Jump to content

I'm sick of hearing about "metrics" in evaluating the Bills


PUNT750

Recommended Posts

A lot of people seem to make excuses  for the LACK OF PERFORMANCE of individual Bill's players.  Ed Oliver doesn't get sacks or make big plays.  Yet, his metrics for causing disruption on the opponents OL is high?  Edmunds has great metrics for covering receivers but he has yet to make an impact in games.  Trent Murphy & Star are similar examples.  We keep waiting for Josh to take over a game with all his great metrics!

KPIs (Key Performance Indicators) are much different from metrics. Literally anything, as long as you can count it, can be a metric. The problem is that a metric doesn’t actually do much for you. It doesn’t give you a true indicator of how someone is performing. It’s just a number and not and indicator of performance.  That number can be divided or multiplied by anything to support a pre determined inferential statistic.

 

If a player is recording sacks, making tackles, throwing for touchdowns or making FG’s that’s a performance level that counts and can be measured!.

 

Link to comment
Share on other sites

7 minutes ago, PUNT750 said:

A lot of people seem to make excuses  for the LACK OF PERFORMANCE of individual Bill's players.  Ed Oliver doesn't get sacks or make big plays.  Yet, his metrics for causing disruption on the opponents OL is high?  Edmunds has great metrics for covering receivers but he has yet to make an impact in games.  Trent Murphy & Star are similar examples.  We keep waiting for Josh to take over a game with all his great metrics!

KPIs (Key Performance Indicators) are much different from metrics. Literally anything, as long as you can count it, can be a metric. The problem is that a metric doesn’t actually do much for you. It doesn’t give you a true indicator of how someone is performing. It’s just a number and not and indicator of performance.  That number can be divided or multiplied by anything to support a pre determined inferential statistic.

 

 

 

If a player is recording sacks, making tackles, throwing for touchdowns or making FG’s that’s a performance level that counts and can be measured!.

 

 

Well, Edmunds was defensive player of the week last game last season.

Link to comment
Share on other sites

10 hours ago, PUNT750 said:

A lot of people seem to make excuses  for the LACK OF PERFORMANCE of individual Bill's players.  Ed Oliver doesn't get sacks or make big plays.  Yet, his metrics for causing disruption on the opponents OL is high?  Edmunds has great metrics for covering receivers but he has yet to make an impact in games.  Trent Murphy & Star are similar examples.  We keep waiting for Josh to take over a game with all his great metrics!

KPIs (Key Performance Indicators) are much different from metrics. Literally anything, as long as you can count it, can be a metric. The problem is that a metric doesn’t actually do much for you. It doesn’t give you a true indicator of how someone is performing. It’s just a number and not and indicator of performance.  That number can be divided or multiplied by anything to support a pre determined inferential statistic.

 

If a player is recording sacks, making tackles, throwing for touchdowns or making FG’s that’s a performance level that counts and can be measured!.

 

So let me get this straight. You stated, if you can count it, it is a metric and doesn't do anything for you because a metric is just a number and not an indicator of true performance. You then say sacks, tackles, touchdowns and FGs are what matters.

 

Can't sacks, tackles, TDs, and FGs all be counted, and therefore, are metrics that can't do anything for you because they are not a true indicator of performance?

 

Or, is it that you just don't really like/get the nuances of player performance analyses and, instead, just prefer bright and shiny things?

Edited by billsfan1959
Link to comment
Share on other sites

14 hours ago, PUNT750 said:

A lot of people seem to make excuses  for the LACK OF PERFORMANCE of individual Bill's players.  Ed Oliver doesn't get sacks or make big plays.  Yet, his metrics for causing disruption on the opponents OL is high?  Edmunds has great metrics for covering receivers but he has yet to make an impact in games.  Trent Murphy & Star are similar examples.  We keep waiting for Josh to take over a game with all his great metrics!

KPIs (Key Performance Indicators) are much different from metrics. Literally anything, as long as you can count it, can be a metric. The problem is that a metric doesn’t actually do much for you. It doesn’t give you a true indicator of how someone is performing. It’s just a number and not and indicator of performance.  That number can be divided or multiplied by anything to support a pre determined inferential statistic.

 

If a player is recording sacks, making tackles, throwing for touchdowns or making FG’s that’s a performance level that counts and can be measured!.

 

The OP may have a point and I may even agree with it, but I think it's clouded by use of the word "metric" in an unusual sense and by lack of clarity about the definition he's using.

The use of the somewhat unusual term "performance level" to describe what the NFL calls "statistics" such as TD passes, TD runs, FG, tackles, sacks, passing yards etc. doesn't help with clarity.  The insertion of a buzzphrase "pre determined inferential statistic" is a nice touch, but it's marooned on a lonely peak isolated from the normal meaning of the term "inferential statistic".  None of the things of which the OP complains (disruption on the OL, covering receivers) are inferential statistics by the usual definition.  If anyone really wants to know what that is, this seems to be a decent explanation.

 

Fundamentally, "metrics" in this context simply means "things that can be quantitated and that are believed to be relevant to the outcome one desires".

If the desired achievement is an NFL Football Championship, wins are a metric.

If the desired achievement is a winning season, wins are an outcome.

 

So the NFL started out by recording statistics that are straightforward to measure: yards gained on a pass, yards gained on a run, first downs achieved or given up, third downs converted or stopped.  These are all, indeed, examples of "metrics".  They are a subset of metrics you can find on the nfl.com "stats" page.

 

Stats geeks (amateur and professional) and other interested parties rummage around in these stats trying to decide which ones are the most important to the desired outcome of wins, and even predict it.  And they noticed something coaches have always known - players who don't show up in some of those stats, seem to make a big difference to the performance of the team and its ability to achieve the desired outcome of wins.  So they looked for ways to quantitate this.

 

I mean, obviously, we've all seen a QB unable to make a good throw because he's being chased around the backfield like a chicken dodging a hungry fox and finally throwing the ball away or maybe throwing an incompletion or even a pick (I'm talking to you, Josh Allen!), but no sack - we would all agree that the players assuming the "hungry fox" role performed in a way that contributed to a bad outcome on that play and eventually a win.

 

Metrics like causing disruption on the opponent's OL can, in fact, be measured and are in fact examples of "descriptive statistics", it's just more time consuming and less straightforward than recording a sack.  It is done by looking on film at where the OLman starts the play and how the DLman impacts his motion.  This is actually pretty straightforward for the OLman's team, because they know what the OLman's assignment is on a play and what he's supposed to do.  It's a little bit more subjective for the DLman's team but the coaches know a good bit about the systems their opponents are running and can make a very educated guess.

 

It takes another step of subjectivity when it's done by external groups who know neither, such as PFF or FO.

 

Then we have what I privately refer to as "Frankenstats", where statisticians weight and combine large quantities of individual metrics in a way that predicts wins or playoffs or championships (their desired outcomes).  I personally don't trust these at all, especially when they're proprietary ("secret sauce"), so if that was what the OP was railing against I'd agree but upon closer examination it doesn't seem to be.

 

Anyway, what the OP seems to be saying is that he doesn't trust data about game contributions he can't quantitate with his own eyeballs or find on the NFL stat page.  He's entitled, but anyone who has coached or played the game (or even talked to people who have coached or played the game) knows that a player's impact on their team in fact can't always be captured by "bright shiny things" like sacks or runs.  And that's just how it is.

I'm sorry to hear you're feeling sick, OP; have you considered the use of hearing protectors?

 

 

 

 

 

  • Like (+1) 1
  • Thank you (+1) 1
Link to comment
Share on other sites

13 hours ago, PUNT750 said:

If a player is recording sacks, making tackles, throwing for touchdowns or making FG’s that’s a performance level that counts and can be measured!.

 

Uh, yeah, okay. So if a team has a bad defense say 25th or worse ranked YPG and 25th or worse in TOP but their MLB is top 5 in tackles I guess he is having a great year because he has so many tackles?

 

Sacks can also be skewed. Team A could play majority veteran QB's who get rid of the ball. Team B could play majority 1st and 2nd year starters who take a lot of sacks in general.

Link to comment
Share on other sites

1 hour ago, Mojo44 said:

Wins are not metrics. They are what metrics try to predict.

 

Maybe.  It depends upon what the desired outcome is.  If the outcome is playoffs, or a championship, then wins are in fact a metric.

Metric in this context really just means "something you can measure".  Passing yards, sacks, rushing yards, completion %, YPA, tackles, etc are all examples of metrics.  They're a subset of metrics that are straightforward to measure or calculate and tracked on the NFL (basic) stats page.

Link to comment
Share on other sites

×
×
  • Create New...