This shows how stats lie

[fedupyid]

Deleted because of formatting.

 

[Riandor]

best stat to use:
Spurs 0-1 WBA

 

[Gassin's finest]

Add formatting, grammar and punctuation. Then come back when it's readable.

 

[Bobbins]

Passing AccuracyWon TackleEffective ClearanceEffective Head ClearanceDuel WonAerial WonInterception Won
Vlad Chiriches97%153216
Danny Rose82%322611
Eric Dier81%154852
Younes Kaboul77%1107433

I think what this actually shows is that you really need to format stuff like this.

 

[SpursManChris]

Passing AccuracyWon TackleEffective ClearanceEffective Head ClearanceDuel WonAerial WonInterception Won
Vlad Chiriches97%153216
Danny Rose82%322611
Eric Dier81%154852
Younes Kaboul77%1107433

This is more of a shower of shyt than our performance against west brom. How can you possibly expect us to make any sense or that at all?

 

[SpursManChris]

Add grammar and punctuation.

Be fair, he could be an English as 2nd language person. I'm sure you can read the non-stats part. Otherwise you'd be telling half the members on here to fix their English.

 

[SpursManChris]

This shows how stats lie too:

SOLDADO

 

[Borks]

Stats don't lie at all, they're facts.

You just can't use stats over a tiny sample size to prove a point. I saw the statement "1 point from a possible 9, that's relegation form". While technically true, that's the same as claiming any defeat is relegation form when in reality the only sample size that matters in the premier league over the course of a season is 38 games.

 

[Riandor]

Stats don't lie at all, they're facts.

You just can't use stats over a tiny sample size to prove a point. I saw the statement "1 point from a possible 9, that's relegation form". While technically true, that's the same as claiming any defeat is relegation form when in reality the only sample size that matters in the premier league over the course of a season is 38 games.

Yes though you should probably do trending analysis in order to not wait until game 38 to wonder why you are being relegated.

 

[Gassin's finest]

I hate the use of stats anyway. You simply cannot understand, or quantify, a game like football; a game with uncountable external variables that can influence the outcome of any second of play; through the use of statistics.

You have to watch it, play it, again and again, to really appreciate this most multi-faceted of sports. You can't break it down with maths.

 

[parklane1]

To much emphisis is put on stats ( by some) they are a tool to be used but are never the best thing since sliced bread.

 

[Ironskullll]

I think what this actually shows is that you really need to format stuff like this.

There ought to be a fuck up stat as well. That would sort out the Titus Brambles of this world. 99% everything combined with an almighty brain fart stands up well in the statistics but the true story would be somewhat different.

 

[fatpiranha]

Stats don't lie but they can be misinterpreted.

 

[Geyzer Soze]

Stats don't lie. the interpretation of them can give mixed results.

 

[hodsgod]

Stats never lie.

The problem with football stats is, it requires someone's assessment of what is a tackle, an interception etc. these are all opinions and not statistics. How many people are actually watching the game and recording each players activities? I have no idea but to minimise error you would probably need 22 each game, and I doubt that is the case.

The only stats you can rely upon is the result, and they often don't reflect the game, like a Newcastle at home last season.

We lost 1-0, it won't change.

 

[hodsgod]

To much emphisis is put on stats ( by some) they are a tool to be used but are never the best thing since sliced bread.

That depends on whether you agree with what the stats are used to imply.

Like others say, stats are always correct, it's the interpretation that needs to be questioned.

 

[HobbitSpur]

Its in Binary, simple to translate

Vlad Chiriches Very Shit
Danny Rose Not Quite as Shit
Eric Dier Slightly the less Shittiest
Younes Kaboul Just Shit

 

[LexingtonSpurs]

Stats are nothing more than (somewhat) unbiased information.

The issue with stats is in the interpretation, and in assigning too much meaning to a small sample size - really no different than the problems with relying on personal observation. We see what we want to see, and we can make stats say what we want them to say.

Really, you need both - collect stats/data in a cold calculated manner, then explain variances with human observations.

For example - if you expect a DM to make 4 tackles per game, and the stats tell you Player X only made 1 - those are hard, unassailable facts. Human observation might tell you the other team played away from Player X, limiting his opportunities to make tackles. Or, human observation may tell you that Player X was frequently out of position and that is why he only made 1 tackle.

 

[goughie1966]

I like this stat.

The Spurs shot on target in the 71st minute against West Brom was the first one on target in just over 175 minutes of football (nearly 3 hours ). None were on target in the previous 90 minutes against Partizan Belgrade. The last shot on target had been in the 75th minute against Sunderland.

 

[aliyid]

Lies, damned lies, and statistics

Stats can be used to prove virtually any point and they can be twisted very easily with a little misinterpretation. The main one that sticks in my head from my uni days was

• 25% of road deaths are caused by drink driving

So 75% were caused by sober driving, meaning you are three times more likely to die while driving if you're sober...