Statistical analysis of football

[billsfan89]

Since Sabermetrics took over baseball there has been a big push to analyze sports differently on a more intellectual level. Do you feel that Safermetrics or the Sabermetrics of football are legit or just over thinking what wasn't broken.

 

I feel that baseball and football are two completely different sports. In baseball you get a more static event in an at bat because every at bat starts the same 0 balls 0 strikes and a pitcher throwing at them. Yeah who is on base and other factors come in to play but you can do more in baseball to adjust for that with splits and other projections. As well as that a player's batting isn't hurt or benefited that much by who their teammates are

 

In football both the defense and offense are more of a symbiotic relationship. A QB Can have his stats altered by his surrounding cast. A secondary can look worse due to a lack of a pass rush or run defense ( like the Bills secondary has suffered ) and vice versa a pass rush could make an average secondary look better ( Like the NY Giants secondary benefited from )

 

So while I look at the new stats I take them with a grain of salt knowing that you have to look at the circumstances that surround those stats and projections.

 

How much do you value what is going on as far as new stats and analysis in football.

[Coach55]

Since Sabermetrics took over baseball there has been a big push to analyze sports differently on a more intellectual level. Do you feel that Safermetrics or the Sabermetrics of football are legit or just over thinking what wasn't broken.

 

I feel that baseball and football are two completely different sports. In baseball you get a more static event in an at bat because every at bat starts the same 0 balls 0 strikes and a pitcher throwing at them. Yeah who is on base and other factors come in to play but you can do more in baseball to adjust for that with splits and other projections. As well as that a player's batting isn't hurt or benefited that much by who their teammates are

 

In football both the defense and offense are more of a symbiotic relationship. A QB Can have his stats altered by his surrounding cast. A secondary can look worse due to a lack of a pass rush or run defense ( like the Bills secondary has suffered ) and vice versa a pass rush could make an average secondary look better ( Like the NY Giants secondary benefited from )

 

So while I look at the new stats I take them with a grain of salt knowing that you have to look at the circumstances that surround those stats and projections.

 

How much do you value what is going on as far as new stats and analysis in football.

I think one should take great heed in what that stats say. But they should be used as a tool as opposed to in baseball, where it has become entirely statistical. Certain items, such as how often you should throw vs. pass on 1st down is valid. However, if a team were to always throw on first down because statistics tell you, the opposing team with ultimately negate it since they will play the appropriate defense.

 

Stats in football could be very dangerous. A team could learn how to exploit weaknesses in the opposing team when analyzing performance metrics. But it still comes down to who shows up that day and whether the players perform.

[billsfan89]

I think one should take great heed in what that stats say. But they should be used as a tool as opposed to in baseball, where it has become entirely statistical. Certain items, such as how often you should throw vs. pass on 1st down is valid. However, if a team were to always throw on first down because statistics tell you, the opposing team with ultimately negate it since they will play the appropriate defense.

 

Stats in football could be very dangerous. A team could learn how to exploit weaknesses in the opposing team when analyzing performance metrics. But it still comes down to who shows up that day and whether the players perform.

 

From a fans perspective we need stats because i cant look at how good every player on each team is but because the nature of the game is one where stats have circumstances surrounding them. I like your approach but I think coaches use film more than ever to gameplan instead of metric stats

[Fingon]

Stats are invaluable to anyone's ability to analyze the pro-game. You can quantify many of the things that players do, and through the compilation of these you can gauge how good a player is. One individual stat may lie, but the probability that a multitude of them do is very slim. All things considered, stats are just another tool that can be effectively used to evaluate players.

[K-9]

When it comes to pro football, there are only 6 stats that lend true insight and are worth caring about:

 

1.) Rushing yards per attempt AGAINST

 

2.) Rushing yards per attempt FOR

 

3.) Yards per play AGAINST

 

4.) Yards per play FOR

 

5.) Rushing attempts per game AGAINST

 

6.) Rushing attempts per game FOR

 

All the other stats are what I classify as glamour stats. I'm not saying they're not indicitive of anything just that glamour stats, when it comes to predicting outcomes don't serve as the foundation for all others. The stats above, the 'nuts and bolts' stats, do.

 

GO BILLS!!!

[berndogg]

To elaborate on the original post, I think for teams, stats can be very telling in football as far as how well they match up against other teams, where they need to improve, how well they're going to do in the future, etc. etc. However, as far as stats for individuals go I think that in football more often than not they are misleading for the reasons mentioned above as well as a variety of others. In baseball, I think stats do a very good job of telling the story, there are a few exceptions, but in my opinion it's largely an individual game.

[KRC]

Stats without context are meaningless. A few good points on this subject (namely the Bill James interview on 60 Minutes on Sunday) have been raised on the PFRA messageboard.

 

For those of you who don't know, the OP of that thread (Bob Carroll) is a legend in the industry. He and Pete Palmer have done extensive work with trying to redefine the passer rating system, as well as looking at other areas of football analysis. Tod Maher, Sean Lahman and others have also done extensive work in this area. I highly suggest everyone read the following books to get some perspective in how to treat football statistics:

 

The Hidden Game of Football

 

The Pro Football Historical Abstract

[billsfan89]

Stats without context are meaningless. A few good points on this subject (namely the Bill James interview on 60 Minutes on Sunday) have been raised on the PFRA messageboard.

 

For those of you who don't know, the OP of that thread (Bob Carroll) is a legend in the industry. He and Pete Palmer have done extensive work with trying to redefine the passer rating system, as well as looking at other areas of football analysis. Tod Maher, Sean Lahman and others have also done extensive work in this area. I highly suggest everyone read the following books to get some perspective in how to treat football statistics:

 

The Hidden Game of Football

 

The Pro Football Historical Abstract

 

I have read the Hidden Game of Football and I took time to read that thread and I think that it rings true. Football stats have to be kept in context to the players surroundings/teammates because the stats are heavily influenced by out side influences. Another good read or glance is pro football prospectus it is the complete guide to all the new stats used in the NFL.

 

I think the best stat never used is the tendencies of the quarterback to what part of the field he completes most of his passes . The best QB's throw all over the field and the worse ones usually throw heavily short passes and screens. Edwards threw more in the middle and longer length of the field more than he did his rookie year. He showed improvement but has a long way to go but TO should help that out.

[stuckincincy]

When it comes to pro football, there are only 6 stats that lend true insight and are worth caring about:

 

1.) Rushing yards per attempt AGAINST

 

2.) Rushing yards per attempt FOR

 

3.) Yards per play AGAINST

 

4.) Yards per play FOR

 

5.) Rushing attempts per game AGAINST

 

6.) Rushing attempts per game FOR

 

All the other stats are what I classify as glamour stats. I'm not saying they're not indicitive of anything just that glamour stats, when it comes to predicting outcomes don't serve as the foundation for all others. The stats above, the 'nuts and bolts' stats, do.

 

GO BILLS!!!

 

Average drive starting position, for your team and the opponent.

 

 

BUF's ST performance the past several years masked a lot of fundamental deficiencies IMO. Along with the Cover-2 stuff, which attenuated things. Finishing up the last 10 games 2 and 8 (albeit with those two rotten weather games) wasn't encouraging. I think they got figured out and dealt with. 0-6 in the division, all after the 5 - 1 start. Ugh.

 

It will be interesting to see how the '09 rule changes affecting the STs will play out, for BUF and everybody else.

[VJ91]

How much do you value what is going on as far as new stats and analysis in football.....

 

I don't value them at all. Every sport already has more then enough stats to show us game in and game out. The only stat that I care about is the one that has been stuck at 7-9-0 for three years now. If the team can somehow improve on that stat by at least 2 or more wins, and sneak into that 6th spot in the AFC, I could care less about all the individual team and player stats they compiled to get there!

[KRC]

I have read the Hidden Game of Football and I took time to read that thread and I think that it rings true. Football stats have to be kept in context to the players surroundings/teammates because the stats are heavily influenced by out side influences. Another good read or glance is pro football prospectus it is the complete guide to all the new stats used in the NFL.

 

I think the best stat never used is the tendencies of the quarterback to what part of the field he completes most of his passes . The best QB's throw all over the field and the worse ones usually throw heavily short passes and screens. Edwards threw more in the middle and longer length of the field more than he did his rookie year. He showed improvement but has a long way to go but TO should help that out.

 

That is a good point. People like to look at QB ratings, etc to judge the quarterback. The QB rating is nothing more than an efficiency rating. If you toss a lot of dump-off passes, you will have a high QB rating. If you have a lot of yards with this, it has nothing to do with the QB, but more the RB or WR catching the ball and making the plays. If you add YPA, it can be influenced by a couple of deep passes and a bunch of short passes. Still, it tells you nothing.

 

That is why I say stats are meaningless without context. You can't just look numbers up in an encyclopedia to find out who are the best players.

 

Here are a few more threads talking about stats:

 

Linky 1

 

Linky 2

 

Linky 3

[Bill from NYC]

BUF's ST performance the past several years masked a lot of fundamental deficiencies IMO. Along with the Cover-2 stuff, which attenuated things.

 

Great points cincy!

 

I always put a lot of stock in time of possession, because the team which dominates the clock is generally running the football successfully, and picking up the first down on 3rd and short, or so it would seem.

[Coach55]

When it comes to pro football, there are only 6 stats that lend true insight and are worth caring about:

 

1.) Rushing yards per attempt AGAINST

 

2.) Rushing yards per attempt FOR

 

3.) Yards per play AGAINST

 

4.) Yards per play FOR

 

5.) Rushing attempts per game AGAINST

 

6.) Rushing attempts per game FOR

 

All the other stats are what I classify as glamour stats. I'm not saying they're not indicitive of anything just that glamour stats, when it comes to predicting outcomes don't serve as the foundation for all others. The stats above, the 'nuts and bolts' stats, do.

 

GO BILLS!!!

I completely disagree. Running and stopping the run is old school football mentality and although correlates to wins it does not produce wins (Stats get skewed at the ends of games when a team goes to run only mode in the 4th quarter after the game is in hand). Ultimately you would like a 50/50 pass/run mix, but in reality it should be more 65/35 with the evening coming out at the end. The stats in order of importance are

 

1. Turnover differential

2. % of negative plays (offense)

3. 3rd down % (both offense and defense)

4. Average drive start

5. Completion % (offense)

6. Yards against per play (defense)

7. Yards per play (offense)

 

The problem is that these are typical stats that typical fans would look at. However, if you were looking at this from a pure statistician point of view, it would be looked at from a more granular perspective examining types of plays in various situations (i.e. how different runs perform in 2nd and 5 yds or less) or looking at individual performances (such as the stats KC Joyner has recently discussed in other posts).

[K-9]

Great points cincy!

 

I always put a lot of stock in time of possession, because the team which dominates the clock is generally running the football successfully, and picking up the first down on 3rd and short, or so it would seem.

 

Once again, related to the "nuts and bolt" stats.

 

GO BILLS!!!

[RLflutie7]

When it comes to pro football, there are only 6 stats that lend true insight and are worth caring about:

 

1.) Rushing yards per attempt AGAINST

 

2.) Rushing yards per attempt FOR

 

3.) Yards per play AGAINST

 

4.) Yards per play FOR

 

5.) Rushing attempts per game AGAINST

 

6.) Rushing attempts per game FOR

 

All the other stats are what I classify as glamour stats. I'm not saying they're not indicitive of anything just that glamour stats, when it comes to predicting outcomes don't serve as the foundation for all others. The stats above, the 'nuts and bolts' stats, do.

 

GO BILLS!!!

 

 

The only stats I look at are team stats. Because it's a team game. The yards per play thing would confuse me. I look at yards gained and yards given up during the entire game. I know Parcells says rushing attempts per game are a huge indicator.

 

I think coaches concentrate on turnovers and I think that's a mistake because turnovers are not predictable. I ignore turnovers because you could be looking at dumb luck.

 

I also don't put much value in passing stats because a QB can have off days or his team runs to ball too much.

[K-9]

I completely disagree. Running and stopping the run is old school football mentality and although correlates to wins it does not produce wins (Stats get skewed at the ends of games when a team goes to run only mode in the 4th quarter after the game is in hand). Ultimately you would like a 50/50 pass/run mix, but in reality it should be more 65/35 with the evening coming out at the end. The stats in order of importance are

 

1. Turnover differential

2. % of negative plays (offense)

3. 3rd down % (both offense and defense)

4. Average drive start

5. Completion % (offense)

6. Yards against per play (defense)

7. Yards per play (offense)

 

The problem is that these are typical stats that typical fans would look at. However, if you were looking at this from a pure statistician point of view, it would be looked at from a more granular perspective examining types of plays in various situations (i.e. how different runs perform in 2nd and 5 yds or less) or looking at individual performances (such as the stats KC Joyner has recently discussed in other posts).

 

Well, since you directly used two of my stats you are off to a good start. Tell you what, don't take my word for it. Ask NFL coaches what they think. Do an historical analysis of teams that win more than they lose. It may surprise you. Here's just a tidbit, teams that have the higher number of rushing attempts per game win 75% of the time.

 

When you use the term "types of plays in various situations" that's just a different way of saying "tendencies." Something coaches ALWAYS have/will look for.

 

Turnover differential, I agree, is a KEY stat. However, it too, is predicated on the 'nuts and bolts' stats. Turnovers 'usually' occur when a team is playing from behind.

 

Favorable drive starts allow offenses to be more selective to be sure but if you're NOT putting yourself in manageable down/distance situations on offense, defenses WILL dictate (there's that Yards per play AGAINST stat again), make you one dimensionsal, and your favorable field position is negated. That's why I put the AGAINST stat first.

 

You may say running and stopping the run is 'old school' but it is NOT. Especially in the NE in November and December when you play half your schedule. It's very much old, new, and every school in between.

 

Again, don't take my word for it. Talk to coaches. Ask them what the first stat they look at for opposing teams is.

 

GO BILLS!!!

[K-9]

Average drive starting position, for your team and the opponent.

 

 

BUF's ST performance the past several years masked a lot of fundamental deficiencies IMO. Along with the Cover-2 stuff, which attenuated things. Finishing up the last 10 games 2 and 8 (albeit with those two rotten weather games) wasn't encouraging. I think they got figured out and dealt with. 0-6 in the division, all after the 5 - 1 start. Ugh.

 

It will be interesting to see how the '09 rule changes affecting the STs will play out, for BUF and everybody else.

 

Hey, I like better field position as well as the next guy. Good drive start position is always preferable as you know because it gives you more options on offense. However, if you don't put yourself in manageable down/distance situations and allow the defense to dictate, it's a wash. There's that Yards Per Play FOR/AGAINST stat I was referring to.

 

The rule change is going to be a critical one, I agree.

 

GO BILLS!!!

[BuffaloBill]

Statistical analysis can be a powerful tool. While I agree with your general tone that there is a big difference between baseball which can be analyzed at an individual level more effectively than football I do not think the process should be set aside for football. In the trenches there are clearly one on one match ups arguably like a one on one match up between a pitcher and hitter. If your line beats their line on a play then it comes down to does your QB complete a pass to a wide receiver and at what distance? these are all "plays within the play" that could potenially be looked at.

 

Numerical analysis is a tool just like watching game film is a tool. Both have their place and both should be used. Coaches talk about players and teams have tendencies stats can prove this out. They can also help go beyond what is a players size and speed. Used wisely they should help get to how effective is a player.

[K-9]

The only stats I look at are team stats. Because it's a team game. The yards per play thing would confuse me. I look at yards gained and yards given up during the entire game. I know Parcells says rushing attempts per game are a huge indicator.

 

I think coaches concentrate on turnovers and I think that's a mistake because turnovers are not predictable. I ignore turnovers because you could be looking at dumb luck.

 

I also don't put much value in passing stats because a QB can have off days or his team runs to ball too much.

 

Total yards per game is not as indicative as the total yards per play because yards per play directly correlates to the number of plays run and the yards gained/given up per each. Total yards in a game are often skewed, especially by losing teams because they often have HUGE passing numbers against prevent defenses. That's why RUSHING yards per attempt for/against is more valuable. Show me a team that have astronomical passing yards every game and I'll show you a losing team more often that not. Show me a team that runs the ball more, for a higher average while holding opponents to lower averages and I'll show you a winning team way more often than not.

 

GO BILLS!!!

[Coach55]

Here's just a tidbit, teams that have the higher number of rushing attempts per game win 75% of the time.

This statistic is not causal. This is a great misnomer in the NFL. Assume the following - in the first half Team A (a power running team - say PIT) and Team B (a wide open offense - say NE). Lets say PIT runs 25 plays of which 15 are runs and NE runs 25 plays 15 of which are passes. They go into the half with NE leading by 14. In the 3rd quarter PIT and NE hold their gameplans and the lead stays the same. PIT running 13 plays (8 runs) and NE running 13 plays (5 runs). In the 4th quarter, PIT trying to claw back runs 12 plays (11 passes) and NE runs 12 plays (11 runs). In the end PIT has run the ball 24 times (48% runs) vs NE's 26 times (52% runs). Now NE has run the ball more and has more attempts, but this stat is a result of their winning not the cause of it.

 

I am not stating that this is how it always plays out, but if you examine those statistics and take out the run only/pass only downs at the end of games, you will get a much different outlook on run/pass ideal ratio.