New (7-12-11) TWU Racquet Finder Tool

[TW Professor]

REVISED: New (7-12-11) TWU Racquet Finder Tool

It's called the "Racquet Selection Guide SPECS-tacular".

http://twu.tennis-warehouse.com/cgi-bin/racquetspecs2.cgi

The only instructions are to be aware of what is clickable or requires input: Clickable buttons, row headers, rows, icons, text boxes, dropdowns etc. appear in each sub-display which take you to more information. The main entry to data (after you choose a search and sort parameter) is by clicking the highlighted (on mouse over) rows. Note: you can open as many simultaneous rows as you want.

It's a searchable, sortable, extravaganza of static and dynamic specs, displays, and tools.

 

[Agent Orynge]

I like how 140 inch head size falls into the 'typical' category.

 

[TW Professor]

I like how 140 inch head size falls into the 'typical' category.

Just a guideline that will include most every racquet ever made. When you are used to thinking in different units you have no idea of what to type in (the "typical" range changes when you change units).

However, you can actually enter any number into the field. For example, if you want to guarantee that you will get a list of every racquet, just enter something like "greater than 1" into the headsize (or whatever) box.

 

[Larrysümmers]

I like it a lot. pretty neat.

 

[travlerajm]

How hard would it be to add the MgR/I values?

http://tt.tennis-warehouse.com/showthread.php?t=387805

 

[TimothyO]

How's the data accuracy and consistency?

The previous version returned (sometimes wildly) inconsistent data that disagreed with the catalog data. Weight, stiffness, etc. could vary.

 

[TimothyO]

Never mind, there are now NEW errors in this tool.

PB10 Mid Stiffness:
Catalog = 59
New Tool = 61

The concern isn't a couple points of stiffness. It's that the data isn't reliable. Maybe other bits are even more wrong. I discovered this error on my FIRST try. TW, you need to go back to the drawing board.

 

[retlod]

Can we tennis geeks have more parameters to narrow the field? For instance, can I see all racquets that weigh 300-330g, have stiffnesses between 62-70, and a balance between 4-8 pts HL?

 

[TimothyO]

Can we tennis geeks have more parameters to narrow the field? For instance, can I see all racquets that weigh 300-330g, have stiffnesses between 62-70, and a balance between 4-8 pts HL?

Yes! I liked the old tool more...you could progressively refine your search or go for it from the start. Too few parameters with this tool.

 

[corners]

Never mind, there are now NEW errors in this tool.

PB10 Mid Stiffness:
Catalog = 59
New Tool = 61

The concern isn't a couple points of stiffness. It's that the data isn't reliable. Maybe other bits are even more wrong. I discovered this error on my FIRST try. TW, you need to go back to the drawing board.

The Professor measures only one frame for his numbers, while TW's retail unit averages the measured values from five frames. Blame Volkl for the variation in their specs from racquet to racquet, not the messenger.

 

[julian]

Wrong swing weight

It's called the "Racquet Selection Guide SPECS-tacular".

http://twu.tennis-warehouse.com/cgi-bin/racquetspecs2.cgi

The only instructions are to be aware of what is clickable or requires input: Clickable buttons, row headers, rows, icons, text boxes, dropdowns etc. appear in each sub-display which take you to more information. The main entry to data (after you choose a search and sort parameter) is by clicking the highlighted (on mouse over) rows. Note: you can open as many simultaneous rows as you want.

It's a searchable, sortable, extravaganza of static and dynamic specs, displays, and tools.

An entry for Aero Pro Drive GT is 316 in the tool.
http://www.tennis-warehouse.com/Babolat_AeroPro_Drive_GT/descpageRCBAB-BAPDGT.html
gives 331
It is too much of a difference

 

[julian]

Please see my post above

The Professor measures only one frame for his numbers, while TW's retail unit averages the measured values from five frames. Blame Volkl for the variation in their specs from racquet to racquet, not the messenger.

Please see my post above

 

[TimothyO]

The Professor measures only one frame for his numbers, while TW's retail unit averages the measured values from five frames. Blame Volkl for the variation in their specs from racquet to racquet, not the messenger.

No, it's definitely a weird DB error. The old finder shows 12 pts HL for the DNX 10. That was different from the online catalog. The new finder shows a balance of 12.7" for the DNX 10. That's two or three significantly different results with two different results coming from the old and new Racquet Finders. Since they're not selling the DNX 10 anymore where the heck is the new finder pulling data from? Certainly not the old finder DB.

Maybe they have three different DBs now?

In any case I gave up relying TW specs. Too many other errors, including in their testing. On the BM 200 Lite they swore they had the right stiffness which was dramatically different from Dunlop and other sites. They confirmed their number online. Whoops! Either a typo or testing screwp caused them to bump it up significantly.

 

[TW Professor]

Fixed a sorting problem

FYI: in case anyone had a sorting problem, it should be fixed now.

 

[TW Professor]

Can we tennis geeks have more parameters to narrow the field? For instance, can I see all racquets that weigh 300-330g, have stiffnesses between 62-70, and a balance between 4-8 pts HL?

A couple of things. First, you can do single range searches by doing, for example, Find1 headsize greater than 299 and Find2 headsize less than 331.

But for your particular example I would do this: Find1: Weight > 299; Find2: Flex > 61. Then I would sort first by weight, then flex, then balance. Then you just go to the portion of the results that are in the ranges you are looking for.

All that being said, I agree that another level or two of search/find might be nice. I will try to make that happen (though it is not a quick fix).The trouble with too many levels is that you can then come up with a lot of empty finds, which can be frustrating.

In the mean time, careful search/sort strategies should help you get the information you want.

 

[TW Professor]

No, it's definitely a weird DB error. The old finder shows 12 pts HL for the DNX 10. That was different from the online catalog. The new finder shows a balance of 12.7" for the DNX 10. That's two or three significantly different results with two different results coming from the old and new Racquet Finders. Since they're not selling the DNX 10 anymore where the heck is the new finder pulling data from? Certainly not the old finder DB.

Maybe they have three different DBs now?

In any case I gave up relying TW specs. Too many other errors, including in their testing. On the BM 200 Lite they swore they had the right stiffness which was dramatically different from Dunlop and other sites. They confirmed their number online. Whoops! Either a typo or testing screwp caused them to bump it up significantly.

To clarify. Yes, there are two different databases for specs -- one for TW and another for TWU. The reason is that in the lab we need to have the exact specs for each racquet being tested because some of the data is calculated based on other measured data. We can't use "published" data and still get internally consistent results in the lab. Nor is it logistically feasible to use the same racquets that are used in retail or for playtests.

There is no such thing as "the real" weight, balance, swingweight, etc. of a racquet. Every racquet is different and, depending on the measurement of interest, the range can be up to +/- 5%. For weight, that would be as much as 30 grams between the lightest and heaviest possible racquets for a 300 gram "spec" racquet. Most racquets fall closer to the middle of the range, but you never know.

That is partly why the dynamic specs (like power potential, sweet zone size, and vibration frequency) are more accurate predictors of performance. They are measurements during impact that are the result of the composite contributions of every other spec (weight, balance, swingweight, RDC flex, pattern, etc.). As such they are performance specs that tend to average out the tolerances in the individual static specs.

 

[TW Professor]

How hard would it be to add the MgR/I values?

http://tt.tennis-warehouse.com/showthread.php?t=387805

Right now, it's too much. Perhaps down the road.

 

[travlerajm]

Right now, it's too much. Perhaps down the road.

No worries. But I'd be curious to hear your thoughts on using the double pendulum model to design a racquet.

I just discovered Cross's recent article on the subject -- he and I seem to have been working independently, but converging toward the same answers from different starting points.

 

[TimothyO]

To clarify. Yes, there are two different databases for specs -- one for TW and another for TWU. The reason is that in the lab we need to have the exact specs for each racquet being tested because some of the data is calculated based on other measured data. We can't use "published" data and still get internally consistent results in the lab. Nor is it logistically feasible to use the same racquets that are used in retail or for playtests.

There is no such thing as "the real" weight, balance, swingweight, etc. of a racquet. Every racquet is different and, depending on the measurement of interest, the range can be up to +/- 5%. For weight, that would be as much as 30 grams between the lightest and heaviest possible racquets for a 300 gram "spec" racquet. Most racquets fall closer to the middle of the range, but you never know.

That is partly why the dynamic specs (like power potential, sweet zone size, and vibration frequency) are more accurate predictors of performance. They are measurements during impact that are the result of the composite contributions of every other spec (weight, balance, swingweight, RDC flex, pattern, etc.). As such they are performance specs that tend to average out the tolerances in the individual static specs.

The retail catalog, old finder, and new finder all differ in specs for some frames. there are three front ends but two databases.

How are we getting three different results on the front ends from two different DBs?

If the old finder managed by the retail store and the new finder from the lab?

 

[TW Professor]

The retail catalog, old finder, and new finder all differ in specs for some frames. there are three front ends but two databases.

How are we getting three different results on the front ends from two different DBs?

If the old finder managed by the retail store and the new finder from the lab?

Question: to help us track this down, what do you mean by "retail catalog"? Do you mean the specs that show up on each racquet description page at TW?

Can you give us an example of a racquet with 3 different values for the same spec?
Thanks.

 

[Darkhors]

The retail catalog, old finder, and new finder all differ in specs for some frames. there are three front ends but two databases.

How are we getting three different results on the front ends from two different DBs?

If the old finder managed by the retail store and the new finder from the lab?

To clarify. Yes, there are two different databases for specs -- one for TW and another for TWU. The reason is that in the lab we need to have the exact specs for each racquet being tested because some of the data is calculated based on other measured data. We can't use "published" data and still get internally consistent results in the lab. Nor is it logistically feasible to use the same racquets that are used in retail or for playtests.

There is no such thing as "the real" weight, balance, swingweight, etc. of a racquet. Every racquet is different and, depending on the measurement of interest, the range can be up to +/- 5%. For weight, that would be as much as 30 grams between the lightest and heaviest possible racquets for a 300 gram "spec" racquet. Most racquets fall closer to the middle of the range, but you never know.

That is partly why the dynamic specs (like power potential, sweet zone size, and vibration frequency) are more accurate predictors of performance. They are measurements during impact that are the result of the composite contributions of every other spec (weight, balance, swingweight, RDC flex, pattern, etc.). As such they are performance specs that tend to average out the tolerances in the individual static specs.

Question: to help us track this down, what do you mean by "retail catalog"? Do you mean the specs that show up on each racquet description page at TW?

Can you give us an example of a racquet with 3 different values for the same spec?
Thanks.

I'm pretty sure that the highlighted part sums up exactly why there's a difference in the measurements. They have to measure EVERY racket they test so that the numbers are legit.

I'm pretty sure they're not going to just throw a bunch of unreliable numbers into a database, spend a ton of time creating a tool and then passing the info along to their users knowing full well that it's not reliable.

DH

 

[TimothyO]

Question: to help us track this down, what do you mean by "retail catalog"? Do you mean the specs that show up on each racquet description page at TW?

Can you give us an example of a racquet with 3 different values for the same spec?
Thanks.

There were a few that I found in the past. The one that sticks out for me most is the DNX 10. The balance is different in the three sources: retail racquet description page, old finder, and new finder.

It's no longer visible from the racquet description page but I remember a huge difference between the old finder and racquet page. The extreme HL balance in the old finder was very exciting. The description page showed a much less HL balance. The new finder shows something like 12.7" IIRC.

 

[retlod]

A couple of things. First, you can do single range searches by doing, for example, Find1 headsize greater than 299 and Find2 headsize less than 331.

But for your particular example I would do this: Find1: Weight > 299; Find2: Flex > 61. Then I would sort first by weight, then flex, then balance. Then you just go to the portion of the results that are in the ranges you are looking for.

All that being said, I agree that another level or two of search/find might be nice. I will try to make that happen (though it is not a quick fix).The trouble with too many levels is that you can then come up with a lot of empty finds, which can be frustrating.

In the mean time, careful search/sort strategies should help you get the information you want.

Thanks, TWP! It's great that you're doing this stuff for us!

 

[Mig1NC]

This is a fantastic tool, but as others have said, we need more filters.

 

[TW Professor]

REVISED: TWU Racquet Finder

The New TWU Racquet Finder has been REVISED and improved.

http://twu.tennis-warehouse.com/cgi-bin/racquetspecs2.cgi

Here's what has been changed:

1. There are six filtering levels instead of two.
2. Default values are inserted for all specs and units.
3. Units conversion occurs automatically for specs with numbers in the Choices/Search area.
4. New Balance Points Conversion Tool added. Change the length or balance input and then click out of the box to see changes.
5. Error location indicator: With so many filter levels, you need to know where your search failed. This is indicated in the error message.
6. Auto highlighting of boxes you may need to edit after you change a drop down selection.

You may have to refresh your page/cache if you don't get this at the top of your display: