|
Post by coachd5085 on Dec 24, 2015 10:16:55 GMT -6
It seems every other day someone starts a thread asking if _______ is a good metric by which to measure an offense or defense. (Usually an offense...seems like a visor thing to do ) I just don't seem to understand this. Maybe I am growing old and testy, maybe my time away from the trenches has me missing something, but I just don't see any reason for this pursuit of metric measurable data when one already exists in the game--THE score board. Not only that, but we already have ways to evaluate the WHY behind the score-- GAMEFILM. I just don't see why Taking all the points scored this year, and then dividing by yards gained, or starting field position, or # of plays, or possessions that started exactly on the minute, or passing plays called under a full moon benefits the team. A coach can see the score board (metric), and he can watch the game film and see little timmy false stepping on his pull or rounding his break, or playing with high pad level. Am I missing something.
|
|
|
Post by natenator on Dec 24, 2015 10:28:24 GMT -6
Humans are inherently lazy so we are continuously seeking out ways to get out of doing work.
Looking for a magic bullet in metrics is a coaches way of getting out of work. Analyzing game film and then determining HOW to translate those mistakes into actionable outcomes to fix (drills, scheme adjustments, etc) takes a lot of work that most just don't want to do.
As someone that makes a living doing statistical analysis for data-driven decision making, I can tell you it takes a boat load of work to uncover the nuggets of wisdom that lead to outcomes. We don't just have one or two metrics that can tell us what we need to know. You have to know what you are looking at which comes from hours and hours of immersing yourself with the data - not unlike game video.
|
|
|
Post by brophy on Dec 24, 2015 10:28:23 GMT -6
to a certain degree it is about control of the unknown. The best way to combat the unknown is with information/data.
The challenge with any data is irrelevant elements.
Sure, we need to hard stick the outside foot after a 2-step stride to catch the slant with the extended arms between our index fingers and thumbs....but the goal is to get 100 yards accumulated on offense. Once we get that we can feel good that we're being productive
Clear, definable metrics to quantify success give folks the roadmap/milestone to shoot for....with the assumption that this path leads to wins. Yes, you can get lost on this road (chasing stats) while losing sight of why it is important (execution)
Lets say your team is coming off of a 4-6 season. If all you came up with was, "man, we've just got to DO BETTER next year". While that is true, what specifically are you going to do better at?
You could acknowledge, "we just have to coach up the double-team technique so that we can consistently get 3 yards on ____ runs". How would you know if you are indeed coaching that technique well enough? I suppose you would reach for a yardage goal or conversion goal, when in actuality it would be player performance grades.
Are stats (or stat chasing) inherently wrong? I don't think so.
How do you lose/gain weight? .....expend energy and get proper nutrition
Easy enough. But how do you know you're doing it right? How do you know you're not screwing this up if you don't immediately see results? You examine the macronutrient composition and measure against the resistance routine. Are these numbers comparable to what is required to get the desired result? You can't expect a greater return than what you've invested.
|
|
|
Post by coachfloyd on Dec 24, 2015 10:59:43 GMT -6
Because my family depends on a dumb game to live. So all things get analyzed.
|
|
|
Post by 33coach on Dec 24, 2015 11:19:55 GMT -6
It seems every other day someone starts a thread asking if _______ is a good metric by which to measure an offense or defense. (Usually an offense...seems like a visor thing to do ) I just don't seem to understand this. Maybe I am growing old and testy, maybe my time away from the trenches has me missing something, but I just don't see any reason for this pursuit of metric measurable data when one already exists in the game--THE score board. Not only that, but we already have ways to evaluate the WHY behind the score-- GAMEFILM. I just don't see why Taking all the points scored this year, and then dividing by yards gained, or starting field position, or # of plays, or possessions that started exactly on the minute, or passing plays called under a full moon benefits the team. A coach can see the score board (metric), and he can watch the game film and see little timmy false stepping on his pull or rounding his break, or playing with high pad level. Am I missing something. with the days of Hudl, and reporting... people have the ability to see the data at work. give people more charts and reports, people will want to find the meaning behind the metric. did you actually lose because your ToP was 30% less then your average? or was it the fumble on the goalline at the end of the game? both are actually arguable. and both are contributors to the loss, but which one means more?
|
|
|
Post by coachd5085 on Dec 24, 2015 12:09:11 GMT -6
... did you actually lose because your ToP was 30% less then your average? or was it the fumble on the goalline at the end of the game? both are actually arguable. and both are contributors to the loss, but which one means more? Neither. You lost because your opponent scored more points than you did.
|
|
|
Post by 33coach on Dec 24, 2015 12:16:43 GMT -6
... did you actually lose because your ToP was 30% less then your average? or was it the fumble on the goalline at the end of the game? both are actually arguable. and both are contributors to the loss, but which one means more? Neither. You lost because your opponent scored more points than you did. thats a very high level and simplistic view. working in the corporate world has taught me one thing, metrics matter. sure, you could answer the question "why did the shop close?" with "they didnt make enough money." - but more accurately you would say "well they lost 300,000 in stolen items". or "a fire broke out in their distribution center, and that forced them to take a reduced amount of orders". is the outcome the same? yes. but knowing WHY is very important....
|
|
|
Post by coachphillip on Dec 24, 2015 12:19:45 GMT -6
I use my experience in the profession to deduce that we didn't get them off the field enough on third down and that's why we lost. I use my data collection to tell me how often we didn't do it, where they targeted us, etc. Experience can help you understand the problem. Data assists you in pinpointing the extent of the problem.
|
|
|
Post by lochness on Dec 24, 2015 12:20:49 GMT -6
It's a feel-good exercise. See how hard I'm working in the off-season? I can't possibly be a bad coach because I analyzed a metric that said every time my TB has a ham samdwich 76 minutes before kickoff, he averages 0.037 yards more per carry. And any time my TB averages 0.03 yards more per carry than his career average, we generally have 1.0097 more explosives. And any time we have more than 3.1415 explosives per game in my career as a HC, we generally win 50.00756% of those games! So I'm trying to generate more explosives (which my metrics dictate is any offensive play that gains 13.375 yards or more).
Metrics in football are generally useless. There are way too many variables with 22 guys on the field at once...and almost infinite combinations of this 22 due to injuries, personnel packages, etc, for you to say "yes! When we line up strong left and run weak right in B gap....that is where we are most productive!" Oh really? What if it's just because for 2 weeks In a row, your opponent had a bag of poop playing DL over there? What if it's because for that one game against a badly coached defense, the opponent was selling out to something else schematically.
Like so many other things (8 hour weekend meetings, endless 7-on-7 participation, practices that are too long, etc) it's more about making us feel like we are hard-working, highly-informed coaches using the latest methodology to give our players an advantage.
If we spent half as much time on developing leaders, building chemistry, and working on getting stronger, faster, and more technically sound as we do on hollow pursuits...we just might have something.
|
|
|
Post by brophy on Dec 24, 2015 12:21:33 GMT -6
|
|
|
Post by lochness on Dec 24, 2015 12:33:09 GMT -6
The thing I'd be interested to hear from people who spend a bunch of time on metrics is "what do you do with them?" What changes in practice or techniques or personnel does it actually drive? And, since your performance on any given particular play depends entirely on the 11 players you had on the field vs. that particular opponent's 11 players, what conclusions can you gain in the off-season, when your roster and your opponents are all going to be very different the following year?
This is where the practice of chasing numbers completely falls apart for me. Your efficiency on "Power" has very little to do with whether you run it from 2-back gun on 2nd and medium or from The I on 3rd and short. It has everything to do with your players vs. your opponent's defensive players in that particular game or in that particular play.
|
|
|
Post by coachd5085 on Dec 24, 2015 12:40:11 GMT -6
Neither. You lost because your opponent scored more points than you did. thats a very high level and simplistic view. working in the corporate world has taught me one thing, metrics matter. sure, you could answer the question "why did the shop close?" with "they didnt make enough money." - but more accurately you would say "well they lost 300,000 in stolen items". or "a fire broke out in their distribution center, and that forced them to take a reduced amount of orders". is the outcome the same? yes. but knowing WHY is very important.... As I stated, the WHY is gained via actually watching the players in the game. lochness 's posts sum up the rest quite nicely. Your corporate analogy is inaccurate because a football team is not a large enough operation to need such analysis. The corporate metrics you mention would be used by a CEO in NY making decisions for 100+ storefronts nationwide because he is not in each store each day. Such is not (or rather SHOULD not) be the case with football success. Every negative piece of data is the product of what you watch on film. That is one reason why I got away from "grading". Watch film, yes. Watch to perform as best as you can, absolutely. But what good does is a grade if one week a player steps with the correct foot, has great pad level, brings his hips and punches and effectively blocks the 3, and the next week he steps with the correct foot, has great pad level, brings his hips, punches and gets stoned by the kid LSU and Bama are fighting over. Data is just a numerical representation of that.
|
|
|
Post by coachphillip on Dec 24, 2015 12:40:19 GMT -6
What if your metrics reveal trends? What if you found that your opponents on average had more yards intermediate over the middle on third down than on the perimeter? Wouldn't that lead you to look more intently at the way you're teaching your ILB how to drop into coverage? Or maybe your defense was geared more towards defending the perimeter. Maybe you need to rethink your scheme. It's information. It's as good as what you do with it, for sure. But, having it doesn't hurt. This is 2015. I can get all of this stuff within a few clicks once I've entered the stuff into Hudl. Some times you use all of it, some time you use some of it, some times the game was so evident that you use none of it.
I really don't understand the looking at statistics like its pseudoscience stance. You're going to equate looking at the average starting position of your offense as it relates to your probability of scoring to something like a running back eating a sandwich? Come on.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 24, 2015 12:42:27 GMT -6
I'm undecided on the value of metrics tbh, I believe that there is enough value that I will take the time to read them just I case. I have also noticed that a lot/all of the useful information that is gained, was also observed watching the game and through an experienced observer. Between myself and my HC(with 30+ experience) we generally know exactly what our issues were, personnel, schematic or otherwise by observation.
So maybe your not a grumpy old man, maybe you just have enough experience to see visually what others may look for in metrics. To each his own.
|
|
|
Post by coachd5085 on Dec 24, 2015 12:58:58 GMT -6
What if your metrics reveal trends? What if you found that your opponents on average had more yards intermediate over the middle on third down than on the perimeter? Wouldn't that lead you to look more intently at the way you're teaching your ILB how to drop into coverage? Or maybe your defense was geared more towards defending the perimeter. Maybe you need to rethink your scheme. It's information. It's as good as what you do with it, for sure. But, having it doesn't hurt. This is 2015. I can get all of this stuff within a few clicks once I've entered the stuff into Hudl. Some times you use all of it, some time you use some of it, some times the game was so evident that you use none of it. I really don't understand the looking at statistics like its pseudoscience stance. You're going to equate looking at the average starting position of your offense as it relates to your probability of scoring to something like a running back eating a sandwich? Come on. I would argue that is not a "metric". That is scouting. There is a big difference between what I would call "scouting" data which is very useful, and the evaluation metrics such as points per play, expected scoring value of starting field position, every 10 yards = point scored etc. etc. " Data that shows what your opponents are doing, or self scouting data showing your tendencies = GOOD. Spending time compiling metrics to evaluate HOW YOU DID... meaningless, because you have a scoreboard and video.
|
|
|
Post by natenator on Dec 24, 2015 13:16:23 GMT -6
KPI's have to be actionable. You have to do something with the metrics otherwise it's data-puke.
|
|
|
Post by wingtol on Dec 24, 2015 14:03:56 GMT -6
The three types of lies: lies, damned lies, and statistics.
|
|
|
Post by rsmith627 on Dec 24, 2015 14:10:40 GMT -6
I don't care about rushing/passing yards, points per play, etc. I care about points on the board being more than what the other teams scored and the W/L column. Those are the metrics I care about.
|
|
jmg999
Junior Member
Posts: 263
|
Post by jmg999 on Dec 24, 2015 15:12:00 GMT -6
Mmm...data analysis. In this thread, there seems to be multiple rationales for utilizing data analysis. If the purpose is to find a magic bullet, I can tell you that if one does it exist, it's going to take people much smarter than us to determine where it resides. But, if the purpose is gto simply improve our team, data analysis is a key component to that. Data analysis answers questions, such as, "Does implementing a particular change make a difference?" "Is this change significant?" If we're measuring quantitative data, we can actually measure degrees of change, but if it's qualitative data, we are looking for patterns (and there are always patterns). From these measurements, we can determine whether the results are effective, efficient, impactful, and/or sustainable. As Arthur Conan Doyle once said, "It is a capital mistake to theorize before one has data."
|
|
|
Post by 33coach on Dec 24, 2015 15:19:38 GMT -6
Mmm...data analysis. In this thread, there seems to be multiple rationales for utilizing data analysis. If the purpose is to find a magic bullet, I can tell you that if one does it exist, it's going to take people much smarter than us to determine where it resides. But, if the purpose is gto simply improve our team, data analysis is a key component to that. Data analysis answers questions, such as, "Does implementing a particular change make a difference?" "Is this change significant?" If we're measuring quantitative data, we can actually measure degrees of change, but if it's qualitative data, we are looking for patterns (and there are always patterns). From these measurements, we can determine whether the results are effective, efficient, impactful, and/or sustainable. As Arthur Conan Doyle once said, "It is a capital mistake to theorize before one has data." im of the opinion that more data means you can make better decisions about how you spend your practice time. and keep in mind, watching game tape is collecting data - just not recording it. but it becomes really interesting when you go through the years, for example: -- over the course of 7 years, we have had winning, losing, and middle of the road records. but one thing i found by logging all the games in hudl and building custom reports is that regardless of year, opponent, or record at the time. we were only 60% on scoring from the 10 yard line. guess what? in my practice plans for next year, im putting in a "going in" period where we work on offense from the 10 yard line. i would have never known that we were neglecting that part of our game had i not had the metrics to prove it.
|
|
|
Post by bignose on Dec 24, 2015 15:25:07 GMT -6
One of the issues I have had with a Hudl style program is how much useful information I can get after breaking down the aggregate of the data. Football is a fluid game, and the situations vary from game to game. Yet I find that many of my opponents do have tendencies.
Do I really need to know that: on the right hash, on third down and 3, between the 30-35 yard line, my opponent has a 50% chance of running a Sweep? 50% probability? That information is almost useless, so you end up with "paralysis by over analysis." If the probability approaches 66%, well, then you've on to something. (I've got a couple of guys on my schedule who are approaching 90% probability of running a certain play in a certain situation. But I've still got to get my 17 year old kid to beat his 17 year old kid).
Since I definitely qualify as being a surly old curmudgeon, having been coaching for the past 42 year, we all look for some useful ditty or convenient little tidbit to apply as a part of the analytical process.
My useful observations: "The team who has the fewest penalties usually wins." "The team that has the fewest unforced errors usually wins"
I'm sure that most of you can add a few of your own. I'd like to hear them.
Broad generalities………qualified by the usually, because there are always exceptions.
However, we all want tilt the odds in our favor.
Bottom line, as my Old Man used to tell me: "Don't ask how, ask how many!"
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 24, 2015 15:30:28 GMT -6
After all the attention that Moneyball got, people started looking for hidden metrics to decypher that magic formula for success and understand the game on a deeper level. Guys are making careers marketing their secret, methodical statistical analysis services to sports teams now.
It's no coincidence that this emphasis on metrics coincides with the rise of computers in coaching and the emphasis on metrics in education, where most of us work.
|
|
jmg999
Junior Member
Posts: 263
|
Post by jmg999 on Dec 24, 2015 16:14:39 GMT -6
Bill James began circulating The Baseball Analyst in June of '82. It took almost two decades for teams to fully buy in, but once they did, the changes were remarkable. Sabermetric ideas literally changed the way the game was played. Basketball is going through this now, and hockey is one of the most data analysis-enlightened sports around. Soon enough, football will be affected.
|
|
|
Post by coachd5085 on Dec 24, 2015 16:57:31 GMT -6
I really don't understand the looking at statistics like its pseudoscience stance. You're going to equate looking at the average starting position of your offense as it relates to your probability of scoring to something like a running back eating a sandwich? Come on. What type of starting field position do you want for each possession? The best possible. Right? That is why I don't understand these types of pursuits. How does having a chart telling me the expected value of starting field position accumulated over a long period of time (or even compiled from OTHER team's data) affect my team on week 9? I want the best possible starting field position. Does it matter to me that starting at the +25 as opposed to the -45 gives me an "expected value" of ____. NO-- because I want to start at the +25 regardless right? Does it matter to me as a coach that a + __ turnover margin has historically resulted in a winning percentage of ____? NO-- because emphasize ball security and creating turnovers already. Does it matter as a coach that my defense is giving up ____ points per possession? NO..because I am coaching them to execute their fundamentals each play. We are giving up points because our pad level is high, our striking is weak, our angles are bad, our reactions are slow etc.
|
|
|
Post by jlenwood on Dec 24, 2015 17:07:15 GMT -6
It seems every other day someone starts a thread asking if _______ is a good metric by which to measure an offense or defense. (Usually an offense...seems like a visor thing to do ) I just don't seem to understand this. Maybe I am growing old and testy, maybe my time away from the trenches has me missing something, but I just don't see any reason for this pursuit of metric measurable data when one already exists in the game--THE score board. Not only that, but we already have ways to evaluate the WHY behind the score-- GAMEFILM. I just don't see why Taking all the points scored this year, and then dividing by yards gained, or starting field position, or # of plays, or possessions that started exactly on the minute, or passing plays called under a full moon benefits the team. A coach can see the score board (metric), and he can watch the game film and see little timmy false stepping on his pull or rounding his break, or playing with high pad level. Am I missing something. You need to slow your roll old man....the grumpy old man title goes to me! I had the thought this season to track missed tackles, a worth pursuit I thought. I was going to track missed, out of position and just plain not as good as an athlete as the guy to be tackled. After about 2 game films it became obvious we were out of position WAY more than any other "metric", so I stopped. Now as a result, I feel I know one of the areas where we need to focus next season. Other than that, I got nothing else to add.
|
|
|
Post by wingtol on Dec 24, 2015 17:07:37 GMT -6
And then you can throw all your metrics out the window when your 16 year old QB's girlfriend break up with him 3 hours before the game.
In HS football there are just way to many factors to consider for me to see where any of this stuff is applicable.
|
|
jmg999
Junior Member
Posts: 263
|
Post by jmg999 on Dec 24, 2015 17:07:40 GMT -6
I really don't understand the looking at statistics like its pseudoscience stance. You're going to equate looking at the average starting position of your offense as it relates to your probability of scoring to something like a running back eating a sandwich? Come on. What type of starting field position do you want for each possession? The best possible. Right? That is why I don't understand these types of pursuits. How does having a chart telling me the expected value of starting field position accumulated over a long period of time (or even compiled from OTHER team's data) affect my team on week 9? I want the best possible starting field position. Does it matter to me that starting at the +25 as opposed to the -45 gives me an "expected value" of ____. NO-- because I want to start at the +25 regardless right? Does it matter to me as a coach that a + __ turnover margin has historically resulted in a winning percentage of ____? NO-- because emphasize ball security and creating turnovers already. Does it matter as a coach that my defense is giving up ____ points per possession? NO..because I am coaching them to execute their fundamentals each play. We are giving up points because our pad level is high, our striking is weak, our angles are bad, our reactions are slow etc. Coach, the problem is that mathematics rules the entire world. Everything we see, hear, taste, smell, and touch can be explained by maths. It's been proven that maths explain everything around us, even the things we don't fully understand, yet. However, what you seem to be saying is that football is unique from the rest of the world. There's no need for mathematically-driven data analysis to occur. I wrote an article for this board in the General Football Topics posting explaining the differences between statistics and analytics and of what use they both are.
|
|
|
Post by coachd5085 on Dec 24, 2015 17:10:10 GMT -6
It seems every other day someone starts a thread asking if _______ is a good metric by which to measure an offense or defense. (Usually an offense...seems like a visor thing to do ) I just don't seem to understand this. Maybe I am growing old and testy, maybe my time away from the trenches has me missing something, but I just don't see any reason for this pursuit of metric measurable data when one already exists in the game--THE score board. Not only that, but we already have ways to evaluate the WHY behind the score-- GAMEFILM. I just don't see why Taking all the points scored this year, and then dividing by yards gained, or starting field position, or # of plays, or possessions that started exactly on the minute, or passing plays called under a full moon benefits the team. A coach can see the score board (metric), and he can watch the game film and see little timmy false stepping on his pull or rounding his break, or playing with high pad level. Am I missing something. You need to slow your roll old man....the grumpy old man title goes to me! I had the thought this season to track missed tackles, a worth pursuit I thought. I was going to track missed, out of position and just plain not as good as an athlete as the guy to be tackled. After about 2 game films it became obvious we were out of position WAY more than any other "metric", so I stopped. Now as a result, I feel I know one of the areas where we need to focus next season. Other than that, I got nothing else to add. Coach- I think THAT was a very valid use of time. In fact it might be one of the best thoughts I have picked up on Huey in the past few years. That isn't worthless data. That is extensive study as to the actual nuts and bolts reason WHY you were missing tackles.
|
|
jmg999
Junior Member
Posts: 263
|
Post by jmg999 on Dec 24, 2015 17:12:04 GMT -6
And then you can throw all your metrics out the window when your 16 year old QB's girlfriend break up with him 3 hours before the game. In HS football there are just way to many factors to consider for me to see where any of this stuff is applicable. The human body performs somewhere in the neighborhood of 400-500 billion chemical reactions per second every second of our lives. Football has nothing on the complexities of the human body, or the Universe for that matter.
|
|
|
Post by coachd5085 on Dec 24, 2015 17:15:58 GMT -6
Coach, the problem is that mathematics rules the entire world. Everything we see, hear, taste, smell, and touch can be explained by maths. It's been proven that maths explain everything around us, even the things we don't fully understand, yet. In all seriousness... Yes, I read that post, and it was solid.
|
|