Announcement

Collapse
No announcement yet.

Statistics and their relevance

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Statistics and their relevance

    For years we have had people quoting statistics for passer rating etc. to argue in the noumerous threads about QB or RB excellence.

    What do you think statistics should/could be used for, and how relevant is it to your evaluation of the performance of the unit that is supposedly measured by that statistic?

    Personally, I think it is a bit ridicolous with all these statistics based arguments, used most agressively in the Cutler vs. Orton or the Cutler vs. Plummer threads.

    As McDaniels said in an interview last week most games are determined by a +/- 7 point margin, and the crucial aspect is to bring the team to the +7 instead of the -7.

    What both Elway and Plummer had was an amazing ability to get the team to give the extra 10% when needed. Elway and Cutler looked so much alike in their abilities, the throws and movement, except when the team really needed those extra 10%. I have seen Elway play some really bad games, but I have never seen him give up.

    Elway was famous for getting the opposing defence to commit penalties, but that is not a QB statistic. Plummer was much better than both to keep the defence guessing, whether he had handed the ball of the the running back, but that is not a QB statistic either. Elway and Cutler were more of a deep thread, probably resulting in more space for the running backs, but that is not a QB statistic either. YAC boosts the QB stat, but he has limited influence on whether the receiver gets 2-3 yards more.

    On running backs Terrell Davis and Clinton Portis are the most regonised RBs we have had in the last 20+ years, and both were great. To me again the stats only tell part of the story. TD in my mind was much better for the specific reason that you could not shut him down. Portis had a game here and there where he could not produce - to me most notably the first Colts Wild Card game where he was worthless - and to me exposed one of the weaknesses Plummer had. In spite of considering him a fantastic QB while in Denver, he was not as able as Elway and Cutler at changing the game in a situation where the running game was completely shut down. Again how the running game functions in a single game is also not a QB statistic, but it sure influences whether the QB plays well and ultimately the W/L.

    It is really boneheaded to ignore the most important statistic the team has -the W/L. A QB is not solely responsible for the W and L and neither are the defence or special team. Bad offences wear out the defences and help give them poor stats. Poor defences and special teams gives the offences poor starting position, which is a double edged sword statistically. A long field makes it really easy to put up amazing passer numbers because you have so much room to work with, but it makes it more difficult to score. Even when the talk comes to Interceptions, they might actually be an ok if it is a calculated risk and your team can contain them. A Hail Mary pass that is intercepted in the redzone, is a whole lot better than most punts IF YOU ARE CERTAIN that you can contain the risk of the return. I think it was actually Favre that said this some years back, that taking an extra chance to score rather than just handing the ball over was a chance they often did not mind.

    W/L is a team statistic, but the QB is the single most influential player on the team, and his play probably counts for some 30-40% - easily the +/-7 points that McDaniels talks about.

    I would like to hear your opinions, and PLEASE lets not turn this into a QB bashing thread.

    "A Bronco Fan Across The Pond"

  • #2
    Fantastic post! A lot more objective than many I have read recently.

    I would just like to add one thing. Basing a players performance by their combine running time means ABSOLUTELY NOTHING. Being able to run in a straight line with no pads on and no pressure to do anything else does not make you a good football player. It makes you a good track star. Unfortunately, being a track star doesn't win football games.
    Is it solipsistic in here, or is it just me?

    Comment


    • #3
      Stats

      Originally posted by BroncoFanDK View Post
      For years we have had people quoting statistics for passer rating etc. to argue in the noumerous threads about QB or RB excellence.

      What do you think statistics should/could be used for, and how relevant is it to your evaluation of the performance of the unit that is supposedly measured by that statistic?

      Personally, I think it is a bit ridicolous with all these statistics based arguments, used most agressively in the Cutler vs. Orton or the Cutler vs. Plummer threads.

      As McDaniels said in an interview last week most games are determined by a +/- 7 point margin, and the crucial aspect is to bring the team to the +7 instead of the -7.

      What both Elway and Plummer had was an amazing ability to get the team to give the extra 10% when needed. Elway and Cutler looked so much alike in their abilities, the throws and movement, except when the team really needed those extra 10%. I have seen Elway play some really bad games, but I have never seen him give up.

      Elway was famous for getting the opposing defence to commit penalties, but that is not a QB statistic. Plummer was much better than both to keep the defence guessing, whether he had handed the ball of the the running back, but that is not a QB statistic either. Elway and Cutler were more of a deep thread, probably resulting in more space for the running backs, but that is not a QB statistic either. YAC boosts the QB stat, but he has limited influence on whether the receiver gets 2-3 yards more.

      On running backs Terrell Davis and Clinton Portis are the most regonised RBs we have had in the last 20+ years, and both were great. To me again the stats only tell part of the story. TD in my mind was much better for the specific reason that you could not shut him down. Portis had a game here and there where he could not produce - to me most notably the first Colts Wild Card game where he was worthless - and to me exposed one of the weaknesses Plummer had. In spite of considering him a fantastic QB while in Denver, he was not as able as Elway and Cutler at changing the game in a situation where the running game was completely shut down. Again how the running game functions in a single game is also not a QB statistic, but it sure influences whether the QB plays well and ultimately the W/L.

      It is really boneheaded to ignore the most important statistic the team has -the W/L. A QB is not solely responsible for the W and L and neither are the defence or special team. Bad offences wear out the defences and help give them poor stats. Poor defences and special teams gives the offences poor starting position, which is a double edged sword statistically. A long field makes it really easy to put up amazing passer numbers because you have so much room to work with, but it makes it more difficult to score. Even when the talk comes to Interceptions, they might actually be an ok if it is a calculated risk and your team can contain them. A Hail Mary pass that is intercepted in the redzone, is a whole lot better than most punts IF YOU ARE CERTAIN that you can contain the risk of the return. I think it was actually Favre that said this some years back, that taking an extra chance to score rather than just handing the ball over was a chance they often did not mind.

      W/L is a team statistic, but the QB is the single most influential player on the team, and his play probably counts for some 30-40% - easily the +/-7 points that McDaniels talks about.

      I would like to hear your opinions, and PLEASE lets not turn this into a QB bashing thread.

      "A Bronco Fan Across The Pond"
      Stats are useful to coaches for analysis. Team stats are much more useful in this regard than individual performances.

      The detailed stats used in preparing a defensive game plan are never seen by regular fans. Tendencies including down and distance, field position, hash mark, etc. Preparing the defense requires much more study of the opponent than offense.

      As far as fans are concerned, generally speaking, stats are for losers trying to find the silver lining in the dark cloud of a loss. As you said, the only stat that really counts is the one on the scoreboard at the end of the game.

      If fans insist on studying stats as a means of analyzing a game, they should concentrate more on team stats: First downs, points scored, time of possession, third down efficiency and stuff like that for Offense; Special Teams are often neglected and the most important stat for ST is average field position (Broncos were last in '08); for Defense the opponent's yardage broken down into categories is a good measurement.

      You are correct in observing that individual performance stats are often used by fans in arguments with each other in such a way that boggles the mind. Their arguments become so convoluted, statistically speaking, that is.

      The narrow margin of victory you cite has much more to do with how a team performs in all three phases of the game (Offense, Defense and Special Teams) which can complement or detract from each other.
      "Stultum est timere quod vitare non potes." ~ Publilius Syrus

      Comment


      • #4
        Whenever you use stats you have to have an established baseline and a way to account for variables to make them useful. This makes comparing football players because there's so many more players and situations on the field and the rules and officiating trends change constantly.

        You also have to be careful, especially with QBs, not to make a few stats appear to be the measure of the position. A RB should be able to pick up a blitz and catch some passes but by and large his job is to run with the football so his rushing numbers weigh heavily in evaluation. But if you try to use passing stats alone to evaluate a QB then you're not telling the whole story of the job. You can get a decent idea of end results but if you don't understand what all the options were then you don't know if the 8 yard gain was the best choice. You also don't know how much of YAC comes from putting the receiver in the best position and what came from the receiver's own efforts. The job is really much much more than just a yardage number. Maybe Bill James will get bored with baseball and figure this out for us.
        \

        Comment


        • #5
          i think that was a very well thought out post with alot of relevance to how misleading stats can be.

          however....EVERYTHING can be measured in statistics. well, maybe with the exception of intangibles like leadership or worth ethic.

          but of the things you mentioned: points and wins are both statistics. true, there are no online researchable stats for false starts committed against a QB, etc, but believe me if you can get in contact with the guys who officially record game stats and stuff, it is there.

          you want to measure TD's unshutdownableness versus clinton portis? well if portis has more game where he fell below [insert yardage total] or maybe below [insert ypc average] then there you go. if you want to measure how well plummer did in games without a running game, then just see how he produced in games where his RB was below that yardage or ypc limit.

          basically, the average person doesn't have time to sit and analyze every detail so general stats like season QB rating and yards can be very misleading. but still, stats can be used to measure anything if you wanted
          disclaimer: if the above post appears to contain outrageously illogical content, ITS PROBABLY SARCASM

          Comment


          • #6
            Originally posted by broncos1997 View Post
            i think that was a very well thought out post with alot of relevance to how misleading stats can be.

            however....EVERYTHING can be measured in statistics. well, maybe with the exception of intangibles like leadership or worth ethic.

            but of the things you mentioned: points and wins are both statistics. true, there are no online researchable stats for false starts committed against a QB, etc, but believe me if you can get in contact with the guys who officially record game stats and stuff, it is there.

            you want to measure TD's unshutdownableness versus clinton portis? well if portis has more game where he fell below [insert yardage total] or maybe below [insert ypc average] then there you go. if you want to measure how well plummer did in games without a running game, then just see how he produced in games where his RB was below that yardage or ypc limit.

            basically, the average person doesn't have time to sit and analyze every detail so general stats like season QB rating and yards can be very misleading. but still, stats can be used to measure anything if you wanted
            You've heard the old adage: "There are Lies, Damn Lies and then there are Statistics"

            Comment


            • #7
              Statistics in my opinion are a very important part of the game. You can't just watch football and think you know everything that's going on. You need to take your knowledge of statistics and combine with your knowledge of what happened on the field.

              Sadly, there are people like Samparnell who apparently think stats are just for losers who sit in their moms basement. They said the same thing about those people in baseball and look what's happen, teams have hired Sabermetric people in the front office and won (Boston Red Sox).

              The basic stats are capable of lying, that's why you dig into situational stats like performance by quarter, score, field position, and passing attempts so that you can compliment your viewing of the game film to the statistics.
              http://goallineblitz.com/game/signup.pl?ref=30412676
              MMO Football Game

              I like McDaniels, he will be a good coach. Just not for this team. Belichick needed his Cleveland stint to become the coach he is now and McDaniels is showing the same signs.

              Comment


              • #8
                Originally posted by Poindexter View Post
                Fantastic post! A lot more objective than many I have read recently.

                I would just like to add one thing. Basing a players performance by their combine running time means ABSOLUTELY NOTHING. Being able to run in a straight line with no pads on and no pressure to do anything else does not make you a good football player. It makes you a good track star. Unfortunately, being a track star doesn't win football games.
                I agree 100% with Poindexter that the combine in mind is just too much information for scouts. Anyone who has read "Blink" By Malcolm Gladwell knows that when Gladwell talked about a poor hospital trying to figure out who needed to stay overnight because they would get tons of people with symptoms that are common with heart attacks (This hospital was poor and trying to save money). So one guy came up with a formula that included only 3 of the risk factors and made a decision tree based on the combination of those 3 risks. The poor hospital decided to put the algorithm to the test and have their own doctors go up against. After collecting data for two years the algorithm trounced the doctors. The algorithm was 70% better than the doctors in figuring out people who weren't having a heart attack. It also was better than the doctors at detecting who was having a heart attack (95% vs 75-89%). The moral of the story is that you don't need all that extra info like 40 time, the cone drill, and wonderlic to determine if someone is a good football player. You look at his game film and his college statistics (Though in college I think game film is more important, unless the Pro's are able to differentiate a 1000 yard receiver at Penn State from a 1000 receiver from Kent State.
                http://goallineblitz.com/game/signup.pl?ref=30412676
                MMO Football Game

                I like McDaniels, he will be a good coach. Just not for this team. Belichick needed his Cleveland stint to become the coach he is now and McDaniels is showing the same signs.

                Comment


                • #9
                  samparnell - Thanks for a very detailed and relevant reply, and for pointing out how detailed it needs to be to be relevant. Bronco1997 states that you can just put in the average YPC to evaluate the running back, but I agree that you need to evaluate the specific situations for it to matter. If a RB has 25 carries and 100 yards, it is a very large difference whether it is 25 3-5 yad carries or two 40 yard carries that does not go to the house and 23 carries for less than one years. A RB needs to do something on basically every down to be a big threat.

                  MJA
                  - Your two posts argues for the importance of the stats, but you dont actually tell how you think it is relevant. We use stats in all walks of life, and they do provide information.

                  How do these stats actualt translate into success for the team over a season?

                  I seem to remember many QBs that had excellent personal stats that did not take the team all the way.

                  Which curiosities have you noted?

                  Comment


                  • #10
                    MJA
                    - Your two posts argues for the importance of the stats, but you dont actually tell how you think it is relevant. We use stats in all walks of life, and they do provide information.

                    How do these stats actualt translate into success for the team over a season?

                    I seem to remember many QBs that had excellent personal stats that did not take the team all the way.

                    Which curiosities have you noted?[/QUOTE]

                    That's a great question.

                    Look at defensive stats, The bottom ten defenses in points per game, only 4 teams finished .500. 2 of those teams were the Broncos and Chargers (Who played in a division with KC, who was ranked 31, and OAK ranked 27).
                    The other 2 teams were Arizona and New Orleans. All of those teams finished exactly .500 (2 made the playoffs). In the top ten meanwhile, every team was at least .500. 7 Teams made the playoffs (11-5 New England did not)

                    Now in regards to QB's doing well themselves but not taking their team all the way. Look at the teams who made the playoffs last year. Their defensive points per game rankings were 1,2,3,4,5,7,9,11,12,13,15,28. Every team but Arizona finished in the top 15, and Arizona finished 8-8. Now look at the QB rankings on those teams based on yards (2,5,6,7,9,13,14,15,17,20,24,27).

                    Something interesting I found was that 2 and 5 in the QB rankings were Kurt Warner and Philip Rivers. The top 2 QB's had the worst 2 worst defense's in points per game. Also, in terms of QB's you have 4 teams that finished outside of the top 15.

                    That shows to me that when it comes to the playoffs, you are more likely to make it if you have a defense in the top 15 points per game versus a QB in the top 15 for yards.

                    Like I've said before, statistics are not the be all end all. You need to watch games AND use statistics to form a good opinion. It's why I get frustrated when people either A) say things that the stats contradict (Plummer had a average defense and a average run game) and B) Talk about leadership and intangibles as being the only important things.
                    http://goallineblitz.com/game/signup.pl?ref=30412676
                    MMO Football Game

                    I like McDaniels, he will be a good coach. Just not for this team. Belichick needed his Cleveland stint to become the coach he is now and McDaniels is showing the same signs.

                    Comment


                    • #11
                      Originally posted by BroncoFanDK View Post
                      samparnell - Thanks for a very detailed and relevant reply, and for pointing out how detailed it needs to be to be relevant. Bronco1997 states that you can just put in the average YPC to evaluate the running back, but I agree that you need to evaluate the specific situations for it to matter. If a RB has 25 carries and 100 yards, it is a very large difference whether it is 25 3-5 yad carries or two 40 yard carries that does not go to the house and 23 carries for less than one years. A RB needs to do something on basically every down to be a big threat.

                      MJA
                      - Your two posts argues for the importance of the stats, but you dont actually tell how you think it is relevant. We use stats in all walks of life, and they do provide information.

                      How do these stats actualt translate into success for the team over a season?

                      I seem to remember many QBs that had excellent personal stats that did not take the team all the way.

                      Which curiosities have you noted?
                      the fact that i used ypc as an example is irrelevant; i don't think that anyone of us thought bell was a better back then anderson just because of his ypc. my point is that anything can be measured with stats, you just have to break it down enough. for instance, the example that you provided about 25 3-5 yard carries or 2 40 yard carries, those ARE statistics, you just broke it down further from my example of ypc. but it is a stat, nevertheless. football is just a collection of data.
                      disclaimer: if the above post appears to contain outrageously illogical content, ITS PROBABLY SARCASM

                      Comment


                      • #12
                        Here's a couple of stats:

                        In the Miami game at Denver, Jay Cutler had more yards passing than Chad Pennington.

                        Miami 26 Denver 17
                        "Stultum est timere quod vitare non potes." ~ Publilius Syrus

                        Comment


                        • #13
                          Originally posted by samparnell View Post
                          In the Miami game at Denver, Jay Cutler had more yards passing than Chad Pennington.

                          Miami 26 Denver 17
                          I don't even know how to respond to such stupidity.
                          http://goallineblitz.com/game/signup.pl?ref=30412676
                          MMO Football Game

                          I like McDaniels, he will be a good coach. Just not for this team. Belichick needed his Cleveland stint to become the coach he is now and McDaniels is showing the same signs.

                          Comment


                          • #14
                            I agree ...

                            Originally posted by MJA View Post
                            I don't even know how to respond to such stupidity.
                            ... that was a pretty stupid game.

                            When you do figure out a response, make sure it's all in stats.:salute!:
                            "Stultum est timere quod vitare non potes." ~ Publilius Syrus

                            Comment


                            • #15
                              Originally posted by samparnell View Post
                              ... that was a pretty stupid game.

                              When you do figure out a response, make sure it's all in stats.:salute!:
                              Glad to see that extremism hasn't died out yet. If you want to ignore stats entirely thats your choosing, just don't be surprised when people think that you have no idea what you are talking about. You used a extremely small sample size, which is never a good idea if you want a accurate measurement.

                              Just look at the situation stats for QB's on NFL.com and tell me that those aren't useful for analyzing a QB's ability. You can look at performance by quarter, half, pass attempts and field position. You can then use your knowledge of watching the game and apply it to the statistics you see in order to come up with a reason for why the player did or did not perform well. For example, Cutler's bad 1-20 statistics show that it's probably due to bad decision making (A 46.4 Completion Percentage and 4 INT's) and also can be the fact that Denver really didn't have a running back who could run through defenders and score (Since Pittman got injured).

                              As I stated before, you need to use both stats and watching of football games to form a intelligent opinion.

                              That's ok though, just call people who use stats "losers". That stereotype isn't old or anything.
                              http://goallineblitz.com/game/signup.pl?ref=30412676
                              MMO Football Game

                              I like McDaniels, he will be a good coach. Just not for this team. Belichick needed his Cleveland stint to become the coach he is now and McDaniels is showing the same signs.

                              Comment

                              Working...
                              X