: Wed Feb 18, 2009 1:44 pm  

The Hollywood Frasier wrote:

So your taking some stats over what you saw with your own eyes.

Yes! Stats are generally more accurate than what my subjective eyes see.

I dont go into an area of the business i work in and think they are great because they all look good, if their stats tell me something else. I use a combination of information available to me to make my mind up.

I said Berro was quiet. Stats look like he wasnt so i stand corrected. Further review might be that the stats reported by SL are incorrect which is why you never take once source of info in isolation to form an opinion!

: Wed Feb 18, 2009 1:53 pm  

Berrigan, as Swain before him will never have a bad game in most Hull fans eyes. I reckon Crocker would've been the same if he got here. Objectivity is a foreign concept when it comes to our 'star' players.

: Wed Feb 18, 2009 2:00 pm  

Big Dave T wrote:

Yes! Stats are generally more accurate than what my subjective eyes see.

More accurate maybe, but more accurate at judging what? Statistics measure what statistics measure. There has to be a strict, immutable definition of what they are measuring. In a subjective area like 'missed tackles', therefore, they are not always (ever?) a true reflection of what has actually taken place. To know that, you have to watch what happens and understand what you are watching. It's still subjective, but it's subjectivity based on your own knowledge/understanding, rather than relying on an arbitrary source, and is something that can be debated/discussed to find true meaning rather than just pointing at a contextless number assigned according to an unknown directive and hoping that that ends discussion.

: Wed Feb 18, 2009 2:05 pm  

trys'r'us wrote:

More accurate maybe, but more accurate at judging what? Statistics measure what statistics measure. There has to be a strict, immutable definition of what they are measuring. In a subjective area like 'missed tackles', therefore, they are not always (ever?) a true reflection of what has actually taken place. To know that, you have to watch what happens and understand what you are watching. It's still subjective, but it's subjectivity based on your own knowledge/understanding, rather than relying on an arbitrary source, and is something that can be debated/discussed to find true meaning rather than just pointing at a contextless number assigned according to an unknown directive and hoping that that ends discussion.

Sorry, nodded off then!

In simply terms yes your defination of measuring a statistic is important, missed tackles have their own defintion as do completed tackles. (does the 3rd man in count or simply the 1st)

However, if your data source is the same place and they have a consistent approach to recording the stats at least they are comparible to give context.

If Berrigan came out statistically better than Alker, whether i think he was poor or not its all relative.

: Wed Feb 18, 2009 2:09 pm  

Berrigan Class Act!

Thats all i want to say!

: Wed Feb 18, 2009 2:15 pm  

Big Dave T wrote:

However, if your data source is the same place and they have a consistent approach to recording the stats at least they are comparible to give context.

They're comparable, but what are you comparing? Say Berrigan genuinely misses one tackle, but has another two added to his count for smashing someone but not technically completing them by being in contact when the referee calls held (is that how a tackle/miss is defined? Who knows?) and Alker genuinely misses two tackles, and doesn't have any others added. Who comes out on top statistically? Who comes out on top in reality?

It's a simple example, but it shows the downfall of relying/placing any trust on the arbitrary figures given.

: Wed Feb 18, 2009 2:18 pm  

trys'r'us wrote:

They're comparable, but what are you comparing? Say Berrigan genuinely misses one tackle, but has another two added to his count for smashing someone but not technically completing them by being in contact when the referee calls held (is that how a tackle/miss is defined? Who knows?) and Alker genuinely misses two tackles, and doesn't have any others added. Who comes out on top statistically? Who comes out on top in reality?

It's a simple example, but it shows the downfall of relying/placing any trust on the arbitrary figures given.

As i said, it depends on consistency of application. You'd like to think that if Berrigan doesnt complete the tackle he gets marked down as 1 missed.

With regard to my comparable comment you are comparing players to each other. If there is the consistent approach to collating the stats and a consistent approach to logging the stats you can compare Alker with Berrigan for example knowing that you are using like for like data which may not give you accurate figure but will allow those inaccurate figure to be compared accurately, telling you which is the better player. (as i said originally)

: Wed Feb 18, 2009 2:32 pm  

Big Dave T wrote:

As i said, it depends on consistency of application. You'd like to think that if Berrigan doesnt complete the tackle he gets marked down as 1 missed.

With regard to my comparable comment you are comparing players to each other. If there is the consistent approach to collating the stats and a consistent approach to logging the stats you can compare Alker with Berrigan for example knowing that you are using like for like data which may not give you accurate figure but will allow those inaccurate figure to be compared accurately, telling you which is the better player. (as i said originally)

No. As my example pointed out, the comparison is in no way guaranteed to be accurately providing the desired result. In that example, Berrigan missed one tackle and Alker missed two, yet Alker's stats would have looked better, despite consistent collation. That's not comparing accurately in any meaningful way.

: Wed Feb 18, 2009 2:48 pm  

trys'r'us wrote:

Berrigan missed one tackle and Alker missed two, yet Alker's stats would have looked better, despite consistent collation.

If there are clear definitions of whats a tackle and what is a missed tackle how would alkers stats have looked better if one is classified as missing 2 and one is classified as missing 1? Consistent collation and clear definitions give the desired clarity here.

Anyhow, Berrigan looked poor and the SL stats are wrong!!

: Wed Feb 18, 2009 2:54 pm  

Big Dave T wrote:

If there are clear definitions of whats a tackle and what is a missed tackle how would alkers stats have looked better if one is classified as missing 2 and one is classified as missing 1?

Because, as I pointed out, what I expect is the definition of a miss would, in the example I gave, result in three misses being 'awarded' against Berrigan.

Big Dave T wrote:

Consistent collation and clear definitions give the desired clarity here.

And, as I said originally, there can't be a clear definition for a subjective action. That is what makes the statistics unusable.