. I've played badminton since I was five years old. I work two jobs(you don't need to know). I might be selling yonex equipment on the side soon.

2.A racquet build consist of flex, balance point, weight, grip, and its string. Are there any other build details that make a racquet model unique from one another?

The length of the racket, not sure if that goes with build. Although most rackets are approximately the same length. Usually longer: offensive players. Shorter: defensive players.

3.From the build details mentioned in the previous question, how would you recommend a suitable racquet for a badminton player based on these racquet builds?

Depends on the player. If they are just starting out, I'd probably recomend something light and flexi, not using the other details.
Not done lab is closing...... will finish later.

I think he clarified that when he said software engineer.he can't be an engineer because:
1. he hasn't yet graduated
2. he taking courses that has no engineering certification or an engineer degree upon graduation.


http://forums.whirlpool.net.au/forum-replies-archive.cfm/703549.html

this exercise is as futile as those tests trying to match up a person's characteristic to a specific career/job. Assigning numerical values to specific human behavior and characteristic is a fantasy. Oh, BTW, another small parameter of caution to consider,

--> human behavior & characteristic, mentally & physically, will change with time <--

LOL

I am a final year computing student majoring in software engineering.

To cooler, please read through the posts in this thread carefully. I have already discontinued the proposal of using specific numerical inputs as possible parameters.

And I have already mentioned I'm designing a racquet recommendation software.

human behavior & characteristic, mentally & physically, will change with time

... but a racquet model's build specifications do not.

he can't be an engineer because:
1. he hasn't yet graduated
2. he taking courses that has no engineering certification or an engineer degree upon graduation.


http://forums.whirlpool.net.au/forum-replies-archive.cfm/703549.html
So you know for a fact he is not taking an engineering course, just because some article on another forum says a lot of people say they are software engineers?

4.What are some of the problems that your customers face when selecting and purchasing a badminton racquet?



Price price price price price price PRICE!!!

5.What are the most common criteria do customers look for when purchasing a badminton racquet? (e.g. price, design, technology)


Light but durable.

6.An expert software system would provide an unbiased and objective recommendation to racquet customers as opposed to human recommendation. Is it feasible (practical) to have such a computerized software system for racquet recommendation in your racquet store? If no, what can be done to make it feasible?

It would certainly narrow it down, it would be better for the customers.

7. Are there any personal comments you would like to add before closing this interview session?



I get a free copy of this right?

Another thing I just thought of, would there be a way for dealers to add rackets to the system? Would come in handy for dealers who sell a lot of brands. Kind of like the golf course gps systems, adding courses....

Thanks chewablemorphin for your responses.


I get a free copy of this right?

You mean the software? Yes, i would release this software for evaluation purposes.


Another thing I just thought of, would there be a way for dealers to add rackets to the system?


As for now, I'm focusing on the core functions of my prototype system. Perhaps I will include it as an enhanced function (as mentioned in the quote) if there's an allowance of time, or if not, after I have submitted this final year project to my supervisor.

I have created a questionnaire to help gather requirements for my project as well as to justify the need of my system. Calling all badminton fans here, to help fill in this questionnaire for me. Your input is very much appreciated :o

http://FreeOnlineSurveys.com/rendersurvey.asp?sid=4fwaudsecopdcz9638089


The survey now is closed. Thank you to all that have participated :)

The interview questions are still opened, so please do feel free to answer them.

..ensure that you copy the question on top of your answer so that I know which question you're answering..
1.Can you provide a brief description about yourself? Mainly about your job description (or area of knowledge).

2.A racquet build consist of flex, balance point, weight, grip, and its string. Are there any other build details that make a racquet model unique from one another?

3.From the build details mentioned in the previous question, how would you recommend a suitable racquet for a badminton player based on these racquet builds?

4.What are some of the problems that your customers face when selecting and purchasing a badminton racquet?

5.What are the most common criteria do customers look for when purchasing a badminton racquet? (e.g. price, design, technology)

6.An expert software system would provide an unbiased and objective recommendation to racquet customers as opposed to human recommendation. Is it feasible (practical) to have such a computerized software system for racquet recommendation in your racquet store? If no, what can be done to make it feasible?

7. Are there any personal comments you would like to add before closing this interview session?