View Full Version : are winner/error ratios that reliable in assessing matches?

Nov 29th, 2003, 06:15 PM
I always wonder how important and realiable winner/(unforced) error rations are when we discuss how good or bad the matches are.
It seems to me every tournament or tournament website uses a different way of counting those numbers. Some seem very strict about counting winners and others are quite lenient.
For example a 20/10 ration by one way of counting could be like a 15/15 by another. Then we are seeing totally dfferent stats from the same match and we still jump to some conclusion relying on the one you have in handy. I know those are important stats but we shouldn't take them too seriously and jump to too quick conclusions.
what d'ya think?

Nov 29th, 2003, 06:32 PM
you have to see the match to say if it was good or not!

Nov 30th, 2003, 09:40 AM

~ The Leopard ~
Nov 30th, 2003, 09:44 AM
It's one indication, but you really need to see the match.

Nov 30th, 2003, 02:39 PM
Ova-rated stat, as there are "semi-forced" errors that don't count as either, cases where your UE's go up because strategy dictates running a tired or unfit opponent around, etc. etc.

Nov 30th, 2003, 02:57 PM
No bc that doesnt tell you about the closeness or drama of the match. The 2001 FO final seemed to have a bad winners/errors ratio but it was exciting bc of the drama.
Also it doesnt tell you about forced errors, which are just like winners but dont count towards that stat.
So you really need to see a match to decide.

Brian Stewart
Dec 1st, 2003, 05:43 AM
Winner/error ratios aren't highly reliable, for a number of reasons.

For one, UE's aren't assessed in a consistent manner. What's called a UE in one match isn't in another. Some counts are way off the mark. The worst I've ever seen was a match in which the UE total was inflated by a factor of 12! :eek:

Another reason is that UEs aren't the only kinds of errors. In addition to forced errors, there are also tactical errors and execution errors. I've seen many matches which are loaded with the latter two, which were really poorly played matches, but they came up quite "clean" on the stats sheet.

And there are the errors themselves. The UE stat equates a shot that just missed the line, or clipped the tape, with one that's sent into the next court. That's hardly the same. Plus, the stats don't tell you when the errors occured. Was it on the 3rd stroke? The 10th? The 20th? And it doesn't measure the quality of the rallies. If two really fast players are facing off, there are often several spectacular shots that should be winners, which are returned into play. Thus, even if the point ends in a UE, in order to assess the quality of the match, you'd have to treat it as if there were a couple of winners hit as well.

Which underscores what the posters above were saying. You really have to see the match to be sure. The stats are no more absolute than the scorelines. I've seen 6-1 and even 6-0 sets that were dead even, and 6-4 sets that weren't competitive at all.

Dec 1st, 2003, 05:47 AM
Yep....nothing that actually watching a match to tell if it was good or not. Numbers are just that: numbers. Donīt tell the whole story.

Dec 1st, 2003, 12:26 PM
I agree that watching a match is the best way to evaluate. For what it's worth, I do read the match stats. The statistic that I find to be most under-reported (or talked-about), is the "total points won". You could win a match 6-3, 6-2 and that would be considered a lop-sided win, yet the number of points won was something like 85-83, which clearly indicate that the match was awfully close. In fact about 2 years ago, Patty Schydner won a match where she scored FEWER total points than the winner of the match. This has happened a few time actually.

Perhaps we should all not pay so much attention to stats anyway.

Dec 1st, 2003, 03:19 PM
There's this Patty Penchant for the unique. ;)