Members Login
Username 
 
Password 
    Remember Me  
Post Info TOPIC: Random Charts & Nonsense


All-time great

Status: Offline
Posts: 5110
Date:
RE: Random 2017 Review snippets


Apologies - this is much better (i.e. not suffering from an epidemic of stupidity):

yI7exnG.png

Week 4, 2017 is obviously still in progress, so those figures may (hopefully) get added to.



__________________

Data I post, opinions I offer, 'facts' I assert, are almost certainly all stupidly wrong.



All-time great

Status: Offline
Posts: 5110
Date:

L8kKKM3.png



__________________

Data I post, opinions I offer, 'facts' I assert, are almost certainly all stupidly wrong.



Challenger level

Status: Offline
Posts: 2442
Date:

That last graph might be easier to read if the y axis were flipped. At present, a move up the rankings shows as a downward slope, and a slide down the rankings looks like an improvement, until you consult the scale on the axis. For example, my first impression was that GB10 had a good week2; whereas the opposite is true.

__________________


All-time great

Status: Offline
Posts: 5110
Date:

wimdledont wrote:

That last graph might be easier to read if the y axis were flipped. At present, a move up the rankings shows as a downward slope, and a slide down the rankings looks like an improvement, until you consult the scale on the axis. For example, my first impression was that GB10 had a good week2; whereas the opposite is true.


It's a fair point, and I'm not happy with my presentation. The variation in the axes is unacceptable and confusing.
I did try reversing the values, but my brain reacted quite severely to having zero at the top of the axis; it seems like a perversion of numbers, somehow. I think the zero needs to be at the bottom.

It does reflect the ongoing logical quandary about what we mean when we say 'higher', or 'lower' ranked. The lower the real numerical value, the better, of course, but we often say lower ranked to mean some one lower down the world ranking, and 'career high' is routinely used to mean the lowest numerical value for the ranking.

I've struggled to find a suitable compromise: this should all be plotted on one chart, but with 20 series - for each of the 10 ranks, a series for each of the two years in question - the result is very cluttered.
For example, Johanna's tiny variation at #1 gets completely washed out when the scale encompasses WR 1-400 to include all the top 10  GB players. It's also very difficult to see the difference year-to-year.

I also tried plotting each GB rank as a single continuous line over 2017-2018, and ading some indicating lines to show the high/low values in that timespan, but I couldn't make it work satisfactorily.

In short, I'm trying to do things which I don't really understand, and are increasingly clearly orders of magnitude beyond my capability to grasp them.

I had such high hopes...

I coulda been a contender...



__________________

Data I post, opinions I offer, 'facts' I assert, are almost certainly all stupidly wrong.



Challenger level

Status: Offline
Posts: 2442
Date:

You're doing great, champ.

Would it be too silly to put the y axis at the top?

__________________


All-time great

Status: Offline
Posts: 7056
Date:

The way the scale is now one definitely feels one has to read it upside down, because the line nearer the bottom is showing the higher (ie better!) ranking for the player, whereas one is used to reading graphs the other way round. In his last post, I think wimdledont's suggestion is to put the x axis (rather than the y axis) at the top of the page, instead of in the normal position at the bottom... That would help with reading the graphs the 'correct' way.

__________________


All-time great

Status: Offline
Posts: 5110
Date:

OK, good suggestions, thank you. I think this is what is meant; it helps a bit:

Gq2pQWJ.png

Uniform scaling, now, too, but the uniform unit of '20' still takes up different amounts of scale. I can force them to conform - i.e. each axis would cover 'x' units, the same as the maximum covered in any one of the charts; in the instances above, that variation is 56 places fro GB#10, so the scale for every chart would cover 80 places (4 * 20 units, 3 *20 not leaving enough 'space' for the display) -  but it's a lot of extra messing about with invisible dummy series.



__________________

Data I post, opinions I offer, 'facts' I assert, are almost certainly all stupidly wrong.



All-time great

Status: Offline
Posts: 7056
Date:

This is better it reads more naturally now (and keep the scale with the uniform unit of 20). For instance Heather's clear ranking improvement is now much more immediately obvious.
The next issue to be remembered of course is that whereas for the first three ranking positions, the players are the same this year and last, from #4 on, the players begin to swap. Thus the #4 ranked player last year ranking between 160 and 150 was Laura.... this year the #4 hovering around 200 is of course Katie B. But as long as one understands this relativity, what the comparisons are showing this early on is interesting. For #s 3-5 so far this year the ranks are lower than last year, but for #s 6-9 the rankings are better than last year. In short there has been a closing up between #s 4-9. And then of course the #10 position is very much up for grabs by the fastest improver....

__________________


All-time great

Status: Offline
Posts: 5110
Date:

FAnf4Gx.png

 



__________________

Data I post, opinions I offer, 'facts' I assert, are almost certainly all stupidly wrong.



All-time great

Status: Offline
Posts: 5110
Date:
RE: Random Charts & Nonsense


84RZLRT.png



__________________

Data I post, opinions I offer, 'facts' I assert, are almost certainly all stupidly wrong.



All-time great

Status: Offline
Posts: 6109
Date:

that 10th place seems to be a problem for GB this season to date...plus no one making any real move yet...

__________________
JonH


Tennis legend

Status: Offline
Posts: 39294
Date:

In our 2018 top 10 predictions each of the current top 9 were selected by at least 31 out of 39 with 6 players selected by all 39.

But finding a 10th brought quite a variety. The most selections were for Freya 12, Sam 11, Tara 10 and Eden 9. Be good if at least one player from outside the current top 9 can make a big move up and of course also hope none of the current top 9 make a big move down.



__________________


Challenger level

Status: Offline
Posts: 2442
Date:

In the latest live rankings, Tara has slipped back further, so Sam is now GB10.

__________________


All-time great

Status: Offline
Posts: 5110
Date:

wimdledont wrote:

In the latest live rankings, Tara has slipped back further, so Sam is now GB10.


In the 29/01/2018 rankings, Tara is GB#10 @ WR 397 - Sam is GB#11 @ WR 423



__________________

Data I post, opinions I offer, 'facts' I assert, are almost certainly all stupidly wrong.



All-time great

Status: Offline
Posts: 7056
Date:

Yeah, Sam needs to start playing this year if she is going to move up. She is down for Loughborough next week though.

__________________
«First  <  1 2 3 4 525  >  Last»  | Page of 25  sorted by
 
Quick Reply

Please log in to post quick replies.

Tweet this page Post to Digg Post to Del.icio.us


Create your own FREE Forum
Report Abuse
Powered by ActiveBoard