Warning, this post is meta-stat nerd.
What is Success Rate, and How Did It Come To Be?
The first question is pretty straightforward and the second I can only guess.
Success Rate is a measure is an attempt to measure how good a player or team is at the traditional concept of “staying ahead of the chains.” There are some slightly different calculations but for the most part a success is defined as at least 40-50% of yards to go on 1st down, at least 50-70% of yards to go on second down and first down achievement on third or fourth down. Typically the target is 50% success rate.
Although I doubt there is any recorded history on how this came to be (I believe its origin or at least its popularization comes from Football Outsiders) I have two theories. The first is that this is how football fans, players, and coaches have been conditioned to think, especially old school, grind-it-out football folks. You still hear it often among clichéd commentators: the offense’s number-one priority is to stay ahead of the chains, don’t put yourself in bad down and distance, stay away from obvious passing downs. All of these things are good things for a football to do.
The second reason I think it came to be is that advanced football stats came to be after advanced metrics for baseball had come a long ways. One of the key tenants of Moneyball/SABR revolution in baseball is that On Base Percentage >>> Batting Average. On top of that, one of the fundamental advanced baseball stats is OPS, On Base Percentage Plus Slugging Percent, a combination of Success and Magnitude. One paralleled by Football Outsiders* in their S&P metric.
*I want to be clear that this is not a critique of Football Outsiders. They do tremendous work and are at the forefront of advanced football analysis.
Why Football is Not Baseball
Good OBP is critical for baseball because you are dealing with a finite, irreplaceable resource, outs. You get 27 of them per game. Once you generate an out there is no way to get it back; you are 1 step closer to the end of your chance to score, and you only have 27 total steps per game. OBP measures a team or individual’s ability to forego outs when they come to the plate. Not getting out will always improve your chances of winning while getting an out will almost always decrease your odds of winning (this is not an article about the sacrifice bunt).
Contrast that with football, where the only finite resource is time. Even if the quarterback gets sacked and loses 10 yards, one play later the effect of that loss can be wiped out. In a sense a set of downs is finite, but not an individual set of downs. If there were a team correlation, first downs converted would be more appropriate and I don’t really see a true individual equivalent.
The Goal Is To Score Points
Consistently being in good down and distances is not a bad thing, but it’s not nearly as important for today’s offenses. Modern offenses have a much greater ability to convert unfriendly down and distances than offenses of old. Plus, the offense’s goal is to score points, not get first downs. Getting first downs obviously helps score points, but a metric like EV/PAN that directly accounts for how each play contributes to scoring is a much stronger measure, not just a complimentary stat like Slugging Percent. In baseball the complimentary stat is needed because of the finite nature of outs. In football, everything is a sliding scale and categorizing plays as pass-fail is simply too black and white for a sport that has more gray.
A couple of examples of how success rate can be misleading (first down gain, second down gain, third down gain):
4,3,2: This is a 67% success rate but is a three and out.
3,3,4: This is a 33% success rate but a first down, plus the first two plays are nearly identical but the first two downs of the first group are both successes and the second group are both failures. Over a large group of data some of these will iron themselves out, but why put such a black and white metric over something that is not. 2nd and 7 is almost the same as 2nd and 6, but 2nd and 1 is very different from 2nd and 6. Success rate completely misses the magnitude of plays.
This is why for football, an Expected Value model is much more valuable. With an enough data, you can get a pretty good description of the expected points based on all down, distance and yardline combinations. Once you have this you can evaluate the shades of gray for each play. A three yard carry on first and ten is nearly as good as a four yard one. A nine yard carry is even better. Expected Value can quantify the subtle and substantial differences between plays. The value difference between first and ten and the twenty and first and ten at the thirty will be the same whether it was one ten yard play or three runs totaling ten yards, although the value per play will justifiably be better. Success rates can vary wildly based on how you get from point A to point B, EV only carries where you start and where you finish.
What is Success Rate Good For?
It is an interesting stat and isn’t totally without value, I just think that it is unnecessary and shouldn’t be a fundamental part of team evaluation. There are lots of stats that fit this characterization. For a lot of teams it’s how they mentally operate, especially in the running game. Success rate does a good job evaluating running backs in traditional ground games. It might not totally align with scoring points and winning games, but it does align well with accomplishing a team's offensive objectives. Running backs often get tightly bunched near the mean in an EV model but success rate can be a way to further separate individual backs. Success rate will hold up between the tackle pounders but knock down the home run threat. EV may consider them the same (or more likely the home run threat will be higher) but the consistency of the old school back will be valued better by success rates.
I don’t think success rate has much value for the passing game. Completion percentage and YPA are more than adequate to indicate both explosiveness and consistency.
Coming Next: The Wisconsin Case Study and Optimal Offense and Defense Response
The underlying context of “ignore success rates” is that the traditional running game is overrated. If your main goal as an offense is to avoid bad third downs, and you are good at it, you will likely end up with a lot of third and short or third and manageable. Even if you they are all “good” third downs, each third down is a chance for the defense to take the field. We all remember the classic drives with multiple third down conversions, but we forget all the ones that could jump the odds and failed after giving the defense one too many chances to get off of the field. Explosive plays are essential to a productive modern offense and unless you are running a Chip Kelly or RichRod style ground attack, explosive plays are much more likely through the air than on the ground.
Next week I will follow up with a detailed look on the relative values of Russell Wilson and Montee Ball to Wisconsin’s 2011 offense. Ball had the TDs and the hype and Wilson was considered a quality second option. I’ll dig deep into the numbers and show why Wilson was the real threat of the Wisconsin offense.
Following that, I’ll have the final article in this series looking at how offenses (and maybe moreso defenses) can effectively maximize their expected points for and against through a better perspective on managing offensive output versus managing each down’s success or failure.
We all know it matters. Otherwise there wouldn’t be four major recruiting sites, countless team-specific recruiting blogs and grown men tweeting and facebooking 17 year old high school males, and breathlessly refreshing message boards for the next 14 days.
The question I want to answer is how much does it matter, and where do the numbers play out the most? How much of team success can be predicted based on recruiting profile of the present roster (not the JUCO-stuffed 38 member SEC class that the majority never shows)? Do recruiting services do a better job of predicting offense or defense? Which is more likely to win you conference and national championships, the 5 star running back or the 5 star linebacker?
I have created a complimentary recruiting database that links into my PBP database. For a source I picked Rivals because I wanted to keep it relatively straightforward and they have a full 10-year history online. I only looked at the players who were ranked at their position. Each year that is about 1,000 players and virtually every signee from a major program. Anyone not ranked for their position was omitted. I only have comprehensive rosters for all teams for the last three years, so for that time period I did my best to link the two DBs together. I am sure there are a few that I am missing but I think I got all the Dee Harts linked up with Demetrius Harts and all the other weird things that happen to a recruit's name between recruitment and the official roster.
Each recruit is given an initial value. The value is roughly
[Percentile within position] * [# of stars] ^ 2
So a 5 star #1 at his position recruit is worth about 25 points and a 50th percentile 3 star would be worth 4.5 pts. The initial value is then adjusted based on how long the player has been in the program.
The recruits are then matched up with the final rosters. Players are only counted if they are still on the roster. So any players that have transferred, left school or gone to the NFL are excluded from the totals. The only major gap is transfers. For ones I knew of right away like Cam Newton or Ryan Mallet, they only count at their final school. Most other transfers will only show up at the original school for their time there and then disappear from the grid. Players are then given a “bonus” multiplier based on their experience. Players' initial values are doubled from their first year to their second year and tripled for every year after that.
That’s a lot fewer words than hours put in but in a nutshell, that’s the background for what I will show you below. The magnitude of the points isn’t relevant, all you need to know is the more points the better.
Answer Your Question Already
When you start talking to yourself within an article on mgoblog, there is only one appropriate response, CHART
Lot’s of variation within the numbers but definitely a strong correlation between recruiting points and team PAN [ed: points above normal, the Mathlete's SOS- and situation-adjusted stat]. For all the charts I put up the data will be BCS schools from 2009-2011. Recruits prior to 2009 will be included, but only the actual seasons of play from 2009 on.
There have been some really good seasons from teams with <1,000 pts like Oklahoma St this past season (896). There have also been some mediocre season from teams with 3,000+ points like Texas in 2010 (3,082 pts). But all in all more recruits is better, but we already knew that. So let’s dig a little deeper and see if recruiting rankings mean more for offense or defense and if any position groups are better indicators than others.
Who To Trust, Offense or Defense
Moving to specifics can become a bit more of a challenge. To ease that, I counted every recruit in the position they play, not the position that they are recruited for. They keep the same point total they would at the original position, it just counts in a different bucket. Whether its a WR moving to DB or an ATH finding a home, the points are set based on the initial group ranking, but they are allocated based on the roster position. On to the offense.
The correlation is still there, but it is much weaker for the offense as opposed to the team as a whole. In fact, most of the best offensive seasons were accomplished with relatively average recruiting talent. The ultimate loaded team, 2009 USC, only managed a 3.3 on offense with 10% pts more than any other team I have measured. Teams like the latest incarnations of Michigan and Oregon were able to achieve double digit offensive PAN without elite offensive recruiting classes.
Defensive recruiting is much more correlated with defensive success than offensive. The slope is nearly double and the R-Squared is much greater as well. There are still exceptions like 2009 Florida St who was almost –10 PAN despite over 1,000 defensive recruiting points. There is still success on the lower range but overall there are fewer failures at the top and less success at the bottom of defensive recruiting rankings.
Based on this data, system, player development and finding diamonds in the rough are more prevalent on offense than defense. On defense there is some variation but for the most part you are who you recruit. Unless you hire Greg Robinson and even your Never Forget roster still has 853 points to “earn” a –7 on the season.
The Best Position To Be In
Since the defense as a whole proved to be the most predictive, let’s look there first.
Being a good defense is all about your weakest link and based on that philosophy, you shouldn’t be surprised to see all positions play out relatively equal. None of the position groups is significantly better or worse than another at predicting defensive success.
Offense is where it really gets muddled. O-Line, tight ends and receivers all are moderate correlations between recruiting and offensive success and running backs (as I’ve stated elsewhere) are the most overrated position in football. Quarterback is far and away the highest correlation to offensive success of any position. Even with that QB, is still below all of the defensive positions when it comes to future success on that side of the ball.
How recruiting matches up with success varies greatly by conference. Rather than throw up six more charts, I just put the R^2 values in a table:
Recruiting has virtually no correlation to success over the last three years in the Big East and the PAC 12 but for the other four conferences it's anywhere from a little (Big 12, land of Red River and everyone else) to a lot (the ACC and the SEC).
The Big Ten is in the middle; Ohio St has dominated at the top of both recruiting and success but Michigan’s underachievement and Wisconsin and Nebraska having strong seasons without top tier recruiting classes have thrown in enough variance to disrupt the correlation.
Your 5 Star Takeaway
Recruiting rankings have a huge correlation to future team success, especially on defense. Great teams can come from average talent, but more talent typically means more success. On defense it is virtually impossible to build an elite defense without elite recruits, and its equally true across all defensive positions. On offense dreams of 5 star skill position players are fun, but coaching, player development, system and luck play a much bigger role in future success than they do on defense. With top 20 and higher recruits at nearly every position on defense, Michigan is poised for a very strong future if they can keep the talent around.
I am trying to build a database of all the head coaches, OC’s and DC’s for D1 teams since 2003. Starting digging in and the head coaches are done but the coordinators are going to take more time than I have. Anyone who is interested I created a public Google Doc that has what I have and anyone who has some time to kill and interested in supporting it would be much appreciated. I have completed all the head coaches and the coordinators available readily on Wikipedia for the Big Ten and Big 12. Everything else is missing.
A couple notes for how to update. Every team and year is listed at least three times, one for head coach, one for OC and one for DC. If there is are multiple OC or DC for a given team, just add a new line with the coach team, year and position. If a coordinator role is held by the head coach, just change the coordinator note for the head coach from “No” to “Yes,” do not add a new line. For any mid or late season changes, list only the coordinator most responsible for the season, do not list both.
Thanks again for anyone who is willing to spend a bit of time and start to fill in blanks. Wikipedia team season pages have most of them listed for bigger schools and more recent years, but some of the older years and smaller schools will take a bit more digging, I’ll take whatever the Mgocommunity can provide.
Back at the real job today after a great two weeks of football, ready to start cheering for the team I love, not just the one I
have money on like the uniforms better.
Virginia Tech Preview
PAN, National Rank (leader)
Michigan: +5, 4th (Oregon)
vs VT: +1, 40th
Michigan: +3, 29th (Baylor)
vs VT: +3, 22nd
Michigan: +1, 46th (Alabama)
vs VT: +2, 28th
Michigan: +1, 39th (Texas)
vs. VT: +2, 31st
Michigan: +0, 60th
vs VT: –0, 73rd
A pretty close match-up in all areas except when Michigan is rushing the ball. That’s likely the best avenue for Michigan to leverage. With arrests, suspensions and a late season Michigan surge, special teams could be an opening as well. This should be a close one but this game is Michigan’s to lose, 31-28 Michigan.
Ron Zook Memorial Dumb Punt of the Bowl Season
Haven’t had a chance to review all of the bowl games, but Wisconsin’s first quarter punt has to be the top contender. Two potent offenses, touchdowns on the first three possessions and the Badgers face 4th and 3 at the Oregon 38. Despite the best scrambling QB in college football and an offense geared to pound the ball on the ground against a defense that hadn’t stopped them in their first two drives, Wisconsin played field position. Against Oregon. Two plays later the Black Mamba is flashing chrome 91 yards for a TD in a 7 point Ducks win. Wisconsin is 22/30 on 3rd and 2-4 yards on the season in competitive situations.
Richt/Shaw/NFL Coaching Conservatory
This was just brutal watching teams play for field goals. It’s not a great strategy in the NFL where nearly all the kickers are money inside of 40 and pretty good from 40-55. College kickers, even the good ones, not so much. Here is the table I use to estimate kicker success. From the 25 even a top-notch college kicker is going to miss 20%, an average one is going to miss nearly half the time. Getting a single first down moves the odds significantly.
One other update to the Game Theory Manifesto that I tweeted about during the MSU/Georgia trying not to lose-fest, if you are in the lead and the other team has time-outs left, don’t run up the middle on third down unless you think that’s your best shot to get the first down. The clock is going to stop after your play no matter what. One extra timeout in your opponent’s pocket has very little chance of deciding the game. A first down in most cases will end or nearly end the game. Don’t be careless but if you have a dependable QB like Aaron Murray, throw the ball and give yourself a chance to end the game.
After a year off, a review of the luckiest teams of the year is back. Two years ago I prepped the concept of “lucky.” It’s been a while so here’s a quick recap of how luck is defined for this exercise.
I define luck as what your record should have been with your full season performance and your schedule versus what your record actually was. It is not about injuries, having an easy schedule or even lucky bounces. Interceptions, punt returns and every relevant rush or pass are included in the full season team score. Fumbles, garbage time, interception returns and a few others play are excluded.
So the annual team score which is an average of up to 12 individual game scores (Championship and Bowl Games are excluded) and I take the schedule for the season and re-simulate it based on the actual quality of all the teams, not the pre-season expectation. I then compare actual records to projected records to find this years luckiest teams.
You could call this over or under-achieving and you would be partially right. The challenge is that over the last several years, I have been unable to find any teams that consistently bias one way by any real margin. You can point to grit, toughness, wanting it more and all the clichés, but with no record of teams being able to repeat, I just call it luck.
80 out of 120 FBS teams perform within 1 game of their projected total, leaving about 20 each in the real lucky and unlucky categories.
|Team||Conf||Proj W||Final W||Vs Exp|
|Kansas St||Big XII||7.5||10||2.5|
|Penn St||Big Ten||7.1||9||1.9|
Ball St is the clear-cut winner for luckiest team of 2011, winning 6 when their performance and schedule predicted less than 3. Kansas St, unsurprisingly is the luckiest BCS team going a full 2.5 games better than they “should” have. Going 9-1 in games decided by a touchdown or less will make you feel pretty lucky.
Sugar Bowl opponent Virginia Tech checks in at #6. As has been discussed elsewhere, the Hokies had a strange season. Got blown out by Clemson twice, several decent wins and some strange close games. Ultimately a weak schedule, blowout losses and narrow wins over four unranked opponents lands VT at 2 games better than expected.
RG3 earned his Heisman carrying the pathetic Baylor defense. Baylor’s overall rating was lower due to their horrendous defense, but the offense was good enough to win games giving 48, 31, 26, 39, 38 and 42 points in the regular season. Of course they cap it off by winning while giving up 56 last night.
Penn St surely doesn’t consider themselves lucky right now but 9 wins was nearly two more than they should have had. The Nittany Lions were the anti-Baylor, doing with a stout defense and a really weak offense. Penn St went 6-2 in Big Ten play despite being outscored and also beat Temple by 4 in a game they had no business pulling out.
|Team||Conf||Proj W||Final W||Vs Exp|
|Texas A&M||Big XII||8.5||6||(2.5)|
|S Florida||Big East||7.2||5||(2.2)|
|Louisiana Mon||Sun Belt||6.1||4||(2.1)|
Some familiar names show up. UCF is tops at 3 games worse than expected. UCF had BCS buster talk heading into the season and failed to even qualify for a bowl. Even the 8 projected wins would have been a disappointment but going 0-6 in games decided by a TD or less really hurts.
Texas A&M’s trials this season are well-documented. They blew leads like crazy, lost five games to bowl teams in OT or by 4 points or less. The Aggies have to feel like in some alternate universe their season and Kansas St’s is switched.
The Big Ten+
|Team||Conf||Proj W||Final W||Vs Exp|
|Ohio St||Big Ten||6.9||6||(0.9)|
|Michigan St||Big Ten||9.3||10||0.7|
|Penn St||Big Ten||7.1||9||1.9|
Michigan checks in at slightly lucky. They were a toss-up between 9 and 10 wins and hit the over. The Wolverines had only 3 games decided by a possession or less and went 2-1 in them.
Nothing is more frustrating for a football fan (especially a math/logic centered one) than to see coaches blow basic strategy elements to the game, many of which are black and white. I can be forgiving on a lot. As a 100-and-nothing-pound Mathlete growing up, I knew I wouldn’t do any better out there on the field but I knew a lot of times I could do better at some of the basic decision making and strategy. Coaches learn a lot about football as they progress through careers, but game-theory type strategy seems to be a common blind spot for many coaches to gain the hidden advantage. Luckily Michigan has a coach that had a pretty sharp first year in this regard, hopefully all the other Zookers out there don’t read this (yeah right) and catch up.
Some of what follows will be backed with hard data from my database, other will be solid and based on strong logic without play data backing (I don’t track timeouts in my database) and some will be things unprovable but backed by 30 years of watching the game from a different vantage point.
Issue 1: How to use your timeouts when you are trailing
Always take your timeouts on defense if there are less than 2-3 minutes left, you are trailing and the opponent is running out the full play clock.
Announcers always like to have have those final timeouts in your pocket for that last drive or to get the kicker out there one last time. Although that is a good to have, saving the time up front is a much better option for end game strategy. If your opponent is working to run out the clock, every timeout you take saves you about 38 seconds. On offense, if you have a reasonably efficient 2 minute offense, each timeout is probably worth 10-15 seconds. One timeout on defense is as valuable as 2-3 on offense. The other advantage you have on offense is you are in control of the play, you can restrict your plays to passing, sidelines and first down distance plays that assist in stopping the clock.
A final note on defensive timeouts, never take them immediately after the offense gains a first down. The clock stops to reset the ball and you will have three more opportunities to stop it later. Taking it after the first down is good for about 25 seconds, waiting will give you nearly 15 seconds more savings.
Issue 2: Should you ever use them when you are ahead
If the opponent is in the red zone pressing for a tying or go-ahead score, don’t be afraid to use your timeouts to ensure a chance to retake the lead.
Once the opponent is inside the 20 (and definitely inside the 10) in the final two minutes, losing the lead is a near certainty (especially if a field goal will do the job). If the opponent is drained or nearly drained of timeouts, all the better for you to use yours. They won’t be able to stop you from taking a knee if you do get a stop and you will have more time to come back if they do score on you.
The flip side is also true. Even if you are trailing but driving, if the opponent is already in a position to run out the clock if you are stopped, first priority is obviously scoring, but second should be not rushing to be faster because the goal is to make your drive the last drive. If it fails the game is over either way, if it succeeds you want to minimize the opponent's chance to score. Even if you started as a two-minute offense, if you get to the red zone fast enough it can make sense to slow down a bit. It will likely be your last possession no matter what, any time you are saving is for them and not you.
Note, this situation also applies to a tie game.
Issue 3: How to use your timeouts when you are tied
Tread lightly after first and second down but pull the trigger fast after third down.
As Bret Bielema and Bob Stoops found out this year, you have to be very careful on this one. Taking one after first down is the riskiest. At least Bielema took his with a Sparty offense facing a 2nd and 20. With two more plays left a lot can happen, unless the opponent is backed up deep in their own territory, it’s best to not get greedy after first down.
After second down can still be a bit risky, but at this point you have a much better idea on what the opponent is faced with and what their strategy is. The worse shape and more conservative the opponent is, the more a timeout makes sense.
After third down in a tied game is usually a straightforward decision. Unless their is a chance the other team will attempt a fourth down conversion, use the timeout right away, it’s your last chance to maximize the value of that timeout.
Issue 4: How soon can you take a knee and run out the clock?
Depends on your opponent timeouts,
0 left: 2 minutes and 6 seconds
1 left: 1 minute and 24
2 left: 46 seconds
3 left: 8 seconds
If there are a couple more seconds than this it gets dicey. If you lead by more than 2 points, you can always have your quarterback sprint backwards and run around to burn some time and take a safety if need be.
2 Point Conversions
Rule #1 of 2 point conversions is don’t even think about them until the fourth quarter. No exceptions. Do not chase points, there is too much variability left in the game to give up a point to get to a “nice number” or even worse to “get back” a point from a missed/blocked/botched PAT. Just don’t do it.
Two point conversion rates are hard to get a true number on. The best numbers I can get is somewhere between 40-45% success. This is backed by a limited sample on actual two point conversions and verified by 43% success on 3rd or 4th and goal from the 3.
Situations to go for 2 any time in the fourth quarter (margin before TD is scored):
Down 22: Prior to scoring this you were down 22, 3 touchdown, and 4 PAT points. That can come as 1/1/2 or 0/2/2. Going for two here is the only way to get even but still leave the door open. Make it and its 14, miss and it’s still a two possession game.
Down 15: In the fourth quarter possessions are limited. Forcing the decision early gives you the information on whether its a one or two possession game. Waiting till the second TD can leave you with a false sense that you are playing a one possession game when you have a less than 50% of hitting the 2 point conversion and may be out of time. Going for it after the first score allows you to make more educated timeout, on-side and fourth down decisions. Waiting may make you feel better about it still be a one possession game, but as Brian has said, it’s only a 40% chance of being a one possession game.
Down 14: This is the genius one. Fail and you still have a chance to get it back on the second TD, just as if you kicked it like a risk-averse NFL coach. But if you make it, you are not playing for overtime but the win. Depending on the 2 point conversion odds, this increases your chances of winning by 10-30% (not percentage points, you were down 2 TD in the fourth quarter, your odds are never great). At a 43% success rate this strategy is a 14% improvement of odds. With a good offense at 50% conversion you jump all the way to 29% improvement by going for 2 on the first TD.
Only if there are 2 or fewer possessions left for each team
Down 8, Down 5, Down 1 and Up 6: Same as above, with 3+ possessions left this is probably a no go but with a possession or two each to go. It’s now a one possession game and with very many possessions left there is too much that can happen to risk giving up the point too early.
Only if your opponent has one possession left
Down 2: You hear there’s no difference between 4 and 6 but there is if there are multiple possessions left. Field goals can really mess with this situation; take the point unless the opponent only has one shot left, in which case you might get a little insurance for a missed PAT if you can make yours. Even in this situation I don’t condone going for 2 when you were down 12 to make the deficit 4 or 6. Chances are your opponent is going to be conservative and a field goal is probably the best case scenario for them. Don’t let a FG end your game with an unnecessary risk.
Wanted to break things up a with a little Mgoblog favorite, a chart.
Outcomes from first possession of overtime period.
|Outcome||Win||Loss||Another OT||Win Odds|
|Fail to score||0||32||3||5%|
|Touchdown + PAT||33||4||31||73%|
This data comes from every overtime period from 2007-2011. What you don’t see here is the strong preference this overtime method has for winning the coin flip. There is a lot of talk about the NFL and its 59-60% advantage for the coin flip winner, but in college the coin flip winner holds a solid 56% advantage for getting to go second and knowing what you have to do.
A touchdown on the first possession puts you in great shape. A field goal attempt is OK if necessary, but you better be confident you can make it. Although there is obviously a greater chance of winning with a TD versus a field goal, the odds don’t support a highly aggressive fourth down strategy, especially inside the 10. Even though the temptation is higher close to the goal line, for most teams going for it on 4th and 1 or 2 make sense. Anything beyond that and the best bet is to give the ball to the kicker. Now a great offense or questionable kicking game quickly changes the calculus, but in close, the odds say kick it. Where it is a bit more interesting is on the first set of downs in the game. The odds actually favor a more aggressive 4th down strategy on 4 and 5 or less from the 16-20 on the first set of downs. At this range most college kickers are good but far from automatic; an aggressive play here can pay out.
Another hotly debated overtime question is going for 2 to win after the other team has scored and kicked. 10 out of 44 teams faced with this proposition have gone for it, their record is 4-6, about in line with the 40-45% 2 point conversion expectation. This would seem like a losing proposition but at 45% the odds would be in line with the chances in the next OT since you have to be on offense first. Not really a clear cut answer here, but either way can be justified and the presence of a great offense from either team can quickly make the decision to go for 2 a good one.
Surprise On-Sides Kicks
Do them more.
OK you need more than that? Advanced NFL stats ran the numbers for the NFL and found that success rates for onside kicks are 20% when expected and 60% when not expected. I found a similar spread for college. Out of 663 expected onside kicks in my database, 23% were recovered by the kicking team. Only 146 (about 1 per week) surprise onsides where tried but 64% of those were recovered. The break-even success rate needed for a surprise onside kick is 46%, the market for surprise on-side kicks is definitely undervalued.
Punting In Opponent Territory
One of the many reasons that punting in opponent territory is dumb is that it is usually couched on the assumption that “we’ll pin them deep.” There are two key problems with this assumption. The first is that 36% of punts from opponent territory result in a touchback or never reach the 20, and that’s before any returns are factored in. The second is that it’s pretty tough to actually down it close to the end zone, and unless you are at the 1 or 2, there is no special advantage.
As discussed previously, it is in an offense’s best interest to go super conservative at the 1 or 2. Outside of that it is nearly business as usual. There is only an 8% chance a punt from the opponent's territory is downed at the 1 or 2. It’s over four times more likely to not even pin an opponent inside the 20 than it is to force the offense’s hand by pinning them at the 1 or 2 yard line.
Another problem with an opponent territory punt is that it’s tough to get an even exchange. Punting into the short side of the field limits the best case scenario and assuming you can force a punt from the opponent, gives them a lot of positive variance opportunity. A long bounce going in brings the ball out to the 20, a long bounce kicking out can quickly turn into a 60-yard punt and a total flip of field position.
And of course, you give up a great scoring opportunity punting in opponent territory.
Red Zone Play-Calling
On a first down Red Zone play, teams are more likely to score if it’s a run than a pass if they are at the 8 yard line or closer. Anything between the 9 and the 20 favors a pass on first down. That doesn’t mean that 100% pass is the optimal strategy, just that the play calling should favor the pass (or run inside the 9). For goal to go situations after first down, second down is the ultimate OC’s choice. From anywhere 10 and in on second and goal running and passing have nearly identical touchdown percents. On third and goal, the run still holds up strongly. A called run is more likely to score a TD on anything from the 6 and in than a pass, which owns 7 and up. Again, not saying the strategy should be 100%, but there is real value to favoring the run inside the 7.
Never take a touchback on a kickoff you don’t have to. The expected starting field position on a return from 9 yards deep is still the 21, plus the opportunity for a big play easily offsets the times when you start from the 10-15, which isn’t a big cost for the opportunity.
In a trash tornado game, the biggest value for the wind goes to the team that has it in the first and third quarters, not the fourth. When the wind is strong it usually takes a possession or two for the field position to level back out. Those possessions occur at the beginning of the 2nd and 4th quarters, essentially giving the team with early field position the wind for about 2/3 of the game.
No numbers on this one but unless it's fourth down, stretching out the football is an extremely dumb move. At the goal line you can make a case for it if its 2nd or 3rd down, but there are very few situations where an incremental yard (nearly worthless) can be offset by the fumble risk of stretching the ball out.
Anything I may have missed here that you want to see, hit me up on twitter or in the comments and I’ll find a spot to address in a future article. I intentionally skipped fourth down decision making for this article. It’s too big of a topic. I previously wrote about it here and have an update to the article coming sometime this offseason.
May your holidays be filled with surprise on-side kicks, fourth down attempts and three wise timeouts.