49ers Best-Coached Team in NFL According to Efficiency Stats
by Tre9er on Oct 26, 2011 2:00 PM PDT
in 49ers Statistical Preview and Review
Email
PrintI read a great article today on Sports Illustrated today, and it wasn't the Swimsuit Edition. This piece was written by Kerry J. Byrne of Cold Hard Football Facts (CHFF) and was focused on our very own San Francisco 49ers.
The title of the piece is "49ers' turnaround thanks to league's best coaching job". I thought at first that it would be a subjective piece; one-man's opinion of how good Jim Harbaugh was at coaching the team based on the same factors the media has been pointing out thus-far: shortened off-season-challenged team beats quality opponents and has top defense, despite much maligned oft-labeled-bust QB, etc, etc.
Once I started reading it I realized that it was actually about quantifying what smart football really is, in terms of statistics. The points made about yards versus points on both sides of the ball are really interesting and things that the national media (and fans) should really consider when evaluating teams based on the numbers.
Join me after the jump for a few examples.
The two key terms the article uses most are "Scoreability" and "Bendability", both of which the 49ers rank first in the NFL.
Scoreability is defined as: offensive efficiency, measured in Yards Per Point Scored (how many yards needed to score a point). The lower the number, the better.
No team has scored more efficiently than the 49ers. San Francisco has produced a meager 1,815 yards of offense. But it has still managed to hang 167 points on the scoreboard. That's an incredible 10.87 Yards Per Point Scored. In other words, the 49ers are good for about one point with each first down.
(Incidentally, the 49ers are on pace to break the record for Scoreability set by New England in 2007, the year the Patriots went 16-0)
Scoreability takes into account ALL points, including special teams, defensive returns (INT's, fumble recoveries), etc.
Well-coached teams score points as many ways as possible. Scoreabilty rewards teams for proficiency in all phases of the game.
The same efficiency applies to the defense as well, making the opposition go as many yards in a game as possible before scoring. Think about it, if you force punts, FG's, or turnovers all game long, the other team is putting up yards, wasting clock, etc. but getting little-to-no points. Red-zone defense is a good example of this. That's why the stat is called "Bendability".
In the context of touchdowns, San Francisco's opponents need 145.3 yards of offense to put 7 points on the board. They have the best "bend-but-don't-break" defense in football. Opponents may generate a lot of yards, but they do very little with all that effort.
Check out the entire article, a great read and good way to look at teams in the NFL on both sides of the ball. It really doesn't matter if you put up 600 yards of offense in a game...if you're only scoring 14 points. Similarly it doesn't matter if you're giving up 500 yards on defense if you're only allowing 10 points.
The 49ers are second in the NFL in total points allowed, despite ranking seventh in yards allowed. Pay no mind to people to speak of the AFC North having the league's "Top Four Defenses". Points is what wins/loses the game.
The same is true with the offense. If you're getting good field position thanks to your special teams and defense, smart football takes advantage of those opportunities and scores without needing a lot of yards.
101 comments
|
1 recs |
Do you like this story?
Comments
so cool it get's double commenting?!?
WHHHHAAAAAATTT?!?!??!
We're as good as we play each week, no better, no worse.
Follow @Tre9er
questions
carried over from previous thread:
1) How are these stats better than points differential? We know that Wins/Losses are highly correlated to points scored vs. allowed (they are in all sports that I know of). Obviously, if yards for/against are normalized, then “Quality” stats reduces directly to points scored/allowed. So what does dividing by yards get you? Basically it weights RZ efficiency and field position higher than moving the ball.
2) What is the evidence that these stats are predictive?
In the descriptive sense – you can explain a lot of football with Red Zone efficiency and Turnover ratio. But that doesn’t mean these stats are predictive. For example, fumble recoveries are not predictive, but they help you win.
To put it another way – show me a team that rates highly on Scorability/Bendability – but LOW on Points differential and has an average W/L record. That would suggest that the team IMPROVES because S/B is measuring “team quality” better than the true “results” (Wins) or secondary results (points scored/allowed)
Because the 49ers are also the top ranking team in http://pro-football-reference.com/ “SRS” (simple rating system) – which is essentially Margin of Victory + Strength of Schedule.
Take something like …oh DVOA. DVOA currently ranks the Jets as the 2nd best team in the league! This is a preposterous ranking…. but it’s based on something. Will the Jets play better than you expect in the next 10 games? If so, this is a big “win” for DVOA.
Finally – There is no evidence whatsoever that high values of S/B have anything to do with Coaching directly.
Now – this doesn’t make the stat bad… I am just not convinced it’s better than what we have.
Brian Sabean: Sing His Praises To The Heavens!
-------
PARPG- Indy post-apocalyptic roleplaying game that seems to have resurrected itself in my absence...
by zenbitz on Oct 26, 2011 2:13 PM PDT reply actions 4 recs
I think the point is putting yourself in manageable situations to score
be it by special teams, playing defense, turnovers, etc…and then capitalizing on it.
It’s not a stat competition, just another aspect to look at. As with many rankings and stats, they are each part of the bigger picture.
We're as good as we play each week, no better, no worse.
Follow @Tre9er
if it's not a competition
why do they charge money for it?
But that’s besides the point. If you want to look at “managable situations to score” – then look at average drive starting position! FO has this in their drive stats btw.
It’s a big mess, and they make outrageous claims – like it measures “good coaching”
Brian Sabean: Sing His Praises To The Heavens!
-------
PARPG- Indy post-apocalyptic roleplaying game that seems to have resurrected itself in my absence...
I think it also suggests a well-rounded team that isn't dependant on any one aspect
We're as good as we play each week, no better, no worse.
Follow @Tre9er
1) how does it suggest this – or rather, how does it suggest this more than margin of victory?
2) what makes you think this is good
What the difference between points scored and points scored / yard. Basically, starting field position. Starting field position is important. Special teams are important. Turnovers are important.
Brian Sabean: Sing His Praises To The Heavens!
-------
PARPG- Indy post-apocalyptic roleplaying game that seems to have resurrected itself in my absence...
Basically
It measures red zone efficiency on offense and defense, as well as special teams efficiency.
We have the best special teams unit in the league, as well as the best red zone defense in the league. Combine that with a pretty good red zone efficiency on offense and you get us as #1 and #2 in “Scoreability” and “Bendability”.
I agree with you that it does not mean much more than what I just stated. We typically have great field position. Our opponents do not. Our opponents put up yards, but we stop them in the red zone. Our offense does not put up yards, party because their scoring drives often start in great field position, but when they get in the red zone, they score.
It may not be very predictive, but if there is a heavy enough correlation between winning and being ranked higher on average in those two stats from CHFF, then it might stand to reason that one could predict game outcomes pretty accurately by relying solely on those two stats. Of course, it’s only as predictive as anything else. Football is played by men, not numbers, etc., etc.
Alex Smith Will Win a Superbowl
http://www.facebook.com/pages/Alex-Smith-Will-Win-a-Superbowl/205058042848290
Red zone efficiency is not very predictive...
…from year to year. I’m not even sure the evidence is that strong that it correlates all that well with winning.
No, but did I only mention RZ efficiency?
Our red zone efficiency, combined with our special teams play, makes us very efficient.
Red zone efficiency is not the only thing. Never was that said. Nor was it stated that red zone efficiency is predictive.
I’m confused on what you’re referring to.
Alex Smith Will Win a Superbowl
http://www.facebook.com/pages/Alex-Smith-Will-Win-a-Superbowl/205058042848290
I bet they also lead the league in penalties per yard of offense. I wonder what that stat proves.
Over
by cybermaldonado on Oct 26, 2011 2:24 PM PDT up reply actions
Holy Lord Balls Negative Nancy
Just enjoy it already!
WE ARE AWESOME!!!
Harbaugh will find a QB and he will succeed.
He is an incredible coach
I just hope he doesn’t burn out…
...and then Singletary opened His Book and it said "thou shalt NOT pass"
Coaches FEAR the "Harbshake"
After Erickson-Nolan-Singletary.....
the Yorks FINALLY got it right with Harbaugh!
One should separate the Papa York nightmarish years from the Jed era
Jed inherited Nolan and made one forgivable mistake of promoting Sing. Most fans were on board with this decision initially. Unlike his Dad, Jed badly wants to win, and so we have the “Jedis”
Kind of erie, the 9ers two best coaches came from Stanford....
and were in our own back yard all along!
Stanford
One aspect of that worth considering: Stanford struggles in recruiting battles given the academic standards. Might require more creativity and mixing up how the team operates to take advantage of what they can.
by David Fucillo on Oct 26, 2011 2:58 PM PDT up reply actions 1 recs
Success at Stanford requires a special kind of coach.
Bob Bowlesby hired Harbaugh after seeing the amazing success the latter had at USD. A coach had to be super-“efficient” to turn the USD program around so promptly. It took longer at Stanford (first winning record only in the third year) because the talent level and team morale was especially low.
Might mean they have more intelligent players.
Many great athletes come from disadvantaged backgrounds because that’s one of the few avenues available to them. There may be a correlation between poor education and great playing, but I don’t see causation per se. All other things being equal (“talent”, motivation, work ethic, skill, etc) I’ll take the intelligent player. And Harbaugh seems to agree, from the coaches on down.
Any NFL team that doesn't fear the Niners really don't know the deal.
And when's the last time a 3-4 had two inside guys who played all three downs?
It's still hard for me to wrap my head around what a fantastic job this team has done
Thanks for the information and thoughts
Come on, Alex, please be good this year..
Follow @JesseReed78
Harbaugh is great BUT.....
He has only shown 6 games to opposing coaching staffs. Other teams have not had enough film and preperation for what a harbaugh team will bring. I am not at all saying hes not awesome, i love him, but i wish everyone would take a step back and relize that other teams will catch up and it may not always be so great.
+1 @helaman
...and then Singletary opened His Book and it said "thou shalt NOT pass"
Coaches FEAR the "Harbshake"
by jimisoursavior on Oct 26, 2011 2:50 PM PDT up reply actions
Stanford
This team is in a way trying to replicate what Stanford does. Stanford game tape isn’t the same, but there are a lot of similarities.
by David Fucillo on Oct 26, 2011 2:59 PM PDT up reply actions
The part you are missing is that Harbaugh and his coaching staff scheme for the team they are playing.
So is Cleveland plays their offense and defense at what they saw the Niners do two weeks ago.. they will lose. If they do guess right, Harbaugh will modify the game plan during the game and Cleveland won’t be prepared for that.
The Niners are not a one dimensional team. Harbaugh allows his OC and DC to game plan where Sing and Nolan would not.
"Whether you think you can, or you think you can't, either way, YOU'RE RIGHT !"
by Eastbayjim on Oct 26, 2011 3:15 PM PDT up reply actions 1 recs
All good coaches overcome this issue pretty much the same way
They adopt one offensive (and defensive) philosophy, but have large playbooks to keep the element of surprise alive, AND tweak their schemes to fit the strengths of the current roster and hide the weaknesses. Coaches watch their own game films carefully to avoid exhibiting predictive tendencies.
Self scouting
I think Harbaugh used that term. It shows that he “gets it”. You have to know yourself as well as your opponent and put your players in the best position to succeed. Sing tried to motivate guys to win their individual battles. Harbaugh and co try to put their guys into winnable battles consistently. Both have merits, but the latter is more successful in the long run.
I don’t think anyone believes Delanie could shut down Suh one on one, heads up consistently. But attacking him from a moving start at a good angle and moving him enough to create a lane… that is very doable and repeatable, particularly when it is novel (i.e. we were the first to really do that to him this season).
But.....
The hallmark of Harbaugh’s offense is that it’s unpredictable. Think back to the detroit game, where Suh and the rest of his linemates had trouble figuring out whether we were going to run or pass. The formations we use allow for diversity and deception in the playcalling. Very tough for defenses to adapt to, even in the long term.
by TotallyBodaciousDude on Oct 26, 2011 3:23 PM PDT up reply actions
this
he get’s that people will see what you do most often, and he will switch it up.
The formations he employs might have birthed a few specific plays last week, then next week he runs totally different plays from that formation.
I think his goal is to be able to punt from goal-line formation, for instance.
We're as good as we play each week, no better, no worse.
Follow @Tre9er
Just the opposite
Theres so much playbook that we haven’t seen yet. He’s going to be showing new plays in the last games of the season which will really be confusing for opponents. Walsh did the same thing, you just never knew what was coming period.
"It's impossible to hide the fire inside" - Bob Seger
I wish you would take a step back...
and realize how to spell realize not relize!
by Ninersincebirth82 on Oct 26, 2011 4:50 PM PDT up reply actions
Really interesting post
At first glimpse, it seems to me that, by far, the largest factor in that offensive efficiency stat would have to be the Niners’ great turnover ratio this season. Alex Smith probably deserves a bit of credit along with the great coaching.
Over
since we are talkin stats
at least I am.
Simple Rating System (SRS) “predicts” the 49ers should be the Browns by 27!! (49ers are +13.5, Browns -12.2… half of which is their Strength of Schedule of -6!!!). Oh and if those don’t add up I give SF 2 for HFA.
(By the same measure, NO over STL by 25 – but the HFA is the other way)
/giant grain of salt
Brian Sabean: Sing His Praises To The Heavens!
-------
PARPG- Indy post-apocalyptic roleplaying game that seems to have resurrected itself in my absence...
Do you know the SRS comparison on Monday night game
Just curious. I think this was the biggest upset of the season so far. I didn’t read or hear anyone who said Jags had a chance of beating the Ravens.
not sure but I am positive it wouldn’t have predicted an upset.
In fact, even AFTER that game, it predicts a 14-point (+/- home field) Ravens victory! However, if they played again I am not sure it would be as close as it was. That’s why they play the games.
Brian Sabean: Sing His Praises To The Heavens!
-------
PARPG- Indy post-apocalyptic roleplaying game that seems to have resurrected itself in my absence...
Counterpoint...
This is a great article, and I fully agree the Niners have become more efficient. Everything Tre said is further confirmed by our 3rd place ranking in team DVOA, another measure of efficiency.
But there’s an important counterpoint to consider that highlights the need for multiple measures of team efficiency. Advanced NFL Stats has us ranked only 16th because of our inconsistency.
As Burke put it, we won the Lions game in the least repeatable way possible — with several long runs and a last second TD on 4th down. Truth be told, this has been happening a lot this season. We’ve gotten some very timely plays that yield big swings in Win Probability.
I disagree with Burke’s ranking and suspect the truth lies somewhere between his and the Football Outsiders’. But he does have a point: we need to become more consistent if we want to keep winning regularly.
I think that this is not quite right
Advanced NFL stats does not rank special teams at all. This is probably most of the difference between them and FO.
FO actually tracks variance… Niners are pretty middle of the pack in overall variance (13th overall, 11th defense) but HIGH variance in offense. Note that if you suck at something, better to have high variance than low!
Advanced NFL stats ranks SF 16th in offense (FO 17th) and 4th in defense (FO same)… but combined much lower, because they ignore #1 special teams ranking.
OSRS (Offense-simple rating system) ranks us 5th and DSRS 1st – (overall 2nd) but these are points-based and the field position (special teams) effect is big – which is why it correlates so highly with scorability/bendability.
Brian Sabean: Sing His Praises To The Heavens!
-------
PARPG- Indy post-apocalyptic roleplaying game that seems to have resurrected itself in my absence...
Good point regarding Special Teams...
..which explains a lot of the difference. But Advanced NFL Stats has us at 18th for offense and 9th for defense in the latest team rankings:
http://www.advancednflstats.com/2011/10/team-rankings-week-eight.html
Also, you make a great point about the wisdom of taking a high-variance strategy when you’re the underdog. But that usually means doing something unconventional like starting Tim Tebow instead of Kyle Orton. We’ve been running a very conservative offense that’s happened to break a few big, timely plays.
huh
I was looking at this page
http://wp.advancednflstats.com/teampage.php – they are ranked 16/4 by EPA and 18/4 by WPA… don’t know what changes them to 9th!
We are not talking about the same kind of variance, although they are related. Variance != Crazy… it just means inconsistent.
Brian Sabean: Sing His Praises To The Heavens!
-------
PARPG- Indy post-apocalyptic roleplaying game that seems to have resurrected itself in my absence...
I understand what "high variance" means...
…I’m just saying not all variance is equal. An underdog does better with a strategy that maximizes risk and reward at the expense of variance. By contrast, we’ve been pursuing a low-risk strategy designed to minimize variance, but that’s actually had the opposite result.
Advanced NFL stats are not very good.
I am convinced by now that statistics that weight contribution towards win probability are not a a very bright idea. Case in point – ESPN’s QB ratings.
Good play is good play, no matter at what point of the game. Attempts to quantify “clutch” are misguided at best.
WPA
is very useful for SITUATIONS – i.e,. how much did a certain play help/hurt the team.
(like: punt or go for it)
For player contributions… I think it’s sketchy.
Brian Sabean: Sing His Praises To The Heavens!
-------
PARPG- Indy post-apocalyptic roleplaying game that seems to have resurrected itself in my absence...
WPA reflects "clutch" factors like situation and score...
…but EPA does not. Each tells you different things.
Yes, and it is a very dumb way to quantify player performance.
They can as well put in moon phases as a parameter.
Situation and score is a global team metric. Not a good play per play performance metric.
It's only dumb...
…if you don’t use it properly. Burke is very clear that WPA is not a predictive stat or the measure of a player’s true ability. But sometimes, it’s useful to know which plays had the biggest impact on the outcome of a game, in which case WPA can be very useful.
I disagree with this basic premise.
If you score 14 points first 7 are numerically equivalent to the last 7.
Yes, there is an additional pressure of the end of the game, but that affects your play only so much. Good players and good teams are demonstrably good for the duration. One could argue that coming out of the gate focused is just as important.
Does not matter how good is your closer if Zito gave up 5 runs in the first inning.
Good metrics do not mix together parameters that have little in common.
Good statistics do agree with common sense. When ESPNs QBR rated Smiths performance in the Eagle’s game in the gutter, it did not make sense to anybody who watched the game. And from purely mathematical perspective it is easy to understand why.
This efficiency statistics on the other hand is clean, makes sense and is certainly revealing.
I don't know what you just said
but I agree with it all
We're as good as we play each week, no better, no worse.
Follow @Tre9er
no, but I think the notion that a TD before the game is closer isn't worth as much as the final TD that wins it is dumb
We're as good as we play each week, no better, no worse.
Follow @Tre9er
It's the same principle!
When you score early in the game, the opposing team has the rest of the game to score themselves. When you score at the end of the game, the opposing team has less time to score themselves.
It’s the difference between bunting for one run early in a game vs. bunting for one run in the 9th inning, with the score tied, and your closer warming up in the bullpen.
Remember, it’s a basic premise of statistical analysis that there are clutch plays, just not clutch players. With all due respect, you and Mindless are confusing the two.
You disagree with the premise..
…because you still are not thinking about these issues correctly.
Scoring a game winning TD is NOT the same thing as scoring a TD in garbage time. It’s not about the “additional pressure at the end of the game” affecting your play. It’s a simple, objective fact that one TD matters more to the outcome of the game than the other. THAT’S ALL WIN PROBABILITY MEASURES.
And I don’t know why you keep mentioning ESPN’s Total QBR , which I have no interest whatsoever in defending. Just because Total QBR abuses the concept doesn’t mean win probability is a useless or irrelevant stat.
You might want to hold up on the "not thinking about these issues correctly" talk
WPA attempts to model football as a Markov chain process, even though the probability of moving to any future state is heavily dependent on previous states and not just the current state. If you try to model a non-Markov process as a Markov chain the further you get down the chain the more inaccuracies creep into the system as value (positive or negative) is carried over into future states instead of being accounted for.
So no, WPA is not that good a statistic for measuring unit level quality, and it’s outright garbage for player performances.
If you just want to measure a drive, it’s interesting, but still very, very flawed.
you had me at
“Markov chain process”
Would it improve (it being “statistical analysis” not “WPA”) as multi-level markov process … with down/distance being sub-states for field position/time/score
Brian Sabean: Sing His Praises To The Heavens!
-------
PARPG- Indy post-apocalyptic roleplaying game that seems to have resurrected itself in my absence...
PS: Garbage Time
I realize you didn’t specifically mention “garbage time,” but the same basic flaw affects your claim about a TD scored at the start of a game’s being equivalent to a TD scored with one second left on the clock. The latter score necessarily has a larger impact on the outcome of the game than the former.
I read the article this morning too ... loved it ... MAYBE WE SHOULD HAVE A WEEKLY POST ON THESE TWO RANKINGS
Given where we are now, think about these two additional factors:
1. With the lack of off-season installation time and need to focus on specific game planning each week during the season, the offense hasn’t had time to thoroughly polish play timing and refinements. What happens when they begin to refine the execution?
2. What percentage of the “total” (I’m sure that they are constantly adding new ideas) Harbaugh / Roman offense has really been installed at this point? What happens next year when the offense picks up the rest?
I dare say that we’re in for an exciting ride in the next several years.
by 49erFanSince1950 on Oct 26, 2011 4:00 PM PDT reply actions
Wow.
Since we began tracking these indicators in 2004, no team has ranked No. 1 in both offensive and defensive efficiency.
I do know it is statistics.
But when you see an outlier like that, there is usually some truth behind it.
that and there are so many statistics that, when combined, show this:
WE ARE A PRETTY DECENT FREAKING TEAM
We're as good as we play each week, no better, no worse.
Follow @Tre9er
everybody keeps telling me
remember Sporano turning around the 1-15 dolphins! remember mcdaniels 6-0 start with the broncos!
i hope we’re different :(
Wow this Kerry Byne guy gets around!! EcERyda69 posted one of his vids last week
First TPN Sports now SportsIllustrated…
Niners,Nets,Reds & USC!!!
The Most Interesting Man In The World---->Mikhail Prokhorov!!!
Kerry Byrne was saying similar things after we beat the Bucs.
Check out this video:
http://www.youtube.com/watch?v=DSvQkrkSsB8
Alex Smith Will Win a Superbowl
http://www.facebook.com/pages/Alex-Smith-Will-Win-a-Superbowl/205058042848290
Nice , more stats , wonder how many coaches and player actually look at this stuff ...
( Harbaugh ) … hey intern Tim , I need some stats … STAT
( Tim ) … Coach , i got your stats …
( Harbaugh ) … what you got Tim …
( Tim ) … Well . DVOA , POQE . DAVE and CHFF .. says we’re sapposta suck , not be 5 – 1
( Harbaugh ) … Nice Tim ,…freaking geek … whats that coach , Oh have a nice week …!!
I'm your " Huckelberry "...it's just my game ...Jimmy Raye your no daisy ...!!
by Edggy on Oct 26, 2011 7:56 PM PDT reply actions 1 recs
Players almost certainly do not, more or less because knowing about advanced stats is the thing that will likely help a player the least. His game improves through repetition, technique, conditioning, etc., and the stats follow.
They may be useful tools for coaches, but the coach is more directly concerned with things like scheme and discipline and adjustments and packages.
The people who should be looking at them are GM-types and contract-types. The guys involved most directly in player acquisition need to be able to evaluate players with as unbiased and wide of a scouting and statistical brush as possible.
Once more, coming to you by proxy.
by howtheyscored on Oct 26, 2011 11:24 PM PDT up reply actions
Does the 9ers front office "fancy" being like the Ravens organization?
Lets see….
Coach Nolan was hired away from the…….Ravens
Coach Singletary got his coaching roots from the….Ravens
Coach Harbaugh brother is the H/C of the……Ravens
Gm Balke was said he repected Ozzie Newsome, the GM of the…..Ravens
Hummmmm, makes one wonder?
and Frank Gore & Ray Rice are very similar in height, weight, running style
Rice of the…….Ravens
They won a superbowl in the last 10-11 years
and are perennial playoff team…I’ll take that for a comparison
We're as good as we play each week, no better, no worse.
Follow @Tre9er
This measure is moronic
If my team goes 3 and out on offense 10 times in a game and on the last punt the opposing kick returner muffs the ball at his own1 yard line and my team then takes 4 downs to run it in for a score, my team would have the highest offensive efficiency in the history of the NFL. One yard of offense for 7 points.
Or 34 offensive plays to make one yard.
I guarantee you that the whole stadium would be calling for my team’s head coach, offensive coordinator & QB to be fired because it was the worst display of offensive football anyone would have seen.
Offensive efficiency my azz.
Or you could take 34 plays to march 79 yards down the field 10 times
And turn the ball over each time to have 790 yards of offense and zero points. Is that offensive efficiency?
Intensity-Attitude-Focus-Execution-Imposition of Will
Time to run the table Fellas. Let's Go Niners!
That is offensive inefficiency.
And opposing defense bendability.
You example would be valid.
If that statistic was applied to a single drive.
Your scenario is not very repeatable. In the long run it will average out. But that play would certainly contribute to the efficiency statistic, as it correlates with special teams play.
As with any measure you need a sufficiently large sample for it to be a useful indicator of the teams performance.
Balony - my example was a complete game. Very repeatable by teams with crap offenses.
“Offensive efficiency” just says you’ve made a habit of short drives and sometimes you score.
Which is pretty much what the Niners have done. Surprise!!!!
How is that better than a team that repeatedly maintains long scoring drives?
Because it ignores the value of excellent special teams play and winning the field position battle
And your example isn’t repeatable — unless you are counting on muffs and other low percentage plays to score. Given how rare those are versus how many possessions you will have your “scorability” over any decent stretch of games is still going to be low.
this
you wont get a muffed punt or TO in your RZ every game, so the model isn’t predictive after one game
We're as good as we play each week, no better, no worse.
Follow @Tre9er
I don't understand
the need or want to combine “offense” with starting field position obtained by defense or ST. Why not just use “points”?
I agree that ST and field position gains by the defense are important for the offense scoring points – but they should be separated because they are achieved by different units.
For example – The 49ers have the #1 scorability in the NFL. Do they have an above average offense?? Can’t tell. Ditto the #1 bendability ( because likewise the defensive component of points allowed is intermingled with field position from offense and ST).
Brian Sabean: Sing His Praises To The Heavens!
-------
PARPG- Indy post-apocalyptic roleplaying game that seems to have resurrected itself in my absence...
It's a truism with statistics that making up an absurd situation to break the statistic is not, generally, a problem with the stat.
You want a stat that functions well over the broad range of plausible real-world results. There are lots of stats that you can break by imagining a situation that has never happened in the history of the league and pointing out how it gives an absurd result.
OK, we suck.
get your info from SportsCenter. They know it all.
/mylife
We're as good as we play each week, no better, no worse.
Follow @Tre9er
My biggest concern with this stat is that it seems self-evidently dependent on unstable things.
Special teams tends to be inconsistent, as does red zone performance. Fumble recoveries are basically random.
Yet all of those things would have a huge affect on these stats.
And because these stats are based on such inconsistent things, you have to ask yourself an important question:
Are we leading the league in these stats because we’re good … or because we’ve been lucky?
(Not necessarily an either-or, mind you).
I mostly agree...
…but aren’t some aspects of special teams (e.g., kickoff distance, as well as coverage and returns) repeatable and thus predictive? Kickoff distance, in particular, is a skill that has a large impact on the outcome of games.
I think they are, yes
That is to say – there is a correlation between ST performance year-to-year and probably by coach as well.
Brian Sabean: Sing His Praises To The Heavens!
-------
PARPG- Indy post-apocalyptic roleplaying game that seems to have resurrected itself in my absence...
Well
Special teams touchdowns are pretty inconsistent. Andy Lee is consistently awesome, and our defense has been doing a pretty good job of forcing three and outs.
Combine those two factors, and we’ve been seeing quite many short fields. Now defensive takeaways in great field position are to some extent unpredictable, however they do become more likely when the play of our defense forces opposing offenses into higher risk/reward plays or just plain out stupid plays (Hello Ronnie Brown) in order to move the ball or score points.
Luck never hurts, but it also is pretty inconsistent. Now it may be that we have had a good deal of luck over a small sample size. I would also feel much better about these efficiency measures over the course of 10 games rather than 6.
My favorite efficiency stat:
Points per minute scored vs. points per minute given up. A football game is only 60 minutes long so you need to focus on how much you score during that time.
Joe, Jerry, Steve, and....Who is next?
Cannot reply to specific comments
Don’t know what’s going on with my computer. Clicking “reply” just takes me straight to the top of the page.
Anyways, to realfan49:
The statistical model is meant to take into account the empirical data in a meaningful way. Coming up with eccentricities is not an argument against the model, because the model does not have to take such events into account, given that they occur rarely (i.e., are not significant enough to matter). To make the point, let me dramatize it:
Every time Brady drove 99 yards down the field one day, on every single drive, his wide receiver broke his leg at the 1 and fumbled the ball. This model would then call Brady’s offense inefficient, when this is clearly not the case – all his receivers broke their legs! Clearly, therefore, the model is wrong.
That’s the argument you’re making, only slightly more realistic (slightly). The truth is the model does not have to put much weight on the Brady broken leg scenario, any more than it does the no first downs all game and then punt return for touchdown scenario, because neither scenario are realistically going to occur enough for the model to need to change to explain it.
Your example would be a problem in the model if your example was empirical fact. But it isn’t.
Alex Smith Will Win a Superbowl
http://www.facebook.com/pages/Alex-Smith-Will-Win-a-Superbowl/205058042848290
Every time Brady drove 99 yards down the field one day, on every single drive, his wide receiver broke his leg at the 1 and fumbled the ball.
Hmmm… curious. Wouldn’t the WR get penalized 15 yards for “touching the Brady?”
Brian Sabean: Sing His Praises To The Heavens!
-------
PARPG- Indy post-apocalyptic roleplaying game that seems to have resurrected itself in my absence...
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Recommended FanPosts
Recent FanPosts
In Association With
Shutdown Corner - NFL - Yahoo! Sports
- Former high-school star Brian Banks does serious time on trumped-up charge, now hopes for an NFL shot
- Former ‘idiot kicker’ Mike Vanderjagt in trouble after accosting junior-high heckler
- Adrian Peterson blows teammates away in footraces; appears to be on the mend
- UPDATE: Charlie Sheen didn’t buy Lawrence Taylor’s Super Bowl ring
- Alex Smith gets blown up by Jon Beason over Cam Newton comments
